Differential Equations
Guest Editors: Fawang Liu, Mark M. Meerschaert, Shaher Momani,
Nikolai N. Leonenko, Wen Chen, and Om P. Agrawal
International Journal of Differential Equations
Fractional Differential Equations
International Journal of Differential
Equations
Fractional Differential Equations
Guest Editors: Fawang Liu, Mark M. Meerschaert, Shaher Momani,
Nikolai N. Leonenko, Wen Chen, and Om P. Agrawal
Copyright q 2010 Hindawi Publishing Corporation. All rights reserved.
This is an issue published in volume 2010 of International Journal of Differential Equations. All articles are open
access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribu
tion, and reproduction in any medium, provided the original work is properly cited.
Editorial Board
Om Agrawal, USA
Bashir Ahmad, Saudi Arabia
Ch erif Amrouche, France
Sabri Arik, Turkey
Dumitru Baleanu, Turkey
Nicola Bellomo, Italy
Vieri Benci, Italy
Elena Braverman, Canada
Alberto Cabada, Spain
Jinde Cao, China
Der Chen Chang, USA
Yang Chen, UK
Gui Qiang Chen, USA
Fengde Chen, China
Charles E. Chidume, Italy
Y. J. E. Cho, Republic of Korea
I. D. Chueshov, Ukraine
Shangbin Cui, China
Toka Diagana, USA
Jinqiao Duan, USA
M. ElGebeily, Saudi Arabia
Ahmed M. ElSayed, Egypt
Khalil Ezzinbi, Morocco
Zhaosheng Feng, USA
Daniel Franco, Spain
Davood Domiri Ganji, Iran
Weigao Ge, China
Fritz Gesztesy, USA
Yoshikazu Giga, Japan
Jaume Gin e, Spain
Jerome A. Goldstein, USA
Said R. Grace, Egypt
John R. Graef, USA
Maurizio Grasselli, Italy
Tasawar K. Hayat, Pakistan
Jihuan He, China
Emmanuel Hebey, France
Helge Holden, Norway
Mayer Humi, USA
Elena I. Kaikina, Mexico
Qingkai Kong, USA
DeXing Kong, China
M. O. Korpusov, Russia
A. M. Krasnoselskii, Russia
Miroslav Krstic, USA
P. A. Krutitskii, Russia
M. R. S. Kulenovc, USA
Karl Kunisch, Austria
Alexander Kurganov, USA
Jos e A. Langa, Spain
Philippe G. Leoch, France
Nikolai N. Leonenko, UK
Yuji Liu, China
Fawang Liu, Australia
Eduardo Liz, Spain
Wen Xiu Ma, USA
Ruyun Ma, China
T. R. Marchant, Australia
Marco Marletta, UK
R. V. N. Melnik, Canada
S. A. Messaoudi, Saudi Arabia
Stanisaw Mig orski, Poland
Andro Mikelic, France
Shaher Momani, Jordan
Gaston M. NGu er ekata, USA
Juan Jose Nieto, Spain
Muhammad A. Noor, Pakistan
Sotiris K. Ntouyas, Greece
Donal ORegan, Ireland
Jong Yeoul Park, Korea
Pablo Pedregal, Spain
Kanishka Perera, USA
Rodrigo Lopez Pouso, Spain
Ramon Quintanilla, Spain
Patrick J. Rabier, USA
Youssef N. Raoul, USA
Themistocles M. Rassias, Greece
Yuri V. Rogovchenko, Sweden
Samir H. Saker, Egypt
Martin Schechter, USA
William E. Schiesser, USA
Leonid Shaikhet, Ukraine
Zhiqiang Shao, China
Qin Sheng, USA
Junping Shi, USA
Panayiotis D. Siafarikas, Greece
Peter L.
Simon, Hungary
Stevo Stevic, Serbia
I. G. Stratis, Greece
JianPing Sun, China
Guido H. Sweers, Germany
Nasser Eddine Tatar, Saudi Arabia
Roger M. Temam, USA
Gunther A. Uhlmann, USA
Jan Van Neerven, USA
Aghalaya S. Vatsala, USA
Pei Guang Wang, China
Mingxin Wang, China
Lihe Wang, USA
Mingxin Wang, Argentina
ZhiQiang Wang, USA
Thomas P. Witelski, USA
Gershon Wolansky, Israel
P.J.Y. Wong, Singapore
Jen Chih Yao, Taiwan
Jingxue Yin, China
Jianshe S. Yu, China
Vyacheslav A. Yurko, Russia
Qi S. Zhang, USA
Sining Zheng, China
Songmu Zheng, China
Yong Zhou, China
Feng Zhou, China
Wenming Zou, China
Xingfu Zou, Canada
Contents
Fractional Dierential Equations, Fawang Liu, Mark M. Meerschaert, Shaher Momani,
Nikolai N. Leonenko, Wen Chen, and Om P. Agrawal
Volume 2010, Article ID 215856, 2 pages
The MWright Function in TimeFractional Diusion Processes: A Tutorial Survey,
Francesco Mainardi, Antonio Mura, and Gianni Pagnini
Volume 2010, Article ID 104505, 29 pages
Stability and Convergence of an Eective Numerical Method for the TimeSpace Fractional
FokkerPlanck Equation with a Nonlinear Source Term, Qianqian Yang, Fawang Liu,
and Ian Turner
Volume 2010, Article ID 464321, 22 pages
Hes Variational Iteration Method for Solving Fractional Riccati Dierential Equation,
H. Jafari and H. Tajadodi
Volume 2010, Article ID 764738, 8 pages
Solitary Wave Solutions for a TimeFraction Generalized HirotaSatsuma Coupled KdV
Equation by a New Analytical Technique, Majid Shateri and D. D. Ganji
Volume 2010, Article ID 954674, 11 pages
TimeOptimal Control of Systems with Fractional Dynamics, Christophe Tricaud and
YangQuan Chen
Volume 2010, Article ID 461048, 16 pages
On the Speed of Spread for Fractional ReactionDiusion Equations, Hans Engler
Volume 2010, Article ID 315421, 16 pages
Riesz Potentials for Kortewegde Vries Solitons and SturmLiouville Problems,
Vladimir Varlamov
Volume 2010, Article ID 193893, 18 pages
Linear Fractionally Damped Oscillator, Mark Naber
Volume 2010, Article ID 197020, 12 pages
Positive Solution to Nonzero Boundary Values Problem for a Coupled System of Nonlinear
Fractional Dierential Equations, Jinhua Wang, Hongjun Xiang, and Zhigang Liu
Volume 2010, Article ID 186928, 12 pages
The Periodic Solutions of the Compound Singular Fractional Dierential System with Delay,
XuTing Wei and XuanZhu Lu
Volume 2010, Article ID 509286, 9 pages
The Use of Fractional BSplines Wavelets in Multiterms Fractional Ordinary Dierential
Equations, X. Huang and X. Lu
Volume 2010, Article ID 968186, 13 pages
Solvability of Nonlinear Langevin Equation Involving Two Fractional Orders with Dirichlet
Boundary Conditions, Bashir Ahmad and Juan J. Nieto
Volume 2010, Article ID 649486, 10 pages
On the Selection and Meaning of Variable Order Operators for Dynamic Modeling,
Lynnette E. S. Ramirez and Carlos F. M. Coimbra
Volume 2010, Article ID 846107, 16 pages
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 215856, 2 pages
doi:10.1155/2010/215856
Editorial
Fractional Differential Equations
Fawang Liu,
1
Mark M. Meerschaert,
2
Shaher Momani,
3
Nikolai N. Leonenko,
4
Wen Chen,
5
and Om P. Agrawal
6
1
School of Mathematical Sciences, Queensland University of Technology, P.O. Box 2434,
Brisbane, Qeensland 4001, Australia
2
Department of Statistics and Probability, Michigan State University, A416 Wells Hall,
East Lansing, MI 48823, USA
3
Department of Mathematics, Mutah University, P.O. Box 7, Mutah 61710, Jordan
4
School of Mathematics, Cardi University, Cardi CF2 4YH, UK
5
Department of Engineering Mechanics, Hohai University, Xikang Road Number 1,
Nanjing, Jiangsu 210098, China
6
Department of Mechanical Engineering and Energy Processes, Southern Illinois University,
Carbondale, IL 62901, USA
Correspondence should be addressed to Fawang Liu, f.liu@qut.edu.au
Received 13 April 2010; Accepted 13 April 2010
Copyright q 2010 Fawang Liu et al. This is an open access article distributed under the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
It is my pleasure to present this special issue of the International Journal of Dierential
Equations dedicated to Fractional Dierential Equations FDEs.
Fractional dierential equations are generalizations of ordinary dierential equations
to an arbitrary noninteger order. Fractional dierential equations have attracted consider
able interest because of their ability to model complex phenomena. These equations capture
nonlocal relations in space and time with powerlaw memory kernels. Due to the extensive
applications of FDEs in engineering and science, research in this area has grown signicantly
all around the world.
This special issue of Fractional Dierential Equations consists of one review article
paper 1 and 12 original articles covering various aspects of FDEs and their applications by
the prominent researchers in the eld.
The review article surveys the properties of a transcendental function of the Wright
type, nowadays known as the MWright function, entering as a probability density
in a relevant class of selfsimilar stochastic processes. The second paper presents a
computationally eective numerical method for the timespace fractional FokkerPlanck
equations with a nonlinear source term, together with a stability and convergence analysis
of the numerical method.
2 International Journal of Dierential Equations
The third paper uses Hes variational iteration method for solving the fractional Riccati
dierential equation. The fourth paper proposes a fractional iteration method and has been
applied to study the HirotaSatsuma coupled KdV of timefractionalorder equation.
The fth paper develops a formulation for fractional timeoptimal control problems.
The sixth paper derives conditions for the speed of spread of solutions of fractional
scalar reactiondiusion equations to be nite or innite. The seventh paper studies the
Riesz fractional derivatives for the Kortewegde Vries solitons and proves that these Riesz
potentials and their Hilbert transforms are linearly independent solutions of a Sturm
Liouville problem.
The eighth paper gives the analytical solution of the linear fractionally damped
oscillator equation by means of Laplace transform. The ninth paper proves the existence
and uniqueness of a positive solution to the nonzero boundary value problem for a coupled
system of fractional dierential equations. The tenth paper gives sucient conditions on
the existence of a periodic solution for a class of compound singular fractional dierential
systems with delay, involving Nishimoto fractional derivative. The eleventh paper discusses
the existence and uniqueness of the solutions of the nonhomogeneous linear dierential
equations of arbitrary positive real order by using the fractional BSplines wavelets and
the MittagLeer function. The twelfth paper proves the existence of solutions of the
fractional Langevin equation with two dierent fractional orders in a Banach space using
the contraction mapping principle and Krasnoselskiis xed point theorem. The nal paper
addresses the selection and meaning of variableorder operators for dynamic modelling.
Thus, this special issue provides a wide spectrum of current research in the area of
FDEs, and I hope that experts in this and related elds would nd it useful.
Finally, I would like to acknowledge the guest editors Mark M. Meerschaert, Shaher
Momani, Nikolai N. Leonenko, We Chen, and Om P. Agrawal for their assistance in
producing this special edition within the specied timeframe.
Fawang Liu
Mark M. Meerschaert
Shaher Momani
Nikolai N. Leonenko
Wen Chen
Om P. Agrawal
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 104505, 29 pages
doi:10.1155/2010/104505
Review Article
The MWright Function in TimeFractional
Diffusion Processes: A Tutorial Survey
Francesco Mainardi,
1
Antonio Mura,
2
and Gianni Pagnini
3
1
Department of Physics, University of Bologna and INFN, Via Irnerio 46, 40126 Bologna, Italy
2
CRESME Ricerche S.p.A, Viale Gorizia 25C, 00199 Roma, Italy
3
CRS4, Centro Ricerche Studi Superiori e Sviluppo in Sardegna, Polaris Building 1,
09010 Pula, Italy
Correspondence should be addressed to Francesco Mainardi, francesco.mainardi@bo.infn.it
Received 13 September 2009; Accepted 8 November 2009
Academic Editor: Fawang Liu
Copyright q 2010 Francesco Mainardi et al. This is an open access article distributed under
the Creative Commons Attribution License, which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
In the present review we survey the properties of a transcendental function of the Wright type,
nowadays known as MWright function, entering as a probability density in a relevant class of
selfsimilar stochastic processes that we generally refer to as timefractional diusion processes.
Indeed, the master equations governing these processes generalize the standard diusion equation
by means of timeintegral operators interpreted as derivatives of fractional order. When these
generalized diusion processes are properly characterized with stationary increments, the M
Wright function is shown to play the same key role as the Gaussian density in the standard and
fractional Brownian motions. Furthermore, these processes provide stochastic models suitable for
describing phenomena of anomalous diusion of both slow and fast types.
1. Introduction
By timefractional diusion processes, we mean certain diusionlike phenomena governed
by master equations containing fractional derivatives in time whose fundamental solution
can be interpreted as a probability density function pdf in space evolving in time. It
is well known that, for the most elementary diusion process, the Brownian motion, the
master equation, is the standard linear diusion equation whose fundamental solution is
the Gaussian density with a spatial variance growing linearly in time. In such case we speak
about normal diusion, reserving the term anomalous diusion when the variance grows
dierently. A number of stochastic models for explaining anomalous diusion have been
introduced in literature; among themwe like to quote the fractional Brownian motion; see, for
example, 1, 2, the Continuous Time Random Walk; see, for example, 36, the L evy ights;
see, for example, 7, the Schneider grey Brownian motion; see 8, 9, and, more generally,
2 International Journal of Dierential Equations
random walk models based on evolution equations of single and distributed fractional order
in time and/or space; see, for example, 1018.
In this survey paper we focus our attention on modications of the standard diusion
equation, where the time can be stretched by a power law t t
z, M
, which allows us to obtain its Laplace and Fourier transforms. In Section 5 we consider
some types of generalized diusion equations containing time partial derivatives of fractional
order and we express their fundamental solutions in terms of the MWright functions
evolving in time with a given selfsimilarity law. In Section 6 we stress how the MWright
function emerges as a natural generalization of the Gaussian probability density for a class
of selfsimilar stochastic processes with stationary increments, depending on two parameters
, . These processes are dened in a unique way by requiring the determination of any
multipoint probability distribution and include the wellknown standard and fractional
Brownian motion. We refer to this class as the generalized grey Brownian motion ggBm,
because it generalizes the grey Brownian motion gBm introduced by Schneider 8, 9.
Finally, a short concluding discussion is drawn. In Appendix A we derive the fundamental
solution of the timefractional diusion equation. In Appendix B we outline the relevance of
the MWright function in timefractional drift processes entering as subordinators in time
fractional diusion.
2. Notions and Notations
2.1. Integral Transforms Pairs
In our analysis we will make extensive use of integral transforms of Laplace, Fourier, and
Mellin types; so we rst introduce our notation for the corresponding transform pairs. We
do not point out the conditions of validity and the main rules, since they are given in any
textbook on advanced mathematics.
Let
fs L
_
fr; r s
_
_
0
e
sr
frdr 2.1
International Journal of Dierential Equations 3
be the Laplace transform of a suciently wellbehaved function fr with r R
, s C, and
let
fr L
1
_
fs; s r
_
1
2i
_
Br
e
sr
fsds 2.2
be the inverse Laplace transform, where Br denotes the socalled Bromwich path, a straight
line parallel to the imaginary axis in the complex splane. Denoting by
L
the justaposition
of the original function fr with its Laplace transform
fs, the Laplace transformpair reads
fr
L
fs.
2.3
Let
f F
_
fx; x
_
_
e
ix
fxdx 2.4
be the Fourier transform of a suciently wellbehaved function fx with x R, R, and
let
fx F
1
_
f; x
_
1
2
_
e
ix
fd 2.5
be the inverse Fourier transform. Denoting by
F
the justaposition of the original function
fx with its Fourier transform
f, the Fourier transform pair reads
fx
F
f.
2.6
Let
f
s M
_
fr; r s
_
_
0
r
s1
frdr 2.7
be the Mellin transform of a suciently wellbehaved function fr with r R
, s C, and let
fr M
1
_
f
s; s r
_
1
2i
_
Br
r
s
f
sds 2.8
be the inverse Mellin transform. Denoting by
M
the justaposition of the original function
fr with its Mellin transform f
s.
2.9
4 International Journal of Dierential Equations
2.2. Essentials of Fractional Calculus with Support in R
Fractional calculus is the branch of mathematical analysis that deals with pseudodierential
operators that extend the standard notions of integrals and derivatives to any positive
noninteger order. The term fractional is kept only for historical reasons. Let us restrict
our attention to suciently wellbehaved functions ft with support in R
. Two main
approaches exist in the literature of fractional calculus to dene the operator of derivative
of noninteger order for these functions, referred to RiemannLiouville and to Caputo. Both
approaches are related to the socalled RiemannLiouville fractional integral dened for any
order > 0 as
J
t
ft :
1
_
_
t
0
t
1
fd. 2.10
We note the convention J
0
t
I Identity and the semigroup property
J
t
J
t
J
t
J
t
J
t
, 0, 0. 2.11
The fractional derivative of order > 0 in the RiemannLiouville sense is dened as the
operator D
t
which is the left inverse of the RiemannLiouville integral of order in analogy
with the ordinary derivative, that is,
D
t
J
t
I, > 0. 2.12
If m denotes the positive integer such that m 1 < m, we recognize, from 2.11 and
2.12, D
t
ft : D
m
t
J
m
t
ft; hence,
D
t
ft
d
m
dt
m
_
1
_
m
_
_
t
0
fd
t
1m
_
, m 1 < < m,
d
m
dt
m
ft, m.
2.13
For completeness we dene D
0
t
I.
On the other hand, the fractional derivative of order > 0 in the Caputo sense is dened
as the operator
D
t
such that
D
t
ft : J
m
t
D
m
t
ft; hence,
t
ft
_
m
_
_
t
0
f
m
d
t
1m
, m 1 < < m,
d
m
dt
m
ft, m.
2.14
International Journal of Dierential Equations 5
We note the dierent behavior of the two derivatives in the limit m 1
. In fact,
m 1
t
ft D
m
t
J
1
t
ft D
m1
t
ft,
t
ft J
1
t
D
m
t
ft D
m1
t
ft D
m1
t
f0
,
2.15
where the limit for t 0
t
t
_
1
_
_
1
_t
,
D
t
t
_
1
_
_
1
_t
,
> 0, > 1, 2.16
we nd the relationship between the two types of fractional derivative
D
_
ft
m1
k0
t
k
k!
f
k
0
_
D
t
ft. 2.17
We note that the Caputo denition for the fractional derivative incorporates the initial values
of the function and of its integer derivatives of lower order. The subtraction of the Taylor
polynomial of degree m 1 at t 0
t
ft; t s
_
fs
s
, 0.
2.18
For the Caputo fractional derivative, we consequently get
L
_
t
ft; t s
_
s
fs
m1
k0
s
1k
f
k
0
, m 1 < m, 2.19
where f
k
0
: lim
t 0
f
k
t. The corresponding rule for the RiemannLiouville fractional
derivative is more cumbersome and it reads
L
_
D
t
ft; t s
_
s
fs
m1
k0
_
D
k
t
J
m
t
_
f0
s
m1k
, m 1 < m, 2.20
6 International Journal of Dierential Equations
where the limit for t 0
t
ft; t s
_
s
fs, m 1 < < m. 2.21
In the special case f
k
0
n0
z
n
n!
_
n
_, > 1, C. 3.1
Originally, Wright assumed that 0, and, only in 1940 32, he considered 1 < < 0.
We note that in Chapter 18 of Vol. 3 of the handbook of the Bateman Project 33, devoted to
Miscellaneous Functions, presumably for a misprint, the parameter of the Wright function is
restricted to be nonnegative. When necessary, we propose to distinguish the Wright functions
in two kinds according to 0 rst kind and 1 < < 0 second kind.
For more details on Wright functions the reader can consult, for example, 3441 and
references therein.
The integral representation of the Wright function reads
W
,
z
1
2i
_
Ha
e
z
d
, > 1, C, 3.2
where Ha denotes the Hankel path. We remind that the Hankel path is a loop that starts from
along the lower side of the negative real axis, encircles the circular area around the origin
with radius 0 in the positive sense, and ends at along the upper side of the negative
International Journal of Dierential Equations 7
real axis. The equivalence of the series and integral representations is easily proved using
Hankel formula for the Gamma function
1
_
Ha
e
u
u
du, C 3.3
and performing a termbyterm integration. In fact,
W
,
z
1
2i
_
Ha
e
z
d
1
2i
_
Ha
e
n0
z
n
n!
n
_
d
n0
z
n
n!
_
1
2i
_
Ha
e
n
d
_
n0
z
n
n!
_
n
_.
3.4
It is possible to prove that the Wright function is entire of order 1/1 ; hence, it is of
exponential type only if 0 which corresponds to Wright function of the rst kind. The
case 0 is trivial since W
0,
z e
z
/, provided that
/
0, 1, 2, . . . .
3.2. The Auxiliary Functions of the Wright Type
Mainardi, in his rst analysis of the timefractional diusion equation 42, 43, aware of the
Bateman handbook 33, but not yet of the 1940 paper by Wright 32, introduced the two
Wrighttype entire auxiliary functions,
F
z : W
,0
z, 0 < < 1,
M
z : W
,1
z, 0 < < 1,
3.5
interrelated through
F
z zM
z. 3.6
As a matter of fact, functions F
z and M
z :
n1
z
n
n! n
1
n1
z
n1
n!
n 1 sinn, 3.7
M
z :
n0
z
n
n! n 1
1
n1
z
n1
n 1!
n sinn. 3.8
The second series representations in 3.73.8 have been obtained by using the reection
formula for the Gamma function 1 /sin .
8 International Journal of Dierential Equations
As an exercise, the reader can directly prove that the radius of convergence of the
power series in 3.73.8 is innite for 0 < < 1 without being aware of Wrights results, as
it was shown independently by Mainardi 42; see also 27.
Furthermore, we have F
0 0 and M
z :
1
2i
_
Ha
e
z
d, 3.9
M
z :
1
2i
_
Ha
e
z
d
1
. 3.10
We note that relation 3.6 can be obtained also from 3.93.10 with an integration by parts.
In fact,
M
z
_
Ha
e
z
d
1
_
Ha
e
1
z
d
d
e
z
_
d
1
z
_
Ha
e
z
d
F
z
z
.
3.11
The equivalence of the series and integral representations is easily proved by using the
Hankel formula for the Gamma function and performing a termbyterm integration.
3.3. Special Cases
Explicit expressions of F
z and M
m0
1
m
_
1
2
_
m
z
2m
2m!
1
exp
_
z
2
4
_
, 3.12
M
1/3
z
1
2/3
m0
_
1
3
_
m
z
3m
3m!
1
1/3
m0
_
2
3
_
m
z
3m1
3m 1!
3
2/3
Ai
_
z
3
1/3
_
,
3.13
where Ai denotes the Airy function.
International Journal of Dierential Equations 9
Furthermore, it can be proved that M
1/q
z satises the dierential equation of order
q 1
d
q1
dz
q1
M
1/q
z
1
q
q
zM
1/q
z 0, 3.14
subjected to the q 1 initial conditions at z 0, derived from 3.14, such that
M
h
1/q
0
1
h
_
1 h 1/q
_
1
h
_
h 1
q
_
sin
_
h 1
q
_
, 3.15
with h 0, 1, . . . , q 2. We note that, for q 4, 3.14 is akin to the hyperAiry dierential
equation of order q 1; see, for example, 44. Consequently, the auxiliary function M
z
could be considered a sort of generalized hyperAiry function. However, in view of further
applications in stochastic processes, we prefer to consider it as a natural fractional
generalization of the Gaussian function, similarly as the MittagLeer function is known
to be the natural fractional generalization of the exponential function.
To stress the relevance of the auxiliary function M
_
1
r
_
r
1
M
_
1
r
_
L
e
s
_
1
r
_
1
r
_
1
r
_
L
e
s
s
1
, 0 < < 1. 4.2
We note that the inversion of the Laplace transform of the exponential exps
is relevant
since it yields for any 0, 1 the unilateral extremal stable densities in probability theory,
denoted by L
; s r
_
1
2i
_
Br
e
srs
ds
1
2i r
_
Ha
e
/r
1
r
F
_
1
r
_
r
1
M
_
1
r
_
.
4.3
Expanding in power series the Laplace transform and inverting term by term, we formally
get
L
1
_
exps
n0
1
n
n!
L
1
s
n
n1
1
n
n!
r
n1
n
1
r
F
_
1
r
_
r
1
M
_
1
r
_
,
4.4
where now we have used the series representation 3.7 for the function F
r when r . Choosing a
variable r/ rather than r, the computation of the desired asymptotic representation by the
saddlepoint approximation is straightforward. Mainardi and Tomirotti 43 have obtained
M
_
r
_
ar
1/2/1
exp
_
br
1/1
_
,
a
1
_
21
> 0, b
1
> 0.
4.5
The above evaluation is consistent with the rst term in the asymptotic series expansion
provided by Wright with a cumbersome and formal procedure for his general function W
,
when 1 < < 0; see 32. In 1999 Wong and Zhao have derived asymptotic expansions of
the Wright functions of the rst and second kind in the whole complex plane following a new
method for smoothing Stokes discontinuities; see 40, 41, respectively.
We note that, for 1/2 as 4.5 provides the exact result consistent with 3.12,
M
1/2
2r
1
e
r
2
M
1/2
r
1
e
r
2
/4
. 4.6
We also note that in the limit 1
the function M
holds
_
0
r
rdr
1
1
, > 1, 0 < 1. 4.7
In order to derive this fundamental result, we proceed as follows on the basis of the integral
representation 3.10:
_
0
r
rdr
_
0
r
_
1
2i
_
Ha
e
r
d
1
_
dr
1
2i
_
Ha
e
__
0
e
r
dr
_
d
1
2i
_
Ha
e
1
d
1
1
.
4.8
12 International Journal of Dierential Equations
Above we have legitimized the exchange between integrals and used the identity
_
0
e
r
dr
1
1
,
4.9
along with the Hankel formula of the Gamma function. Analogously, we can compute all the
moments of F
r in R
.
4.4. The Laplace Transform of the MWright Function
Let the MittagLeer function be dened in the complex plane for any 0 by the following
series and integral representation; see, for example, 33, 60:
E
n0
z
n
n 1
1
2i
_
Ha
1
e
z
d, > 0, z C. 4.10
Such function is entire of order 1/ for > 0 and reduces to the function expz for > 0 and
to 1/1 z for 0. We recall that the MittagLeer function for > 0 plays fundamental
roles in applications of fractional calculus like fractional relaxation and fractional oscillation;
see, for example, 25, 26, 61, 62, so that it could be referred as the Queen function of fractional
calculus. Recently, numerical routines for functions of MittagLeer type have been provided,
e.g., by Freed et al. 63, Goreno et al. 64 with MATHEMATICA, Podlubny 65 with
MATLAB, Seybold and Hilfer 66.
We now point out that the MWright function is related to the MittagLeer function
through the following Laplace transform pair:
M
r
L
E
s, 0 < < 1.
4.11
For the readers convenience, we provide a simple proof of 4.11 by using two dierent
approaches. We assume that the exchanges between integrals and series are legitimate in
view of the analyticity properties of the involved functions. In the rst approach we use the
integral representations of the two functions obtaining
_
0
e
sr
M
rdr
1
2i
_
0
e
s r
__
Ha
e
r
d
1
_
dr
1
2i
_
Ha
e
1
__
0
e
rs
dr
_
d
1
2i
_
Ha
e
s
d E
s.
4.12
International Journal of Dierential Equations 13
In the second approach we develop in series the exponential kernel of the Laplace transform
and we use the expression 4.7 for the absolute moments of the MWright function arriving
to the following series representation of the MittagLeer function:
_
0
e
sr
M
rdr
n0
s
n
n!
_
0
r
n
M
rdr
n0
s
n
n!
n 1
n 1
n0
s
n
n 1
E
s.
4.13
We note that the transformation term by term of the series expansion of the MWright
function is not legitimate because the function is not of exponential order; see 67. However,
this procedure yields the formal asymptotic expansion of the MittagLeer function E
s
as s in a sector around the positive real axis; see, for example, 33, 60, that is,
n0
_
0
e
sr
r
n
dr
n!n 1
n0
1
n
n 1
1
s
n1
m1
1
m1
m 1
1
s
m
E
s, s .
4.14
4.5. The Fourier Transform of the Symmetric MWright Function
The MWright function, extended on the negative real axis as an even function, is related to
the MittagLeer function through the following Fourier transform pair:
M
x
F
2E
2
_
2
_
, 0 < < 1. 4.15
Below, we prove the equivalent formula
_
0
cosrM
rdr E
2
_
2
_
. 4.16
For the proof it is sucient to develop in series the cosine function and use formula 4.7 for
the absolute moments of the MWright function:
_
0
cosrM
rdr
n0
1
n
2n
2n!
_
0
r
2n
M
rdr
n0
1
n
2n
2n 1
E
2
_
2
_
.
4.17
14 International Journal of Dierential Equations
0
0.2
0.4
0.6
0.8
1
5 4 3 2 1 0 1 2 3 4 5
M
x
0
1/8
1/4
3/8
1/2
a
0
10
2
10
1
10
0
5 4 3 2 1 0 1 2 3 4 5
M
x
0
1/8
1/4
3/8
1/2
b
Figure 1: Plots of the symmetric M
r
M
s
s 1 1
, 0 < < 1. 4.18
4.7. Plots of the Symmetric MWright Function
It is instructive to show the plots of the symmetric MWright function on the real axis for
some rational values of the parameter . In order to have more insight of the eect of the
parameter itself on the behavior close to and far from the origin, we adopt both linear and
logarithmic scale for the ordinates.
In Figures 1 and 2 we compare the plots of the M
expx
2
for 1/2, whereas
in Figure 2 we see the transition from 1/
expx
2
to the delta functions x1 for 1.
Because of the two symmetrical humps for 1/2 < 1, the M
x
1/2
5/8
3/4
1
a
0
10
2
10
1
10
0
5 4 3 2 1 0 1 2 3 4 5
M
x
1/2
5/8
3/4
1
b
Figure 2: Plots of the symmetric MWright function with 1/2, 5/8, 3/4, 1 for5 x 5: a linear scale,
b logarithmic scale.
4.8. The MWright Function in Two Variables
In view of the timefractional diusion processes that will be considered in the next sections,
it is worthwhile to introduce the function in the two variables
M
x, t : t
_
xt
_
, 0 < < 1, x, t R
, 4.19
which denes a spatial probability density in x evolving in time t with selfsimilarity
exponent H . Of course for x R we have to consider the symmetric version obtained
from 4.19 multiplying by 1/2 and replacing x by x.
Hereafter we provide a list of the main properties of this function, which can be
derived from Laplace and Fourier transforms of the corresponding MWright function in
one variable.
From 4.2 we derive the Laplace transform of M
x, t with respect to t R
as
L{M
x, t; t s} s
1
e
xs
. 4.20
From 4.10 we derive the Laplace transform of M
x, t with respect to x R
as
L{M
x, t; x s} E
st
. 4.21
From 4.15 we derive the Fourier transform of M
x, t; x } 2E
2
_
2
t
_
. 4.22
Moreover, using the Mellin transform, Mainardi et al. 71 derived the following integral
formula:
M
x, t
_
0
M
x, M
, td, . 4.23
16 International Journal of Dierential Equations
Special cases of the MWright function are simply derived for 1/2 and 1/3 from the
corresponding ones in the complex domain; see 3.123.13. We devote particular attention
to the case 1/2 for which we get from 4.6 the Gaussian density in R:
M
1/2
x, t
1
2
t
1/2
e
x
2
/4t
. 4.24
For the limiting case 1 we obtain
M
1
x, t
1
2
x t x t. 4.25
5. Fractional Diffusion Equations
Let us now consider a variety of diusionlike equations starting from the standard diusion
equation whose fundamental solutions are expressed in terms of the MWright function
depending on space and time variables. The two variables, however, turn out to be related
through a selfsimilarity condition.
5.1. The Standard Diffusion Equation
The standard diusion equation for the eld ux, t with initial condition ux, 0 u
0
x is
u
t
K
1
2
u
x
2
, < x < , t 0,
5.1
where K
1
is a suitable diusion coecient of dimensions K
1
L
2
T
1
cm
2
/sec. This
initialboundary value problem can be easily shown to be equivalent to the Volterra integral
equation
ux, t u
0
x K
1
_
t
0
2
ux,
x
2
d. 5.2
It is well known that the fundamental solution usually referred as the Green function, which
is the solution corresponding to u
0
x x, is the Gaussian probability density evolving
in time with variance mean square displacement proportional to time. In our notation we
have
G
1
x, t
1
2
_
K
1
t
1/2
e
x
2
/4K
1
t
,
5.3
2
1
t :
_
x
2
G
1
x, tdx 2K
1
t. 5.4
This variance law characterizes the process of normal diusion as it emerges from Einsteins
approach to Brownian motion Bm; see, for example, 72.
International Journal of Dierential Equations 17
In view of future developments, we rewrite the Green function in terms of the M
Wright function by recalling 3.12, that is,
G
1
x, t
1
2
1
_
K
1
t
1/2
M
1/2
_
x
_
K
1
t
1/2
_
. 5.5
From the selfsimilarity of the Green function in 5.3 or 5.5, we are led to write
G
1
x, t
1
_
K
1
t
H
G
1
_
x
_
K
1
t
H
, 1
_
, 5.6
where H 1/2 is the similarity or Hurst exponent and x/
_
K
1
t
1/2
acts as the
similarity variable. We refer to the onevariable function G
1
as the reduced Green function.
5.2. The StretchedTime Standard Diffusion Equation
Let us now stretch the time variable in 5.1 by replacing t with t
2
u
x
2
, < x < , t 0, 5.7
where K
L
2
T
cm
2
/sec
. It is easy to recognize that such equation is akin to the standard diusion equation
but with a diusion coecient depending on time; K
1
t t
1
K
1
t
1
t
, 5.8
we have
u
t
t
1
K
2
u
x
2
, < x < , t 0.
5.9
The integral form corresponding to 5.75.9 reads
ux, t u
0
x K
_
t
0
2
ux,
x
2
1
d. 5.10
The corresponding fundamental solution is the stretchedtime Gaussian
G
x, t
1
2
_
K
t
/2
e
x
2
/4K
1
2
1
_
K
t
/2
M
1/2
_
x
_
K
t
/2
_
. 5.11
18 International Journal of Dierential Equations
The corresponding variance
t :
_
x
2
G
x, tdx 2K
5.12
is characteristic of a general process of anomalous diusion, precisely of slow diusion for 0 <
< 1, and of fast diusion for 1 < < 2.
5.3. The TimeFractional Diffusion Equation
In literature there exist two forms of the timefractional diusion equation of a single order,
one with RiemannLiouvile derivative and one with Caputo derivative. These forms are
equivalent if we refer to the standard initial condition ux, 0 u
0
x, as shown in 73.
Taking a real number 0, 1, the timefractional diusion equation of order in the
RiemannLiouville sense reads
u
t
K
D
1
t
2
u
x
2
,
5.13
whereas in the Caputo sense reads
t
u K
2
u
x
2
, 5.14
where K
L
2
T
cm
2
/sec
. Like for diusion equations of integer order 5.1 and 5.75.9, we consider the
equivalent integral equation corresponding to our fractional diusion equations 5.135.14
as
ux, t u
0
x K
_
_
t
0
t
1
2
ux,
x
2
d. 5.15
The Green function G
x, t
1
2
1
_
K
t
/2
M
/2
x
_
K
t
/2
. 5.16
The corresponding variance can be promptly obtained from the general formula 5.5 for
the absolute moment of the MWright function. In fact, using 5.5 and 5.16 and after an
obvious change of variable, we obtain
t :
_
x
2
G
x, tdx
2
_
1
_K
. 5.17
International Journal of Dierential Equations 19
As a consequence, for 0 < < 1 the variance is consistent with a process of slow diusion with
similarity exponent H /2. For further reading on timefractional diusion equations and
their solutions the reader is referred, among others, to 38, 51, 53 and 74, 75.
5.4. The Stretched TimeFractional Diffusion Equation
In the fractional diusion equation 5.13, let us stretch the time variable by replacing t with
t
/
where 0 < < 2 and 0 < 1. We have
u
t
/
K
D
1
t
/
2
u
x
2
,
5.18
namely,
u
t
t
/1
K
D
1
t
/
2
u
x
2
, 5.19
where K
L
2
T
cm
2
/sec
that reduces to K
if 1 and to K
_
t
0
/1
_
t
/
/
_
1
2
ux,
x
2
d, 5.20
whose Green function G
x, t is
G
x, t
1
2
1
_
K
t
/2
M
/2
x
_
K
t
/2
, 5.21
with variance
2
,
t :
_
x
2
G
,
x, tdx
2
_
1
_K
. 5.22
As a consequence, the resulting process turns out to be selfsimilar with Hurst exponent H
/2 and a variance law consistent with both slow diusion if 0 < < 1 and fast diusion if
1 < < 2. We note that the parameter does explicitly enter in the variance law 5.22 only
as in the determination of the multiplicative constant.
20 International Journal of Dierential Equations
It is straightforward to note that the evolution equations of this process reduce to those
for timefractional diusion if < 1, for stretched diusion if
/
1 and 1, and nally
for standard diusion if 1.
6. Fractional Diffusion Processes with Stationary Increments
We have seen that any Green function associated to the diusionlike equations considered
in the previous section can be interpreted as the timeevolving onepoint pdf of certain
selfsimilar stochastic processes. However, in general, it is not possible to dene a unique
selfsimilar stochastic process because the determination of any multipoint probability
distribution is required; see, for example, 22.
In other words, starting from a master equation which describes the dynamic
evolution of a probability density function fx, t, it is always possible to dene an
equivalence class of stochastic processes with the same marginal density function fx, t.
All these processes provide suitable stochastic representations for the starting equation. It is
clear that additional requirements may be stated in order to uniquely select the probabilistic
model.
For instance, considering 5.19, the additional requirement of stationary increments,
as shown by Mura et al.; see 1922, can lead to a class {B
,
t, t 0}; called generalized
grey Brownian motion ggBm, which, by construction, is made up of selfsimilar processes with
stationary increments and Hurst exponent H /2. Thus {B
,
t, t 0} is a special class of
Hsssi processes, which provide nonMarkovian stochastic models for anomalous diusion,
of both slow type 0 < < 1 and fast type 1 < < 2. According to a common terminology,
Hsssi stands for Hselfsimilarstationaryincrements, see for details 2.
The ggBm includes some well known processes, so that it denes an interesting
general theoretical framework. The fractional Brownian motion fBm appears for 1
and is associated with 5.7; the grey Brownian motion gBm, dened by Schneider 8, 9,
corresponds to the choice , with 0 < < 1, and is associated to 5.13, 5.14, or 5.15;
nally, the standard Brownian motion Bm is recovered by setting 1 being associated
to 5.1. We should note that only in the particular case of Bm the corresponding process is
Markovian.
In Figure 3 we present a diagram that allows to identify the elements of the ggBm
class. The top region 1 < < 2 corresponds to the domain of fast diusion with longrange
dependence. We remind that a selfsimilar process with stationary increments is said to possess
longrange dependence if the autocorrelation function of the increments tends to zero like a
power function and such that it does not result integrable, see for details 2. In this domain
the increments of the process B
,
t are positively correlated, so that the trajectories tend to
be more regular persistent. It should be noted that longrange dependence is associated to
a nonMarkovian process which exhibits longmemory properties. The horizontal line 1
corresponds to processes with uncorrelated increments, which model various phenomena of
normal diusion. For 1 we recover the Gaussian process of the standard Brownian
motion. The Gaussian process of the fractional Brownian motion is identied by the vertical
line 1. The bottom region 0 < < 1 corresponds to the domain of slow diusion. The
increments of the corresponding process B
,
t turn out to be negatively correlated and this
implies that the trajectories are strongly irregular antipersistent motion; the increments form
a stationary process which does not exhibit longrange dependence. Finally, the diagonal line
represents the Schneider grey Brownian motion gBm.
International Journal of Dierential Equations 21
0
0.5
1
1.5
2
0 0.5 1
0.5
0
1
H
B
,1
t
B
,
t
Brownian
motion
Purely random
Long range dependence
Persistent
Anti 
persistent
fBm
gBm
B
,
t
B
2,
t
Figure 3: Parametric class of generalized grey Brownian motion.
Here we want to dene the ggBm by making use of the Kolmogorov extension
theorem and the properties of the MWright function. According to Mura and Pagnini 21,
the generalized grey Brownian motion B
,
t is a stochastic process dened in a certain
probability space such that its nitedimensional distributions are given by
f
,
_
x
1
, x
2
, . . . , x
n
;
,
_
2
n1/2
_
2
_
1
_
n
det
,
_
0
1
n/2
M
1/2
_
1/2
_
M
d,
6.1
with
2
_
1
_
1
n
i,j1
x
i
,
1
_
t
i
, t
j
_
x
j
1/2
6.2
and covariance matrix
,
_
t
i
, t
j
_
1
_
1
_
_
t
i
t
j
_
_
t
i
t
j
_
_
_
, i, j 1, . . . , n.
6.3
The covariance matrix 6.3 characterizes the typical dependence structure of a selfsimilar
process with stationary increments and Hurst exponent H /2; see, for example, 2.
Using 4.23, for n 1, 6.1 reduces to
f
,
x, t
1
4t
0
M
1/2
_
xt
/2
,
_
M
, 1d
1
2
t
/2
M
/2
_
xt
/2
_
. 6.4
This means that the marginal density function of the ggBm is indeed the fundamental
solution 5.21 of 5.185.19 with K
1. Moreover, because M
1
1, for 1,
22 International Journal of Dierential Equations
putting
,1
M
1/2
2
n
i,j1
x
i
t
i
, t
j
x
j
1/2
, 6.5
which nally reduces to the standard Gaussian distribution of Brownian motion as 1.
By the denition used above, it is clear that, xed , B
,
t is characterized only by its
covariance structure, as shown by Mura et al. 20, 21. In other words, the ggBm, which is
not Gaussian in general, is an example of a process dened only through its rst and second
moments, which indeed is a remarkable property of Gaussian processes. Consequently, the
ggBm appears to be a direct generalization of Gaussian processes, in the same way as the
MWright function is a generalization of the Gaussian function.
7. Concluding Discussion
In this review paper we have surveyed a quite general approach to derive models for
anomalous diusion based on a family of timefractional diusion equations depending on
two parameters: 0, 2, 0, 1.
The unifying topic of this analysis is the socalled MWright function by which the
fundamental solutions of these equations are expressed. Such function is shown to exhibit
fundamental analytical properties that were properly used in recent papers for characterizing
and simulating a general class of selfsimilar stochastic processes with stationary increments
including fractional Brownian motion and grey Brownian motion.
In this respect, the MWright function emerges to be a natural generalization of the
Gaussian density to model diusion processes, covering both slow and fast anomalous
diusion and including nonMarkovian property. In particular, it turns out to be the main
function for the special Hsssi class of stochastic processes which are selfsimilar with
stationary increments governed by a master equation of fractional type.
Appendices
A. The Fundamental Solution of the TimeFractional
Diffusion Equation
The fundamental solution G
x, 0
u
0
x x. Let us devote our attention
to the integral form 5.15 using nondimensional variables by setting K
1 and adopting
the notation J
t
for the fractional integral 2.10. Then, our Cauchy problem reads
G
x, t x J
2
G
x
2
x, t.
A.1
International Journal of Dierential Equations 23
In the FourierLaplace domain, after applying formula 2.18 for the Laplace transform of the
fractional integral and observing
1; see, for example, 76, we get
, s
1
s
2
s
, s,
A.2
from which
, s
s
1
s
2
, 0 < 1, Rs > 0, R.
A.3
To determine the Green function G
a
2b
1/2
e
xb
1/2
, a, b > 0, A.4
and setting a s
1
, b s
, we get
x, s
1
2
s
/21
e
xs
/2
. A.5
Strategy (S2)
Recalling the Laplace transform pair
s
1
s
c
L
E
_
ct
_
, c > 0,
A.6
and setting c
2
, we have
, t E
2
t
_
. A.7
Both strategies lead to the result
G
x, t
1
2
M
/2
x, t
1
2
t
/2
M
/2
_
x
t
/2
_
, A.8
24 International Journal of Dierential Equations
consistent with 5.16. Here we have used the MWright function, introduced in Section 4,
and its properties related to the Laplace transform pair 4.20 for inverting A.5 and the
Fourier transform pair 4.22 for inverting A.7.
B. The Fundamental Solution of the TimeFractional Drift Equation
Let us nally note that the MWright function does appear also in the fundamental solution
of the timefractional drift equation. Writing this equation in nondimensional form and
adopting the Caputo derivative, we have
t
ux, t
x
ux, t, < x < , t 0,
B.1
where 0 < < 1 and ux, 0
u
0
x. When u
0
x x, we obtain the fundamental solution
Green function that we denote by G
x, t
_
x
t
_
, x > 0,
0, x < 0,
B.2
which for 1 reduces to the right running pulse x t for x > 0.
In the FourierLaplace domain, after applying formula 2.19 for the Laplace transform
of the Caputo fractional derivative and observing
1; see, for example, 76, we get
s
, s s
1
i
, s,
B.3
from which
, s
s
1
s
i
, 0 < 1, Rs > 0, R.
B.4
Like in Appendix A, to determine the Green function G
a
b
e
xb
, a, b > 0, x > 0, B.5
and setting a s
1
, b s
, we get
x, s s
1
e
xs
. B.6
Strategy (S2)
Recalling the Laplace transform pair
s
1
s
c
L
E
_
ct
_
, c > 0,
B.7
and setting c i, we have
, t E
_
it
_
. B.8
Both strategies lead to the result B.2. In view of 4.1 we also recall that the MWright
function is related to the unilateral extremal stable density of index L
x, t
t
x
11/
L
_
tx
1/
_
, B.9
To conclude this Appendix, let us briey discuss the above results in view of their relevance
in fractional diusion processes following the recent paper by Goreno and Mainardi 77.
Equation B.1 describes the evolving sojourn probability density of the positively oriented
timefractional drift process of a particle, starting in the origin at the instant zero. It has
been derived in 77 as a properly scaled limit for the evolution of the counting number
of the MittagLeer renewal process the fractional Poisson process. It can be given in
several forms, and often it is cited as the subordinator producing the operational time from
the physical time for spacetimefractional diusion as in the form B.9. For more details see
3, where simulations of spacetimefractional diusion processes have been considered as
composed by timefractional and spacefractional diusion processes.
This analysis can be compared to that described with a dierent language in papers
by Meerschaert et al. 4, 78. Recently, a more exhaustive analysis has been given by Goreno
79.
26 International Journal of Dierential Equations
Acknowledgments
This work has been carried out in the framework of the research project Fractional Calculus
Modelling http://www.fracalmo.org/. The authors are grateful to V. Kiryakova, R. Goreno
and the anonymous referees for useful comments.
References
1 B. B. Mandelbrot and J. W. Van Ness, Fractional Brownian motions, fractional noises and
applications, SIAM Review, vol. 10, pp. 422437, 1968.
2 M. S. Taqqu, Fractional Brownian motion and longrange dependence, in Theory and Applications
of LongRange Dependence, P. Doukan, G. Oppenheim, and M. S. Taqqu, Eds., pp. 538, Birkh auser,
Boston, Mass, USA, 2003.
3 R. Goreno, F. Mainardi, and A. Vivoli, Continuoustime random walk and parametric subordina
tion in fractional diusion, Chaos, Solitons & Fractals, vol. 34, no. 1, pp. 87103, 2007.
4 M. M. Meerschaert, D. A. Benson, H.P. Scheer, and B. Baeumer, Stochastic solution of spacetime
fractional diusion equations, Physical Review E, vol. 65, no. 4, Article ID 041103, 4 pages, 2002.
5 R. Metzler and J. Klafter, The random walks guide to anomalous diusion: a fractional dynamics
approach, Physics Reports A, vol. 339, no. 1, pp. 177, 2000.
6 E. Scalas, R. Goreno, and F. Mainardi, Uncoupled continuoustime random walks: solution and
limiting behavior of the master equation, Physical Review E, vol. 69, no. 1, Article ID 011107, 8 pages,
2004.
7 A. A. Dubkov, B. Spagnolo, and V. V. Uchaikin, L evy ight superdiusion: an introduction,
International Journal of Bifurcation and Chaos, vol. 18, no. 9, pp. 26492672, 2008.
8 W. R. Schneider, Grey noise, in Stochastic Processes, Physics and Geometry (Ascona and Locarno, 1988),
S. Albeverio, G. Casati, U. Cattaneo, D. Merlini, and R. Moresi, Eds., pp. 676681, World Scientic,
Teaneck, NJ, USA, 1990.
9 W. R. Schneider, Grey noise, in Ideas and Methods in Mathematical Analysis, Stochastics, and
Applications (Oslo, 1988), S. Albeverio, J. E. Fenstad, H. Holden, and T. Lindstrm, Eds., pp. 261282,
Cambridge University Press, Cambridge, UK, 1992.
10 A. V. Chechkin, R. Goreno, and I. M. Sokolov, Retarding subdiusion and accelerating
superdiusion governed by distributedorder fractional diusion equations, Physical Review E, vol.
66, no. 4, Article ID 046129, 7 pages, 2002.
11 A. V. Chechkin, R. Goreno, I. M. Sokolov, and V. Yu. Gonchar, Distributed order time fractional
diusion equation, Fractional Calculus & Applied Analysis, vol. 6, no. 3, pp. 259279, 2003.
12 A. V. Chechkin, V. Yu. Gonchar, R. Goreno, N. Korabel, and I. M. Sokolov, Generalized fractional
diusion equations for accelerating subdiusion and truncated L evy ights, Physical Review E, vol.
78, no. 2, Article ID 021111, 13 pages, 2008.
13 R. Goreno, F. Mainardi, D. Moretti, and P. Paradisi, Time fractional diusion: a discrete random
walk approach, Nonlinear Dynamics, vol. 29, no. 14, pp. 129143, 2002.
14 R. Goreno, F. Mainardi, D. Moretti, G. Pagnini, and P. Paradisi, Discrete random walk models for
spacetime fractional diusion, Chemical Physics, vol. 284, no. 12, pp. 521541, 2002.
15 F. Liu, S. Shen, V. Anh, and I. Turner, Analysis of a discrete nonMarkovian random walk
approximation for the time fractional diusion equation, The ANZIAM Journal, vol. 46, pp. C488
C504, 2005.
16 F. Liu, P. Zhuang, V. Anh, I. Turner, and K. Burrage, Stability and convergence of the dierence
methods for the spacetime fractional advectiondiusion equation, Applied Mathematics and
Computation, vol. 191, no. 1, pp. 1220, 2007.
17 Y. Zhang, D. A. Benson, M. M. Meerschaert, and H.P. Scheer, On using random walks to solve
the spacefractional advectiondispersion equations, Journal of Statistical Physics, vol. 123, no. 1, pp.
89110, 2006.
18 P. Zhuang and F. Liu, Implicit dierence approximation for the time fractional diusion equation,
Journal of Applied Mathematics & Computing, vol. 22, no. 3, pp. 8799, 2006.
19 A. Mura, NonMarkovian stochastic processes and their applications: from anomalous diusion to time series
analysis, Ph.D. thesis, Department of Physics, University of Bologna, March 2008, Supervisor Prof. F.
Mainardi, http://www.fracalmo.org/mura/.
International Journal of Dierential Equations 27
20 A. Mura and F. Mainardi, A class of selfsimilar stochastic processes with stationary increments to
model anomalous diusion in physics, Integral Transforms and Special Functions, vol. 20, no. 34, pp.
185198, 2009.
21 A. Mura and G. Pagnini, Characterizations and simulations of a class of stochastic processes to model
anomalous diusion, Journal of Physics A, vol. 41, no. 28, Article ID 285003, 22 pages, 2008.
22 A. Mura, M. S. Taqqu, and F. Mainardi, NonMarkovian diusion equations and processes: analysis
and simulations, Physica A, vol. 387, no. 21, pp. 50335064, 2008.
23 M. Caputo, Linear models of dissipation whose Q is almost frequency independentpart II,
Geophysical Journal of the Royal Astronomical Society, vol. 13, pp. 529539, 1967.
24 M. Caputo, Elasticity and Dissipation, Zanichelli, Bologna, Italy, 1969.
25 R. Goreno and F. Mainardi, Fractional calculus: integral and dierential equations of fractional
order, in Fractals and Fractional Calculus in Continuum Mechanics (Udine, 1996), A. Carpinteri and F.
Mainardi, Eds., vol. 378 of CISM Courses and Lectures, pp. 223276, Springer, Vienna, Austria, 1997.
26 F. Mainardi and R. Goreno, Timefractional derivatives in relaxation processes: a tutorial survey,
Fractional Calculus & Applied Analysis, vol. 10, no. 3, pp. 269308, 2007.
27 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, San Diego, Calif, USA, 1999.
28 A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo, Theory and Applications of Fractional Dierential
Equations, vol. 204 of NorthHolland Mathematics Studies, Elsevier Science B.V., Amsterdam, The
Netherlands, 2006.
29 E. M. Wright, On the coecients of power series having exponential singularities, Journal of the
London Mathematical Society, vol. 8, pp. 7179, 1933.
30 E. M. Wright, The asymptotic expansion of the generalized Bessel function, Proceedings of the London
Mathematical Society, vol. 38, pp. 257270, 1935.
31 E. M. Wright, The asymptotic expansion of the generalized hypergeometric function, Journal of the
London Mathematical Society, vol. 10, pp. 287293, 1935.
32 E. M. Wright, The generalized Bessel function of order greater than one, The Quarterly Journal of
Mathematics, vol. 11, pp. 3648, 1940.
33 A. Erd elyi, W. Magnus, F. Oberhettinger, and F. G. Tricomi, Higher Transcendental Functions, vol. 3,
chapter 18, McGrawHill, NewYork, NY, USA, 1954.
34 R. Goreno, Yu. Luchko, and F. Mainardi, Analytical properties and applications of the Wright
function, Fractional Calculus & Applied Analysis, vol. 2, no. 4, pp. 383414, 1999.
35 R. Goreno, Yu. Luchko, and F. Mainardi, Wright functions as scaleinvariant solutions of the
diusionwave equation, Journal of Computational and Applied Mathematics, vol. 118, no. 12, pp. 175
191, 2000.
36 A. A. Kilbas, M. Saigo, and J. J. Trujillo, On the generalized Wright function, Fractional Calculus &
Applied Analysis, vol. 5, no. 4, pp. 437460, 2002.
37 V. Kiryakova, Generalized Fractional Calculus and Applications, vol. 301 of Pitman Research Notes in
Mathematics Series, Longman Scientic & Technical, Harlow, UK, 1994.
38 F. Mainardi, Fractional calculus: some basic problems in continuum and statistical mechanics, in
Fractals and Fractional Calculus in Continuum Mechanics (Udine, 1996), A. Carpinteri and F. Mainardi,
Eds., vol. 378 of CISM Courses and Lectures, pp. 291348, Springer, Vienna, Austria, 1997.
39 B. Stankovi c, On the function of E. M. Wright, Publications de lInstitut Math` ematique, vol. 10 24,
pp. 113124, 1970.
40 R. Wong and Y.Q. Zhao, Smoothing of Stokess discontinuity for the generalized Bessel function,
Proceedings of the Royal Society A, vol. 455, no. 1984, pp. 13811400, 1999.
41 R. Wong and Y.Q. Zhao, Smoothing of Stokess discontinuity for the generalized Bessel function.
II, Proceedings of the Royal Society A, vol. 455, no. 1988, pp. 30653084, 1999.
42 F. Mainardi, On the initial value problem for the fractional diusionwave equation, in Waves and
Stability in Continuous Media (Bologna, 1993), S. Rionero and T. Ruggeri, Eds., vol. 23 of Series on
Advances in Mathematics for Applied Sciences, pp. 246251, World Scientic, River Edge, NJ, USA, 1994.
43 F. Mainardi and M. Tomirotti, On a special function arising in the time fractional diusionwave
equation, in Transform Methods and Special Functions, Soa 1994, P. Rusev, I. Dimovski, and V.
Kiryakova, Eds., pp. 171183, Science Culture Technology, Singapore, 1995.
44 C. M. Bender and S. A. Orszag, Advanced Mathematical Methods for Scientists and Engineers, McGraw
Hill, Singapore, 1987.
45 A. Hanyga, Multidimensional solutions of spacetimefractional diusion equations, Proceedings of
the Royal Society A, vol. 458, no. 2018, pp. 429450, 2002.
28 International Journal of Dierential Equations
46 R. Balescu, VLangevin equations, continuous time random walks and fractional diusion, Chaos,
Solitons & Fractals, vol. 34, no. 1, pp. 6280, 2007.
47 G. Germano, M. Politi, E. Scalas, and R. L. Schilling, Stochastic calculus for uncoupled continuous
time random walks, Physical Review E, vol. 79, no. 6, Article ID 066102, 12 pages, 2009.
48 V. Kiryakova, The multiindex MittagLeer functions as an important class of special functions of
fractional calculus, Computers and Mathematics with Applications. In press.
49 V. Kiryakova, The special functions of fractional calculus as generalized fractional calculus operators
of some basic functions, Computers and Mathematics with Applications, vol. 59, no. 3, pp. 11281141,
2010.
50 F. Mainardi, R. Goreno, and A. Vivoli, Renewal processes of MittagLeer and Wright type,
Fractional Calculus & Applied Analysis, vol. 8, no. 1, pp. 738, 2005.
51 F. Mainardi and G. Pagnini, The Wright functions as solutions of the timefractional diusion
equation, Applied Mathematics and Computation, vol. 141, no. 1, pp. 5162, 2003.
52 F. Mainardi, The fundamental solutions for the fractional diusionwave equation, Applied
Mathematics Letters, vol. 9, no. 6, pp. 2328, 1996.
53 F. Mainardi, Yu. Luchko, and G. Pagnini, The fundamental solution of the spacetime fractional
diusion equation, Fractional Calculus & Applied Analysis, vol. 4, no. 2, pp. 153192, 2001.
54 W. Feller, An Introduction to Probability Theory and Its Applications. Vol. II, John Wiley & Sons, New
York, NY, USA, 1st edition, 1966.
55 W. Feller, An Introduction to Probability Theory and Its Applications. Vol. II, John Wiley & Sons, New
York, NY, USA, 2nd edition, 1971.
56 P. Humbert, Nouvelles correspondances symboliques, Bulletin des Sciences Math ematiques, vol. 69,
pp. 121129, 1945.
57 H. Pollard, The representation of e
x
x
m
1
K
1
2
x
2
_
ux, t, 1.1
where m is the mass of the diusing test particle,
1
denotes the fraction constant
characterising the interaction between the test particle and its embedding, and the force is
2 International Journal of Dierential Equations
related to the external potential through Fx dVx/dx. The FPE 1.1 is well studied for
a variety of potential types, and the respective results have found wide application. In many
studies of diusion processes where the diusion takes place in a highly nonhomogeneous
medium, the traditional FPE may not be adequate 2, 3. The nonhomogeneities of the
medium may alter the laws of Markov diusion in a fundamental way. In particular,
the corresponding probability density of the concentration eld may have a heavier tail
than the Gaussian density, and its correlation function may decay to zero at a much
slower rate than the usual exponential rate of Markov diusion, resulting in longrange
dependence. This phenomenon is known as anomalous diusion 4. Fractional derivatives
play a key role in modeling particle transport in anomalous diusion including the space
fractional FokkerPlanck advectiondispersion equation describing L evy ights, the time
fractional FokkerPlanck equation depicting traps, and the timespace fractional equation
characterizing the competition between L evy ights and traps 5, 6. Dierent assumptions
on this probability density function lead to a variety of timespace fractional FokkerPlanck
equations TSFFPEs.
TSFFPE has been successfully used for modeling relevant physical processes. When
the fractional dierential equation is used to describe the asymptotic behavior of continuous
time random walks, its solution corresponds to the L evy walks, generalizing the Brownian
motion to the L evy motion. The following space fractional FokkerPlanck equation has been
considered 2, 3, 7:
ux, t
t
v
ux, t
x
K
_
c
a
D
x
ux, t c
x
D
b
ux, t
_
,
1.2
where v is the drift of the process, that is, the mean advective velocity; K
is the coecient of
dispersion;
a
D
x
and
x
D
b
are the left and right RiemannLiouville space fractional derivatives
of order given by
a
D
x
ux, t
1
_
2
_
2
x
2
_
x
a
u, td
x
1
,
x
D
b
ux, t
1
_
2
_
2
x
2
_
b
x
u, td
x
1
;
1.3
c
and c
1/2 /2 and c
ux, t
x
, 1.4
where
/x
a
D
x
and a right RiemannLiouville derivative
x
D
b
, namely,
x
ux, t c
_
a
D
x
x
D
b
_
ux, t. 1.5
As a model for subdiusion in the presence of an external eld, a time fractional
extension of the FPE has been introduced as the time fractional FokkerPlanck equation
TFFPE 5, 9:
ux, t
t
0
D
1
t
_
x
V
x
m
2
x
2
_
ux, t, 1.6
where the RiemannLiouville operator
0
D
1
t
, 0 < < 1 is dened through its operation:
0
D
1
t
ux, t
1
t
_
t
0
u
_
x,
_
_
t
_
1
d.
1.7
Yuste and Acedo 10 proposed an explicit nite dierence method and a new von
Neumanntype stability analysis for the anomalous subdiusion equation 1.6 with V
x
0. However, they did not give a convergence analysis and pointed out the diculty of this
task when implicit methods are considered. Langlands and Henry 11 also investigated
this problem and proposed an implicit numerical L1approximation scheme and discussed
the accuracy and stability of this scheme. However, the global accuracy of the implicit
numerical scheme has not been derived and it seems that the unconditional stability for
all in the range 0 < 1 has not been established. Recently, Chen et al. 12 presented
a Fourier method for the anomalous subdiusion equation, and they gave the stability
analysis and the global accuracy analysis of the dierence approximation scheme. Zhuang
et al. 13 also proposed an implicit numerical method and an analytical technique for the
anomalous subdiusion equation. Chen et al. 14 proposed implicit and explicit numerical
approximation schemes for the Stokes rst problem for a heated generalized second grade
uid with fractional derivatives. The stability and convergence of the numerical scheme
are discussed using a Fourier method. A Richardson extrapolation technique for improving
the order of convergence of the implicit scheme is presented. However, eective numerical
methods and error analysis for the timespace fractional FokkerPlanck equation with a
nonlinear source term are still in their infancy and are open problems.
Equation 1.6 can be written as the following equivalent form 15:
0
D
t
ux, t
ux, 0t
1
_
x
V
x
m
2
x
2
_
ux, t. 1.8
By noting that 15
ux, t
t
0
D
t
ux, t
ux, 0t
1
, 1.9
4 International Journal of Dierential Equations
we arrive at
ux, t
t
_
x
V
x
m
2
x
2
_
ux, t, 1.10
where
ux, t/t
ux, t
t
1
1
_
t
0
u
_
x,
_
d
_
t
_
. 1.11
The timespace fractional FokkerPlank equation TSFFPE, which describes the competition
between subdiusion and L evy ights, is given by 5
ux, t
t
0
D
1
t
_
x
V
x
m
x
_
ux, t, 1.12
or
ux, t
t
_
x
V
x
m
x
_
ux, t, 1.13
where K
ux, t
t
x
Fxux, t K
x
ux, t
_
t
0
r
_
t
_
u
_
x,
_
d, 1.14
where Fx is the external force and rt is a timedependent absorbent term, which may be
related to a reaction diusion process.
The fractional FokkerPlanck equations FFPEs have been recently treated by many
authors and are presented as a useful approach for the description of transport dynamics in
complex systems that are governed by anomalous diusion and nonexponential relaxation
patterns. The analytical solution of FFPE is only possible in simple and special cases 2, 3, 18
and the analytical solution provides a general representation in terms of Greens functions.
We note that the representation of Greens functions is mostly expressed as convergent
expansions in negative and positive power series. These special functions are not suitable
for numerical evaluation when x is suciently small or suciently large. Therefore, a new
numerical strategy is important for solving these equations. Although numerical methods
for the time fractional FokkerPlanck type equation, the space fractional FokkerPlank type
equation, and the timespace fractional FokkerPlanck type equation have been considered
7, 15, 19, numerical methods and stability and convergence analysis for the FFPE are quite
International Journal of Dierential Equations 5
limited and dicult. In fact, published papers on the numerical methods for the FFPE are
sparse. We are unaware of any other published work on numerical methods for the time
space fractional FokkerPlanck type equation with a nonlinear source term. This motivates
us to consider an eective numerical method for the timespace fractional FokkerPlanck
equation with a nonlinear source term and to investigate its stability and convergence.
In this paper, we consider the following timespace fractional FokkerPlanck equation
with a nonlinear source term TSFFPENST:
ux, t
t
_
x
V
x
m
x
_
ux, t su, x, t 1.15
subject to the boundary and initial conditions:
ua, t ub, t 0, 0 t T,
ux, 0 u
0
x, a x b,
1.16
where px V
x/m
ux, t
n1
2
n
j0
b
j
_
u
_
x, t
n1j
_
u
_
x, t
nj
__
O
_
1
_
,
2.1
where b
j
j 1
1
j
1
, j 0, 1, 2, . . . , N 1.
6 International Journal of Dierential Equations
For the symmetric Riesz space fractional derivative, we use the following shifted
Gr unwald approximation 22:
ux
l
, t
x
2 cos
_
/2
_
_
l1
i0
w
i
ux
li1
, t
Ml1
i0
w
i
ux
li1
, t
_
Oh
q
, 2.2
where the coecients are dened by
w
0
1, w
i
1
i
_
1
_
_
i 1
_
i!
, i 1, 2, . . . , M.
2.3
This formula is not unique because there are many dierent valid choices for w
i
that lead to
approximations of dierent orders q 23. The denition 2.2 provides order q 1.
The rstorder spatial derivative can be approximated by the backward dierence
scheme if px < 0, otherwise, the forward dierence scheme can be used if px > 0:
x
px
l
ux
l
, t
px
l
ux
l
, t px
l1
ux
l1
, t
h
Oh.
2.4
The nonlinear source term can be discretised either explicitly or implicitly. In this
paper, we use an explicit method and evaluate the nonlinear source term at the previous
time step:
sux, t
n1
, x, t
n1
sux, t
n
, x, t
n
O. 2.5
In this way, we avoid solving a nonlinear system at each time step and obtain an
unconditionally stable and convergent numerical scheme, as shown in Section 3. However,
the shortcoming of the explicit method is that it generates additional temporal error, as shown
in 2.5.
Thus, using 2.12.5, we have
2
n
j0
b
j
_
u
_
x
l
, t
n1j
_
u
_
x
l
, t
nj
__
px
l
ux
l
, t
n1
px
l1
ux
l1
, t
n1
2 cos
_
/2
_
_
l1
i0
w
i
ux
li1
, t
n1
Ml1
i0
w
i
ux
li1
, t
n1
_
sux
l
, t
n
, x
l
, t
n
O
_
1
h
_
.
2.6
International Journal of Dierential Equations 7
After some manipulation, 2.6 can be written in the following form:
ux
l
, t
n1
b
n
ux
l
, t
0
n1
j0
_
b
j
b
j1
_
u
_
x
l
, t
nj
_
0
h
_
px
l
ux
l
, t
n1
px
l1
ux
l1
, t
n1
_
0
r
0
_
l1
i0
w
i
ux
li1
, t
n1
Ml1
i0
w
i
ux
li1
, t
n1
_
0
sux
l
, t
n
, x
l
, t
n
R
n1
l
,
2.7
where
0
2 > 0, r
0
K
1
h
_
. 2.8
Let u
n
l
be the numerical approximation of ux
l
, t
n
, and let s
n
l
be the numerical
approximation of sux
l
, t
n
, x
l
, t
n
. We obtain the following eective numerical method
ENM of the TSFFPENST 1.151.16:
u
n1
l
b
n
u
0
l
n1
j0
_
b
j
b
j1
_
u
nj
l
0
h
_
p
l
u
n1
l
p
l1
u
n1
l1
_
0
r
0
_
l1
i0
w
i
u
n1
li1
Ml1
i0
w
i
u
n1
li1
_
0
s
n
l
2.9
for l 1, 2, . . . , M 1, n 0, 1, 2, . . . , N 1. The boundary and initial conditions can be
discretised using
u
n
0
u
n
M
0, n 0, 1, 2, . . . , N,
u
0
l
u
0
lh, l 0, 1, 2, . . . , M.
2.10
Remark 2.1. If we use the implicit method to approximate the nonlinear source term, the
numerical method of the TSFFPENST can be written as
u
n1
l
b
n
u
0
l
n1
j0
_
b
j
b
j1
_
u
nj
l
0
h
_
p
l
u
n1
l
p
l1
u
n1
l1
_
0
r
0
_
l1
i0
w
i
u
n1
li1
Ml1
i0
w
i
u
n1
li1
_
0
s
n1
l
,
2.11
that is, replace s
n
l
in 2.9 with s
n1
l
. This numerical method is stable and convergent when
the source term sux, t, x, t satises the Lipschitz condition 1.17 see, e.g., 24.
8 International Journal of Dierential Equations
Lemma 2.2 see 19. The coecients b
j
satisfy
1 b
j
> 0 for j 0, 1, 2, . . . , n;
2 1 b
0
> b
1
> > b
n
, b
n
0 as n ;
3 when 0 < < 1,
lim
j
b
1
j
j
lim
j
j
1
_
1 j
1
_
1
1
1
1
. 2.12
Thus, there is a positive constant C
2
such that
b
1
j
C
2
j
, j 0, 1, 2, . . . . 2.13
Lemma 2.3 see 25. The coecients w
i
satisfy
1 w
0
0, w
1
< 0, and w
i
> 0 for i 2, 3, . . . , M;
2
i0
w
i
0, and
n
i0
w
i
< 0 for n N.
3. Stability of the Effective Numerical Method
In this section, we analyze the stability of the ENM 2.92.10. Firstly, we rewrite 2.9 in the
following form:
u
n1
l
0
h
_
p
l
u
n1
l
p
l1
u
n1
l1
_
0
r
0
_
l1
i0
w
i
u
n1
li1
Ml1
i0
w
i
u
n1
li1
_
b
n
u
0
l
n1
j0
_
b
j
b
j1
_
u
nj
l
0
s
n
l
.
3.1
Let u
n
l
be the approximate solution of the ENM 3.1, and let s
n
l
be the approximation
of s
n
l
. Setting
n
l
u
n
l
u
n
l
, we obtain the following roundo error equation:
n1
l
0
h
_
p
l
n1
l
p
l1
n1
l1
_
0
r
0
_
l1
i0
w
i
n1
li1
Ml1
i0
w
i
n1
li1
_
b
n
0
l
n1
j0
_
b
j
b
j1
_
nj
l
0
_
s
n
l
s
n
l
_
3.2
for l 1, 2, . . . , M 1; n 0, 1, . . . , N 1.
We suppose that px 0 and that px decreases monotonically on a, b. This is
based on the fact that physical considerations and stability dictate that p
max
1lM1

n
l
, and using mathematical induction, we obtain the
following theorem.
International Journal of Dierential Equations 9
Theorem 3.1. Suppose that
n
l
(l 1, 2, . . . , M 1, n 1, 2, . . . , N) is the solution of the roundo
error equation 3.2, and the nonlinear source term sux, t, x, t satises the Lipschitz condition
1.17, then there is a positive constant C
0
, such that
_
_
n
_
_
C
0
_
_
_
0
_
_
_
, n 1, 2, . . . , N. 3.3
Proof. When n 1, assume that 
1
l
0
 max{
1
1
, 
1
2
, . . . , 
1
M1
}. Because px 0 and
decreases monotonically on a, b, we have
0
0
h
_
p
l
0
p
l
0
1
_
_
_
_
1
l
0
_
_
_
0
h
p
l
0
_
_
_
1
l
0
_
_
_
0
h
p
l
0
1
_
_
_
1
l
0
1
_
_
_. 3.4
Using the properties of
i
in Lemma 2.3, we have
0
0
r
0
_
l
0
1
i0
w
i
_
_
_
1
l
0
_
_
_
Ml
0
1
i0
w
i
_
_
_
1
l
0
_
_
_
_
0
r
0
_
l
0
1
i0
w
i
_
_
_
1
l
0
i1
_
_
_
Ml
0
1
i0
w
i
_
_
_
1
l
0
i1
_
_
_
_
.
3.5
Combining 3.4 with 3.5, using the Lipschitz condition 1.17 and smooth solution
condition, we obtain
_
_
_
1
l
0
_
_
_
_
_
_
1
l
0
_
_
_
0
h
p
l
0
_
_
_
1
l
0
_
_
_
0
h
p
l
0
1
_
_
_
1
l
0
1
_
_
_
0
r
0
_
l
0
1
i0
w
i
_
_
_
1
l
0
i1
_
_
_
Ml
0
1
i0
w
i
_
_
_
1
l
0
i1
_
_
_
_
_
_
_
_
_
1
l
0
0
h
p
l
0
1
l
0
0
h
p
l
0
1
1
l
0
1
0
r
0
_
l
0
1
i0
w
i
1
l
0
i1
Ml
0
1
i0
w
i
1
l
0
i1
__
_
_
_
_
_
_
_b
0
0
l
0
0
_
s
0
l
0
s
0
l
0
__
_
_
b
0
_
_
_
0
l
0
_
_
_
0
L
_
_
_
0
l
0
_
_
_
_
1
0
L
_
_
_
_
0
l
0
_
_
_.
3.6
Let C 1
0
L. Thus, we obtain
_
_
_
1
_
_
_
C
_
_
_
0
_
_
_
. 3.7
Now, suppose that
_
_
_
k
_
_
_
C
_
_
_
0
_
_
_
, k 2, . . . , n. 3.8
10 International Journal of Dierential Equations
By assuming 
n1
l
0
 max{
n1
1
, 
n1
2
, . . . , 
n1
M1
}, we have that
_
_
_
n1
l
0
_
_
_
_
_
_
_
_
n1
l
0
0
h
_
p
l
0
n1
l
0
p
l
0
1
n1
l
0
1
_
0
r
0
_
l
0
1
i0
w
i
n1
l
0
i1
Ml
0
1
i0
w
i
n1
l
0
i1
__
_
_
_
_
_
_
_
_
_
_
b
n
0
l
0
n1
j0
_
b
j
b
j1
_
nj
l
0
0
_
s
n
l
0
s
n
l
0
_
_
_
_
_
_
_
b
n
_
_
_
0
l
0
_
_
_
n1
j0
_
b
j
b
j1
_
_
_
_
nj
l
0
_
_
_
0
L
_
_
_
n
l
0
_
_
_.
3.9
Using 3.7 and 3.8, we have
_
_
_
n1
l
0
_
_
_ b
n
_
_
_
0
l
0
_
_
_ C
n1
j0
_
b
j
b
j1
_
_
_
_
0
l
0
_
_
_ C
0
L
_
_
_
0
l
0
_
_
_
b
n
_
_
_
0
l
0
_
_
_ Cb
0
b
n
_
_
_
0
l
0
_
_
_ C
0
L
_
_
_
0
l
0
_
_
_
_
b
n
0
L C
2
__
_
_
0
l
0
_
_
_.
3.10
Let C
0
b
n
0
L C
2
. Hence, we have
_
_
_
n1
_
_
_
C
0
_
_
_
0
_
_
_
. 3.11
The proof of Theorem 3.1 is completed.
Applying Theorem 3.1, the following theorem of stability is obtained.
Theorem 3.2. Assuming that the nonlinear source term sux, t, x, t satises the Lipschitz
condition 1.17 and that the drift coecient px 0 decreases monotonically on a, b, the ENM
dened by 2.92.10 is stable.
Remark 3.3. If px > 0 and decreases monotonically on a, b, we can use the forward
dierence method to approximate the rstorder spatial derivative and apply a similar
analysis of stability.
Remark 3.4. In fact, for the case px does not decrease monotonically, we can still obtain a
stable numerical scheme by a minor change in our current ENM. We can expand the rst
term on the RHS of 1.15 as /xpxux, t dp/dxux, t pxux, t/x, which
enables us to group dp/dxux, t together with the nonlinear source termsu, x, t to obtain
a new nonlinear source term s
T
. Subtracting 2.9 from 2.7 leads to
n1
l
0
h
_
p
l
n1
l
p
l1
n1
l1
_
0
r
0
_
l1
i0
w
i
n1
li1
Ml1
i0
w
i
n1
li1
_
b
n
0
l
n1
j0
_
b
j
b
j1
_
nj
l
0
_
sux
l
, t
n
, x
l
, t
n
s
n
l
_
R
n1
l
,
4.1
where l 1, 2, . . . , M 1; n 0, 1, . . . , N 1.
Assuming that Y
n
max
1lM1

n
l
 and using mathematical induction, we obtain
the following theorem.
Theorem 4.1. Assuming the nonlinear source term sux, t, x, t satises the Lipschitz condition
1.17, and the drift coecient px 0 decreases monotonically on a, b, the ENM dened by
2.92.10 is convergent, and there exists a positive constant C
, such that
Y
n
1
h
_
, n 1, 2, . . . , N. 4.2
Proof. Assume R
k
0
l
0
 max
1lM1,1nN
R
n
l
. Following a similar argument to that presented
above for the stability analysis of the ENM 2.92.10, when n 1, assuming that 
1
l
0

max{
1
1
, 
1
2
, . . . , 
1
M1
}, we have
_
_
_
1
l
0
_
_
_
_
_
_b
0
0
l
0
0
_
sux
l
0
, t
0
, x
l
0
, t
0
s
0
l
0
_
R
1
l
0
_
_
_. 4.3
Utilising Y
0
0, the Lipschitz condition 1.17, and smooth solution condition, we obtain
_
_
_
1
l
0
_
_
_ b
0
_
_
_
0
l
0
_
_
_
0
L
_
_
_
0
l
0
_
_
_
_
_
_R
k
0
l
0
_
_
_
_
_
_R
k
0
l
0
_
_
_. 4.4
Thus,
_
_
_Y
1
_
_
_
_
_
_R
k
0
l
0
_
_
_. 4.5
Now, suppose that
_
_
_Y
k
_
_
_
b
1
k1
_
_
_R
k
0
l
0
_
_
_, k 1, 2, . . . , n. 4.6
12 International Journal of Dierential Equations
Using Lemma 2.2, b
k
> b
k1
, we have
_
_
_Y
k
_
_
_
b
1
n
_
_
_R
k
0
l
0
_
_
_. 4.7
Similarly, assuming 
n1
l
0
 max{
n1
1
, 
n1
2
, . . . , 
n1
M1
}, we have
_
_
_
n1
l
0
_
_
_
_
_
_
_
_
_
b
n
0
l
0
n1
j0
_
b
j
b
j1
_
nj
l
0
0
_
sux
l
0
, t
n
, x
l
0
, t
n
s
n
l
0
_
R
n1
l
0
_
_
_
_
_
_
. 4.8
Utilising Y
0
0, the Lipschitz condition 1.17, and smooth solution condition, we obtain
_
_
_
n1
l
0
_
_
_ b
1
n
b
0
b
n
_
_
_R
k
0
l
0
_
_
_
0
Lb
1
n
_
_
_R
k
0
l
0
_
_
_
_
_
_R
k
0
l
0
_
_
_
b
1
n
_
b
0
b
n
0
L b
n
_
_
_
_R
k
0
l
0
_
_
_
b
1
n
_
b
0
0
L
_
_
_
_R
k
0
l
0
_
_
_ Cb
1
n
_
_
_R
k
0
l
0
_
_
_.
4.9
Hence,
_
_
_Y
n1
_
_
_
Cb
1
n
_
_
_R
k
0
l
0
_
_
_. 4.10
Finally, utilising 2.8 and Lemma 2.2, b
1
n
C
2
n
CC
1
C
2
n
1
h
_
C
1
h
_
4.11
for n 1, 2, . . . , N.
Remark 4.2. If we use an implicit method to approximate the nonlinear source term, as shown
in Remark 2.1, we can prove that the numerical method dened in 2.11 is convergent when
1
0
L > 0, which is independent of the spatial step. In fact, when the time step is small, the
condition 1
0
L > 0 is generally satised.
5. Numerical Results
In this section, we present four numerical examples of the TSFFPE to demonstrate the
accuracy of our theoretical analysis. We also use our solution method to illustrate the changes
in solution behavior that arise when the exponent is varied from integer order to fractional
order and to identify the dierences between solutions with and without the external force
term.
International Journal of Dierential Equations 13
Table 1: Maximum error behavior versus grid size reduction for Example 5.1 at T 1.0.
h Maximum error
1/10 4.8148E2
1/20 1.0111E2
1/40 2.0587E3
1/80 7.3019E4
T 5
T 1
1 0.8 0.6 0.4 0.2 0
x
ENM
Exact
0
0.5
1
1.5
2
2.5
3
u
x
,
t
Figure 1: Comparison of the numerical solution with the exact solution for Example 5.1 at T 1.0, 3.0, 5.0.
T increases in the direction of the arrow.
Example 5.1. Consider the following TSFFPE:
ux, t
t
x
K
x
_
ux, t fx, t,
ua, t ub, t 0, 0 t T,
ux, 0 K
x a
2
b x
2
, a x b,
5.1
where
fx, t 1 1 tx a
2
b x
2
_
K
t
1
_
2 cos
_
/2
_
_
gx a gb x
_
2
_
K
t
1
_
x ab xa b 2x,
gx
4!
_
5
_x
4
2b a
3!
_
4
_x
3
b a
2
2
_
3
_x
2
.
5.2
14 International Journal of Dierential Equations
The exact solution of the TSFFPE 5.1 is found to be
ux, t
_
K
t
1
_
x a
2
b x
2
, 5.3
which can be veried by direct fractional dierentiation of the given solution, and
substituting into the fractional dierential equation.
In this example, we take a 0, b 1, K
ux, t
t
x
ux, t
_
_
t
0
t
1
ux, d,
u5, t u5, t 0, 0 t T,
ux, 0 x.
5.4
This example is a TSFFPENST without the external force term. In fact, it reduces to
the fractional diusion equation with an absorbent term. The formulae to approximate the
absorbent term are presented in the appendix. Here, we take 0.5, 1, and K
1.
Figures 24 show the changes in the solution proles of the TSFFPENST 5.4 when and
are changed from integer to fraction at dierent times T. We see that the solution prole
of the fractional order model is characterized by a sharp peak and a heavy tail. The peak
height in Figure 2 1.0 and 2.0 decreases more rapidly than that in Figure 3 0.8
and 1.8. Furthermore, when we choose 0.5 and 1.5, a more interesting result
can be observed; that is, the peak height in Figure 2 decreases more slowly than that shown
in Figure 4 at the early time T 0.1, but this trend reverses for the later times T 0.5 and
T 1.0. Hence, the TSFFPENST 5.4 may be useful to investigate several physical processes
in the absence of an external force eld by choosing appropriate and .
Example 5.3. Consider the following TSFFPENST:
ux, t
t
_
x
px K
x
_
ux, t
_
_
t
0
t
1
ux, d,
u5, t u5, t 0, 0 t T,
ux, 0 x.
5.5
This example of the TSFFPENST incorporates the external force term with px
1 and an absorbent term. The formula to approximate the absorbent term is presented in
the appendix. Here, we take 0.5, 1, and K
x
,
t
Figure 2: Numerical solutions for Example 5.2 with 1.0 and 2.0 at dierent times T 0.1, 0.5, 1.0.
5 4 3 2 1 0 1 2 3 4 5
x
T 0.1
T 0.5
T 1
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
0.2
u
x
,
t
Figure 3: Numerical solutions for Example 5.2 with 0.8 and 1.8 at dierent times T 0.1, 0.5, 1.0.
in the solution proles of the TSFFPENST 5.5 when and are changed from integer
order to fractional order at dierent times T. Again, we see that the solution prole of the
fractional order model is characterized by a sharp peak and a heavy tail. Furthermore, due
to the presence of the external force term with px 1, the solution proles are shifted to
the right. It is worthwhile to note that the peak of the integer order model in Figure 5 1.0
16 International Journal of Dierential Equations
5 4 3 2 1 0 1 2 3 4 5
x
T 0.1
T 0.5
T 1
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
u
x
,
t
Figure 4: Numerical solutions for Example 5.2 with 0.5 and 1.5 at dierent times T 0.1, 0.5, 1.0.
and 2.0 moves to the right as time increases, but the peak of the fractional order model
in Figure 6 0.8 and 1.8 and Figure 7 0.5 and 1.5 does not move.
We also see that the peak heights in Figures 5 and 6 remain almost the same for
increasing time. The peak height in Figure 5 decreases more slowly than that shown in
Figure 7 at the early time T 0.1, but this trend reverses for the later times T 0.5 and
T 1.0. Hence, the TSFFPENST 5.5 may be useful to investigate several physical processes
within an external force eld by choosing appropriate and .
Example 5.4. Consider the following TSFFPENST:
ux, t
t
_
x
px K
x
_
ux, t rux, t
_
1
ux, t
K
_
,
u0, t u5, t 0, 0 t T,
ux, 0 x
2
5 x
2
, 0 x 5.
5.6
In applications to population biology, ux, t is the population density at location x R and
time t > 0. The nonlinear source termsux, t, x, t rux, t1ux, t/K is Fishers growth
term that models population growth, where r is the intrinsic growth rate of a species and K
is the environmental carrying capacity, representing the maximum sustainable population
density 20, 28, 29.
In this example, we take r 0.2, K 1. Figure 8 shows the solution behavior when
0.8, 1.6 at dierent times T 0.1, 0.5, 1.0, while Figure 9 shows the solution behavior
with dierent values of between 0 and 1 and xed value of 1.8 at time T 1.0. Figure 9
also shows that the system exhibits anomalous diusion behavior and that the solution
International Journal of Dierential Equations 17
5 4 3 2 1 0 1 2 3 4 5
x
T 0.1
T 0.5
T 1
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
u
x
,
t
Figure 5: Numerical solutions for Example 5.3 with 1.0 and 2.0 at dierent times T 0.1, 0.5, 1.0.
5 4 3 2 1 0 1 2 3 4 5
x
T 0.1
T 0.5
T 1
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
u
x
,
t
Figure 6: Numerical solutions for Example 5.3 with 0.8 and 1.8 at dierent times T 0.1, 0.5, 1.0.
continuously depends on the time and space fractional derivatives. Although the source
term for Fishers equation sux, t, x, t rux, t1 ux, t/K is not globally Lipschitz
continuous, the solution of the discrete numerical method still yields bounds on the solution
of the continuous problem and the solution of the numerical method ENM converges to
the unique solution of the continuous problem 5.6 as the time and space steps tend to zero
30.
18 International Journal of Dierential Equations
5 4 3 2 1 0 1 2 3 4 5
x
T 0.1
T 0.5
T 1
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
u
x
,
t
Figure 7: Numerical solutions for Example 5.3 with 0.5 and 1.5 at dierent times T 0.1, 0.5, 1.0.
5 4 3 2 1 0
x
T 0.1
T 0.5
T 1
0
2
4
6
8
10
12
14
16
18
20
u
x
,
t
Figure 8: Numerical solutions for Example 5.4 with 0.8 and 1.6 at dierent times T 0.1, 0.5, 1.0.
6. Conclusions
In this paper, we have proposed an eective numerical method to solve the TSFFPENST
and proved that the ENM is stable and convergent provided that the nonlinear source term
satises the Lipschitz condition, the solution of the continuous problem satises the smooth
International Journal of Dierential Equations 19
5 4 3 2 1 0
x
0.3
1
0
2
4
6
8
10
12
u
x
,
t
Figure 9: Numerical solutions for Example 5.4 with xed 1.8 at time T 1.0, and dierent values of
0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0.
solution condition, and p
i0
w
i
u
n1
li1
Ml1
i0
w
i
u
n1
li1
_
b
n
u
0
l
n1
j0
_
b
j
b
j1
_
u
nj
l
0
s
n
l
.
A.1
Now setting s
n
l
/
_
t
n
0
t
n
1
ux
l
, d, then we have
s
n
l
_
n1
j0
_
t
j1
t
j
t
n
1
ux
l
, d. A.2
20 International Journal of Dierential Equations
Applying the Mean Value Theorem M.V.T for integration yields
s
n
l
_
n1
j0
u
_
x
l
,
j
_
_
t
j1
t
j
t
n
1
d, where t
j
<
j
< t
j1
_
n1
j0
u
j
l
u
j1
l
2
_
t
n
t
j
_
_
t
n
t
j1
_
_
2
n1
j0
_
u
j
l
u
j1
l
__
_
n j
_
_
n j 1
_
2
_
1
_
n1
j0
_
u
nj1
l
u
nj
l
__
_
j 1
_
_
1
n1
j0
q
j
_
u
nj1
l
u
nj
l
_
,
A.3
where
1
/2 1, q
j
j 1
, j 0, 1, . . . .
Also, we have
l1
i0
w
i
u
n1
li1
Ml1
i0
w
i
u
n1
li1
M1
i0
li
u
n1
i
, A.4
where
li
w
li1
, 1 i l 2,
w
0
w
2
, i l 1,
2w
1
, i l,
w
0
w
2
, i l 1,
w
il1
, l 2 i M 1.
A.5
Now, substituting A.3 and A.4 into A.1, we obtain the numerical scheme for
Example 5.2 as
u
n1
l
0
r
0
M1
i0
li
u
n1
i
b
n
u
0
l
n1
j0
_
_
b
j
b
j1
_
u
nj
l
1
q
j
_
u
nj1
l
u
nj
l
__
, A.6
International Journal of Dierential Equations 21
and the numerical scheme for Example 5.3 as
_
1
0
p
l
h
_
u
n1
l
0
p
l1
h
u
n1
l1
0
r
0
M1
i0
li
u
n1
i
b
n
u
0
l
n1
j0
_
_
b
j
b
j1
_
u
nj
l
1
q
j
_
u
nj1
l
u
nj
l
__
.
A.7
Acknowledgments
This research has been supported by a Ph.D. Fee Waiver Scholarship and a School of
Mathematical Sciences Scholarship, QUT. The authors also wish to thank the referees for their
constructive comments and suggestions.
References
1 H. Risken, The FokkerPlanck Equation: Methods of Solution and Applications, vol. 18 of Springer Series in
Synergetics, Springer, Berlin, Germany, 2nd edition, 1989.
2 D. A. Benson, S. W. Wheatcraft, and M. M. Meerschaert, Application of a fractional advection
dispersion equation, Water Resources Research, vol. 36, no. 6, pp. 14031412, 2000.
3 D. A. Benson, S. W. Wheatcraft, and M. M. Meerschaert, The fractionalorder governing equation of
L evy motion, Water Resources Research, vol. 36, no. 6, pp. 14131423, 2000.
4 J.P. Bouchaud and A. Georges, Anomalous diusion in disordered media: statistical mechanisms,
models and physical applications, Physics Reports, vol. 195, no. 45, pp. 127293, 1990.
5 R. Metzler and J. Klafter, Fractional FokkerPlanck equation: dispersive transport in an external force
eld, Journal of Molecular Liquids, vol. 86, no. 1, pp. 219228, 2000.
6 G. M. Zaslavsky, Chaos, fractional kinetics, and anomalous transport, Physics Reports, vol. 371, no.
6, pp. 461580, 2002.
7 F. Liu, V. Anh, and I. Turner, Numerical solution of the space fractional FokkerPlanck equation,
Journal of Computational and Applied Mathematics, vol. 166, no. 1, pp. 209219, 2004.
8 R. Goreno and F. Mainardi, Random walk models for spacefractional diusion processes,
Fractional Calculus & Applied Analysis, vol. 1, no. 2, pp. 167191, 1998.
9 C. W. Chow and K. L. Liu, FokkerPlanck equation and subdiusive fractional FokkerPlanck
equation of bistable systems with sinks, Physica A, vol. 341, no. 14, pp. 87106, 2004.
10 S. B. Yuste and L. Acedo, An explicit nite dierence method and a newvon Neumanntype stability
analysis for fractional diusion equations, SIAM Journal on Numerical Analysis, vol. 42, no. 5, pp.
18621874, 2005.
11 T. A. M. Langlands and B. I. Henry, The accuracy and stability of an implicit solution method for the
fractional diusion equation, Journal of Computational Physics, vol. 205, no. 2, pp. 719736, 2005.
12 C.M. Chen, F. Liu, I. Turner, and V. Anh, A Fourier method for the fractional diusion equation
describing subdiusion, Journal of Computational Physics, vol. 227, no. 2, pp. 886897, 2007.
13 P. Zhuang, F. Liu, V. Anh, and I. Turner, New solution and analytical techniques of the implicit
numerical method for the anomalous subdiusion equation, SIAM Journal on Numerical Analysis,
vol. 46, no. 2, pp. 10791095, 2008.
14 C.M. Chen, F. Liu, and V. Anh, A Fourier method and an extrapolation technique for Stokes
rst problem for a heated generalized second grade uid with fractional derivative, Journal of
Computational and Applied Mathematics, vol. 223, no. 2, pp. 777789, 2009.
15 S. Chen, F. Liu, P. Zhuang, and V. Anh, Finite dierence approximations for the fractional Fokker
Planck equation, Applied Mathematical Modelling, vol. 33, no. 1, pp. 256273, 2009.
16 I. Podlubny, Fractional Dierential Equations: An Introduction to Fractional Derivatives, Fractional
Dierential Equations, to Methods of Their Solution and Some of Their Applications, vol. 198 of Mathematics
in Science and Engineering, Academic Press, New York, NY, USA, 1999.
22 International Journal of Dierential Equations
17 A. Schot, M. K. Lenzi, L. R. Evangelista, L. C. Malacarne, R. S. Mendes, and E. K. Lenzi, Fractional
diusion equation with an absorbent term and a linear external force: exact solution, Physics Letters
A, vol. 366, no. 45, pp. 346350, 2007.
18 E. K. Lenzi, L. C. Malacarne, R. S. Mendes, and I. T. Pedron, Anomalous diusion, nonlinear
fractional FokkerPlanck equation and solutions, Physica A, vol. 319, pp. 245252, 2003.
19 P. Zhuang, F. Liu, V. Anh, and I. Turner, Numerical treatment for the fractional FokkerPlanck
equation, The ANZIAM Journal, vol. 48, pp. C759C774, 20062007.
20 B. Baeumer, M. Kov acs, and M. M. Meerschaert, Numerical solutions for fractional reactiondiusion
equations, Computers & Mathematics with Applications, vol. 55, no. 10, pp. 22122226, 2008.
21 K. B. Oldham and J. Spanier, The Fractional Calculus: Theory and Applications of Dierentiation and
Integration to Arbitrary Order, vol. 11 of Mathematics in Science and Engineering, Academic Press, New
York, NY, USA, 1974.
22 M. M. Meerschaert and C. Tadjeran, Finite dierence approximations for twosided spacefractional
partial dierential equations, Applied Numerical Mathematics, vol. 56, no. 1, pp. 8090, 2006.
23 R. Lin and F. Liu, Fractional high order methods for the nonlinear fractional ordinary dierential
equation, Nonlinear Analysis: Theory, Methods & Applications, vol. 66, no. 4, pp. 856869, 2007.
24 F. Liu, C. Yang, and K. Burrage, Numerical method and analytic technique of the modied
anomalous subdiusion equation with a nonlinear source term, Journal of Computational and Applied
Mathematics, vol. 231, no. 1, pp. 160176, 2009.
25 M. M. Meerschaert and C. Tadjeran, Finite dierence approximations for fractional advection
dispersion owequations, Journal of Computational and Applied Mathematics, vol. 172, no. 1, pp. 6577,
2004.
26 W. Deng, Finite element method for the space and time fractional FokkerPlanck equation, SIAM
Journal on Numerical Analysis, vol. 47, no. 1, pp. 204226, 20082009.
27 J. I. Ramos, Damping characteristics of nite dierence methods for onedimensional reaction
diusion equations, Applied Mathematics and Computation, vol. 182, no. 1, pp. 607609, 2006.
28 D. delCastilloNegrete, B. A. Carreras, and V. E. Lynch, Front dynamics in reactiondiusion
systems with Levy ights: a fractional diusion approach, Physical Review Letters, vol. 91, no. 1,
Article ID 018302, 4 pages, 2003.
29 V. E. Lynch, B. A. Carreras, D. delCastilloNegrete, K. M. FerreiraMejias, and H. R. Hicks,
Numerical methods for the solution of partial dierential equations of fractional order, Journal of
Computational Physics, vol. 192, no. 2, pp. 406421, 2003.
30 B. Baeumer, M. Kov acs, and M. M. Meerschaert, Fractional reproductiondispersal equations and
heavy tail dispersal kernels, Bulletin of Mathematical Biology, vol. 69, no. 7, pp. 22812297, 2007.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 764738, 8 pages
doi:10.1155/2010/764738
Research Article
Hes Variational Iteration Method for Solving
Fractional Riccati Differential Equation
H. Jafari and H. Tajadodi
Department of Mathematics and Computer Science, University of Mazandaran,
P.O. Box 4741695447, Babolsar, Iran
Correspondence should be addressed to H. Jafari, jafari@umz.ac.ir
Received 10 August 2009; Accepted 28 January 2010
Academic Editor: Shaher M. Momani
Copyright q 2010 H. Jafari and H. Tajadodi. This is an open access article distributed under
the Creative Commons Attribution License, which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
We will consider Hes variational iteration method for solving fractional Riccati dierential
equation. This method is based on the use of Lagrange multipliers for identication of optimal
value of a parameter in a functional. This technique provides a sequence of functions which
converges to the exact solution of the problem. The present method performs extremely well in
terms of eciency and simplicity.
1. Introduction
The fractional calculus has found diverse applications in various scientic and technological
elds 1, 2, such as thermal engineering, acoustics, electromagnetism, control, robotics,
viscoelasticity, diusion, edge detection, turbulence, signal processing, and many other
physical processes. Fractional dierential equations FDEs have also been applied in
modeling many physical, engineering problems, and fractional dierential equations in
nonlinear dynamics 3, 4.
The variational iteration method was proposed by He 5 and was successfully
applied to autonomous ordinary dierential equation 6, to nonlinear partial dierential
equations with variable coecients 7, to SchrodingerKdV, generalized Kd and shallow
water equations 8, to linear Helmholtz partial dierential equation 9, recently to
nonlinear fractional dierential equations with Caputo dierential derivative 10, 11, and
to other elds, 12. The variational iteration method gives rapidly convergent successive
approximations of the exact solution if such a solution exists; otherwise a fewapproximations
can be used for numerical purposes. The method is eectively used in 68, 1315 and the
references therein. Jafari et al. applied the variational iteration method to the Gas Dynamics
2 International Journal of Dierential Equations
Equation and Stefan problem 13, 14. We consider here the following nonlinear fractional
Riccati dierential equation:
D
yt At Bty Cty
2
, 1.1
subject to the initial conditions
y
k
0 c
k
, k 0, 1, . . . , n 1, 1.2
where is fractional derivative order, n is an integer, At, Bt, and Ct are known real
functions, and c
k
is a constant. There are several denitions of a fractional derivative of order
> 0. The two most commonly used denitions are the RiemannLiouville and Caputo.
Each denition uses RiemannLiouville fractional integration and derivatives of whole order.
The dierence between the two denitions is in the order of evaluation. RiemannLiouville
fractional integration of order a is dened as
I
fx
1
x
0
x t
1
ftdt, > 0, x > 0. 1.3
The following two equations dene RiemannLiouville and Caputo fractional derivatives of
order , respectively:
D
fx
d
m
dx
m
I
m
fx
,
1.4
D
fx I
m
d
m
dx
m
fx
, 1.5
where m 1 < m and m N. We have chosen to use the Caputo fractional
derivative because it allows traditional initial and boundary conditions to be included in
the formulation of the problem, but for homogeneous initial condition assumption, these
two operators coincide. For more details on the geometric and physical interpretation for
fractional derivatives of both the RiemannLiouville and Caputo types, see 1.
2. Analysis of the Variational Iteration Method
We consider the fractional dierential equation
D
yt At Bty Cty
2
, 0 < 1, 2.1
with initial condition y0 0, where D
/dt
y
n
d
At Bty
n
Cty
2
n
. 2.2
International Journal of Dierential Equations 3
To identify the multiplier, we approximately write 2.2 in the form
y
n1
y
n
t
0
y
n
d
At Bt y
n
Ct y
2
n
d, 2.3
where is a general Lagrange multiplier, which can be identied optimally via the variational
theory, and y
n
is a restricted variation, that is, y
n
0.
The successive approximation y
n1
, n 0 of the solution yt will be readily obtained
upon using Lagranges multiplier, and by using any selective function y
0
. The initial value
y0 and y
t
0 are usually used for selecting the zeroth approximation y
0
. To calculate the
optimal value of , we have
y
n1
y
n
t
0
dy
n
d
d 0. 2.4
This yields the stationary conditions
y
n
d
At Bty
n
Cty
2
n
, 2.6
and nally the exact solution is obtained by
yt lim
n
y
n
t.
2.7
3. Applications and Numerical Results
To give a clear overview of this method, we present the following illustrative examples.
Example 3.1. Consider the following fractional Riccati dierential equation:
d
y
dt
y
2
t 1, 0 < 1,
3.1
subject to the initial condition y0 0.
The exact solution of 3.1 is yt e
2t
1/e
2t
1, when 1.
In view of 2.6 the correction functional for 3.1 turns out to be
y
n1
y
n
I
y
n
d
y
2
n
1
d. 3.2
4 International Journal of Dierential Equations
0.2
0.4
0.6
0.8
y
0.2 0.4 0.6 0.8 1 1.2 1.4
t
Figure 1: Dashed line: Approximate solution.
Beginning with y
0
t t
1
1 2t
3
1
2
1 3
,
y
2
t
t
1
1 2t
3
1
2
1 3
2
32
41/2 t
5
1 1 31 5
64
1 2
2
1/2 3t
7
1
4
1 31 7
,
.
.
.
3.3
and so on. The nth Approximate solution of the variational iteration method converges to the
exact series solution. So, we approximate the solution yt lim
n
y
n
t.
In Figure 1, Approximate solution yt
y
3
t of 3.4 using VIM and the exact
solution have been plotted for 1. In Figure 2, Approximate solution yt
y
3
t of 3.4
using VIM and the exact solution have been plotted for 0.98.
Comment. This example has been solved using HAM, ADM, and HPM in 1618. It should
be noted that these methods have given same result after applying the Pad e approximants on
yt.
International Journal of Dierential Equations 5
0.2
0.4
0.6
0.8
y
0.2 0.4 0.6 0.8 1 1.2 1.4
t
Figure 2: Dashed line: Approximate solution.
Example 3.2. Consider the following fractional Riccati dierential equation:
d
y
dt
2yt y
2
t 1, 0 < 1,
3.4
subject to the initial condition y0 0.
The exact solution of 3.4 is yt 1
2tanh
2t 1/2 log
2 1/
2 1,
when 1.
Expanding yt using Taylor expansion about t 0 gives
yt t t
2
t
3
3
t
4
3
7t
5
15
7t
6
45
53t
7
315
71t
8
315
.
3.5
The correction functional for 3.4 turns out to be
y
n1
y
n
I
y
n
d
2y
n
y
2
n
1
d. 3.6
6 International Journal of Dierential Equations
0.5
1
1.5
2
y
0.2 0.4 0.6 0.8 1 1.2 1.4
t
Figure 3: Dashed line: Approximate solution.
Beginning with y
0
t t
2
12
t
2
1 1/2
4
t
3
1/2
13 1
,
y
2
t
t
2
12
t
2
1 1/2
2
12
1/2t
4
14
4
3
3 1
1 2t
3
1
2
3 1
123t
4
2 14 1
2
42
4t
5
1/22 15 1
2
32
4 1/2t
5
13 15 1
205t
6
13 16 1
1024
1/2
2
3 1/2t
7
3
1
2
3 17 1
.
.
.
3.7
and so on. In Figure 3, Approximate solution yt
y
3
t of 3.4 using VIM and the exact
solution have been plotted for 1. In Figure 4, Approximate solution yt
y
3
t of 3.4
using VIM and the exact solution have been plotted for 0.98.
International Journal of Dierential Equations 7
0.5
1
1.5
2
y
0.2 0.4 0.6 0.8 1 1.2 1.4
t
Figure 4: Dashed line: Approximate solution.
4. Conclusion
In this paper the variational iteration method is used to solve the fractional Riccati dierential
equations. We described the method, used it on two test problems, and compared the results
with their exact solutions in order to demonstrate the validity and applicability of the method.
Acknowledgments
The authors express their gratitude to the referees for their valuable suggestions and
corrections for improvement of this paper. Mathematica has been used for computations in
this paper.
References
1 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, San Diego, Calif, USA, 1999.
2 K. B. Oldham and J. Spanier, The Fractional Calculus, Academic Press, New York, NY, USA, 1974.
3 H. Jafari and V. DaftardarGejji, Solving a system of nonlinear fractional dierential equations using
Adomian decomposition, Journal of Computational and Applied Mathematics, vol. 196, no. 2, pp. 644
651, 2006.
4 J. G. Lu and G. Chen, A note on the fractionalorder Chen system, Chaos, Solitons & Fractals, vol. 27,
no. 3, pp. 685688, 2006.
5 J. He, A new approach to nonlinear partial dierential equations, Communications in Nonlinear
Science and Numerical Simulation, vol. 2, no. 4, pp. 230235, 1997.
6 J.H. He, Variational iteration method for autonomous ordinary dierential systems, Applied
Mathematics and Computation, vol. 114, no. 23, pp. 115123, 2000.
8 International Journal of Dierential Equations
7 J.H. He, Variational principles for some nonlinear partial dierential equations with variable
coecients, Chaos, Solitons & Fractals, vol. 19, no. 4, pp. 847851, 2004.
8 M. A. Abdou and A. A. Soliman, New applications of variational iteration method, Physica D, vol.
211, no. 12, pp. 18, 2005.
9 S. Momani and S. Abuasad, Application of Hes variational iteration method to Helmholtz
equation, Chaos, Solitons & Fractals, vol. 27, no. 5, pp. 11191123, 2006.
10 Z. M. Odibat and S. Momani, Application of variational iteration method to nonlinear dierential
equations of fractional order, International Journal of Nonlinear Sciences and Numerical Simulation, vol.
7, no. 1, pp. 2734, 2006.
11 S. Momani and Z. Odibat, Numerical approach to dierential equations of fractional order, Journal
of Computational and Applied Mathematics, vol. 207, no. 1, pp. 96110, 2007.
12 J.H. He, Some asymptotic methods for strongly nonlinear equations, International Journal of Modern
Physics B, vol. 20, no. 10, pp. 11411199, 2006.
13 H. Jafari, H. Hosseinzadeh, and E. Salehpoor, A new approach to the gas dynamics equation: an
application of the variational iteration method, Applied Mathematical Sciences, vol. 2, no. 48, pp. 2397
2400, 2008.
14 H. Jafari, A. Golbabai, E. Salehpoor, and Kh. Sayehvand, Application of variational iteration method
for Stefan problem, Applied Mathematical Sciences, vol. 2, no. 60, pp. 30013004, 2008.
15 H. Jafari and A. Alipoor, A new method for calculating General Lagranges multiplier in the
variational iteration method, Numerical Method for Partial Dierential Equations, In press, 2010.
16 J. Cang, Y. Tan, H. Xu, and S.J. Liao, Series solutions of nonlinear Riccati dierential equations with
fractional order, Chaos, Solitons & Fractals, vol. 40, no. 1, pp. 19, 2009.
17 S. Momani and N. Shawagfeh, Decomposition method for solving fractional Riccati dierential
equations, Applied Mathematics and Computation, vol. 182, no. 2, pp. 10831092, 2006.
18 Z. Odibat and S. Momani, Modied homotopy perturbation method: application to quadratic Riccati
dierential equation of fractional order, Chaos, Solitons & Fractals, vol. 36, no. 1, pp. 167174, 2008.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 954674, 11 pages
doi:10.1155/2010/954674
Research Article
Solitary Wave Solutions for a TimeFraction
Generalized HirotaSatsuma Coupled KdV
Equation by a New Analytical Technique
Majid Shateri and D. D. Ganji
Department of Mechanical Engineering, Babol University of Technology,
P.O. Box 484, 47148 71167 Babol, Iran
Correspondence should be addressed to D. D. Ganji, ddg davood@yahoo.com
Received 17 May 2009; Accepted 7 July 2009
Academic Editor: Shaher Momani
Copyright q 2010 M. Shateri and D. D. Ganji. This is an open access article distributed under
the Creative Commons Attribution License, which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
A new iterative technique is employed to solve a system of nonlinear fractional partial dierential
equations. This new approach requires neither Lagrange multiplier like variational iteration
method VIM nor polynomials like Adomians decomposition method ADM so that can be more
easily and eectively established for solving nonlinear fractional dierential equations, and will
overcome the limitations of these methods. The obtained numerical results show good agreement
with those of analytical solutions. The fractional derivatives are described in Caputo sense.
1. Introduction
In recent years, it has been turned out that fractional dierential equations can be
used successfully to model many phenomena in various elds such as uid mechanics,
viscoelasticity, physics, chemistry, and engineering. For instance, the uiddynamics trac
model with fractional derivatives 1 is able to eliminate the deciency arising from the
assumption of continuum trac ow, and the nonlinear oscillation of earthquakes can be
modeled by fractional derivatives 2. Fractional dierentiation and integration operators can
also be used for extending the diusion and wave equations 3. Most of fractional dierential
equations do not have exact analytical solutions, hence considerable heed has been focused
on the approximate and numerical solutions of these equations. Although variational
iteration method 48 and Adomians decomposition method 914 are approaches that
have been utilized extensively to provide analytical approximations of linear and nonlinear
problems, they have limitations due to complicated algorithms of calculating Adomian
polynomials for nonlinear fractional problems, and an inherent inaccuracy in determining
2 International Journal of Dierential Equations
the Lagrange multiplier for fractional equations. In this study, a new alternative procedure
that needs no Lagrange multiplier or Adomian polynomials is used to obtain an analytical
approximate solution of a system of nonlinear fractional partial dierential equations 1.1 to
illustrate the eectiveness, accuracy, and convenience of this method.
In this work, we consider the solution of generalized HirotaSatsuma coupled KdV
of timefractional order which is presented by a system of nonlinear partial dierential
equations, of the form:
D
t
u
1
2
u
xxx
3uu
x
3vw
x
,
D
t
v v
xxx
3uv
x
,
D
t
w w
xxx
3uw
x
,
0 < < 1, 1.1
subject to the following initial conditions:
ux, 0
2k
2
3
2k
2
tanh
2
kx,
vx, 0
4k
2
c
0
_
k
2
_
3c
2
1
4k
2
_
k
2
_
tanhkx
3c
1
,
wx, 0 c
0
c
1
tanhkx,
1.2
where k, c
0
, c
1 /
0, and are arbitrary constants. The HirotaSatsuma system of equations
15 was introduced to describe the interaction of two long waves with dierent dispersion
relations. The case of 1 in system 1.1 was solved by Wu et al. 16. If c , then in one
case the ux, t 2k
2
/3 2k
2
tanh
2
kx ct, vx, t 4k
2
c
0
k
2
/3c
2
1
4k
2
k
2
/3c
1
tanhkx ct and wx, t c
0
c
1
tanhkx ct are travellingwave solutions of
system 1.1 when 1.
2. Basic Denitions
In this section, there are some basic denitions and properties of the fractional calculus theory
which are used in this paper.
Denition 2.1. A real function fx, x > 0, is mentioned to be in the space C
, R if there
exists a real number p> such that fx x
p
f
1
x, where f
1
x C0, , and it is said to
be in the space C
m
if f
m
C
, m N.
Denition 2.2. The leftsided RiemannLiouville fractional integral operator of order 0,
of a function f C
, 1 is dened as
D
fx
1
_
x
0
x t
1
ftdt, > 0, x > 0,
D
0
fx fx.
2.1
International Journal of Dierential Equations 3
The properties of the operator D
, 1, , 0 and > 1:
D
fx D
fx,
D
fx D
fx,
D
_
1
_
_
1
_x
.
2.2
The RiemannLiouville derivative has certain disadvantages, in trying way to model
realworld phenomena with fractional dierential equations. Therefore, we will employ a
modication of fractional dierential operator D
fx D
m
D
m
fx
1
m
_
x
0
x t
m1
f
m
tdt,
2.3
for m 1 < m, m N, x > 0, f C
m
1
Also, we need two of its basic properties.
Lemma 2.4. If m 1 < m, m N, and f C
m
fx fx,
D
fx fx
m1
k0
f
k
0
x
k
k!
, x > 0.
2.4
The Caputo fractional derivative is considered here, because it allows traditional initial
and boundary conditions to be included in the formulation of the problem.
In this work, we consider the onedimensional linear nonhomogeneous fractional
partial dierential equations in uid mechanics, where the unknown function ux, t is
assumed to be a causal function of time, that is, vanishing for t < 0.
Denition 2.5. For m as the smallest integer that exceeds , the Caputo timefractional
derivative operator of order > 0 is dened as
D
t
ux, t
ux, t
t
1
m
_
t
0
t
m1
m
ux,
m
d, m 1 < < m,
m
ux, t
t
m
, m N.
2.5
For more information on the mathematical properties of fractional derivatives and
integrals, one can consult the mentioned references.
4 International Journal of Dierential Equations
0.4
0.3
0.2
0.1
0
u1, 0.6
v1, 0.6
w1, 0.6
3 2 1 0 1
h
Figure 1: hcurves of ux, t: dash, vx, t: dash dot, and wx, t: solid when k 0.1, 1, 1.5, c
0
1.5,
c
1
0.1 at point 1,0.6.
3. Basic Ideas of Fractional Iteration Method (FIM)
As pointed in 19, to illustrate fractional iteration method, we consider the following
nonlinear fractional dierential equation more general form can be considered without loss
of generality:
D
yx f
_
x, yx
_
0, y
k
0 a
k
, k 0, 1, . . . , m 1, 3.1
where the fractional dierential operator D
_
Hx
_
D
yx f
_
x, yx
___
0. 3.3
Let h
/
0 denote the socalled auxiliary parameter. Multiplying 3.3 by h and then adding y,
the solution of 3.1, on both sides of the resulted term yields
yx yx hD
_
Hx
_
D
yx f
_
x, yx
___
. 3.4
International Journal of Dierential Equations 5
0.493
0.498
0.503
0.508
0.513
40
20
0
20
40
4
3
2
1
0
t
x
a
0.493
0.498
0.503
0.508
0.513
40
20
0
20
40
4
3
2
1
0
t
x
b
Figure 2: The solitary wave solution of ux, t, FIM result a and exact solution b, when k 0.1, 1,
1.5, c
0
1.5, c
1
0.1.
3.22
3.12
3.02
2.92
2.82
40
20
0
20
40
4
3
2
1
0
t
x
a
3.22
3.12
3.02
2.92
2.82
40
20
0
20
40
4
3
2
1
0
t
x
b
Figure 3: The solitary wave solution of vx, t, FIM result a and exact solution b, when k 0.1, 1,
1.5, c
0
1.5, c
1
0.1.
However 3.4 can be solved iteratively as follows:
y
n1
x y
n
x hD
_
Hx
_
D
y
n
x f
_
x, y
n
x
___
. 3.5
In 3.5, the subscript n denotes the nth iteration, and provided that the right hand of
it, that is, yxhD
HxD
4k
2
_
k
2
_
tanhkx
3c
1
,
w
0
x, t c
0
c
1
tanhkx.
4.1
Choosing and Hx, t 1, we can construct the iterative scheme 3.5 for
investigation of the traveling wave solution of 1.1 as follows:
u
n1
x, t u
n
x, t hD
t
_
D
t
u
n
x, t
1
2
3
x
3
u
n
x, t 3u
n
x, t
x
u
n
x, t 3
x
v
n
x, tw
n
x, t
_
,
v
n1
x, t v
n
x, t hD
t
_
D
t
v
n
x, t
3
x
3
v
n
x, t 3u
n
x, t
x
v
n
x, t
_
,
w
n1
x, t w
n
x, t hD
t
_
D
t
w
n
x, t
3
x
3
w
n
x, t 3u
n
x, t
x
w
n
x, t
_
.
4.2
Substituting the initial approximations, 4.1, into 4.2, for the case 1, yields
u
1
0.493 0.02 tanh0.1x
2
h
_
0.016
_
0.1 0.1 tanh0.1x
2
_
2
tanh0.1xt
0.0008 tanh0.1x
3
_
0.1 0.1 tanh0.1x
2
_
t 0.12
_
0.493 0.02 tanh0.1x
2
_
tanh0.1x
_
0.1 0.1 tanh0.1x
2
_
t 3
_
0.001342 0.001342 tanh0.1x
2
_
1.5 1.5 tanh0.1xt30.001342 0.001342 tanh0.1x
_
0.15 0.15 tanh0.1x
2
_
t
_
,
v
1
0.001342 0.001342 tanh0.1x h
_
0.002684
_
0.1 0.1 tanh0.1x
2
_
2
t
0.0005368 tanh0.1x
2
_
0.1 0.1 tanh0.1x
2
_
t 3
_
0.493 0.02 tanh0.1x
2
_
_
0.001342 0.001342 tanh0.1x
2
_
t
_
,
w
1
1.5 1.5 tanh0.1x h
_
0.3
_
0.1 0.1 tanh0.1x
2
_
2
t
0.06 tanh0.1x
2
_
0.1 0.1 tanh0.1x
2
_
3
_
0.493 0.02 tanh0.1x
2
_
_
0.15 0.15 tanh0.1x
2
_
t
_
.
4.3
International Journal of Dierential Equations 9
5. Result and Discussion
In this section, four gures are presented corresponding to FIM results and exact solutions
for the solitary wave solutions ux, t, vx, t, and wx, t with the initial conditions 1.2,
when k 0.1, 1, 1.5, c
0
1.5, c
1
0.1. Furthermore, numerical values for the case
0.5, 0.75, 1.0, and k 0.1, 1.5, c
0
1.5, c
1
0.1 are obtained for ux, t, vx, t, and
wx, t.
Demonstrating the exactness of FIM, the numerical results are presented and only
few iterations are required to achieve accurate solutions. The convergence of FIM for the
generalized fractionalorder HirotaSatsumacoupled KdV equation is controllable, using
the socalled hcurves presented in Figure 1 which are obtained based on the fourthorder
FIM approximate solutions. In general, by the means of the socalled hcurve, it is straight
forward to choose a proper value of h which ensures that the solution series is convergent.
This proper value of h corresponds to the curve segment nearly parallel to the horizontal axis.
Both exact results and approximate solutions obtained for the rst four approximations are
plotted in Figures 2, 3, and 4. There are no visible dierences in two solutions of each pair of
diagrams.
Tables 1, 2, and 3 show the numerical values by FIM when 0.5, 0.75, 1.0 and
k 0.1, 1.5, c
0
1.5, c
1
0.1 for ux, t, vx, t, and wx, t respectively.
6. Conclusion
In this paper, the fractional iteration method FIM has been successfully applied to study
HirotaSatsumacoupled KdV of timefractionalorder equation. FIM results are compared
with the exact solutions and those obtained by Homotopy perturbation method 21.
The results show that fractional iteration method is a powerful and ecient technique
in nding exact and approximate solutions for nonlinear partial dierential equations of
fractional order. The method provides the user with more realistic series solutions that
converge very rapidly in real physical problems.
Compared with the ADM and VIM, the FIM has following advantages, 19.
1 The auxiliary parameter h provides us with a convenient way to modify and control
the convergence region of the solution.
2 The solution of a given nonlinear problemcan be expressed by an innite number of
solution series and thus can be more eciently approximated by a better selection
of the auxiliary parameter values.
3 Unlike the ADM, the FIM method is free from the need to use Adomian
polynomials.
4 This method has no need for the Lagrange multiplier, correction functional,
stationary conditions, the variational theory, and so forth, which eliminates the
complications that exist in the VIM.
5 The fractional iteration method can be easily comprehended with only a basic
knowledge of fractional calculus.
6 Compared to the ADM and VIM, the presented method proves simpler in its
principles and more convenient for computer algorithms.
In this work, we used Maple Package to calculate the series obtained by fractional
iteration method.
10 International Journal of Dierential Equations
References
1 I. Podlubny, Fractional Dierential Equations, Academic Press, New York, NY, USA, 1999.
2 F. Mainardi, Fractional Calculus Some Basic Problems in Continuum and Statistical Mechanics, Springer,
New York, NY, USA, 1997.
3 W. R. Schneider and W. Wyss, Fractional diusion and wave equations, Journal of Mathematical
Physics, vol. 30, pp. 134144, 1989.
4 J. H. He, Approximate analytical solution for seepage ow with fractional derivatives in porous
media, Computer Methods in Applied Mechanics and Engineering, vol. 167, pp. 5768, 1998.
5 J. H. He, Variational iteration methodsome recent results and new interpretations, Journal of
Computational and Applied Mathematics, vol. 207, no. 1, pp. 317, 2007.
6 J. H. He and X. H. Wu, Variational iteration method: newdevelopment and applications, Computers
and Mathematics with Applications, vol. 54, no. 78, pp. 881894, 2007.
7 H. Tari, D. D. Ganji, and H. Babazadeh, The application of Hes variational iteration method to
nonlinear equations arising in heat transfer, Physics Letters, Section A, vol. 363, no. 3, pp. 213217,
2007.
8 S. Abbasbandy and A. Shirzadi, The variational iteration method for a family of fthorder boundary
value dierential equations, International Journal of Nonlinear Dynamics in Engineering and Sciences,
vol. 1, no. 1, pp. 3946, 2009.
9 G. Adomian, Solving Frontier Problems of Physics: The Decomposition Method, Kluwer Academic
Publishers, Boston, Mass, USA, 1994.
10 G. Adomian, A review of the decomposition method in applied mathematics, Journal of
Mathematical Analysis and Applications, vol. 135, no. 2, pp. 501544, 1988.
11 G. Adomian, Solutions of nonlinear P.D.E, Applied Mathematics Letters, vol. 11, no. 3, pp. 121123,
1998.
12 Q. Esmaili, A. Ramiar, E. Alizadeh, and D. D. Ganji, An approximation of the analytical solution of
the JeeryHamel ow by decomposition method, Physics Letters A, vol. 372, no. 19, pp. 34343439,
2008.
13 A. M. Wazwaz, A new algorithm for calculating adomian polynomials for nonlinear operators,
Applied Mathematics and Computation, vol. 111, no. 1, pp. 5369, 2000.
14 S. Momani and Z. Odibat, Analytical solution of a timefractional NavierStokes equation by
Adomian decomposition method, Applied Mathematics and Computation, vol. 177, pp. 488494, 2006.
15 R. Hirota and J. Satsuma, Soliton solutions of a coupled Kortewegde Vries equation, Physics Letters
A, vol. 85, no. 89, pp. 407408, 1981.
16 Y. Wu, X. Geng, X. Hu, and S. Zhu, A generalized HirotaSatsuma coupled Kortewegde vries
equation and Miura transformations, Physics Letters A, vol. 255, pp. 259264, 1999.
17 I. Podlubny, Numerical solution of ordinary fractional dierential equations by the fractional
dierence method, in Advances in Dierence Equations, S. Elaydi, I. Gyori, and G. Ladas, Eds., Gordon
and Breach, Amsterdam, The Netherlands, 1997.
18 M. Caputo, Linear models of dissipation whose Q is almost frequency independent, Journal of the
Royal Astronomical Society, vol. 13, pp. 529539, 1967.
19 A. Ghorbani, Toward a new analytical method for solving nonlinear fractional dierential
equations, Computer Methods in Applied Mechanics and Engineering, vol. 197, pp. 41734179, 2008.
20 M. Reed and B. Simon, Methods of Modern Mathematical Physics, I: Functional Analysis, Academic Press,
New York, NY, USA, 1980.
21 Z. Z. Ganji, D. D. Ganji, and Y. Rostamiyan, Solitary wave solutions for a timefraction generalized
HirotaSatsuma coupled KdV equation by an analytical technique, Applied Mathematical Modelling,
vol. 33, no. 7, pp. 31073113, 2009.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 461048, 16 pages
doi:10.1155/2010/461048
Research Article
TimeOptimal Control of Systems with
Fractional Dynamics
Christophe Tricaud and YangQuan Chen
Center for SelfOrganizing and Intelligent Systems (CSOIS), Department of Electrical and
Computer Engineering, Utah State University, 41260 Old Main Hill, Logan, UT 843224160, USA
Correspondence should be addressed to Christophe Tricaud, c.tricaud@aggiemail.usu.edu
Received 1 August 2009; Accepted 5 December 2009
Academic Editor: Wen Chen
Copyright q 2010 C. Tricaud and Y. Chen. This is an open access article distributed under the
Creative Commons Attribution License, which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
We introduce a formulation for the timeoptimal control problems of systems displaying fractional
dynamics in the sense of the RiemannLiouville fractional derivatives operator. To propose a
solution to the general timeoptimal problem, a rational approximation based on the Hankel
data matrix of the impulse response is considered to emulate the behavior of the fractional
dierentiation operator. The original problem is then reformulated according to the new model
which can be solved by traditional optimal control problem solvers. The timeoptimal problem is
extensively investigated for a double fractional integrator and its solution is obtained using either
numerical optimization or timedomain analysis.
1. Introduction
In the world surrounding us, the physical laws of dynamics are not always followed by all
systems. When the considered systems are complex or can only be studied on a macroscopic
scale, they sometimes divert from the traditional integer order dynamic laws. In some cases,
their dynamics follow fractionalorder laws meaning that their behavior is governed by
fractionalorder dierential equations 1. As an illustration, it can be found in the literature
2 that materials with memory and hereditary eects, and dynamical processes, such as
gas diusion and heat conduction, in fractal porous media can be more precisely modeled
using fractionalorder models than using integerorder models. Another vein of research has
identied the dynamics of a frogs muscle to display fractionalorder behavior 3.
Optimal Control Problems OCPs or IntegerOrder Optimal Controls IOOCs can
be found in a wide variety of research topics such as engineering and science of course, but
economics as well. The eld of IOOCs has been investigated for a long time and a large
collection of numerical techniques has been developed to solve this category of problems 4.
2 International Journal of Dierential Equations
The main objective of an OCP is to obtain control input signals that will make a given
system or process satisfy a given set of physical constraints either on the systems states
or control inputs while extremizing a performance criterion or a cost function. Fractional
Optimal Control Problems FOCPs are OCPs in which the criterion and/or the dierential
equations governing the dynamics of the system display at least one fractional derivative
operator. The rst record of a formulation of the FOCP was given in 5. This formulation
was general but includes constraints on the systems states or control inputs. Later, a general
denition of FOCP was formulated in 6 that is similar to the general denition of OCPs. The
number of publications linked to FOCPs is limited since that problem has only been recently
considered.
Over the last decade, the framework of FOCPs has hatched and grown. In 5, Agrawal
gives a general formulation of FOCPs in the RiemannLiouville RL sense and proposes
a numerical method to solve FOCPs based on variational virtual work coupled with the
Lagrange multiplier technique. In 7, the fractional derivatives FDs of the system are
approximated using the Gr unwaldLetnikov denition, providing a set of algebraic equations
that can be solved using numerical techniques. The problem is dened in terms of the
Caputo fractional derivatives in 8 and an iterative numerical scheme is introduced to solve
the problem numerically. Distributed systems are considered in 9 and an eigenfunction
decomposition is used to solve the problem.
Ozdemir et al. 10 also use eigenfunction
expansion approach to formulate an FOCP of a 2dimensional distributed system. Cylindrical
coordinates for the distributed systemare considered in 11. Amodied Gr unwaldLetnikov
approach is introduced in 12 which leads to a central dierence scheme. Frederico and
Torres 1315, using similar denitions of the FOCPs, formulated a Noethertype theorem
in the general context of the fractional optimal control in the sense of Caputo and studied
fractional conservation laws in FOCPs. In 6, a rational approximation of the fractional
derivatives operator is used to link FOCPs and the traditional IOOCs. Anewsolution scheme
is proposed in 16, based on a dierent expansion formula for fractional derivatives.
In this article, we introduce a formulation to a special class of FOCP: the Fractional
TimeOptimal Control Problem FTOCP. Timeoptimal control problems are also referred
to in the literature as minimumtime control problems, free nal timeoptimal control, or
bangbang control problems. These dierent denominations dene the same kind of optimal
control problem in which the purpose is to transfer a system from a given initial state
to a specied nal state in minimum time. So far, this special class of FOCPs has been
disregarded in the literature. In 6, such a problemwas solved as an example to demonstrate
the capability and generality of the proposed method but no thorough studies were done.
The article is organized as follows. In Section 2, we give the denitions of fractional
derivatives in the RL sense and FOCP and introduce the formulation of FTOCP. In Section 3,
we consider the problem of the timeoptimal control of a fractional double integrator and
propose dierent schemes to solve the problem. In Section 4, the solution to the problem
is obtained for each approach for a given system. Finally, we give our conclusions in
Section 5.
2. Formulation of the Fractional TimeOptimal Control Problem
2.1. The Fractional Derivative Operator
There exist several denitions of the fractional derivative operator: RiemannLiouville,
Caputo, Gr unwaldLetnikov, Weyl, as well as Marchaud and Riesz 1720. Here, we are
International Journal of Dierential Equations 3
interested in the RiemannLiouville denition of the fractional derivatives for the formulation
of the FOCP.
The Left RiemannLiouville Fractional Derivative LRLFD of a function ft is
dened as
a
D
t
ft
1
n
_
d
dt
_
n
_
t
a
t
n1
fd, 2.1
where is the Gamma function dened for any complex number z as
z
_
0
t
z1
e
t
dt 2.2
and where the order of the derivative satises n 1 < n. The Right RiemannLiouville
Fractional Derivative RRLFD is dened as
t
D
b
ft
1
n
_
d
dt
_
n
_
b
t
t
n1
fd. 2.3
2.2. Fractional Optimal Control Problem Formulation
With the RL denition of the fractional derivatives operator given in 2.1 and 2.3, we can
specify a general FOCP: nd the optimal control ut for a fractional dynamical system that
minimizes the following performance criterion:
Ju G
o
xa, xb
_
b
a
L
o
x, u, tdt 2.4
subject to the following system dynamics:
a
D
t
xt Hx, u, t 2.5
with initial condition
xa x
a
2.6
and with the following constraints:
u
min
t ut u
max
t,
x
min
a xa x
max
a,
L
ti
t, xt, ut 0,
G
ei
xa, xb 0,
G
ee
xa, xb 0,
2.7
4 International Journal of Dierential Equations
where x is the state variable, t a, b stands for the time, and L, G, and H are arbitrary given
nonlinear functions. The subscripts o, ti, ei, and ee on the functions G, and L, , stand
for, respectively, objective function, trajectory constraint, endpoint inequality constraint and
endpoint equality constraint.
2.3. Fractional TimeOptimal Control Problem Formulation
A FTOCP is dened by a performance index of the form
Ju
_
b
a
1 dt b a, 2.8
which appears when we desire to minimize the time required to reach a given target x
b
given
some initial conditions x
a
.
Under such conditions, the problem is to transfer the system whose dynamics are
given by
a
D
t
xt Hx, u, t 2.9
from a given initial state xa x
a
to the desired state xb x
b
. The minimum time required
to reach the target is dened as t
.
To ensure that the problem has a solution, the control variables are required to be
constrained in the following way:
u
min
ut u
max
. 2.10
In the following, we will make use of the minimumprinciple to determine the optimal control
law u
t for the previously dened problem. We dene the state associated with the optimal
control law as x
t.
We dene the Hamiltonian H for the problem described by the dynamic system 2.9
and the criterion 2.8 as
Hx, u, t 1
_
Hx, u, t
a
D
t
x
_
, 2.11
where t stands for the costate variable. The optimal costate variable is dened as
t.
In the case of constrained systems, the results given in 5 do not apply as the control
function ut is constrained and does not have arbitrary variations. Indeed, if the control ut
lies on the boundary in 2.10, then the variations are not arbitrary. Instead, we need to use
Pontryagins minimum principle 4. According to the proof of the theorem in 21, we use
for demonstration arbitrary variations in the control signal ut u
t ut. We dene
both the increment J and the rst variation J of the performance index J as
Ju
, u Ju Ju
0 for minimum
Ju
, u O
2
,
2.12
International Journal of Dierential Equations 5
where the rst variation is dened as
J
J
u
ut.
2.13
With the constraints 2.10 and making the assumption that all the admissible variations
on the control ut are small enough to ensure that the sign of the increment J can
be determined by the sign of the variation J, the necessary condition on u
to minimize
J becomes
Ju
t, ut 0. 2.14
According to 21, Chapter 2, the rst variation can be dened as
Ju
t, ut
_
b
a
_
_
H
x
t
_
xt
_
H
u
_
ut
_
H
xt
_
t
_
dt
_
S
x
t
_
b
x
b
_
H
S
t
_
b
b.
2.15
In the previous equation,
1 if the optimal state x
, u
_ _
H
u
_
udt. 2.16
The integrand in the previous relation is the rstorder approximation to change in the
Hamiltonian H due to a change in u alone. This means that by denition
_
H
u
x
, u
, t
_
u Hx
, u
u,
, t Hx
, u
, t 2.17
combining the previous two equations leads us to
Ju
, u
_
b
a
Hx
, u
u,
, t Hx
, u
, tdt. 2.18
6 International Journal of Dierential Equations
Now, using the above, the necessary condition becomes
_
b
a
Hx
, u
u,
, t Hx
, u
, tdt 0, 2.19
for all admissible u less than a small value. The relation becomes
Hx
, u
u,
, t Hx
, u
, t. 2.20
Replacing u
, u
, t Hx
, u,
, t. 2.21
When applied to our problem, we obtain
1
_
Hx
, u
, t
a
D
t
x
_
1
_
Hx
, u, t
a
D
t
x
_
, 2.22
which can be simplied to
_
Hx
, u
, t
a
D
t
x
_
Hx
, u, t
a
D
t
x
_
,
Hx
, u
, t Hx
, u, t.
2.23
The state and costate equations can be retrieved from 5 and give
a
D
t
x Hx, u, t,
t
D
b
Hx, u, t
x
,
2.24
with
xa x
a
, xb x
b
, 2.25
where we again note that b is free. We can notice that u
t
x Ax Bu. 2.26
The state and costate equations become
a
D
t
x Ax Bu,
t
D
b
A,
2.27
International Journal of Dierential Equations 7
with
xa x
a
, xb x
b
. 2.28
Using Pontryagins minimum principle, we have
u
uB
. 2.29
Dening q
gives us
u
uq
, 2.30
u
min
u
min
uu
max
_
uq
_
.
2.31
We can now derive the optimal control sequence u
t. Given 2.31,
1 if q
_
q
_
_
q
_
_
,
2.32
2 and if q
_
q
_
_
q
_
_
.
2.33
Combining 2.32 and 2.33 gives us the following control law:
u
u
max
if q
t < 0,
u
min
if q
t > 0,
undetermined if q
t 0.
2.34
In Section 3, we provide several numerical methods to obtain the control u, state x, and
costate for a specic problem from the literature.
3. Solution of the TimeOptimal Control of
a Fractional Double Integrator
In this section, we consider the following FTOCP:
min
u,T
Ju, T
.
T
, 3.1
8 International Journal of Dierential Equations
subject to
x
1
x
2
, x
1
0 0, x
1
T A,
0
D
x
2
u, x
2
0 0, x
2
T 0,
u
min
ut u
max
, t 0, T.
3.2
3.1. Solution Using Rational Approximation of the Fractional Operator
It is possible for timeoptimal control problems to be reformulated into traditional optimal
control problems by augmenting the system dynamics with additional states one additional
state for autonomous problems. For that purpose, the rst step is to specify a nominal time
interval, a b, for the problem and to dene a scaling factor, adjustable by the optimization
procedure, to scale the system dynamics, and hence, in eect, scale the duration of the time
interval accordingly. This scale factor and the scaled time are represented by the extra states.
The problem dened by 3.13.2 can accordingly be reformulated as follows: nd
the control ut satisfying u
min
ut u
max
over the time interval 01, which minimizes
the quadratic performance index
Ju T, 3.3
subject to the following dynamics:
x
1
Tx
2
,
0
D
t
x
2
Tu,
T 0,
3.4
where the initial conditions are
x
1
0 0,
x
2
0 0,
T0 n,
3.5
where T0 is the initial value chosen by the user. Final state constraints are
x
1
1 A,
x
2
1 0.
3.6
According to 22, we can approximate the operator
a
D
t
using a state space denition
a
D
t
x u
_
z Az bu
x cz
_
. 3.7
International Journal of Dierential Equations 9
Such approximation is called rational approximation of the fractional operator. To ensure the
applicability of our method, we need to dene a new state vector yt such that
yt
x
1
t
zt
T
, 3.8
where zt is the state vector of the rational approximation for the fractionalorder system
described by
0
D
t
x
2
u.
Using the methodology proposed in 6, we reformulate the problem dened by 3.1
3.2. Find the control ut satisfying u
min
ut u
max
, which minimizes the quadratic
performance index
Ju T, 3.9
subjected to the following dynamics:
y
c
_
y
2
t y
N1
t
_
T
A
_
y
2
t y
N1
t
_
T
but
0
, 3.10
the initial condition
y0
_
0 0 0 T
_
T
,
3.11
and the nal state constraints given by
y
1
T 300,
c
_
y
2
T y
N1
T
_
T
0.
3.12
Such a denition allows the problem to be solved by any traditional optimal control
problem solver.
10 International Journal of Dierential Equations
3.2. Solution Using BangBang Control Theory
The solution to the problemdened by 3.13.2 can also be solved using bangbang control
theory. In the integer case 1, the solution in the time domain is well documented and is
given by
T
1
_
2Au
min
u
max
u
max
u
min
,
T T
1
_
2Au
max
u
min
u
max
u
min
,
3.13
where T
1
is the switching time.
Such a solution is obtained by deriving the systems dynamic equations in the time
domain: for t 0, T
1
acceleration
ut u
max
,
x
2
t u
max
t,
x
1
t u
max
t
2
2
,
3.14
and for t T
1
, T deceleration
ut u
min
,
x
2
t u
min
t T
1
x
2
T
1
,
x
1
t u
min
t T
1
2
2
x
2
T
1
t T
1
x
1
T
1
.
3.15
Finding the switching time T
1
and the nal time T can be done by solving the system of nal
time conditions:
x
1
T
2
A,
x
2
T
2
0.
3.16
We can apply a similar technique to solve the fractional problem. For t 0, T
1
acceleration
ut u
max
,
x
2
t u
max
t
1
,
x
1
t u
max
t
1
2
,
3.17
International Journal of Dierential Equations 11
and for t T
1
, T
2
deceleration
ut u
min
,
x
2
t u
min
t T
1
1
x
2
T
1
t T
1
,
x
1
t u
min
t T
1
1
2
x
2
T
1
t T
1
1
x
1
T
1
.
3.18
Finding the switching time T
1
and the nal time T can be achieved by solving the system of
nal time conditions:
x
1
T
2
A,
x
2
T
2
0.
3.19
When expanded, it becomes
T
2
T
1
1
2
u
min
T
1
1
T
2
T
1
1
u
max
T
1
1
2
u
max
A,
T
2
T
1
1
u
min
T
1
1
T
2
T
1
u
max
0.
3.20
Renaming T
21
T T
1
, we get
T
1
21
2
u
min
T
1
T
21
1
2
u
max
T
1
1
2
u
max
A,
T
21
1
u
min
T
1
T
1
21
1
u
max
0.
3.21
The solution to the system of equations is as follows: solve for T
21
u
min
2
T
1
21
u
min
2
_
u
min
u
max
_
1/
T
1/
21
A,
T
1
_
u
min
u
max
T
21
_
1/
,
T T
21
T
1
.
3.22
12 International Journal of Dierential Equations
0 5 10 15 20 25
Time s
0
50
100
150
200
250
300
P
o
s
i
t
i
o
n
Figure 1: States xt as functions of time t for the BangBang control problem for 1.
4. Results
In this section, we nd the solution to the problem dened in Section 3 for the following
parameter values:
i u
min
2,
ii u
max
1,
iii A 300.
The analytical solution for this system for 1 the traditional double integrator is
given in 23 by T
30 as
ut
t
2
2
for 0 t < 20,
t
2
60t 600 for 20 t 30,
xt
t
ux, t
2
x
ux, t gux, t 1.1
has been the subject of much study, beginning with the celebrated paper 1. The authors
of 1 proposed this equation, with g being positive and concave on 0, 1 such that g0
g1 0, as a model for a population that undergoes logistic growth and Brownian diusion.
If ux, 0 Hx, the Heaviside function, and gu u u
2
, the equation in fact has
an exact probabilistic interpretation, given in 2. Consider a population of particles that
undergo independent Brownian motion and branching processes, with each child particle
again following the same behavior. Then ux, t is the probability that there is a particle to
the left of position x at time t, assuming that there was exactly one particle at position x 0
at time t 0. Equation 1.1 also can be derived heuristically for the mean behavior of an
2 International Journal of Dierential Equations
interacting particle process in which two types of particles call them A and Bparticles
simultaneously undergo Brownian diusion and conversion reactions
A B 2A, A B 2B 1.2
with suitable reaction rates. Then the volume density fraction ux, t of Aparticles in the
hydrodynamic limit of large particle numbers per unit volume formally satises 1.1 with
gu cu1 u where c depends on the reaction rates in 1.2; see 3 for a discussion of the
underlying limit procedure and an exact connection to a stochastic version of 1.1. If instead
the conversion reactions are
2A B 3A, 2A B A 2B, 1.3
then the equation for u becomes 1.1 with gu cu
2
1u. Other polynomial reaction terms
g occur in similar ways.
For 1.1 with gu u1 u, it is known that solutions approach a wave prole in
the sense that
ux mt, t x t , 1.4
where mt is the median, umt, t 1/2. It turns out that mt c
such that for fairly general initial data u, 0 that are nonnegative and supported on 0, ,
limsup
t
sup
x<ct
ux, t 0
1.5
whenever c > c
and
liminf
t
inf
x<ct
ux, t 1
1.6
whenever c < c
0 for some u
0, 1, and g < 0 on 0, u
,
and g > 0 on u
t
ux, t Aux, t gux, t. 1.7
Here A A
that is homogeneous of
degree 0, 2, such that p p and p1 1. Following the presentation in 8, we
write p in the RieszFeller form:
p e
i/2 sign
, 1.8
where the skewness parameter must also satisfy min, 2 . The operator A is the
innitesimal generator of a stable L evy process, that is, a continuous time stochastic process
that has c adl ag paths and independent stationary increments with stable distributions; see
9. Paths of such a process must have jumps, and the variance of the displacement must be
innite.
For 0 we obtain fractional powers of the usual negative onedimensional
Laplacian, abbreviated often by
/2
. There are various real variable representations of
such operators, for example, as singular integral operators or as limits of suitable dierence
operators; see 8. In the special case where 1 and 1 1, there is the representation
A
1,
ux
d
dx
_
cos
2
Jux sin
2
ux
_
,
1.9
where Ju is the Hilbert transform of u. In particular, for 1, this is an ordinary rstorder
derivative, not a fractional derivative.
The function g is always assumed to satisfy g0 g1 0. We are interested in both
the KPPcase, that is, g 0 for 0 < < 1, and the AllenCahn case, that is, g < 0 for
near 0 and g > 0 for near 1.
The equation occurs in the heuristic hydrodynamic limit of interacting particle
populations in which conversion reactions such as 1.2 or 1.3 occur together with motion
by a stable L evy process. It has been proposed in, for example, 1015. It should be noted
that the term anomalous diusion is also used for situations in which the rstorder time
derivative is replaced by a fractional order derivative; see 8 where the present case of a
rstorder time derivative is called spacefractional diusion. Another generalization of 1.1
consists in allowing time delays; see 16. These further generalizations will not be discussed
here.
There is strong evidence that 1.7 does not admit traveling wave solutions if 0 < < 2
and g is positive and concave on 0, 1. Rather, numerical results in 11, 17 suggest that for
initial data that are supported on the positive half axis and increase there from 0 to 1, the
median satises mt e
ct
for some c > 0. In 18, the estimates
limsup
t
sup
x<e
ct
ux, t 0, liminf
t
inf
x>e
dt
ux, t 1
1.10
are shown to hold for such initial data whenever c > c
g
/
0/. Thus the
support of a solution grows asymptotically like an exponential. For the case where g < 0 on
4 International Journal of Dierential Equations
some interval 0, u
, the results in 14, 19 suggest on the other hand that there exist wave
prole solutions that move with constant speed, although no rigorous proofs are given there.
The main results of this note are concerned with the existence and nonexistence of a
nite speed of spread and take the form 1.5 and 1.6. It is shown that for a large class of
righthand sides g that are nonnegative on 0, 1, the speed of spread is innite; that is, the
estimate 1.6 holds for all positive speeds c. It is not necessary to assume that g
/
0 > 0, and
if < 1, one does not even have to assume that g is strictly positive on 0, 1. On the other
hand, if g
/
0 < 0 and therefore g is negative near 0, then it will be shown that there exists a
nite speed of spread; that is, 1.5 holds for some nite positive c. These results are stated
and proven in Section 3. Some basic existence and comparison results for 1.7 are sketched
in Section 4.
To prove these results, comparison arguments are employed which follow from
integral representations of solutions of 1.7 and which are therefore extensions of similar
arguments for the study of 1.1. The challenge then is to come up with suitable comparison
solutions. Since 1.7 is nonlocal in nature, techniques from ordinary dierential equations
cannot be employed to construct such solutions. Instead, in this paper a set of travelling wave
solutions is used that comes directly fromthe fundamental solution of the linear problemsee
2.1 below. These solutions are discussed in Section 2 and may be of independent interest.
2. A Class of Travelling Wave Solutions
In this section, it will be shown that fundamental solutions of fractional diusion equations
without reaction terms lead to traveling wave solutions ux, t Ux ct of 1.7, for
suitable functions g. A twoparameter family will be constructed for each possible choice
of and , one parameter being the speed c. The main contribution of this section is the
characterization of the nonlinear function g that is required to make the equation hold.
Throughout this section, let 0, 2 and < min, 2 .
Consider the free equation
t
ux, t A
ux, t 0, 2.1
where A
x, t t
1/
f
_
xt
1/
_
2.2
with initial data W
x, 0
0
x, the delta distribution; see 8. In particular, W
x, 1
f
x. Here f
x, t F
_
xt
1/
_
, 2.4
where F
x
_
x
is positive,
innitely dierentiable, and unimodal. Also, as x , there are expansions
1 F
j1
c
j
x
j
,
f
j1
c
j
x
1j
,
2.5
and as x
F
j1
d
j
x
j
,
f
j1
d
j
x
1j
.
2.6
These are convergent expansions if 0 < < 1 and asymptotic expansions if 1 < < 2; see 8.
For the remainder of this section, we suppress the subscripts and in most formulae
that involve A
, f
, and F
F
_
1/
_
. 2.7
Set u
c
x, t U
x ct, then
Au
c
x, t
F
_
x ct
1/
_
_
x ct
1/1
_
f
_
x ct
1/
_
t
u
c
x, t c
1/
f
_
x ct
1/
_
,
2.8
and therefore
t
u
c
x, t Au
c
x, t
_
1
_
x ct
1/1
_
c
1/
_
f
_
x ct
1/
_
. 2.9
Now x ct
1/
F
1
u
c
x, t and consequently
t
u
c
x, t Au
c
x, t c
1/
g
0
u
c
x, t
1
g
1
u
c
x, t 2.10
6 International Journal of Dierential Equations
with
g
0
f
_
F
1
_
, g
1
F
1
f
_
F
1
_
. 2.11
Equation 2.10 is of the form 1.7, with g c
1/
g
0
1/g
1
.
In the case 1 and 1 < < 1, everything is explicit. Let cos/2 and
sin/2. Then by results in 21,
Fx
1
2
1
arctan
x
,
F
1
cot,
fx
1
_
1 x
2
/
2
_,
g
0
1
sin
2
,
g
1
sin
2
1
cos sin.
2.12
It remains to characterize the functions g
0
, g
1
in the general case.
Proposition 2.1. Let 0 < < 2, < min, 2 . The functions g
0
, g
1
have the following
properties.
a g
0
and g
1
are innitely dierentiable on 0, 1.
b The function g
0
is positive on 0, 1. The function g
1
is negative on 0, F
0 and positive
on F
0, 1. The function
c
1/
g
0
1
g
1
2.13
is negative on 0, u
and positive on u
, 1, where u
c
1/1
.
c As 0, g
0
O
11/
and g
0
1 O
11/
.
d As 0, g
1
O
11/
. As 1, g
1
1 O1
11/
.
e The functions g
0
and g
1
can be represented as
g
0
d
d
_
F
1
f
2
sds,
g
1
d
d
_
F
1
sf
2
sds.
2.14
Proof. Property a follows since f and F together with its inverse are innitely dierentiable.
Property b is obvious. Properties c and d follow from the asymptotic expansions 2.5
and 2.6. Finally e can be checked by dierentiation.
International Journal of Dierential Equations 7
Property e will not be used in what follows. It should be noted that g
0
and g
1
are of
class C
1
on 0, 1 but are not innitely dierentiable at the interval endpoints, except if 1.
Clearly g
0
and g
1
do not depend on c or . We are therefore free to form fairly arbitrary linear
combinations of g
0
and g
1
by choosing c and .
The construction provides travelling wave solutions for 1.7 for a special class of
functions for which g C
1
0, 1, g0 gu
g1 for some u
0, 1, and g
/
0 <
0, g
/
u
> 0, g
/
1 < 0. This suggests that 1.7 possesses travelling wave solutions for more
general functions g with these properties.
If the same construction is attempted for the case 2, it turns out that g
0
and g
1
are
merely continuous on 0, 1, with derivatives that have logarithmic singularities near 0
and 1. Therefore the arguments in the next section cannot be extended to the case 2,
and indeed the results of the next section do not hold in that case.
3. Results on the Speed of Spread
This section contains the main results of this paper. As before, the operator A has symbol
1.8 with 0 < < 2 and < min, 2 . We always assume that u is a solution of 1.7
and that g C
1
0, 1, R with g0 g1 0. Initial data u
0
will be assumed to satisfy
u
0
CR, R, 0 u
0
x 1, lim
x
u
0
x 1, suppu
0
0, .
3.1
The results in Section 4 then imply that 1.7 has a unique mild solution u that exists for all
x R, t > 0, and this solution satises 0 ux, t 1 for all x, t. The notation of that section
will also be used here.
We rst discuss the case where g 0 on 0, 1. The main result in this case is the
following.
Theorem 3.1. Let u be the solution of 1.7 with u
0
satisfying 3.1.
a Let > 1. Assume that g > 0 on 0, 1 and that there are 0 < < / 1 , c
0
> 0 such
that for all 0, 1/2
g c
0
. 3.2
Then for all c > 0
liminf
t
inf
xct
ux, t 1.
3.3
b Let 1. Assume that g > 0 on 0, 1. Then for all c > 0
liminf
t
inf
xct
ux, t 1.
3.4
c Let < 1. Assume that g 0 on 0, 1 and g > 0 for /2, 1. Then for all
c > 0
liminf
t
inf
xct
ux, t 1.
3.5
8 International Journal of Dierential Equations
The result shows that the speed of spread is unbounded i.e., 1.6 holds for all c > 0,
and it exhibits dierent mechanisms for this phenomenon. Recall that in the interpretation
of 1, the function g is responsible for the growth of a substance whose density is given by
u, while A describes the spread of this substance. If 0, 2, the substance spreads with a
jump process, not with Brownian diusion, and jumps of magnitude exceeding Moccur with
a probability that is OM
and f f
, fx c
2
x
1
. 3.11
This is possible by 2.6. Let c > 0 be given, then we may increase M further such that also
c
0
c
1
M
r
c
2
1
c
/1
, 3.12
where r /1 > 0. Now set FM. Then for 0 < Fx , that is, x < M,
and for all M/c
/1
c
1/
g
0
Fx
1
g
1
Fx
_
c
1/
1
x
_
fx
c
/1
x
1/1
fx
c
/1
x
1/1
c
2
x
1
c
/1
c
2
x
2
/1
,
3.13
where a standard calculus argument has been used to see that the expression c
1/
1
x is maximal for x/c
/1
. We estimate further, using the choice of M,
c
1/
g
0
Fx
1
g
1
Fx c
0
c
1
M
r
x
2
/1
c
0
c
1
x
r
x
2
/1
c
0
_
c
1
x
c
0
Fx
gFx.
3.14
10 International Journal of Dierential Equations
Therefore, for all M/c
/1
and all < FM,
c
1/
g
0
1
g
1
g. 3.15
Since g > 0 on , 1 by assumption, this inequality can be achieved also on , 1 by increasing
even further. This proves part a.
The proof of part b is straight forward: given c > 0, note that cg
0
g
1
0 on the
interval 0, Fc. Then cg
0
g
1
/ g is true if is suciently large.
To prove part c, let again c > 0 be given. Let u
,
c
1/
g
0
1
g
1
g 3.16
since the lefthand side is nonpositive there by Proposition 2.1. By increasing further, we
can obtain this estimate also for u
t
u A u g u. 3.18
Then g c
0
t
u A u g u. 3.24
By Lemma 3.3, part d, there exist c > 0 and > 0 such that
g c 1
1/
g
0
1
g
1
. 3.25
Since lim
x
ux, 0 1/2 and ux, 0 0 for x < 0, we can nd d > 0 such that Fx
d
1/
ux, 0 for all x R. Using Proposition 4.1, one sees that
F
_
x d c 1t
1/
_
ux, t 3.26
for all x R, t > 0. Therefore for t > 0 and x ct,
ux, t F
_
x d c 1t
1/
_
F
_
ct d c 1t
1/
_
F
_
t d
1/
_
.
3.27
12 International Journal of Dierential Equations
The righthand side tends to 0 as t . In terms of u, this implies
limsup
t
sup
xct
ux, t 0.
3.28
This concludes the proof.
4. Facts about Fractional ReactionDiffusion Equations
In this section we summarize some basic theory about 1.7 that is needed in this note. A
broader and deeper discussion may be found in 10.
We work in the Banach space
C
lim
_
w CR lim
x
wx and lim
x
wx exist
_
4.1
equipped with the supremum norm . Let A A
_
x y, t
_
_
y
_
dy. 4.2
For xed and and C
lim
, we therefore dene
Stx ux, t, 4.3
where u is given by 4.2. Then St
t0
is a positive C
0
semigroup on C
lim
and a Feller
semigroup on the subspace of functions in C
lim
that vanish at . If is a continuous function
from 0, T to C
lim
, then solutions of the inhomogeneous equation
t
ux, t Aux, t x, t, u, 0 4.4
can be written with the variationofconstants formula:
u, t St
_
t
0
St s, s ds. 4.5
A continuous curve u : 0, T C
lim
that satises 4.5 is commonly called a mild solution
of 4.4. Next let g : 0, R R be locally Lipschitz continuous in both variables and let
C
lim
. Then the equation
t
ux, t Aux, t gt, ux, t for which 1.7 is a special case
has a unique mild solution u C0, T, C
lim
, where 0 < T is maximal. Either T , or
International Journal of Dierential Equations 13
u, t as t T. The solution can be obtained as the locally in time uniform limit of the
iteration scheme:
u
n1
, t St
_
t
0
St sgs, u
n
, s ds n 0, 1, . . . 4.6
with u
0
being arbitrary, for example, u
0
, t St. It is possible to set up a more general
solution theory, but this is not needed for the purposes of this paper.
Solutions of 1.7 satisfy comparison theorems. Results of this type are true for all
Feller semigroup. A systematic study of such semigroups and their generators was carried
out in 22, following the seminal work on this topic in 23. For the sake of completeness, a
comparison result is stated here, and its proof is sketched.
Proposition 4.1. Let u, v C0, T, C
lim
be mild solutions of the equations
t
u Au gu,
t
v Av hv, 4.7
where g, h : R R are locally Lipschitz continuous. If
g h, R,
u, 0 v, 0,
4.8
then
ux, t vx, t x, t R 0, T. 4.9
Proof. Let M max
0,T
u, t v, t 1. Let > g
/
h
/
for all M. Without
loss of generality we may assume that g and h are constant outside M, M. Set
Ux, t e
t
ux, t, Vx, t e
t
vx, t, 4.10
and observe that U and V satisfy
t
U AU gt, U,
t
V AV
ht, V
4.11
with gt, e
t
ge
t
and
ht, dened similarly. Clearly, gt,
ht, for all .
The function g is nondecreasing in its second argument, since for almost all
gt, g
/
_
e
_
0. 4.12
Consider the iteration scheme:
U
n1
, t Stu, 0
_
t
0
St s gs, U
n
, sds 4.13
14 International Journal of Dierential Equations
and similarly for V
n
and
h. The scheme for the U
n
converges to the limit U, and the scheme
for the V
n
converges to the limit V.
We now employ a standard induction argument to show that U
n
V
n
on R0, T for
all n. Let U
0
, t Stu, 0 and V
0
, t Stv, 0, then U
0
V
0
on R 0, T since S is a
positive semigroup and u, 0 v, 0. Suppose U
n
V
n
on R 0, T, then
U
n1
, t Stu, 0
_
t
0
St s gs, U
n
, sds
Stv, 0
_
t
0
St s gs, V
n
, sds
Stv, 0
_
t
0
St s
hs, V
n
, sds
V
n1
, t,
4.14
which completes the induction step. Taking the limit, this implies that U V and therefore
also u v on R 0, T. This proves the proposition.
Corollary 4.2. Consider a mild solution u C0, T, C
lim
of 1.7 and assume that g is locally
Lipschitz continuous. If g 0 for some and u, 0 , then u, t for all t. If g
0 g for some < and u, 0 , then u, t for all t, and the solution can be
continued to R 0, .
The proof consists in observing that the constant functions vx, t and wx, t
solve 1.7 with righthand sides 0 and therefore must be pointwise bounds for the
solution, by Proposition 4.1. If the solution remains bounded between two constants, then
its supremum norm remains bounded and it can be continued to R 0, .
Also required is a comparison result for solutions whose initial data are step functions.
Since such initial data are not in C
lim
, a separate argument is required.
Proposition 4.3. Let u C0, T, C
lim
be a mild solution of 1.7, with locally Lipschitz
continuous g. Assume that
ux, 0 v
0
x a
0
N
j1
a
j
H
_
x c
j
_
x R, 4.15
where a
j
R, c
1
< c
2
< c
N
, and H is the Heaviside function. Let min
R
v
0
x and
max
R
v
0
x. Assume also that g 0 on , . Then
ux, t a
0
N
j1
a
j
V
_
x c
j
, t
_
x R, 0 < t T, 4.16
where V
is dened in 2.4.
International Journal of Dierential Equations 15
Proof. We knowthat ux, t and thus may assume that g 0 also for < . For arbitrary
, > 0, we set
v
x
1
0
v
0
x zdz . 4.17
Then v
C
lim
. Solving 2.1 with initial data v
,
gives the solution
G
x, t
1
0
N
j1
a
j
V
_
x z c
j
, t
_
a
0
. 4.18
Given > 0, it is possible to nd > 0 such that v
< u
0
x on R, since u
0
is uniformly
continuous. Clearly, G
s, a and
/2
with R. They
are well known for their role in investigating the solvability of nonlinear partial dierential
equations, and the Kortewegde Vries equation KdV henceforth in particular see, e.g., 7
11 and the references therein. In the current work, Riesz potentials of KdV solitons are
computed and their relation to ordinary dierential equations is established.
2 International Journal of Dierential Equations
We continue the study of Riesz fractional derivatives of solutions to Kortewegde
Vriestype equations started in 12. After appropriate rescaling, KdV can be written in the
form
u
t
u
xxx
3
_
u
2
_
x
0, x R, t > 0. 1.1
It is well known that the fundamental solution of the Cauchy problem for the linearized
KdV is expressed in terms of the Airy function of the rst kind Aix and its Hilbert
transformconjugate in terms of the Scorer function Gix. The papers 1214 were devoted
to the study of fractional properties of the Airy functions and their conjugates and to the
establishing of related properties for KdVtype equations.
Although there exists extensive literature on solitons, as far as we know, a study of
their fractional properties is still missing. A preliminary investigation of Riesz potentials for
a KdV soliton was carried out in 15. In this paper the emphasis was put on the issue of
whether solitons inherit fractional properties of fundamental solutions. Riesz potentials of a
soliton, u
X D
u
0
X, where u
0
X 2sech
2
X, X x 4t, and D
2
x
/2
, and their
Hilbert transforms, v
X Hu
X and v
s, a
s, as, 1a and
X u
X iv
d
dX
_
P
X
dw
dX
_
Q
Xw
Xw, X R, 1.2
lim
X
wX 0,
1.3
for 1. Here P
X > 0,
X > 0, and Q
X and v
X and
X and v
, v
and v
are studied in
Section 3.4. In Section 4, the inequality 1.4 is discussed.
2. Preliminaries
Introduce the Fourier transform of the function f : R R by
f
_
f
_
_
e
ix
fxdx, 2.1
and the inverse Fourier transform by
fx
1
_
f
_
x
1
2
_
e
ix
fd. 2.2
For real and x R, dene Riesz potentials of a function fx by the formula see 7,
page 117
D
fx
1
2
_
fe
ix
d, 2.3
provided that the integral in the righthand side exists. Dene derivatives of D
x
fx with
respect to by
x
fx
1
2
_
ln
n
fe
ix
d, n N, 2.4
provided that these integrals exist.
Introduce the Hilbert transform of the function f by see 20, page 120
Hfx
1
P.V.
_
f
_
y
_
y x
dy, 2.5
4 International Journal of Dierential Equations
where x R and P.V. denotes the Cauchy principal value of the integral. According to our
choice of the Fourier transform
Hf i sgn
/
z
d
2
dz
2
ln z
_
0
te
zt
1 e
t
dt, Rz > 0. 2.6
Notice that see 21, page 260, 6.4.7
/
z
/
1 z
d
dz
cotz.
2.7
Also, the following asymptotic expansion holds:
/
z
1
z
1
2z
2
1
6z
3
O
_
1
z
5
_
for z ,
_
_
arg z
_
_
< . 2.8
The Hurwitz generalized Zeta function is dened by see 22, page 88
s, a
n0
1
n a
s
for Rs > 1, a C, a
/
0, 1, 2, . . . . 2.9
This implies that
a
s, a ss 1, a. 2.10
This function has the integral representation
s, a
1
s
_
0
x
s1
e
ax
1 e
x
dx for Rs > 1, Ra > 0, 2.11
where x is the Gamma function. In two particular cases, we have see 22 that
_
s,
1
2
_
2
s
1s, 2.12
2, a
/
a, 2.13
where s is the Riemann Zeta function.
The singularity of s, a as s 1 is given by the relation
lim
s 1
_
s, a
1
s 1
_
/
a
a
a. 2.14
International Journal of Dierential Equations 5
The asymptotic expansion of s, a for large a is see 23
s, a
1
2
a
s
a
1s
s 1
1
s
k1
B
2k
2k!
2k s 1
a
2ks1
, a < ,
_
_
arg a
_
_
< , 2.15
where B
n
are the Bernoulli numbers.
Introduce the fullrange forms of Hurwitz Zeta functions see 19
s, a
n
1
_
n a
2
_
s/2 , 2.16
s, a
n0
1
_
n a
2
_
s/2
1
n
1
_
n a
2
_
s/2
.
2.17
Representations 2.16 and 2.17 imply that
s, a s, a s, 1 a,
s, a s, a s, 1 a. 2.18
Hence, follow the symmetric and antisymmetric properties of the functions,
s, a
s, 1 a,
s, a
s, 1 a. 2.19
It follows from 2.16 that
s, 1/2
0. These functions satisfy the functional dierential equations
s, a s
s 1, a,
a
s, a s
s 1, a. 2.20
Consider now that s R and denote by a the complex conjugate of a. Then s, a
s, a. It implies that, for a a
1
ia
2
,
s, a 2Rs, a
2
s
_
0
x
s1
e
a
1
x
cos a
2
x
1 e
x
dx. 2.21
In a similar way,
s, a 2iIs, a
2i
s
_
0
x
s1
e
a
1
x
sina
2
x
1 e
x
dx. 2.22
Denote by Wu, v the Wronskian of the functions ux and vx, that is, Wu, v
_
_
_
u v
u
/
v
/
_
_
_. For readers convenience, we present here 24, Theorem 5.3.
6 International Journal of Dierential Equations
Theorem 2.1. Let pt > 0, qt be real valued and continuous for 0 t T. Let u and v be real
valued solutions of the equation
_
pu
/
_
/
qu 0, 2.23
satisfying the condition
ptWut, vt C
0
> 0. 2.24
Let N be the number of zeros of ut on 0, T. Then
_
_
_
_
_
N C
0
_
T
0
dt
ptv
2
t u
2
t
_
_
_
_
_
. 2.25
In conclusion of this section, we would like to quote an interesting result concerning
integrals over the real axis see 25.
Theorem 2.2. For any integrable function fx and gx x c
2
/x with c const R,
P.V.
_
f
_
gx
_
dx P.V.
_
fxdx. 2.26
Moreover, the above formula holds true if
gx x
j1
b
j
x c
j
,
2.27
where b
j
 is any sequence of positive constants, c
j
are any real constants, and the series is convergent.
3. Fractional Derivatives of A KdV Soliton and Their Conjugates
In this section, we consider Riesz fractional derivatives of a KdV soliton and their Hilbert
transforms and establish their properties. We notice that all the graphs were obtained with
the Mathematica 6 software.
We take the soliton solution of 1.1 in the form u
0
X 2sech
2
X with X x 4t see
Figure 1 and introduce the function
w
X u
X iv
X, 3.1
where
u
X D
u
0
X, v
Hu
0
X. 3.2
International Journal of Dierential Equations 7
4 2 2 4
0.5
1
1.5
2
2.5
Figure 1: Graph of the soliton u
0
X.
Notice that the functions u
X and v
X Hv
X, v
X Hu
X. 3.3
The next statement was proved in 15. Using the functions 2.16 and 2.17, we
rewrite it in a more convenient form.
Theorem 3.1. The functions u
X and v
X
22
_
2 ,
1
2
i
X
_
, 3.4
v
X i
22
_
2 ,
1
2
i
X
_
, 3.5
where
s, a and
s, a are the fullrange Hurwitz Zeta functions (see 2.16 and 2.17) and s
is the Gamma function.
3.1. Properties of the Functions u
and v
and v
and v
1 The functions u
X and v
X v
1
X, v
/
X u
1
X, 3.6
where > 1, X R, and the prime denotes dierentiation with respect to X. This follows
from 3.3 and the relation d/dX H D.
8 International Journal of Dierential Equations
2 The functions u
XdX 0,
_
XdX 0. 3.7
These properties are reected on the graphs see Figures 3 and 4.
5 For all > 0 and n N,
XdX 0,
n
XdX 0. 3.8
These relations follow from the dierentiation of the identities in 3.7.
6 For all > 0 and c R,
P.V.
_
_
X
c
2
X
_
dX 0, P.V.
_
_
X
c
2
X
_
dX 0. 3.9
Moreover,
P.V.
_
j1
b
j
X c
j
dX 0,
P.V.
_
j1
b
j
X c
j
dX 0,
3.10
where b
j
is any sequence of positive constants, c
j
are any real constants, and the series
converges. These relations follow from 3.7 and Theorem 2.2 .
7 The functional sequence of Riesz potentials u
u
0
XdX 4, P.V.
_
v
0
XdX 0 3.11
see Figures 1 and 2. Here the conjugate soliton v
0
X is given by
v
0
X
2i
2
_
/
_
1
2
i
X
_
/
_
1
2
i
X
__
. 3.12
Equation 3.12 can be recovered from 3.5 with 0 thanks to 2.13.
Remark 3.2. The conjugate soliton 3.12 is an algebraic solitary wave for extended KdV:
v
t
v
xxx
3
_
2v Hv H
_
v
2
__
x
0, 3.13
obtained by applying the Hilbert transformto 1.1 and setting v Hu. The termalgebraic
solitary wave is explained by the fact that v
0
X has a 1/X decay for large X. More precisely
see 15,
v
0
X
2
1
X
O
_
1
X
2
_
for X . 3.14
8 For 0, the functions u
and v
1
Xv
2
XdX 0. 3.15
10 International Journal of Dierential Equations
4 2 2 4
10
10
20
Figure 3: Graph of the fractional derivative u
3.8
X.
Equations 3.12 and 3.14 imply that v
0
L
2
R. The fact that u
and v
X
41
2
cos1 arctan2X/
_
1/4 X/
2
_
1/2
O
_
1
X
2
_
,
v
X
42
2
2 sin1 arctan2X/
1
_
1/4 X/
2
_
1/2
O
_
1
X
2
_
.
3.16
Orthogonality of u
and v
0
42
2
_
2
2
1
_
2 > 0, 3.17
v
0 0, 3.18
where s is the Riemann Zeta function see 15.
3.2. SturmLiouville Problem
It is convenient to represent w
X R
X expi
X, 3.19
International Journal of Dierential Equations 11
4 2 2 4
20
10
10
20
Figure 4: Graph of the conjugate fractional derivative v
3.8
X.
where
R
X
_
u
X
2
v
X
2
,
X, u
X > 0,
X , u
X < 0, v
X 0,
X , u
X < 0, v
X < 0,
X arctan
v
X
u
X
.
3.20
Exponential representation 3.19 allows one to deduce the boundary value problem for the
functions w
X
dw
dX
_
Q
Xw
Xw 0, X R, 3.21
where
P
X
C
Wu
X, v
X
, C const,
Q
X
P
XR
/
X
/
R
X
,
X
/
X
2
,
3.22
X
Wu
X, v
X
R
2
X
. 3.23
Here Wu
X, v
X is the Wronskian of u
X and v
X Wu
X, v
X.
12 International Journal of Dierential Equations
4 2 2 4
2
1.5
1
0.5
0.5
1
1.5
2
Figure 5: Graph of the arctangent function
3.8
X.
Taking into account the behavior of w
d
dX
_
P
X
dw
dX
_
Q
Xw
Xw, X R, 3.24
lim
X
wX 0,
3.25
corresponding to 1. Without loss of generality, we can choose the constant in 3.22 to be
positive. In the next subsection, we shall prove that W
X > 0 and
X > 0 in 3.22.
It follows from 3.7 that for > 0 the functions w
XdX 0. 3.26
This reects the oscillatory behavior of w
X is shown in Figure 5.
X conveniently
serves as a zero counter for both functions: u
X and v
X has zeros.
Remark 3.3. Observe that a general solution of 3.24 can be written in the form
w
g
X C
1
u
X C
2
u
X
_
X
0
W
_
y
_
u
2
_
y
_ dy, 3.27
where C
1
, C
2
const.
International Journal of Dierential Equations 13
Remark 3.4. We would like to point out that the dierential equation given by 3.21 can be
factored in the following way see 26, page 269:
d
dX
_
P
X
dw
dX
_
Q
Xw
Xw
W
X
u
X
d
dX
_
u
2
X
W
X
d
dX
_
wX
u
X
_
_
. 3.28
3.3. Wronskian of u
and v
X W
X, W
/
0 0, 3.29
W
X > 0. 3.30
Proof. We start with 3.29. Taking into account relations in 3.3 and the fact that
X
H D,
we can write
W
X u
Xv
/
X v
Xu
/
X u
Xu
1
X v
Xv
1
X. 3.31
Since the functions u
X is even
for X R. Dierentiation of 3.31 with the help of 3.6 and 3.18 yields W
/
0 0. Next,
we turn to the proof of 3.30. Since the functions u
X and v
X
/
0 for all > 1 and x R. It remains to establish
the sign of the Wronskian. In view of 3.17 and 3.18,
W
0 u
0u
1
0
162 3
52
_
2
2
1
__
2
3
1
_
2 3 > 0. 3.32
By Abels formula, for all X R,
W
X W
0 exp
_
_
X
0
P
/
_
P
_ d
_
, 3.33
where the function
P
/
X
P
X
is 1
s, z
s 2, z
s, z
s 2, z
s, z
s 1, z
s, z
s 1, z
,
s 2 , z 1/2 iX/,
3.34
14 International Journal of Dierential Equations
0
20
40
0
1
2
4
2
0
2
4
Figure 6: Threedimensional graph of the Wronskian W
X.
is continuous on R. After some simplication, we have that
W
X W
0 exp
_
s 1
X is given in Figure 6.
Remark 3.6. What does the positivity of the Wronskian yield for the soliton and its conjugate?
For 0, 3.30 simplies to read
W
0
X 2sech
2
X
_
v
/
0
X tanhX v
0
X
_
> 0. 3.36
We notice that v
0
X > 0 for X > 0, v
0
X v
0
X for X R, and v
0
0 0 see Figure 2.
Integrating the inequality v
/
0
X/v
0
X tanhX > 0 over the interval , X for X > 0 and
the inequality v
/
0
X/v
0
X tanhX < 0 over , X for X < 0 yields the estimate
v
0
X > v
0
cosh
coshX
for X > 0.
3.37
3.4. Zeros of the Functions u
X and v
X
This subsection is devoted to the estimates of the number of zeros for the functions in
question. By a strictly monotone change of variable
y
_
X
b
d
P
_, b const R, 3.38
International Journal of Dierential Equations 15
Equation 3.21 can be reduced to the equation
d
2
1
dy
2
Q
_
y
_
1 0, 3.39
where 1y wXy and
Q
y P
XyQ
X on the interval 0, X
0
, where
X
0
< . Then the following inequality holds:
I
1 N
1, 3.40
where
I
arctan
i
2 , 1/2 iX
0
/
2 , 1/2 iX
0
/
. 3.41
Proof. Since we chose C 1 in 3.22,
P
XWu
X, v
X 1. 3.42
Therefore,
X
v
/
u
/
v
2
> 0. 3.43
Therefore, by Theorem 2.1 of Section 2, for the interval 0, X
0
we have the estimate
1 I
1 I
, 3.44
where
I
_
X
0
0
W
_
d
v
2
_
u
2
_
_
X
0
0
_
d arctan
i
2 , 1/2 iX
0
/
2 , 1/2 iX
0
/
. 3.45
Here we have used the fact that
X.
Proof. This follows from Sturms Separation Theorem see 24, page 335.
16 International Journal of Dierential Equations
4. Inequality for Hurwitz Zeta Functions
Here we discuss a new inequality for the Hurwitz Zeta functions which follows from
Lemma 3.5. The next statement is a corollary of this lemma.
Corollary 4.1. For s > 1 and z 1/2 iX/ with X R, the following inequality holds:
Js, z s, zs 1, z s, zs 1, z > 0. 4.1
This inequality can also be written in another form:
Js, z Rs, zRs 1, z Is, zIs 1, z > 0. 4.2
Proof. Dropping positive terms in front of the fullrange Hurwitz Zeta functions in 3.4 and
3.5 and using 3.30 lead to
s, z
s 1, z
s, z
s 1, z
2s, zs 1, z s, zs 1, z
4Rs, zRs 1, z Is, zIs 1, z > 0.
4.3
Remark 4.2. Setting s, z r
s
ze
i
s
z
and s 1, z r
s1
e
i
s1
z
, we can rewrite 4.1 in
the form
J 2r
s
r
s1
_
exp
_
i
_
s1
s
__
exp
_
i
_
s1
s
___
4r
s
r
s1
cos
_
s1
s
_
. 4.4
It shows that, for z 1/2 iX/, X R,
cos
_
s1
z
s
z
_
> 0. 4.5
Introduce the scalar product for the complexvalued functions f f
1
if
2
and g g
1
ig
2
by
the formula
_
f g
_
f g f
1
f
2
f
2
g
2
i
_
f
2
g
1
f
1
g
2
_
. 4.6
Then for s > 1 and z 1/2 iX/,
Js, z 2Rs, z s 1, z) > 0. 4.7
Remark 4.3. The proof of 4.1 becomes quite dicult when one approaches it from the point
of view of special functions. For example, the use of integral representations 2.21 and 2.22
yields
J
1
2ss 1
__
0
t
s1
s
cosX/t
sinht/2 sinh/2
dtd. 4.8
International Journal of Dierential Equations 17
The change of variables t /2, t /2 leads to
J
2
ss 1
_
0
cos
_
2
X
_
d
_
_
_
s1
_
_
s
cosh cosh
d.
4.9
It is not clear at all that the integral 4.9 is positive for all s > 1 and X R. However, 4.1
shows that it is. Multiplication of the series representations 2.16 and 2.17 does not make
the proof any easier.
References
1 R. Hilfer, Ed., Applications of Fractional Calculus in Physics, World Scientic, River Edge, NJ, USA, 2000.
2 A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo, Theory and Applications of Fractional Dierential Equa
tions, vol. 204 of NorthHolland Mathematics Studies, Elsevier Science, Amsterdam, The Netherlands,
2006.
3 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, San Diego, Calif, USA, 1999.
4 I. Podlubny, Geometric and physical interpretation of fractional integration and fractional
dierentiation, Fractional Calculus & Applied Analysis, vol. 5, no. 4, pp. 367386, 2002.
5 Y. Luchko, Maximum principle for the generalized timefractional diusion equation, Journal of
Mathematical Analysis and Applications, vol. 351, no. 1, pp. 218223, 2009.
6 J. Duoandikoetxia, Fourier Analysis, American Mathematical Society, Providence, RI, USA, 2001.
7 E. M. Stein, Singular Integrals and Dierentiability Properties of Functions, Princeton Mathematical Series,
no. 30, Princeton University Press, Princeton, NJ, USA, 1970.
8 T. Kato, On the Cauchy problem for the generalized Kortewegde Vries equation, in Studies in
Applied Mathematics, vol. 8 of Advances in Mathematics. Supplementary Studies, pp. 93128, Academic
Press, New York, NY, USA, 1983.
9 C. E. Kenig, G. Ponce, and L. Vega, On the generalized Kortewegde Vries equation, Duke
Mathematical Journal, vol. 59, no. 3, pp. 585610, 1989.
10 C. E. Kenig, G. Ponce, and L. Vega, Wellposedness and scattering results for the generalized
Kortewegde Vries equation via the contraction principle, Communications on Pure and Applied
Mathematics, vol. 46, no. 4, pp. 527620, 1993.
11 J. C. Saut and R. Temam, Remarks on the Kortewegde Vries equation, Israel Journal of Mathematics,
vol. 24, no. 1, pp. 7887, 1976.
12 V. Varlamov, Semiinteger derivatives of the Airy functions and related properties of the Korteweg
de Vriestype equations, Zeitschrift f ur Angewandte Mathematik und Physik, vol. 59, no. 3, pp. 381399,
2008.
13 V. Varlamov, Fractional derivatives of products of Airy functions, Journal of Mathematical Analysis
and Applications, vol. 337, no. 1, pp. 667685, 2008.
14 V. Varlamov, Dierential and integral relations involving fractional derivatives of Airy functions and
applications, Journal of Mathematical Analysis and Applications, vol. 348, no. 1, pp. 101115, 2008.
15 V. Varlamov, Riesz potentials for KortewegdeVries solitons, Zeitschrift f ur Angewandte Mathematik
und Physik. In press.
16 R. C. Sharpley and V. Vatchev, Analysis of the intrinsic mode functions, Constructive Approximation,
vol. 24, no. 1, pp. 1747, 2006.
17 V. Vatchev and R. Sharpley, Decomposition of functions into pairs of intrinsic mode functions,
Proceedings of the Royal Society A, vol. 464, no. 2097, pp. 22652280, 2008.
18 O. Vall ee and M. Soares, Airy Functions and Applications to Physics, Imperial College Press, London,
UK, 2004.
19 R. C. McPhedran, L. C. Botten, N.A. P. Nicorovici, and I. J. Zucker, Symmetrization of the Hurwitz
zeta function and Dirichlet L functions, Proceedings of the Royal Society A, vol. 463, no. 2077, pp. 281
301, 2007.
20 E. C. Titchmarsh, Introduction to the Theory of Fourier Integrals, Oxford University Press, Glasgow, UK,
2nd edition, 1962.
18 International Journal of Dierential Equations
21 M. Abramowitz and I. Stegun, Eds., Handbook of Mathematical Functions with Formulas, Graphs and
Mathematical Tables, National Bureau of Standards, Washington, DC, USA, 1964.
22 H. M. Srivastava and J. Choi, Series Associated with the Zeta and Related Functions, Kluwer Academic
Publishers, Dordrecht, The Netherlands, 2001.
23 W. Magnus, F. Oberhettinger, and R. P. Soni, Formulas and Theorems for the Special Functions of
Mathematical Physics, Die Grundlehren der Mathematischen Wissenschaften, Band 52, Springer, New
York, NY, USA, 1966.
24 P. Hartman, Ordinary Dierential Equations, John Wiley & Sons, Baltimore, Md, USA, 1973.
25 M. L. Glasser, A remarkable property of denite integrals, Mathematics of Computation, vol. 40, no.
162, pp. 561563, 1983.
26 D. Zwillinger, Handbook of Dierential Equations, Academic Press, New York, NY, USA, 1998.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 197020, 12 pages
doi:10.1155/2010/197020
Research Article
Linear Fractionally Damped Oscillator
Mark Naber
Department of Mathematics, Monroe County Community College, Monroe, MI 481619746, USA
Correspondence should be addressed to Mark Naber, mnaber@monroeccc.edu
Received 8 July 2009; Accepted 11 August 2009
Academic Editor: Mark M. Meerschaert
Copyright q 2010 Mark Naber. This is an open access article distributed under the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The linearly damped oscillator equation is considered with the damping term generalized to a
Caputo fractional derivative. The order of the derivative being considered is 0 v 1. At the
lower end v 0 the equation represents an undamped oscillator and at the upper end v 1 the
ordinary linearly damped oscillator equation is recovered. A solution is found analytically, and a
comparison with the ordinary linearly damped oscillator is made. It is found that there are nine
distinct cases as opposed to the usual three for the ordinary equation damped, overdamped, and
critically damped. For three of these cases it is shown that the frequency of oscillation actually
increases with increasing damping order before eventually falling to the limiting value given by
the ordinary damped oscillator equation. For the other six cases the behavior is as expected, the
frequency of oscillation decreases with increasing order of the derivative damping term.
1. Introduction
In this paper the linearly damped oscillator equation is considered with the damping term
replaced by a fractional derivative 1 whose order, , will be restricted to, 0 1,
D
2
t
x
0
D
t
x
2
x 0. 1.1
Burov and Barkai 2, 3 examined such an equation in connection with critical behavior. They
were able to determine a solution in terms of generalized MittagLeer functions. Nonlinear
fractional oscillators have been studied numerically by Zaslavsky et al. 4. He was primarily
interested in chaotic behavior. It is hoped that a careful study of the analytic solution to
the linear fractionally damped equation will help shed light on properties of the nonlinear
equation and be of use for direct applications of fractionally damped oscillations see, e.g.,
5, 6.
2 International Journal of Dierential Equations
In this paper the Caputo formulation of the fractional derivative will be used. The
Caputo derivative is preferred over the RiemannLiouville derivative for physical reasons.
Consider the Laplace transform of the two formulations of the fractional derivative for 0 <
< 1
L
_
RL
0
D
t
ft
_
s
Fs
RL
0
D
1
t
ft
_
_
0
,
L
_
C
0
D
t
ft
_
s
Fs ft
_
_
0
.
1.2
The constant term arising from the Laplace transform of the Caputo derivative is merely
the initial value of the function. For the RiemannLiouville derivative this is not the case.
The constant term arising from the Laplace transform currently has no simple physical
interpretation. Hence the Caputo fractional derivative seems to be more useful for modeling
physical systems.
If the order of the fractional damping term is allowed to become 3/2 outside the
range of values considered in this paper the equation is usually referred to as the Bagley
Torvik equation see, e.g., 1, 7, 8. The solution of this equation exhibits damped oscillatory
behavior similar to what we expect to nd for the equation studied in this paper. The Bagley
Torvik equation was originally derived to study the motion of a rigid plate in a Newtonian
uid 7.
The analytic solution to the fractionally damped equation is found by means of
Laplace transform. For the sake of clarity, and for pointing out some unique diculties with
the factional equation, a comparison with the Laplace transform method as applied to the
nonfractional case is made. It is found that there are nine distinct cases for the fractionally
damped equation as opposed to the usual three cases for the Nonfractional equation. In six
of the nine cases the results are as expected; increasing the order of the factional derivative
increases the eects of the damping i.e., the frequency of the damping slows as the order
of the derivative increases. However, in three cases, the frequency of the damping actually
increases as the order of the fractional derivative increases until a peak value is reached after
which the frequency falls to its Nonfractional limit. The physical reason for this increase in
the oscillation frequency is not yet clear.
2. The Nonfractional Case
Before a solution to the linear fractionally damped oscillator equation is constructed it will be
useful to review the Laplace transform method of solution for the linearly damped oscillator
equation
D
2
t
x D
t
x
2
x 0. 2.1
The constants and are taken to be real and positive. is the damping force per unit
mass.
2
is the restoring force per unit mass. In both cases fractional and Nonfractional the
following initial conditions will be used:
x0 x
0
, D
t
x0 x
1
. 2.2
International Journal of Dierential Equations 3
Transforming 2.1 together with the initial conditions 2.2 gives the following:
s
2
Xs sx
0
x
1
sXs x
0
2
Xs 0, 2.3
or
Xs
sx
0
s
2
s
2
x
1
x
0
s
2
s
2
. 2.4
Equation 2.4 can be inverted using tables, however, to shed light on a problem that will
happen later, 2.4 will be inverted via the complex inversion integral. The exponents on the
s variable in both terms are whole numbers, hence there will not be a branch cut in the contour
integral and the Bromwich contour can be used
xt Residue
1
2i
_
e
st
Xsds. 2.5
Recall that the Bromwich contour begins at i goes vertically up to i where is
chosen so that all poles will lie to the left of the vertical contour line and thus all poles will be
captured within the contour and then travels in a half circle to the left, counter clockwise
back to i. For this problem there is no contribution from the contour integral. The only
contribution comes from the residue. The residue is generated from the roots of the following
quadratic equation
s
2
s
2
0. 2.6
There are three dierent cases.
1 > 2; 2 unequal real roots that are negative
s
1,2
2
4
2
2
. 2.7
2 2; 2 repeated real roots that are negative
s
3
2
. 2.8
3 < 2; 2 complex roots whose real parts are negative
s
4,5
i
4
2
2
2
. 2.9
See Figure 1 for a graphical representation of the location of the roots.
Note that in cases one and three the poles will be of order one and in case two the
pole will be of order two. Note also that if there were no damping the poles would be on the
imaginary axis at i.
4 International Journal of Dierential Equations
s
2
s
3
s
1
s
4
s
5
i
i
Im
Re
Figure 1
Computing the residue for case one gives
Residue lim
s s
1,2
s s
1,2
e
st
_
sx
0
s
2
s
2
x
1
x
0
s
2
s
2
_
, 2.10
or
xt
e
s
1
t
2s
1
s
1
x
0
x
1
x
0
e
s
2
t
2s
2
s
2
x
0
x
1
x
0
. 2.11
As s
1
and s
2
are both negative this solution will decay exponentially. This is usually referred
to as the overdamped case.
Case three is computed the same way as case one. Now the poles are complex so the
exponential function can be expressed using sine and cosine with an overall exponential
damping factor
xt e
t
_
x
0
cos
_
t
_
2x
1
x
0
2
sin
_
t
_
_
, 2.12
where
_
2
2
/4 and /2 Notice that the presence of damping causes the eective
angular frequency, , to be smaller than the undamped angular frequency; that is, the
oscillations go slower, as one might expect if there were damping to impede the motion. This
is usually referred to as the underdamped case. By comparison, cases one and two could be
viewed as having a zero frequency or an innite period.
For case two the pole is of order 2, and the residue is given by
lim
s
d
ds
_
s
2
e
st
_
sx
0
x
1
x
0
s
2
s
2
__
. 2.13
Recall that 2, this allows the denominator to be factored. The limit then becomes
lim
s
d
ds
_
e
st
sx
0
x
1
2x
0
_
, 2.14
or
xt e
t
tx
0
x
1
x
0
. 2.15
International Journal of Dierential Equations 5
This is usually called the critically damped case. Graphs of sample solutions to these three
cases can be found in any introductory book on dierential equations.
3. The Fractional Case
Now consider 1.1. In this case, has units of time raised to the power 2. Hence the
over all units of the second term remain the same as in 2.1. There are two cases to consider,
0 < < 1 and 1 < < 2. We will consider the former in this paper. The Laplace transform of
1.1 is
s
2
X sx
0
x
1
_
s
X s
1
x
0
_
2
X 0, 3.1
or
X
sx
0
x
1
s
1
x
0
s
2
s
2
. 3.2
If 3.2 is inverted using a contour integral a branch cut is needed on the negative real axis
due to the fractional exponents on the complex variable s. Hence, a Hankel contour will be
used. This contour starts at i and goes vertically up to i where is again chosen
so that all poles will lie to the left of the vertical contour line then travels in a quarter circle
arc to the left to just above the negative real axis i.e., . The contour then has a cut that
goes into the origin following the negative real axis, around the origin in a clockwise sense
to just below the negative real axis and then back out to . The contour then has another
quarter circle arc to i.
Now the question is, where are the poles? This is a somewhat more involved question
than in the standard linearly damped model. To nd the poles the following equation needs
to be solved:
s
2
s
2
0, 3.3
Which, for an arbitrary , is not a trivial problem. To determine if there are solutions, and if
so how many, let s re
i
then 3.3 breaks into 2 equations, a real and an imaginary part
r
2
cos2 r
cos
2
0,
r
2
sin2 r
sin 0.
3.4
Could there be a solution on the positive real axis? No, in this case 0 and the rst equation
of 3.4 would be the sum of three positive nonzero terms, which would never be zero. Could
there be a solution on the negative real axis? No, in this case and the second term
of the second equation of 3.4 would never be zero. Using similar arguments we can show
that there are no solutions on the positive or negative imaginary axes, recall 0 < < 1. It
can also be shown that no solutions are in the right half plane both terms of the second
equation would always be positive. If there are solutions, they should be in pairs, complex
conjugates, with /2 < < and /2 > > . To attempt to nd a solution rst solve the
6 International Journal of Dierential Equations
second equation of 3.4 for r and substitute this into the rst equation only look for positive
values rst
_
sin
sin2
_
2/2
cos2
_
sin
sin2
_
/2
cos
2
0. 3.5
The reader may be worried about the negative sign and the fractional exponent in 3.5,
however, for the restricted angular range being considered, /2 < < , sin2 is always
negative. So, the argument of the root will always be positive.
Given values for , , and it would appear to be impossible to solve 3.5 for .
Equation 3.5 can be simplied to a more aesthetically pleasing form
_
sin
sin2
2
_
1/2
sin2
_
1/2
_
2
. 3.6
Now it needs to be seen if there is a value that will satisfy 3.6. For this equation
to be true sin2 needs to be positive. This will only happen on the restricted domain
/2 < < /2 . Now the question becomes, on this restricted domain can we pick a
value that will make the lefthand side of 3.6 as large or as small as we wish? Thus ensuring
that no matter what the values of , , and we are given we can always nd a value that
will satisfy 3.6. Consider the two limits
lim
/2
_
sin
sin2
2
_
1/2
sin2 ,
lim
/2
_
sin
sin2
2
_
1/2
sin2 0.
3.7
Since the lefthand side of 3.6 is continuous in and we have the two limits above, 3.7,
it is guaranteed that there will be at least one solution to 3.6 and hence there will be at
least two poles for the residue calculation. If we can show that the left hand side of 3.6
decreases monotonically in over the restricted domain then we know that there will be only
one solution to 3.6, and thus only two poles in the residue calculation. To show that the left
hand side of 3.6 decreases monotonically in we need to show that the derivative of the
left hand side of 3.6 with respect to is always negative, that is,
_
sin
sin2
2
_
1/2
sin2
< 0. 3.8
Computing the derivative, doing some algebra, and throwing away over all factors that are
always positive we have
2
sin
2
2 4 sin2 sin cos2 4sin
2
> 0. 3.9
International Journal of Dierential Equations 7
2
2
s
6
s
5
2
2
Im
Re
Figure 2
On the restricted domain sin > 0, sin2 < 0, and cos2 1. This reduces 3.9 to
2
sin
2
2 4 sin2 sin 4sin
2
> 0. 3.10
Equation 3.10 can now be factored into a perfect square and prove the assertion made
in3.8
sin2 2 sin
2
> 0. 3.11
Hence the left hand side of 3.6 will decrease monotonically on the restricted domain with
the upper bound being and the lower bound being 0. To summarize, it has just been shown
that there is always one solution, with a positive angle, to 3.6 and this solution must be such
that /2 < < /2 . Consequently there will be two poles for the residue calculation
and they will be complex conjugates of each other. Notice that for the fractionally damped
equation repeated roots are not possible. Repeated roots can only happen when the order of
the derivative becomes one. See Figure 2 for a graphical representation of the location of the
roots.
Now the question of the poles that has been settled the solution to 1.1 can be
generated. Denote the two poles as
s
6,7
i re
i
, 3.12
where and are determined from the r and values that satisfy 3.6 in the usual way,
r
_
2
2
and tan /. Note that is negative, the two solutions are in the second
and third quadrants and s
7
is the complex conjugate of s
6
. Note also that plays the role of
/2 from the Nonfractional case. The poles are of order one and the residue is given by
Residue lim
s s
6
s s
6
e
st
_
sx
0
x
1
s
1
x
0
s
2
s
2
_
lim
s s
7
s s
7
e
st
_
sx
0
x
1
s
1
x
0
s
2
s
2
_
e
s
6
t
_
s
6
x
0
x
1
x
0
s
1
6
2s
6
s
1
6
_
e
s
6
t
_
s
6
x
0
x
1
x
0
s
1
6
2s
6
s
1
6
_
,
3.13
8 International Journal of Dierential Equations
where s
7
has been replaced by s
6
. After some algebra this can be reduced to
2e
t
cost
x
0
_
2r
2
2
r
22
r
2 cos 2
_
A
4r
2
4
cos2
2
2
r
22
2e
t
sint
x
0
r
2 sin 2 x
1
_
2r sin r
1
sin 1
_
4r
2
4r
cos2
2
2
r
22
,
3.14
where A denote to x
1
2r cos r
1
cos 1.
For the contour integral the only contributions come fromthe paths along the negative
real axis
0
Rx
0
x
1
sin x
0
/R
_
R
2
2
_
sin 1
R
2
2
2
2R
R
2
2
cos R
2
e
Rt
R
dR. 3.15
The solution to 1.1 is then 3.15 subtracted from 3.14. This may look overly complicated
but the solution does have the general form of
xt Ae
t
cost Be
t
sint Decay function. 3.16
Notice that the decay function, 3.15, goes to zero if goes to zero or one; that is, if 1.1 goes
to its Nonfractional limits the decay function goes away, as expected. The damping factor
e
t
is similar to the damping factor for the Nonfractional case, e
t/2
. Notice that since the
poles, for the residue calculation, have nonzero imaginary and nonzero real parts we will not
have the same three distinct cases as we did for the Nonfractional case critically damped,
overdamped, and underdamped.
4. The Oscillation Frequency
Consider the frequency of the oscillation component of the solution, Ims
6
. One question
we might ask is: how does the frequency change as we change the order of the fractional
damping? When is set to zero we have an undamped oscillator with frequency
_
2
. 4.1
When is set to one we have the three cases given in Section 2: overdamped, critically
damped, and underdamped. So, the frequency may be zero or nonzero, that is,
0 2,
4
2
2
2
< 2.
4.2
For 0 < < 1 there will always be a nonzero frequency. Note that 0
4
2
2
/2 <
2
.
International Journal of Dierential Equations 9
In the Nonfractional case increasing causes the frequency of oscillation to become
smaller, monotonically, until the critical cases are reached and the oscillation period becomes
innite these are the critical and overdamped cases. In the fractional case the frequency of
oscillation, Ims
6
, now depends on the order of the derivative, , as well as and . To
try to determine how depends on these three parameters consider s to be a function of ,
on 0 1, implicitly dened by
s
2
s
2
0, 4.3
for xed values of and both being positive. Let us restrict our attention to the upper half
plane for s. As such s will be one to one on 0 < 1. Due to the fractional exponent causing
a branch cut on the negative real axis s will not be one to one at 1.
Nowthe question arises, does fall monotonically with respect to ? To get the answer
to this consider the derivative of 4.3 with respect to and isolate ds/d remember, and
are being held xed
ds
d
s
lnss
2s
2
s
. 4.4
The imaginary part of this equation is
d
d
Im
_
ds
d
_
. 4.5
Specically, consider this equation at 0
d
d
_
_
_
_
0
Im
__
s
2
2
_
lnss
2s
2
s
2
2
__
_
_
_
_
0
_
ln
_
2
__
4
2
. 4.6
This gives three initial slopes for the rate of change of with respect to .
2
> 1 The frequency initially increases with increasing damping order.
2
1 The frequency initially is not changing with increasing damping order.
2
< 1 The frequency initially decreases with increasing damping order.
This is not entirely what might have been expected. In the rst case the oscillation
frequency actually increases before falling. Hence there will be some values of for which
the fractional damping will actually cause the oscillations to go faster than the undamped
oscillator the damping will still cause the amplitude to decrease. Each of the above three
cases can become any of the three Nonfractionally damped cases by letting 1 2.7,
2.8, and 2.9. Hence, there are nine cases for the linear fractionally damped oscillator.
There are some graphs of solutions to the imaginary part of 4.3 the oscillation
frequency for various values of , , and . In all three graphs the oscillation frequency is on
the vertical axis and the order of the derivative is on the horizontal axis. The three graphs for
each case correspond to what would be underdamped, critically damped, and overdamped
for a damped oscillator with whole order derivatives.
10 International Journal of Dierential Equations
0 0.2 0.4 0.6 0.8 1
Derivative order
0.9
1
1.1
1.2
1.3
1.4
O
s
c
i
l
l
a
t
i
o
n
f
r
e
q
u
e
n
c
y
Figure 3
0 0.2 0.4 0.6 0.8 1
Derivative order
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
O
s
c
i
l
l
a
t
i
o
n
f
r
e
q
u
e
n
c
y
Figure 4
Figure 3 is a representative graph of case one,
2
> 1. The green graph is for
1, the black graph is for 2 and 1, and the red graph is for 3 and 1.
Figure 4 is a representative graph from case two,
2
1, a at start. The red graph
is for 2
yt
1
_
t
0
t s
1
ysds, 2.1
provided that the right side is pointwise dened on 0, .
Denition 2.2 see 6, 7. The fractional derivative of order > 0 of a continuous function
y : 0, R is given by
D
yt
1
n
_
d
dt
_
n
_
t
0
t s
n1
ysds, 2.2
where n 1 provided that the right side is pointwise dened on 0, .
Remark 2.3 see 3. The following properties are useful for our discussion:
1 I
ut ut
N
k1
C
k
t
k
, D
ut C0, 1
L0, 1, C
k
R, N 1,
2 D
ut ut,
3 D
1/ 1 t
ut yt 0, 0 < t < 1,
u0 0, u1 au,
2.3
then one has the following lemma.
Lemma 2.5. Let y C0, 1 and 1 < < 2, then ut is a solution of BVP 2.3 if and only if ut is
a solution of the integral equation:
ut
_
1
0
G
1
t, sysds, 2.4
where
G
1
t, s
t1 s
1
at
1
s
1
t s
1
_
1 a
1
_
_
1 a
1
_
, 0 s t 1, s ,
t1 s
1
t s
1
_
1 a
1
_
_
1 a
1
_
, 0 < s t 1,
t1 s
1
at
1
s
1
_
1 a
1
_
, 0 t s 1,
t1 s
1
_
1 a
1
_
, 0 t s 1, s.
2.5
Proof. Assume that ut is a solution of BVP 2.3, then by Remark 2.3, we have
ut I
yt C
1
t
1
C
2
t
2
_
t
0
t s
1
ysds C
1
t
1
C
2
t
2
.
2.6
By 2.3, we have
C
2
0, C
1
_
1
0
1 s
1
_
1 a
1
_ysds a
_
0
s
1
_
1 a
1
_ysds. 2.7
4 International Journal of Dierential Equations
Therefore, we obtain
ut
_
t
0
t s
1
ysds
_
1
0
t
1
1 s
1
_
1 a
1
_ysds a
_
0
t
1
s
1
_
1 a
1
_ysds
_
1
0
G
1
t, sysds.
2.8
Conversely, if ut is a solution of integral equation 2.4, using the relation D
t
m
ut D
_
_
t
0
t s
1
ysds
_
D
t
1
_
_
1
0
1 s
1
_
1 a
1
_ysds a
_
0
s
1
_
1 a
1
_ysds
_
D
yt yt.
2.9
A simple computation showed u0 0, u1 au. The proof is complete.
Let
G
2
t, s
t1 s
1
bt
1
s
1
t s
1
_
1 b
1
_
_
1 b
1
_
_ , 0 s t 1, s ,
t1 s
1
t s
1
_
1 b
1
_
_
1 b
1
_
_ , 0 < s t 1,
t1 s
1
bt
1
s
1
_
1 b
1
_
_ , 0 t s 1,
t1 s
1
_
1 b
1
_
_, 0 t s 1, s,
2.10
we call Gt, s G
1
t, s, G
2
t, s Greens function of the boundary value problem 1.1.
Lemma 2.6. Let 0 a, b 1, then the function Gt, s is continuous and satises
1 Gt, s > 0, for t, s 0, 1,
2 Gt, s Gs, s, for t, s 0, 1.
International Journal of Dierential Equations 5
Proof. It is easy to prove that Gt, s is continuous on 0, 1 0, 1, here we omit it. Now we
prove G
1
t, s > 0. Let
g
1
t, s
t1 s
1
at
1
s
1
t s
1
_
1 a
1
_
_
1 a
1
_
, 0 < s t 1, s ,
g
2
t, s
t1 s
1
t s
1
_
1 a
1
_
_
1 a
1
_
, 0 < s t 1,
g
3
t, s
t1 s
1
at
1
s
1
_
1 a
1
_
, 0 < t s 1,
g
4
t, s
t1 s
1
_
1 a
1
_
, 0 < t s 1, s.
2.11
We only need to prove g
1
t, s > 0, 0 < s t 1, s . Since
t1 s
1
at
1
s
1
t s
1
_
1 a
1
_
t
1
1 s
1
a s
1
_
1
s
t
_
1
_
1 a
1
_
,
2.12
set gt 1 s
1
a s
1
1 s/t
1
1 a
1
, we have
g
t 1
_
1
s
t
_
2
s
t
2
_
1 a
1
_
0, for 0 < s < t 1, s . 2.13
Then gt is decreasing on 0, 1. Meanwhile,
g1 1 s
1
a s
1
1 s
1
_
1 a
1
_
a
1
_
1 s
1
_
1
s
_
1
_
> 0, 0 < s < t 1, s .
2.14
Therefore, g
1
t, s > 0, for 0 < s < t 1, s . Clearly g
1
t, s > 0, t s, so g
1
t, s >
0, s, t 0, 1. It is easy to show that g
2
t, s > 0, g
3
t, s > 0, g
4
t, s > 0. Hence, G
1
t, s >
0, s, t 0, 1.
Similarly, G
2
t, s > 0, s, t 0, 1. The proof of 1 is completed.
Let
g
2
t
t1 s
1
t s
1
_
1 a
1
_
_
1 a
1
_
, 0 < s t 1, 2.15
6 International Journal of Dierential Equations
then,
g
2
t
1t
2
1 s
1
1 s/t
2
_
1 a
1
_
_
1 a
1
_
, 0 < s < t 1,
1 s
1
1 s/t
2
_
1 a
1
_
1 s
1
1 s
2
_
1 a
1
_
1 s
2
_
a
1
s
_
0, 0 < s < t 1,
2.16
therefore,
g
2
t 0, 0 < s < t 1. 2.17
So, g
2
t, s is decreasing with respect to t. Similarly, g
1
t, s is decreasing with respect to t.
Also g
3
t, s and g
4
t, s are increasing with respect to t. We obtain that G
1
t, s is decreasing
with respect to t for s t and increasing with respect to t for t s.
With the use of the monotonicity of G
1
t, s, we have
max
0t1
G
1
t, s G
1
s, s
s1 s
1
as s
1
_
1 a
1
_ , s 0, ,
s1 s
1
_
1 a
1
_, s , 1.
2.18
Similarly,
max
0t1
G
2
t, s G
2
s, s
s1 s
1
bs s
1
__
1 b
1
_ , s 0, ,
s1 s
1
__
1 b
1
_, s , 1.
2.19
The proof of 2 is completed.
3. Main Result
In this section, we will discuss the existence and uniqueness of positive solution for boundary
value problem 1.1.
We dene the space X {ut  ut C0, 1} endowed with u
X
max
0t1
ut,
Y {vt  vt C0, 1} endowed with u
Y
max
0t1
vt.
For u, v X Y, let u, v
XY
max{u
X
, v
Y
}.
Dene P {u, v X Y  ut 0, vt 0} , then the cone P X Y.
International Journal of Dierential Equations 7
From Lemma 2.5 in Section 2, we can obtain the following lemma.
Lemma 3.1. Suppose that ft, v and gt, u are continuous, then u, v X Y is a solution of
BVP 1.1 if and only if u, v X Y is a solution of the integral equations
ut
_
1
0
G
1
t, sfs, vsds,
vt
_
1
0
G
2
t, sgs, usds.
3.1
Let T : X Y X Y be the operator dened as
Tu, vt
_
_
1
0
G
1
t, sfs, vsds,
_
1
0
G
2
t, sgs, usds
_
: T
1
vt, T
2
ut,
3.2
then by Lemma 3.1, the xed point of operator T coincides with the solution of system 1.1.
Lemma 3.2. Let ft, v and gt, u be continuous on 0, 1 0, 0, , then T : P P
dened by 3.2 is completely continuous.
Proof. Let u, v P, in view of nonnegativeness and continuity of functions Gt, s, f, and g,
we conclude that T : P P is continuous.
Let P be bounded, that is, there exists a positive constant h > 0 such that u, v
h for all u, v .
Let
M max
__
_
ft, vt
_
_
1 : 0 t 1, 0 v h
_
,
N max
__
_
gt, ut
_
_
1 : 0 t 1, 0 u h
_
,
3.3
then we have
T
1
vt
_
_
_
_
_
_
1
0
G
1
t, sfs, vsds
_
_
_
_
_
M
_
1
0
G
1
s, sds,
T
2
ut
_
_
_
_
_
_
1
0
G
2
t, sgs, usds
_
_
_
_
_
N
_
1
0
G
2
s, sds.
3.4
Hence, Tu, v max{M
_
1
0
G
1
s, sds, N
_
1
0
G
2
s, sds}. T is uniformly bounded.
Since G
1
t, s is continuous on 0, 1 0, 1, it is uniformly continuous on 0, 1 0, 1.
Thus, for xed s 0, 1 and for any > 0, there exists a constant > 0, such that any
t
1
, t
2
0, 1 and t
1
t
2
 < ,
G
1
t
1
, s G
1
t
2
, s < /M. 3.5
8 International Journal of Dierential Equations
Then
T
1
vt
2
T
1
vt
1
 M
_
1
0
G
1
t
2
, s G
1
t
1
, sds < . 3.6
Similarly,
T
2
ut
2
T
2
ut
1
 N
_
1
0
G
2
t
2
, s G
2
t
1
, sds < . 3.7
For the Euclidean distance d on R
2
, we have that if t
1
, t
2
0, 1 are such that t
2
t
1
 < , then
dTu, vt
2
, Tu, vt
1
_
T
1
vt
2
T
1
vt
1
2
T
2
ut
2
T
2
ut
1
2
<
2. 3.8
That is to say, TP is equicontinuous. By the means of the ArzelaAscoli theorem, we have
T : P P is completely continuous. The proof is completed.
Theorem 3.3. Assume that ft, v and gt, u are continuous on 0, 10, 0, , and there
exist two positive functions mt, nt that satisfy
H
1
ft, v
2
ft, v
1
 mtv
2
v
1
, for t 0, 1, v
1
, v
2
0, ,
H
2
gt, u
2
gt, u
1
 ntu
2
u
1
, for t 0, 1, u
1
, u
2
0, .
Then system 1.1 has a unique positive solution if
_
1
0
G
1
s, smsds < 1,
_
1
0
G
2
s, snsds < 1. 3.9
Proof. For all u, v P, by the nonegativeness of Gt, s and ft, v, gt, u, we have
Tu, vt 0. Hence, TP P.
T
1
v
2
T
1
v
1
max
t0,1
T
1
v
2
T
1
v
1

max
t0,1
_
_
_
_
_
_
1
0
G
1
t, s
_
fs, v
2
s fs, v
1
s
_
ds
_
_
_
_
_
_
1
0
G
1
s, smsdsv
2
v
1
v
2
v
1
.
3.10
Similarly,
T
2
u
2
T
2
u
1
u
2
u
1
. 3.11
International Journal of Dierential Equations 9
We have,
Tu
2
, v
2
Tu
1
, v
1
max
_
,
_
u
2
, v
2
u
1
, v
1
. 3.12
From Lemma 3.2, T is completely continuous, by Banach xed point theorem, the operator
T has a unique xed point in P, which is the unique positive solution of system 1.1. This
completes the proof.
Theorem 3.4. Assume that ft, v and gt, u are continuous on 0, 1 0, 0, and
satisfy
H
3
ft, vt a
1
t a
2
tvt,
H
4
gt, ut b
1
t b
2
tut,
H
5
A
1
_
1
0
G
1
s, sa
2
sds < 1, 0 < B
1
_
1
0
G
1
s, sa
1
sds < ,
H
6
A
2
_
1
0
G
2
s, sb
2
sds < 1, 0 < B
2
_
1
0
G
2
s, sb
1
sds < .
Then the system 1.1 has at least one positive solution u, v in
C
_
u, v P  u, v < min
_
B
1
1 A
1
,
B
2
1 A
2
__
. 3.13
Proof. Let C {u, v X Y : u, v < r} with r minB
1
/1 A
1
, B
2
/1 A
2
, dene
the operator T : C P as 3.2.
Let u, v C, that is, u, v < r. Then
T
1
v max
t0,1
_
_
_
_
_
_
1
0
G
1
t, sfs, vsds
_
_
_
_
_
_
1
0
G
1
s, sa
1
s a
2
svsds
_
1
0
G
1
s, sa
1
sds
_
1
0
G
1
s, sa
2
sdsv
B
1
A
1
v r.
3.14
Similarly, T
2
u r, so Tu, v r, Tu, v C. From Lemma 3.2 T : C C is
completely continuous.
Consider the eigenvalue problem
u, v Tu, v, 0, 1. 3.15
10 International Journal of Dierential Equations
Under the assumption that u, v is a solution of 3.15 for a 0, 1, one obtains
u T
1
v
max
t0,1
_
_
_
_
_
_
1
0
G
1
t, sfs, vsds
_
_
_
_
_
<
_
1
0
G
1
s, sa
1
s a
2
svsds
_
1
0
G
1
s, sa
1
sds
_
1
0
G
1
s, sa
2
sdsv
B
1
A
1
v r.
3.16
Similarly, v T
2
u < r, so u, v < r, which shows that u, v / C. By Lemma 2.4, T
has a xed point in C. We complete the proof of Theorem 3.4.
Example 3.5. Consider the problem
D
7/4
ut ft, vt 0, 0 < t < 1,
D
3/2
vt gt, ut 0, 0 < t < 1,
u0 0, u1
1
2
u
_
1
2
_
,
v0 0, v1
3
4
v
_
1
2
_
,
3.17
where
ft, vt
tvt
1 t1 vt
, gt, ut arctan
t
1 t
sinut. 3.18
Set v
1
t, v
2
t, u
1
t, u
2
t 0, and t 0, 1, then we have
_
_
ft, v
2
t ft, v
1
t
_
_
t
1 t
v
2
t v
1
t,
_
_
gt, u
2
t gt, u
1
t
_
_
arctan
t
1 t
u
2
t u
1
t.
3.19
International Journal of Dierential Equations 11
Therefore,
_
1
0
G
1
s, smsds
_
1
0
G
1
s, sds
1
7/4
_
1 1/2
7/4
_
_
_
1/2
0
s1 s
3/4
ds
_
1/2
0
1
2
_
s
_
1
2
s
__
3/4
ds
_
1
1/2
s1s
3/4
ds
_
2
_
1 1/2
7/4
_
5
3/4
1/2
<
4
5
< 1,
_
1
0
G
2
s, snsds
4
_
1
0
G
2
s, sds
4
1
3/2
_
13/41/2
1/2
_
_
_
1/2
0
s1s
1/2
ds
_
1/2
0
3
4
_
s
_
1
2
s
__
1/2
ds
_
1
1/2
s1s
1/2
ds
_
4
_
1 3/41/2
2
_
_
1 3/41/2
1/2
_
3/2
3
0.6018 < 1.
3.20
With the use of Theorem 3.3, BVP 3.17 has a unique positive solution.
Example 3.6. Consider the problem
D
7/4
ut ft, vt 0, 0 < t < 1,
D
3/2
vt gt, ut 0, 0 < t < 1,
u0 0, u1
1
2
u
_
1
2
_
,
v0 0, v1
3
4
v
_
1
2
_
,
3.21
where
ft, vt t
2
t
1 t
ln1 vt, gt, ut 10
t
2
20
ut. 3.22
We have
_
_
ft, vt
_
_
t
2
t
1 t
vt,
_
_
gt, ut
_
_
_
10
t
2
20
_
ut. 3.23
12 International Journal of Dierential Equations
Hence,
A
1
_
1
0
G
1
s, sa
2
sds
_
1
0
G
1
s, sds
2
_
1 1/2
7/4
_
5
3/4
1/2
< 1,
B
1
_
1
0
G
1
s, sa
1
sds
_
1
0
G
1
s, s s
2
ds < ,
A
2
_
1
0
G
2
s, sb
2
sds
_
1
0
G
2
s, sds 0.7666 < 1,
B
2
_
1
0
G
2
s, sb
1
sds
_
1
0
G
2
s, s
_
10
s
2
20
_
ds < .
3.24
By Theorem 3.4, BVP 3.21 has at least one positive solution in
C
_
u, v P  u, v < min
_
B
1
1 A
1
,
B
2
1 A
2
__
. 3.25
Acknowledgments
This work was jointly supported by the Natural Science Foundation of Hunan Provincial
Education Department under Grants 07A066 and 07C700, the Construct Program of the Key
Discipline in Hunan Province, Aid Program for Science and Technology Innovative Research
Teamin Higher Educational Institutions of Hunan Province, and the Foundation of Xiangnan
University.
References
1 C. Bai and J. Fang, The existence of a positive solution for a singular coupled system of nonlinear
fractional dierential equations, Applied Mathematics and Computation, vol. 150, no. 3, pp. 611621,
2004.
2 X. Su, Boundary value problem for a coupled system of nonlinear fractional dierential equations,
Applied Mathematics Letters, vol. 22, no. 1, pp. 6469, 2009.
3 Z. Bai and H. L u, Positive solutions for boundary value problem of nonlinear fractional dierential
equation, Journal of Mathematical Analysis and Applications, vol. 311, no. 2, pp. 495505, 2005.
4 S. Zhang, Existence of solution for a boundary value problem of fractional order, Acta Mathematica
Scientia, vol. 26, no. 2, pp. 220228, 2006.
5 S. Q. Zhang, Positive of solution for boundaryvalue problems of nonlinear fractional dierential
equations, Electronic Journal of Dierential Equations, vol. 36, pp. 112, 2006.
6 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, San Diego, Calif, USA, 1999.
7 A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo, Theory and Applications of Fractional Dierential Equations,
vol. 204 of NorthHolland Mathematics Studies, Elsevier Science, Amsterdam, The Netherlands, 2006.
8 E. Zeidler, Nonlinear Functional Analysis and Its ApplicationsI: FixedPoint Theorems, Springer, New
York, NY, USA, 1986.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 509286, 9 pages
doi:10.1155/2010/509286
Research Article
The Periodic Solutions of the Compound Singular
Fractional Differential System with Delay
XuTing Wei and XuanZhu Lu
School of Mathematics and Computer Science, Fuzhou Universiy, Fuzhou 350108, China
Correspondence should be addressed to XuTing Wei, wxt1985wan@163.com
Received 31 July 2009; Revised 16 November 2009; Accepted 1 December 2009
Academic Editor: Fawang Liu
Copyright q 2010 X. Wei and X. Lu. This is an open access article distributed under the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The paper gives sucient conditions on the existence of periodic solution for a class of
compound singular fractional dierential systems with delay, involving Nishimoto fractional
derivative. Furthermore, for the particular functions, the necessary conditions on the existence of
periodic solution are also derived. Especially, for twodimensional compound singular fractional
dierential equation with delay, the criteria of existence of periodic solution are obtained. Finally,
two examples are presented to verify the validity of criteria.
1. Introduction
In real life, there are many phenomena with time delay. The mathematical model derived
from engineering, physics, mechanics, control theory, chemical reactions, biology, and
medicine was made with a signicant amount of delay, such as the limited signal
transmission speed human reaction time to the outside world. Therefore, the delay is
widespread in nature and society, in the introduction of timedelay dierential equations
there can be a more accurate description and explanation of various phenomena and
processes.
Fractional calculus is the promotion of classical calculus. The study found that
fractional calculus was very suitable to describe long memory and hereditary properties
of various materials and processes 1, 2. In the recent years, fractional calculus becomes a
research hotspot, its eld of concern has become wide, such as the numerical method of the
equation in 3, the existence and uniqueness of equations in 4, fractional Brownian motion,
fractional reactiondiusion equation and random walk 5, 6, fractional wavelet transform
7, and fractional control 8.
Most of the abovementioned studies, utilize the Riemannliouville fractional deriva
tive denition, which due to its nature of its denition is simple and relatively good. But
Nishimoto denition of fractional calculus 9, 10, has not received a lot of attention, this may
2 International Journal of Dierential Equations
be part of the naturalization due to the complexity of its denition, but compared to Riemann
liouville fractional calculus, it has a better nature, relevant results more concise useful.
The existence of periodic solutions of dierential equations is one of the important
research directions of biomathematics 1115, which has a wide range of applications, such
as the existence of periodic orbits of celestial movement and its stability.
In 12, the author discussed the following system:
E xt Axt Bxt
1
Cxt
2
, 1.1
and obtained sucient and necessary conditions for the existence of periodic solutions for the
system. Taking into account the periodic solutions of the fractional timedelay system will be
a very important practical signicance; we are tried to generalize the corresponding results
to the case of fractional order.
For the above reasons we consider the following compound singular fractional
dierential system with delay:
HD
xt Axt Bxt
1
Cxt
2
, 1.2
where D
, C
}, C
is a contour along the cut joining the points z and iIz, which starts from
the point at , encircles the point z once counterclockwise, and returns to the point at ,
C
is a contour along the cut joining the points z and iIz, which starts from the point
at , encircles the point z once counterclockwise, and returns to the point at ,
f
z
_
fz
_
:
1
2i
_
c
f
z
1
d
_
R
Z
; Z
: {1, 2, 3, . . .}
_
,
f
n
z : lim
n
_
f
z
_
n N : {1, 2, 3, . . .},
2.1
where , arg z , for C
.
Then f
z <
R.
Let us recall the following useful properties associated with the denition introduced
above 9.
International Journal of Dierential Equations 3
Property 1. For a constant ,
_
e
z
_
e
z
/
0; R; z C. 2.2
Property 2. For a constant ,
_
e
z
_
e
i
e
z
/
0; R; z C. 2.3
Property 3. If the function fz is singlevalued and analytic in some domain C, then
_
f
z
_
z
_
f
z
_
. 2.4
Property 4. For a constant ,
_
z
e
i
_
R; z C;
_
_
_
_
_
_
_
_
<
_
. 2.5
In the following section of this paper, we let D
H A Be
Ce
_
0. 3.1
Proof. Assume that i is pure imaginary root of 3.1, let xt Ke
it
K R
n
, substituting
xt in 1.2, then
_
_
i
_
H A Be
1
i
Ce
2
i
_
K 0. 3.2
As i is pure imaginary roots of 3.1, note that
det
_
_
i
_
H A Be
1
i
Ce
2
i
_
0. 3.3
So 3.1 exists nonzero solution K, then, xt Ke
it
is the nonconstant periodic solution of
1.2.
If the system have nonconstant periodic solutions, then we may wonder whether
the solution satisfy 3.1, in fact, as you will see it holds when the function satisfy some
4 International Journal of Dierential Equations
conditions. We know that if the function ft is a continuous smooth periodic function with
period 2l, then it can be expressed as its fourier series form
ft
k
C
k
e
ikt/l
, 3.4
where C
k
1/2l
_
l
l
fe
ikt/l
d.
And as we also know that its fourier series expansion has the following properties:
f
k
C
k
ik
l
e
ikt/l
, 3.5
we can even get the following relation if ft satisfy some more strictly condition:
f
k
t
k
C
k
_
ik
l
_
k
e
ikt/l
. 3.6
To obtain the similar property of our fractional derivative, what conditions the function
should satisfy? We give the following function space.
Denition 3.2. If the periodic function ft is continuous and smooth on R, its > 1 order
Nishimoto derivative exists, then we let t denote the corresponding function space whose
elements have the following property:
D
ft D
k
C
k
e
ikt/l
_
k
C
k
D
_
e
ikt/l
_
k
C
k
_
ikt
l
_
e
ikt/l
, 3.7
it is easy to know from the denition that e
it
t R, and so t is nonempty.
Theorem 3.3. If xt is the nonconstant periodic solution of 1.2, and further xt t, one can
obtain the necessity of Theorem 3.1.
Proof. Suppose the period of xt is 2l, xt is continuous and dierentiable because of > 1,
then we can denote it in the form of its fourier series:
xt
k
C
k
e
ikt/l
. 3.8
Since xt t, we have
D
xt
k
C
k
_
ikt
l
_
e
ikt/l
, 3.9
where C
k
1/2l
_
l
l
xe
ikt/l
d.
International Journal of Dierential Equations 5
We put 3.8 and 3.9 into 1.2, and obtain
k
__
ik
l
_
H A Be
ik
1
/l
Ce
ik
2
/l
_
C
k
e
ikt/l
0, 3.10
then multiply 1/2le
imt/l
m 0, 1, 2, . . . on both sides of 3.10 and integrate it froml
to l, hence
k
__
ik
l
_
H A Be
ik
1
/l
Ce
ik
2
/l
_
C
k
1
2l
_
l
l
e
ikmt/l
dt 0. 3.11
It is easy to deduce that
1
2l
_
l
l
e
ikmt/l
dt
1, k m,
0, k
/
m,
3.12
recalling 3.10, it reduces to
__
ik
l
_
H A Be
ik
1
/l
Ce
ik
2
/l
_
C
m
0 m 0, 1, 2, . . .. 3.13
Thus, if there are no pure imaginary roots in 3.1, then for every k we have C
k
0, according
to 3.8, we conclude that xt cons tan t vector which conicts the suppose that xt is the
nonconstant periodic solution of 1.2.
4. TwoDimensional Case
For the case of the twodimensional compound singular fractional dierential system with
delay, there is
HD
xt Axt Bxt
1
Cxt
2
, 4.1
where > 0, H
_
1 0
0 0
_
, A
_
a
11
a
12
a
21
a
22
_
, B
_
b
11
b
12
b
21
b
22
_
, C
_
c
11
c
12
c
21
c
22
_
, xt
_
x
1
t
x
2
t
_
,
and x
1
t, x
2
t is scalar function.
Using Theorem 3.1, we obtained the following theorem.
6 International Journal of Dierential Equations
Theorem 4.1. If one of the following equations exists nonzero real root, then system4.1 exists non
constant periodic solution, further more, if xt t, then the conclusion is sucient and necessary
A B cos
_
2
1
y
_
C cos
_
2
2
y
_
Ecos
_
1
y
_
F cos
_
2
y
_
Gcos
_
2
1
y
_
b
22
_
_
y
_
_
cos
_
1
y
2
_
c
22
_
_
y
_
_
cos
_
2
y
2
_
a
22
_
_
y
_
_
cos
_
2
_
0
B sin
_
2
1
y
_
C sin
_
2
2
y
_
Esin
_
1
y
_
F sin
_
2
y
_
Gsin
_
2
1
y
_
b
22
_
_
y
_
_
sin
_
1
y
2
_
c
22
_
_
y
_
_
sin
_
2
y
2
_
a
22
_
_
y
_
_
sin
_
2
_
0,
4.2
or
A B cos
_
2
1
y
_
C cos
_
2
2
y
_
Ecos
_
1
y
_
F cos
_
2
y
_
Gcos
_
2
1
y
_
b
22
_
_
y
_
_
cos
_
1
y
2
_
c
22
_
_
y
_
_
cos
_
2
y
2
_
a
22
_
_
y
_
_
cos
_
2
_
0
B sin
_
2
1
y
_
C sin
_
2
2
y
_
Esin
_
1
y
_
F sin
_
2
y
_
Gsin
_
2
1
y
_
b
22
_
_
y
_
_
sin
_
1
y
2
_
c
22
_
_
y
_
_
sin
_
2
y
2
_
a
22
_
_
y
_
_
sin
_
2
_
0,
4.3
where E
_
_
_
a
11
a
12
b
21
b
22
_
_
_
_
_
_
b
11
b
12
a
21
a
22
_
_
_, F
_
_
_
a
11
a
12
c
21
c
22
_
_
_
_
_
_
c
11
c
12
a
21
a
22
_
_
_, G
_
_
_
b
11
b
12
c
21
c
22
_
_
_
_
_
_
c
11
c
12
b
21
b
22
_
_
_.
Proof. First of all we know that
_
yi
_
e
i/2
, y 0,
_
_
y
_
_
e
i/2
, y < 0,
4.4
according to Theorem 3.1, we have
h
_
yi
_
_
_
_yiI A Be
1
yi
Ce
2
yi
_
_
_
_
_
_
_
_
_
yi
_
a
11
b
11
e
1
yi
c
11
e
2
yi
a
12
b
12
e
1
yi
c
12
e
2
yi
a
21
b
21
e
1
yi
c
21
e
2
yi
a
22
b
22
e
1
yi
c
22
e
2
yi
_
_
_
_
_
A Be
1
yi
Ce
2
2
yi
Ee
1
yi
Fe
2
yi
Ge
1
yi
b
22
_
yi
_
1
yi
c
22
_
yi
_
2
yi
a
22
_
yi
_
.
4.5
International Journal of Dierential Equations 7
then
R
_
h
_
yi
__
A B cos
_
2
1
y
_
C cos
_
2
2
y
_
Ecos
_
1
y
_
F cos
_
2
y
_
Gcos
_
2
1
y
_
b
22
_
_
y
_
_
cos
_
1
y
2
_
c
22
_
_
y
_
_
cos
_
2
y
2
_
a
22
_
_
y
_
_
cos
_
2
_
0,
I
_
h
_
yi
__
B sin
_
2
1
y
_
C sin
_
2
2
y
_
Esin
_
1
y
_
F sin
_
2
y
_
Gsin
_
2
1
y
_
b
22
_
_
y
_
_
sin
_
1
y
2
_
c
22
_
_
y
_
_
sin
_
2
y
2
_
a
22
_
_
y
_
_
sin
_
2
_
0,
4.6
where R z and I z denote the real and imaginary parts of z, respectively. Then using
Theorem 3.1, if there exists y R, y
/
0, hyi 0, we obtain that
R
_
h
_
yi
__
0,
I
_
h
_
yi
__
0,
4.7
hence system 4.1 exists nonconstant periodic solution. This proved the theorem.
5. Examples
In this section we give some concrete examples to illustrate our conclusions.
Example 5.1. We consider the following twodimensional compound singular fractional
dierential system with delay:
D
x
1
t x
2
t x
1
t 1 x
2
t 1 x
2
t 1,
0 x
1
t x
1
t 1 x
1
t 1,
5.1
where H
_
1 0
0 0
_
, A
_
0 1
1 0
_
, B
_
1 1
1 0
_
, C
_
0 1
1 0
_
,
1
1,
2
1, > 1. so we have
A 1, B 1, C 1, E 0, F 0, G 2.
Using the discriminant of Theorem 4.1, we have
2 cos
_
2y
_
1,
0 0,
5.2
the solution is y k /3 k 0, 1, 2, . . ..
8 International Journal of Dierential Equations
According to Theorem 4.1 , system 5.1 has nonconstant periodic solution. We
suppose that y /3, as
_
_
yi
_
I A Be
yi
Ce
yi
_
K 0 5.3
exists nonzero solution, it means that
_
_
3
_
cos
_
2
_
3
2
_
_
_
3
_
sin
_
2
_
3
2
_
i 0
0 0
_
k
1
k
2
_
_
0
0
_
. 5.4
So we have k
1
0, k
2
for any real number. Supposed that k
2
1, then we obtained a non
constant periodic solution of system 5.1:
xt e
/3it
_
0
1
_
. 5.5
We can verify that xt is a nonconstant periodic solution of system 5.1.
Example 5.2. Consider the following twodimension compounded with singular fractional
dierential equation delay system:
D
1/2
x
1
t x
2
t x
1
t 1 x
2
t 1,
0 x
1
t x
2
t x
1
t 1 x
2
t 1 x
1
t 1 x
2
t 1.
5.6
where H
_
1 0
0 0
_
, A
_
0 1
1 1
_
, B
_
0 1
1 1
_
, C
_
0 1
1 1
_
,
1
1,
2
1, 1/2.
So we have A 1, B 1, C 1, E 0, F 0, G 2.
Using the discriminant of Theorem 4.1, we obtained
_
2 cos y 1
__
_
y
_
_
1/2
cos
4
2 cos 2y 1 0,
_
2 cos y 1
__
_
y
_
_
1/2
sin
4
2 sin2y.
5.7
Through the simplication of this equation, we have
2 sin
_
2y /4
_
1/2, 5.8
and the solution is y k 1/2arc sin1/2
2 /8, k 0, 1, 2, . . ..
According to Theorem 4.1, there exists nonconstant periodic solution in the system.
International Journal of Dierential Equations 9
References
1 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, London, UK, 1999.
2 K. S. Miller and B. Ross, An Introduction to the Fractional Calculus and Fractional Dierential Equations,
John Wiley & Sons, New York, NY, USA, 1993.
3 Xuanzhu Lu and Fawang Liu, The explicit and implicit nite dierence approximations for the space
fractional advection diusion equation, Computational Mechanics, September 2004, Beijing, China.
4 A. Arikoglu and I. Ozkol, Solution of fractional dierential equations by using dierential transform
method, Chaos, Solitons and Fractals, vol. 34, no. 5, pp. 14731481, 2007.
5 M. M. Meerschaert, Stochastic solution of spacetime fractional diusion equations, Physical Review
E, vol. 65, no. 4, Article ID 041103, 4 pages, 2002.
6 F. Liu, V. V. Anh, and I. Turner, Time fractional advectiondispersion equation, Journal of Applied
Mathematics & Computing, vol. 13, no. 12, pp. 233245, 2003.
7 M. Unser and T. Blu, Fractional splines and wavelets, SIAM Journal, vol. 42, no. 1, pp. 4367, 2000.
8 P. Varshney, M. Gupta, and G. S. Visweswaran, New switched capacitor fractional order integrator,
Journal of Active and Passive Electronic Devices, pp. 187197, 2007.
9 T.M. Hsieh, S.D. Lin, and H. M. Srivastava, Some relationships between certain families of ordinary
and fractional dierential equations, Computers and Mathematics with Applications, vol. 46, no. 1011,
pp. 14831492, 2003.
10 K. Nishimoto, Fractional Calculus: Integrations and Dierentiations of Arbitrary Order, vol. 1, Descartes
Press, Koriyama, Japan, 1984.
11 M. L. Hbid and R. Qesmi, Periodic solutions for functional dierential equations with periodic delay
close to zero, Electronic Journal of Dierential Equations, vol. 141, pp. 112, 2006.
12 Z.X. Zhang and W. Jiang, The periodic solution of compound singular dierential equation with
delay, Journal of Anhui University Natural Science Edition, vol. 30, pp. 47, 2006.
13 V. Kevantuo and C. Lizama, A characterization of periodic solutions for timefractional dierential
equations in UMD spaces and applications, Mathematische Nachrichten, pp. 2528, 2008.
14 J. L. Kaplan and J. A. Yorke, Ordinary dierential equations which yield periodic solutions of
dierential delay equations, Journal of Mathematical Analysis and Applications, vol. 48, pp. 317324,
1993.
15 W. Deng, C. Li, and J. L u, Stability analysis of linear fractional dierential system with multiple time
delays, Nonlinear Dynamics, vol. 48, no. 4, pp. 409416, 2007.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 968186, 13 pages
doi:10.1155/2010/968186
Research Article
The Use of Fractional BSplines
Wavelets in Multiterms Fractional Ordinary
Differential Equations
X. Huang and X. Lu
School of Mathematical and Computer Sciences, Fuzhou University, Fuzhou 350002, China
Correspondence should be addressed to X. Huang, hxdjwan@163.com
Received 31 July 2009; Revised 2 November 2009; Accepted 4 November 2009
Academic Editor: Fawang Liu
Copyright q 2010 X. Huang and X. Lu. This is an open access article distributed under the Creative
Commons Attribution License, which permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
We discuss the existence and uniqueness of the solutions of the nonhomogeneous linear dierential
equations of arbitrary positive real order by using the fractional BSplines wavelets and the Mittag
Leer function. The dierential operators are taken in the RiemannLiouville sense and the initial
values are zeros. The scheme of solving the fractional dierential equations and the explicit
expression of the solution is given in this paper. At last, we show the asymptotic solution of the
dierential equations of fractional order and corresponding truncated error in theory.
1. Introduction
Recently, there have been several schemes devoted to the solution of fractional dierential
equations. These schemes can be broadly classied into two classes, numerical and analytical
1. As we know, with the help of some special functions, such as MittagLeer function
and Green function, Miller and Ross have obtained the explicit representations of solutions of
some classes of homogeneous linear fractional dierential equations FDEs in 2; through
the use of the technique of Laplace and Fourier Transform, the analytical solutions have been
given by Podlubny in 3. The numerical scheme we have encountered can be divided into
two groups. In the rst group, the solution is approximated over the entire domain using
approximating functions such as polynomials and orthogonal functions. In the second group,
the entire domain is divided into several small domains like in a nite element technique, and
the solution is obtained for variables at the node points 1. As Edwards et al. declaimed in
4, several forms of fractional dierential equations have been proposed in standard models,
and there has been signicant interest in developing numerical schemes for their solution.
Thus, several papers have been presented in dealing with approximate numerical techniques
2 International Journal of Dierential Equations
for FDEs. Among these, the papers of Diethelm, Edwards, Ford, Freed, and Simpson are
noteworthy see, e.g., 1, 46. One of these schemes which should be mentioned is the
use of a PredictorCorrector more precisely, PECE method in 6. In order to obtain higher
precision, they have replaced the PECE method by a PEC
m
E method with m 2 in 7.
In particular, the PECE method is an important numerical scheme which has been applied
in many elds; for example, Yang and Liu have applied the PECE method for simulating
fraction order dynamical control system in 8. In addition, several numerical schemes have
also been proposed by other authors see 9, 10. In this paper, we present a new scheme
which contains the features of both two groups by using the fractional Bsplines wavelet.
In paper 11, Unser and Blu have proved that the fractional Bsplines generates a valid
multiresolution analyses of L
2
for > 1/2, which means that the orthogonal fractional B
splines could be obtained by applying the standard technique described in 12. And they
also obtained that the fractional Bsplines deriving the asymptotic development of the L
2
have a fractional order of approximation.
In fact, as the theories of wavelets analyses improve day by day, the wavelet has
become a powerful mathematical tool which widely used in signal processing, image
compression and enhancement, pattern recognition, control systems, and other elds in the
past two decades. But almost no papers or books have applied the theories of wavelets
to solve the fractional dierential equations. And our fundamental purpose of this paper
is applying the fractional BSplines wavelets to prove the existence and uniqueness of the
solution of the nonhomogeneous linear fractional dierential equations also socalled linear
multiterm fractional dierential equations with its initial conditions. Let us begin to discuss
the solution of multiterms fractional ordinary dierential equations with the following form:
b
n
D
n
b
n1
D
n1
b
1
D
1
yt ft. 1.1
For convenience, we consider the initial values:
_
D
k
1
yt
_
t0
0,
_
D
k
2
yt
_
t0
0, . . . ,
_
D
k
r
k
yt
_
t0
0, 1.2
where D
, b
k
R, k 1, . . . , n. The function ft belongs
to the space L
2
; without loss of generality, in this paper, we consider the interval as
0, T, T R
.
The plan of this paper is as follows. In Section 2, we recall the denitions of fractional
derivative and integral and related properties which will be used in this paper, give the
representation of MittagLeer function and generalized MittagLeer function, and then
introduce the fractional Bsplines and some related properties of wavelets. In Section 3,
by applying the technique of the Laplace Transforms, and considering the proprieties of
the generalized MittagLeer function, we prove the lemma of the dierential equations
of arbitrary positive real order, which make sure the solution belongs to the space L
2
; by
virtue of the solution that can be expressed as the form of wavelet series and the basis
function that is the orthogonal fractional Bsplines wavelet which yields the Riesz basis for
the space L
2
, we can prove the uniqueness of the coecients of the representation of the
solution, which gain the uniqueness of the solution of the fractional dierential equations
and validate the representation of solution. Thus, we have nished the proof of the theorem.
International Journal of Dierential Equations 3
In Section 4, the asymptotic solution of the dierential equations of fractional order Q
t
yt :
1
m
d
m
dt
m
_
t
0
y
t
m1
d, 2.1
where m is the integer dened by m 1 < m see 3, and is gamma function.
Similarly, the left RiemannLiouville integral operators of order > 0 is dened as
0
D
t
yt :
1
_
t
0
t
1
yd. 2.2
And then we should give the following expression for the Laplace transform of the Riemann
Liouville dierential operators of the order > 0, which is
L
_
0
D
t
yt; s
_
s
Ys
m1
k0
s
k
_
0
D
k1
t
yt
_
t0
, m 1 < m. 2.3
Let us now introduce the case of the Caputo dierential operators of arbitrary order
> 0, which is dened as
C
0
D
t
ft
1
n
_
t
0
f
n
d
t
1n
, n 1 < n. 2.4
And the formula of its Laplace transform can be expressed as
L
_
C
0
D
t
ft; s
_
s
Fs
n1
k0
s
k1
f
k
0, n 1 < n. 2.5
2.2. MittagLefer Functions and Generalized
The MittagLeer functions and its generalized forms have played a special role in solving
the fractional dierential equations. In this section, we just give the denition of the following
4 International Journal of Dierential Equations
series of representation of the MittagLeer function E
z :
n0
z
n
n 1
, > 0, z C. 2.6
And for the generalized MittagLeer function, we use the following denition:
E
,
z :
n0
z
n
_
n
_, > 0, > 0, z C. 2.7
Let us now consider the Laplace transforms of the function t
k1
E
,
at
see 3, which
is dened by
_
0
e
st
t
k1
E
k
,
at
dt
k!s
a
k1
,
_
Res > a
1/
_
. 2.8
2.3. Fractional BSplines Wavelet
Splines have had a signicant impact on the early development of the theory of the wavelet
transformsee 13. And Unser and Blu have rst mentioned the fractional BSplines in 13,
who extended Schoenbergs family of polynomial splines to all fractional degrees > 1 and
dened the fractional causal Bsplines by taking the 1th fractional dierence of the one
sided power function:
x :
1
1
1
1
k0
1
k
_
1
k
_
x k
, 2.9
where the oneside power function x
is dened as follows:
x
, x 0,
0, otherwise.
2.10
Then we introduce the fractional Bsplines autocorrelation sequence as follows:
21
k :
_
x,
x k
_
. 2.11
From 13, we know the explicit form of the fractional Bsplines wavelet:
_
x
2
_
kZ
1
k
2
lZ
_
1
l
_
21
l k 1
x k, 2.12
International Journal of Dierential Equations 5
which yields a Riesz basis for L
2
. Thus, using the standard orthogonalize technique
described in 12, we have
w
_
k0
_
_
2k
_
_
2
_
1/2
, 2.13
where
k
ck
t k, 2.14
where the coecients ck are constants.
3. The Existence and Uniqueness of the Solution of
Multiterms Fractional Ordinary Differential Equations
In this section, we will prove the existence and uniqueness properties of the solutions of the
nonhomogeneous linear dierential equations of arbitrary real order > 0.
3.1. An Important Lemma
Under the hypothesis of the existence for the solutions of 1.1, we have the following lemma.
Lemma 3.1. Let
n
>
n1
> >
1
, and
n
1, and ft L
2
, then the solution of the initial
value problem 1.1 and 1.2 is also in L
2
.
Proof. In order to prove this lemma, we divide the proof into two steps.
Firstly, we consider the case of n 1, then 1.1 can be rewritten as
0
D
1
t
yt ft. 3.1
Taking the transform of 3.1 to both sides, we obtain
yt
0
D
1
t
ft. 3.2
By virtue of the functions ft L
2
and D
: L
2
L
2
that is a bounded linear
operator see 15, we can easily derive the function yt L
2
.
And then, let us consider the case of n > 1.
To 1.1, using the formula 2.2 and taking the Laplace transform to both sides, we
conclude that
b
1
s
1
b
2
s
2
b
n
s
n
Ys Fs, 3.3
6 International Journal of Dierential Equations
and then
Ys
Fs
b
1
s
1
b
2
s
2
b
n
s
n
FsGs. 3.4
To prove the functions yt L
2
, we change the functions Gs into the following form:
Gs
1
b
n
s
n
b
n1
s
n1
1
1
_
n2
k0
b
k
s
k
/b
n
s
n
b
n1
s
n1
_. 3.5
Then replacing the factor 1/b
n
s
n
b
n1
s
n1
with b
1
n
s
n1
/s
n1
b
n1
/b
n
in 3.5, and
expanding the second factor into a formal of series, we have
Gs
m0
1
m
b
1
n
s
n1
s
n1
b
n1
/b
n
m1
_
n2
k0
_
b
k
b
n
_
s
n1
_
m
. 3.6
Using the method described in 3, we obtain the expression of Gs, which is
Gs
1
b
n
m0
1
m
k
0
k
1
k
n2
m
k
0
,k
1
,...,k
n2
0
m; k
0
, k
1
, . . . , k
n2
n2
i0
_
b
i
b
n
_
k
i
s
n1
n2
i0
i
n1
k
i
s
n1
b
n1
/b
n
m1
,
3.7
Substituting Gs in 3.4, considering the formula 2.8, and taking the inverse Laplace
transform termbyterm, we obtain that the analytical solution of initial values problem 1.1
and 1.2 in the following form:
yt ft gt 3.8
with
gt
1
b
n
m0
1
m
m!
k
0
k
1
k
n2
m
k
0
,k
1
,...,k
n2
0
m; k
0
, k
1
, . . . , k
n2
n2
i0
_
b
i
b
n
_
k
i
t
n1
m
n
n2
j0
n1
j
k
j
1
E
m
n1
,
n
n2
j0
n1
j
k
j
_
b
n1
b
n
t
n1
_
,
3.9
where m; k
0
, k
1
, . . . , k
n2
m!/
n2
i0
k
i
! is the multinomial coecient, the representation
of ft gt is the convolution of functions ft and gt, and E
k
,
is the kth derivative of the
MittagLeer function with parameters and see 16.
International Journal of Dierential Equations 7
From the representation 3.9, the function E
m
,
z is bounded in z C, and the index
satises
n
n1
m
n
n2
j0
n1
j
k
j
1 0; so we can easily gain that the function
gt is bounded in . And then
_
_
yt
_
_
L
2
_
_
ft gt
_
_
L
2
_
_
ft
_
_
L
2
_
_
gt
_
_
L
2
C
_
_
ft
_
_
L
2
, 3.10
it yields that the functions yt L
2
, and so we have nished the proof of the lemma.
3.2. The Proof of Existence and Uniqueness
Considering the solution of the fractional dierential equation yt L
2
and the fractional
Bspline wavelet
k
ck
t k, 3.11
where the index satises > 0.
Then substituting 2.14 into 1.1, and taking the Fourier transform to both sides, we
obtain
c
_
b
1
i
1
b
2
i
2
b
n
i
n
_
w
f, 3.12
where c
k
cke
ikw
, and 3.12 is equivalent to the following form:
c
f
_
b
1
i
1
b
2
i
2
b
n
i
n
_. 3.13
And then, by taking the inverse Fourier transform to 3.13, we have
k
ck
t k ft gt. 3.14
Because the function
t k
_
, 3.15
8 International Journal of Dierential Equations
where , ) is the inner product. Then substituting 3.15 in 2.14, we obtain the solution of
the initial value problems 1.1 and 1.2.
Let us suppose that the initial value problems 1.1 and 1.2 have another solution,
which can be expressed as
y
1
t
n
dn
t n. 3.16
Utilizing the similar scheme to ck, we acquire the representation of dn, which is
dn
_
ft gt,
t n
_
. 3.17
Taking 3.17 into 3.16, we obtain the representation of y
1
t. Obviously, we have
yt y
1
t
k
ck
t k
n
dn
t n 0; 3.18
it derives that yt y
1
t, which means that the solution of the initial value problems 1.1
and 1.2 is unique. Then substituting y
1
t with its coecient in 1.1, we can easily check
that the equation is correct. It indicates that the function y
1
t is a solution of the dierential
equation 1.1, which yields the existence of solution. Finally, we have completed the proof of
the solution of existence and uniqueness of the nonhomogeneous linear dierential equations
of arbitrary order.
4. The Asymptotic Solution of the Fractional Differential Equations
and Error Estimation
4.1. The Asymptotic Solution of the Fractional Differential Equations
The purpose of this section is to discuss the case of
i
Q
, r
k
N, k 0, 1, . . . , m, have a unique solution.
International Journal of Dierential Equations 9
In Section 3, we have proved the existence and uniqueness of the nonhomogeneous
linear fractional dierential equations of arbitrary positive real order. Obviously, it also
satises the case of positive rational order. To 1.1, let q Z be the least common multiple
of the denominators of
i
Q, i 1, 2, . . . , n, so
i
q
i
/q, where q
i
Z, i 1, 2, . . . , n. And
to 4.1, let v 1/q and m i
0
q
n
, m i
1
q
n1
, . . . , m i
n1
q
1
, then considering the
conditions of a
mi
j
b
nj
, a
k
0, k
/
m i
j
, where i
j
< m, i
j
Z, j 1, 2, . . . , n 1. Hence,
the initial value problems 4.1 and 4.2 have been changed into of the problems 1.1 and
1.2 with the order
i
Q
k
ck
t k, 4.3
where ck ft gt,
j
are distinct zeros of Px with the order
of l
1
, l
2,...,
l
j
, respectively; by applying the theories of polynomials, the Px can be rewritten in
the following form
Px a
m
_
x
1
_
l
1
_
x
2
_
l
2
_
x
j
_
l
j
, 4.5
where
j
i1
l
i
m, j, l
i
N, i 1, 2, . . . , j, Then, substituting the function Px in 4.2 and
expanding it to the sum of partial fractions, we have
Gs
j
i1
l
i
1
n0
A
i,l
i
n
_
s
v
i
_
l
i
n
, 4.6
where the coecients A
i,l
i
n
, i 1, 2, . . . , j; n 0, 1, . . . , l
i
1, are constants.
Thus, taking the inverse Laplace Transform to 4.6 and using the formulation 2.8,
we obtain explicit formulation of the function gt, dened by
gt
j
i1
l
i
1
n0
t
i1v1
l
i
n!A
i,l
i
n
E
l
i
n1
v,v
_
i
t
v
_
. 4.7
10 International Journal of Dierential Equations
According to Unser and Blu in 13, the wavelet base generated by fractional Bsplines
wavelet
, where
w
_
k0
_
_
_
2k
_
_
_
2
_
1/2
. 4.8
With the help of the theories of wavelet analyses, we know that the functions
x are an
orthogonal fractional Bsplines wavelet and dense in L
2
with > 0. Thus the solutions of
initial value problems 4.1 and 4.2 have the following form:
yt
k
ck
t k. 4.9
Then substituting yt in 4.1, we obtain the representation of the coecients ck, dened
by
ck
_
ft gt,
t k
_
. 4.10
Finally, by combining 4.9 and 4.10, the obtained function yt is the asymptotic solution
of initial value problems 4.1 and 4.2 that we are looking for.
4.2. Order of the Error Estimation
To estimate the error of asymptotic solution of initial value problems 4.1 and 4.2, we
introduce the following properties of the fractional Bspines
x
2 sin
x
2
n1
e
2nix
2ni
1
o
_
1
x
2
_
, 4.11
when x tends to .
To 4.8, the function a
k0
2k
2
is 2periodic and symmetric, and
so we can restrict its study to 0, . In particular, one has
a
_
_
_sinc
2
_
_
_
22
_
2
_
22
, 4.12
since sinc/2 is strictly decreasing over 0, see 13.
International Journal of Dierential Equations 11
Combining 4.12 and 4.8 and taking the inverse of Fourier Transform, we obtain
t k
_
2
_
1
t k. 4.13
Then according to 4.10, 4.11, and 4.13, the asymptotic solution yt can be dened by
yt
k
_
ft gt,
t k
_
t k
_
_
ft gt
_
_
L
2
k
_
_
t k
_
_
2
. 4.14
To calculate the truncated error of the asymptotic solution, let y
N
t be the truncated sum
k N corresponding to asymptotic solution 4.8, where
y
N
t
N
kN
_
ft gt,
t k
_
t k. 4.15
Thus the truncated error y
t yt y
N
t
k>N
_
ft gt,
t k
_
t k
_
_
ft gt
_
_
k>N
_
_
t k
_
_
2
,
4.16
where the function gt is bounded in and ft L
2
; with the help of 4.12, the
inequality of 4.16 can be amplied, which means that
y
t
_
2
_
22
_
_
ft gt
_
_
k>N
_
_
t k
_
_
2
. 4.17
Then substituting 4.10 in 4.16, we have
y
t
_
2
_
22
_
_
ft gt
_
_
k>N
_
_
_
_
_
2 sin
t k
2
n1
e
2inx
2ni
1
o
_
1
t k
2
__
_
_
_
_
2
.
4.18
To the right side of 4.17, we divided the representation into three parts for discussion.
Firstly, because the function gt is bounded in and ft L
2
, there exists a constant C
1
which holds
_
_
ft gt
_
_
L
2
c
_
_
ft
_
_
L
2
C
1
. 4.19
12 International Journal of Dierential Equations
Secondly, we consider the series function
_
_
_
_
_
n1
e
2inx
2ni
1
_
_
_
_
_
2
n1
_
_
_
_
_
e
2inx
2ni
1
_
_
_
_
_
2
n1
1
2n
22
, 4.20
which is convergent in ; thus there exists a constant C
2
which satises
_
_
_
_
_
n1
e
2inx
2ni
1
_
_
_
_
_
2
C
2
. 4.21
At last, we denote a constant C
3
which dened by
C
3
2
22
2
2sin
2
24
. 4.22
Then combining 4.19, 4.21, and 4.22, the inequality of 4.18 can be amplied to the
following form:
y
t C
1
C
2
2
C
3
k>N
_
_
_
_
_
1
t k
2
o
_
1
t k
2
__
_
_
_
_
2
C
1
C
2
2
C
3
k>N
_
_
_
_
_
1 C
0
t k
2
_
_
_
_
_
2
, 4.23
when t k tends to .
Thus, from the above discussion, it is evident to derive that the truncated error
converges as
k>N
C/k t
24
, where C C
1
C
2
2
C
3
1 C
0
2
. Note that we should choose
a suitable N which should be much more greater than T, where, t 0, T. Finally, we
have obtained the error order of the asymptotic truncated sum in theory.
Remark 4.3. Noting that the process of the proof of the existence and uniqueness of the
solution of the initial value problems 1.1 and 1.2 in the case of RiemannLiouville
fractional dierential operator, and considering the formula of the Laplace transform of
Caputo dierential operator, we can be easily replaced the case of RiemannLiouville
fractional dierential operator by Caputo sense with its initial values. It means the following
corollary is correct.
Corollary 4.4. Let
n
>
n1
> >
1
> 0, and
n
1, and ft L
2
, then the multiterms
fractional ordinary dierential equations
b
n
D
n
b
n1
D
n1
b
1
D
1
yt ft 4.24
with its initial values
y
k
0 0, k 0, 1, 2, . . . , ]
n
] 1, 4.25
where ]
n
] maxm m
n
, m Z, have a unique solution.
International Journal of Dierential Equations 13
5. Conclusion
In this paper, we have proved the existence and uniqueness of the solution of the dierential
equations of arbitrary positive real order. And the representation of the solution of 1.1
has been given in the process of proof. We have obtained the asymptotic solution of the
dierential equations of fractional order Q
c
D
1
,
2
X. Here, X, is a Banach space and C C0, 1, X denotes the Banach space of
all continuous functions from 0, 1 X endowed with a topology of uniform convergence
with norm dened by x sup{xt, t 0, 1}.
In Section 1, we prove a new result for linear dierential equations involving
two fractional orders. Section 2 deals with the theory of nonlinear dierential equations
with two fractional orders. We rst use the contraction mapping principle to prove the
existence and uniqueness of the solution of problem 1.1 in a Banach space. We then
employ Krasnoselskiis xed point theorem to establish another new existence result for
problem 1.1. We also give an example for the illustration of the theory established in this
paper.
A function x C with its Caputo derivative of fractional order existing on 0, 1 is a
solution of 1.1 if it satises 1.1.
Relative to 1.1, we now introduce the following linear problem:
c
D
c
D
_
_
u
0
u s
1
_ sds xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_ sds xu
_
du
_
2
1
_
t
1
.
1.3
Proof. As argued in 2, Section 5.4, the general solution of
c
D
c
D
xt t 1.4
can be written as
xt
_
t
0
t u
1
_
_
u
0
u s
1
_ sds xu
_
du
c
0
1
t
c
1
. 1.5
Using the boundary conditions for 1.2, we nd that
c
1
1
,
c
0
1
_
1
0
1 u
1
_
_
u
0
u s
1
_ sds xu
_
du
2
1
.
1.6
Substituting 1.6 in 1.5, we obtain the solution given by 1.3. This completes the proof.
Now, we state a known result due to Krasnoselskii see 20 which is needed to prove
the existence of at least one solution of 1.1.
4 International Journal of Dierential Equations
Theorem 1.2. Let M be a closed convex and nonempty subset of a Banach space X. Let A, B be the
operators such that i Ax By M whenever x, y M; ii A is compact and continuous; iii B
is a contraction mapping. Then there exists z M such that z Az Bz.
2. Existence of Solutions
Theorem 2.1. Let f : 0, 1 X X be a jointly continuous function satisfying the condition
_
_
ft, x f
_
t, y
__
_
L
_
_
x y
_
_
, t 0, 1, x, y X, 2.1
Then the boundary value problem 1.1 has a unique solution provided < 1, where
2L
_
1
_
2
1
. 2.2
Proof. Dene : C C by
xt
_
t
0
t u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
_
2
1
_
t
1
, t 0, 1.
2.3
Let us set sup
t0,1
ft, 0 M and choose
r
1
1
_
2M
_
1
_
__
_
2
_
_
2
_
_
1
_
_
_
_
, 2.4
International Journal of Dierential Equations 5
where is such that < 1. Now we show that B
r
B
r
, where B
r
{x C : x r}.
For x B
r
, we have
xt
sup
t0,1
_
_
_
_
_
_
t
0
t u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
_
2
1
_
t
1
_
_
_
_
_
sup
t0,1
_
_
t
0
t u
1
_
_
u
0
u s
1
_
__
_
fs, xs fs, 0
_
_
_
_
fs, 0
_
_
_
ds xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_
__
_
fs, xs fs, 0
_
_
_
_
fs, 0
_
_
_
ds
xu
_
du
_
__
_
2
_
_
_
_
1
_
_
_
t
_
_
1
_
_
_
sup
t0,1
_
_
t
0
t u
1
_
_
u
0
u s
1
_
_
Lxs
_
_
fs, 0
_
_
_
ds xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_
_
Lxs
_
_
fs, 0
_
_
_
ds xu
_
du
_
__
_
2
_
_
_
_
1
_
_
_
t
_
_
1
_
_
_
sup
t0,1
_
_
t
0
t u
1
_
_
u
0
u s
1
_ ds
_
Lxu
_
_
fu, 0
_
_
_
xu
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_ ds
_
Lxu
_
_
fu, 0
_
_
_
xu
_
du
_
__
_
2
_
_
_
_
1
_
_
_
t
_
_
1
_
_
_
sup
t0,1
_
t
0
t u
1
_
u
0
u s
1
dux
_
1
0
1 u
1
_
u
0
u s
1
_ dsduLx M
_
1
0
1 u
1
dux
_
_
2
_
_
2
_
_
1
_
_
6 International Journal of Dierential Equations
2Lr M
_
1
0
1 u
1
_
u
0
u s
1
_ dsdu
2r
_
1
0
1 u
1
du
__
_
2
_
_
2
_
_
1
_
_
_
2Lr M
_
1
_
_
1
0
1 u
1
u
du
2r
1
__
_
2
_
_
2
_
_
1
_
_
_
.
2.5
Using 2.2, 2.4, and the relation for Beta function B, :
B
_
1,
_
_
1
0
1 u
1
u
du
_
1
_
_
1
_ , 2.6
we nd that
xt 1 r r. 2.7
Now, for x, y C and for each t 0, 1, we obtain
_
_
xt
_
y
_
t
_
_
sup
t0,1
_
_
xt
_
y
_
t
_
_
sup
t0,1
_
_
t
0
t u
1
_
_
u
0
u s
1
_
_
_
fs, xs f
_
s, ys
__
_
ds
_
du

_
t
0
t u
1
__
_
xu ys
_
_
_
du
t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_
_
_
fs, xs f
_
s, ys
__
_
ds
_
du

_
1
0
1 u
1
_
_
xu yu
_
_
du
__
International Journal of Dierential Equations 7
sup
t0,1
_
L
_
t
0
t u
1
_
u
0
u s
1
_ dsdu
_
_
xt yt
_
_

_
t
0
t u
1
du
_
_
xt yt
_
_
_
L
_
1
0
1 u
1
_
u
0
u s
1
_ dsdu
_
_
x y
_
_

_
1
0
1 u
1
du
_
_
x y
_
_
_
_
x y
_
_
_
2L
_
1
0
1 u
1
_
u
0
u s
1
_ dsdu 2
_
1
0
1 u
1
du
_
_
_
x y
_
_
,
2.8
where
2L
_
1
_
2
1
, 2.9
which depends only on the parameters involved in the problem. As < 1, then is
a contraction. Thus, the conclusion of the theorem follows by the contraction mapping
principle. This completes the proof.
Theorem 2.2. Assume that f : 0, 1 X X is a jointly continuous function and maps bounded
subsets of 0, 1 X into relatively compact subsets of X. Furthermore, assume that
H
1
ft, x ft, y Lx y, for all t 0, 1, x, y X;
H
2
ft, x t, for all t, x 0, 1 X, and L
1
0, 1, R
.
If
_
L
_
1
_

1
_
< 1, 2.10
then the boundary value problem 1.1 has at least one solution on 0, 1.
Proof. Let us x
r
_
2
_
_
_
_
L
1
/
_
1
_
_
_
2
_
_
2
_
_
1
_
_
1 2/ 1
_
2.11
8 International Journal of Dierential Equations
and consider B
r
{x C : x r}. We dene the operators and on B
r
as
xt
_
t
0
t u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du,
xt t
_
_
1
0
1 u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
_
2
1
_
t
1
.
2.12
For x, y B
r
, we nd that
_
_
x y
_
_
_
2
_
_
_
_
L
1
_
1
_
2r
1
_
_
2
_
_
2
_
_
1
_
_
_
r. 2.13
Thus, x y B
r
. From the assumption
_
L
_
1
_

1
_
< 1, 2.14
it follows that is a contraction mapping. The continuity of f implies that the operator is
continuous. Also, is uniformly bounded on B
r
as
x
_
_
_
_
L
1
_
1
_
r
1
. 2.15
Now we prove the compactness of the operator . Setting 0, 1 B
r
, we dene
sup
t,x
ft, x f, and consequently we have
xt
1
xt
2
_
_
_
_
_
_
t
1
0
t
1
u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
_
t
2
0
t
2
u
1
_
_
u
0
u s
1
_ fs, xsds xu
_
du
_
_
_
_
_
_
1
_
_
_
_t
1
t
2
_
_
_
r
1
_
_
t
1
t
2
_
_
,
2.16
which is independent of x. Thus, is equicontinuous. Using the fact that f maps bounded
subsets into relatively compact subsets, we have that At is relatively compact in X for
every t, where A is a bounded subset of C. So is relatively compact on B
r
. Hence, by the
Arzela Ascoli theorem, is compact on B
r
. Thus all the assumptions of Theorem 1.2 are
satised and the conclusion of Theorem 1.2 implies that the boundary value problem 1.1
has at least one solution on 0, 1. This completes the proof.
International Journal of Dierential Equations 9
Example 2.3. Consider the boundary value problem
c
D
1/4
_
c
D
1/2
1
4
_
xt
1
t 3
2
x
1 x
, 0 < t < 1,
x0
1
, x1
2
.
2.17
Here, ft, x 1/t 3
2
x/1 x, 1/2, 1/4 and 1/4. Clearly ft, x
ft, y 1/9x y with L 1/9. Further,
8
273/4
1
< 1. 2.18
Thus, by Theorem 2.1, the boundary value problem 2.17 has a unique solution on 0, 1.
3. Conclusions
The existence of solutions for a Dirichlet boundary value problem involving Langevin
equation with two dierent fractional orders has been discussed. We apply the concepts of
fractional calculus together with xed point theorems to establish the existence results. First
of all, we nd the unique solution for a linear Dirichlet boundary value problem involving
Langevin equation with two dierent fractional orders, which in fact provides the platform
to prove the existence of solutions for the associated nonlinear fractional Langevin equation
with two dierent orders. Our approach is simple and is applicable to a variety of real world
problems.
Acknowledgments
The research of J. J. Nieto has been supported by Ministerio de Educacion y Ciencia
and FEDER, project MTM200761724, and by Xunta de Galicia and FEDER, project
PGIDIT06PXIB207023PR.
References
1 A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo, Theory and Applications of Fractional Dierential
Equations, vol. 204 of NorthHolland Mathematics Studies, Elsevier Science B. V., Amsterdam, The
Netherlands, 2006.
2 V. Lakshmikantham, S. Leela, and J. Vasundhara Devi, Theory of Fractional Dynamic Systems,
Cambridge Academic, Cambridge, UK, 2009.
3 I. Podlubny, Fractional Dierential Equations, Academic Press, San Diego, Calif., USA, 1999.
4 B. Ahmad and J. J. Nieto, Existence results for nonlinear boundary value problems of fractional
integrodierential equations with integral boundary conditions, Boundary Value Problems, vol. 2009,
Article ID 708576, 11 pages, 2009.
5 B. Ahmad and J. J. Nieto, Existence of solutions for nonlocal boundary value problems of higher
order nonlinear fractional dierential equations, Abstract and Applied Analysis, vol. 2009, Article ID
494720, 9 pages, 2009.
10 International Journal of Dierential Equations
6 B. Ahmad and J. J. Nieto, Existence results for a coupled system of nonlinear fractional dierential
equations with threepoint boundary conditions, Computers & Mathematics with Applications, vol. 58,
no. 9, pp. 18381843, 2009.
7 B. Ahmad and J. J. Nieto, Existence of solutions for antiperiodic boundary value problems involving
fractional dierential equations via LeraySchauder degree theory, to appear in Topological Methods
in Nonlinear Analysis.
8 B. Ahmad, Existence of solutions for irregular boundary value problems involving nonlinear
fractional dierential equations, Applied Mathematics Letters, 2009.
9 B. Ahmad and J. J. Nieto, Existence of solutions for impulsive antiperiodic boundary value problems
of fractional order, to appear in Taiwanese Journal of Mathematics.
10 Y.K. Chang and J. J. Nieto, Some new existence results for fractional dierential inclusions with
boundary conditions, Mathematical and Computer Modelling, vol. 49, no. 34, pp. 605609, 2009.
11 V. DaftardarGejji and S. Bhalekar, Boundary value problems for multiterm fractional dierential
equations, Journal of Mathematical Analysis and Applications, vol. 345, no. 2, pp. 754765, 2008.
12 R. Hilfer, Ed., Applications of Fractional Calculus in Physics, World Scientic, River Edge, NJ, USA, 2000.
13 S. Z. Rida, H. M. ElSherbiny, and A. A. M. Arafa, On the solution of the fractional nonlinear
Schr odinger equation, Physics Letters A, vol. 372, no. 5, pp. 553558, 2008.
14 A. Arara, M. Benchohra, N. Hamidi, and J. J. Nieto, Fractional order dierential equations on an
unbounded domain, Nonlinear Analysis, vol. 72, pp. 580586, 2010.
15 D. Baleanu, A. K. Golmankhaneh, and A. K. Golmankhaneh, Fractional Nambu mechanics,
International Journal of Theoretical Physics, vol. 48, no. 4, pp. 10441052, 2009.
16 M. R. Ubriaco, Entropies based on fractional calculus, Physics Letters A, vol. 373, no. 30, pp. 2516
2519, 2009.
17 W. T. Coey, Yu. P. Kalmykov, and J. T. Waldron, The Langevin Equation: With Applications to Stochastic
Problems in Physics, Chemistry and Electrical Engineering, vol. 14 of World Scientic Series in Contemporary
Chemical Physics, World Scientic, River Edge, NJ, USA, 2nd edition, 2004.
18 S. C. Lim, M. Li, and L. P. Teo, Langevin equation with two fractional orders, Physics Letters A, vol.
372, no. 42, pp. 63096320, 2008.
19 S. C. Lim and L. P. Teo, The fractional oscillator process with two indices, Journal of Physics A, vol.
42, no. 6, Article ID 065208, 34 pages, 2009.
20 D. R. Smart, Fixed Point Theorems, Cambridge University Press, Cambridge, UK, 1980.
Hindawi Publishing Corporation
International Journal of Dierential Equations
Volume 2010, Article ID 846107, 16 pages
doi:10.1155/2010/846107
Research Article
On the Selection and Meaning of Variable Order
Operators for Dynamic Modeling
Lynnette E. S. Ramirez and Carlos F. M. Coimbra
School of Engineering, University of California, P.O. Box 2039, Merced, CA 95344, USA
Correspondence should be addressed to Carlos F. M. Coimbra, ccoimbra@ucmerced.edu
Received 4 August 2009; Accepted 8 October 2009
Academic Editor: Nikolai Leonenko
Copyright q 2010 L. E. S. Ramirez and C. F. M. Coimbra. This is an open access article distributed
under the Creative Commons Attribution License, which permits unrestricted use, distribution,
and reproduction in any medium, provided the original work is properly cited.
We reviewthe application of dierential operators of noninteger order to the modeling of dynamic
systems. We compare all the denitions of Variable Order VO operators recently proposed in
literature and select the VO operator that has the desirable property of continuous transition
between integer and noninteger order derivatives. We use the selected VOoperator to connect the
meaning of functional order to the dynamic properties of a viscoelastic oscillator. We conclude that
the order of dierentiation of a single VO operator that represents the dynamics of a viscoelastic
oscillator in stationary motion is a normalized phase shift. The normalization constant is found by
taking the dierence between the order of the inertial term 2 and the order of the spring term
0 and dividing this dierence by the angular phase shift between acceleration and position in
radians , so that the normalization constant is simply 2/.
1. Introduction
The integer order dierential operators of classical calculus such as the rst or second order
derivatives are familiar to anyone who has an active interest in understanding dynamic
systems. These dierential operators are used to formulate models that accurately describe
the majority of physical phenomena and are ubiquitous in the mathematical description of
dynamic behavior. However eective these integer order dierential operators are in general,
there are more complex systems that are better characterized by dynamic behavior that lies in
between the normal integer order description. A case in point is the socalled viscoelastic
behavior, which has characteristics of both elastic order zero and viscous order one
elements. It is thus natural to assume that dierential operators of noninteger order, such
as a 0.25, 0.50, or 0.75 would provide a convenient mathematical description to analyze these
intermediate behaviors. The study of these noninteger dierential operators falls under the
general subject of what became known as Fractional Calculus, though the orders studied are
not strictly limited to rational numbers.
2 International Journal of Dierential Equations
A further generalization of the concept of noninteger order derivatives that is
applicable to more complex systems is that of a derivative of varying order. One can nd
systems where the order of dynamics associated with each element is a function of time or
frequency or position or any derivative of the position vector 13. The objective of this work
is to identify the most appropriate denition of a variableorder VO operator for modeling
dynamic systems and to assign to the order of the derivative a physical meaning that will
facilitate the understanding of its use in problems of vibration and control. First, we compare
all VO operator denitions recently proposed in literature in order to select a denition that
is better suited for modeling purposes. We then use the familiar example of viscoelastic
harmonic oscillators to connect the order of the derivative to the dynamic properties of the
oscillators.
Unlike ordinary derivatives, noninteger order derivatives are integrodierential
operators with either a powerlaw in the case of fractional derivatives or a variable
exponent for VOderivatives kernel. Thus, noninteger derivatives are nonlocal by denition
and are ideally suited for modeling systems characterized by nonlocal or memoryladen
behavior. Multiple denitions for a fractional derivative have been proposed, and there
is no straightforward geometric or physical interpretation for the meaning of a fractional
derivative, although a few interpretations have been proposed 4. The subject of Fractional
Calculus has developed rapidly, especially since in the last four decades, with quite a few
recent books dedicated exclusively to the subject see, e.g., 48. Some applications that
involve fractional derivatives include particle motion at small but nite Reynolds numbers
912, viscoelastic constitutive equations 1316, transport dynamics with anomalous
diusion 17, and fractional order control systems 4, 18.
The concept of a VO operator is a much more recent development and is less widely
known. The order of the derivative/integral is allowed to vary over the domain of interest,
thus they are suitable for modeling systems with evolving dynamics. Such systems include
deformation of viscoelastic materials 1922 and the mechanics of variable viscoelastic
oscillators 1, 3. Similar to the state of aairs in Fractional Calculus, multiple denitions
of a VO derivative have been suggested 1, 2, 19, 23, 24, each preferred for dierent reasons.
Since the kernel of the VO operators have a variableexponent, analytical solutions to VO
dierential equations VODEs are more dicult to obtain, and have not been the focus of
much attention. However, numerical solutions for the VODEs that have been formulated
with the various VO operators have been developed 13. In addition, rather than seeking
a constant or multiple constants for the order of the derivative in the model that would
accurately represent the data, in the VO case a function needs to be determined through
mathematical and/or physical arguments. Previous applications of VO derivatives have
either explicitly chosen the function 13 or found an approximation numerically 19, 20, 22.
This complication can be lessened if the functional form of the order is determined through
physical arguments as in 21, where a model based on statistical mechanics was developed
to describe the compression of a viscoelastic material.
The aim of this work is to compare the Variable Order operator denitions that have
been proposed and to select the operator with the characteristics that are critical for the
success of a dynamic model. We compare the various VO operators based on a very simple
criteria: the VO operator must return the correct fractional derivative that corresponds to
the argument of the functional order. In other words, if the argument qt is equal to,
say, 0, 0.50 or 1, then the function itself zeroorder derivative, the halfderivative, or the
rst derivative must be returned as the output of the operator. We will see that only two
denitions previously proposed satisfy this elementary requirement. Of the two operator
International Journal of Dierential Equations 3
denitions that satisfy this property, one is more ecient from the numerical standpoint,
and is therefore adopted in the remainder of this work. The appropriate operator then is used
to study the somewhat familiar problem of a harmonically forced oscillator with viscoelastic
damping. The goal of this second part of this work is to illustrate how a familiar problem in
dynamics can be used to understand the meaning of a VO operator, and to understand how
the dynamics in this familiar problem is aected by the physical parameters of the system
using a VO analysis.
The next section presents an overview of the various VO operator denitions and a
brief comparison of the VO operators applied to a harmonic and other bounded function.
Subsequent to selecting the operator, we propose a VO model for the harmonically forced
oscillator with viscoelastic damping of order p 0 < p < 1 and conduct a stationary analysis
that yields a very concrete meaning to the order of the operator.
2. Variable Order Operators
The VO operator denitions that have been proposed are either direct extensions of
the fractional calculus denitions or generalizations that arise from Laplace or Fourier
transformations. In the direct extension approach, the constant exponent in the fractional
operator is replaced with a function. For example, a VO integral is dened in 23 as
D
qt
c
ft
1
_
qt
_
_
t
c
t
qt1
fd. 2.1
When qt constant, then the thorder fractional integral is recovered. Other
denitions can be formulated by changing the form of the argument of the exponent to be
q q or q qt and considering the Gamma function under the integral sign 24, 25:
1
D
qt
0
ft
_
t
0
t
qt1
_
qt
_ fd, 2.2
2
D
qt
0
ft
_
t
0
t
q1
_
q
_ fd, 2.3
3
D
qt
0
ft
_
t
0
t
qt1
_
qt
_ fd. 2.4
In the cases above, the lower terminal is set equal to 0, and it is assumed that f0 0 for
t < 0. Since 2.3 and 2.4 involve the variable of integration within the exponent, then this
implies memory in the order, with the past states having a stronger eect on the order for
denition 2.4 24. Also, the full convolution form of 2.4 enables use of the convolution
properties to study the operator.
Similarly, a VO derivative denition can be obtained by directly substituting q qt
in the RiemannLiouville fractional derivative denition 23 valid for 0 < q < 1:
D
qt
c
ft
1
_
1 qt
_
d
dt
_
t
c
f
t
qt
d. 2.5
4 International Journal of Dierential Equations
Further denitions are obtained by taking mrstorder derivatives of the integrals dened in
2.22.4
D
qt
0
ft
d
m
dt
m
_
t
0
t
mqt,1
_
qt,
_ fd, 2.6
where m 1 < qt < m. Caputotype VO operators are dened by taking the derivative of
the function under the integrals
D
qt
0
ft
_
t
0
t
mqt,1
_
qt,
_ f
m
d, 2.7
where f
m
denotes the mth integer order derivative of ft. VO operators based on
other fractional derivative denition forms have also been proposed. Samko and Ross 23
introduce a VO operator based on the Marchaud fractional derivative:
D
qt
c
ft
ft
_
1 qt
_
t c
qt
qt
_
1 qt
_
_
t
c
ft f
t
1qt
d, 2.8
where 0 < Re qt < 1.
Using a dierent approach, Coimbra 1 begins with the Laplace transform of the
Caputo operator to obtain a VO operator denition. Treating qt as a running parameter
and inverting back into the time domain yields the following denition valid for qt < 1:
D
qt
ft
1
_
1 qt
_
_
t
0
t
qt
f
1
d
_
f0 f0
_
t
qt
_
1 qt
_ . 2.9
Extension of denition 2.9 to values of qt larger than unity is possible as long as the higher
order derivatives are dened and integrable. Although this denition began with the Laplace
transform of the Caputo operator, it is not strictly a Caputo generalization because of the
addition of the initial condition term. As a result, D
q0
f0
/
0 for any function ft as is the
the case for a Caputobased operator. Soon et al. 3 show that a properly weighted sum of
fractional order derivatives terms approximates this single term VO operator when a large
number of terms are used, which implies convergence of both the operator and the numerical
method.
Although denitions 2.8 and 2.9 were dened independently through dierent
methods, they are similar. After integrating 2.8 by parts and simplifying, we arrive at
1
_
1 qt
_
_
t
c
f
1
t
qt
d
fc
_
1 qt
_
t c
qt
. 2.10
The dierence between 2.9 and 2.10 lies in the termthat is evaluated at the lower terminal.
For situations in which the system is assumed to be in dynamical equilibrium for t < 0 such
that D
2
ft < 0 0 for any ft, denitions 2.9 and 2.10 return the same result. However,
International Journal of Dierential Equations 5
if f0 is a true constant, such that f0
f0
_
1 qt
_
d
dt
_
t
0
f
t
qt
d, 2.11
2
D
qt
0
ft
ft
_
1 qt
_
t
qt
qt
_
1 qt
_
_
t
0
ft f
t
1qt
d, 2.12
35
D
qt
0
ft
d
dt
_
_
t
0
t
qt,1
_
qt,
_ fd
_
, 2.13
68
D
qt
0
ft
_
t
0
t
qt,1
_
qt,
_ f
1
d, 2.14
9
D
qt
0
ft
1
_
1 qt
_
_
t
0
t
qt
f
1
d
_
f0 f0
_
t
qt
_
1 qt
_ , 2.15
where qt, in denitions 2.13 and 2.14 signify the three arguments: qt, qt,
qt, q, and qt, qt. Each of the above denitions is dened for real derivative
orders between 0 and 1. The value at all the lower terminals is set to 0, since we are interested
in applying the operators to physical processes not necessarily at steady state we examine
a stationary problem in the next section. As is the case with fractional derivatives, there is
no single VO derivative or integral denition that is widely considered to be the correct
denition. Samko and Ross prefer denition 2.12 because the operator retains the symmetry
on power functions that is found in the case of constant orders, that is:
D
qt
c
t c
_
1 qt
_t c
qt
2.16
for 0 < Re qt < 1 and > 1 6. Lorenzo and Hartley prefer the full convolution VO
integral denition 2.4 because it satises the index rule for certain functions 25 and is
timeinvariant 24.
The approach chosen here is to determine which operator when acting upon a
function returns the fractional derivative of the function at the corresponding time. This is
an important characteristic from the aspect of physical modeling because it signies that the
operator yields a continuous transition of all orders of dierentiation between integer orders.
Thus, a smooth transition from zeroorder dynamics to rstorder dynamics is possible. For
a qualitative comparison of the the VO operators, we look at the qt t derivative of
two bounded functions: sin2t, and erfct. The qt t derivative of both functions is
computed numerically using a product trapezoidal rule to evaluate the convolution integrals
3, 26. Plots of the t derivative of sin2t and erfct are shown in Figures 12. The 0, 0.25,
0.50, 0.75, and 1storder derivatives of the sinusoidal function are shown for comparison
since both the RiemannLiouville and Caputo fractional derivatives return the same result.
We show only the 0 and 1storder derivatives of the erfc function since the dierent
6 International Journal of Dierential Equations
4
2
0
2
4
6
D
t
s
i
n
f0
, the operator in 2.15 returns zero for all values of qt. However,
if ft is not continuous between t 0
and t 0
n
, 3.2
International Journal of Dierential Equations 9
where L
c
is a characteristic length or amplitude of the motion and
n
_
k/m is the
undamped natural frequency of the system. The dimensionless equation of motion is
D
2
x
D
p
x x
F
o
cos
_
t
_
, 3.3
where
F
0
F
0
/m
2
n
L
c
is the dimensionless amplitude of the forcing, and is the damping
ratio and is
ck
p/21
m
p/2
.
3.4
Since we are concerned only with stationary behavior, and q are not functions of
time, so we look for a relationship such as
_
p,
,
_
D
qp,
,
x
_
D
2
D
p
D
0
_
x Re
_
F
0
exp
_
i
t
__
. 3.5
The stationary VO derivative of e
i
t
is
D
qp,
,
Ae
i
_
i
_
qp,
,
Ae
i
t
,
3.6
where denition 2.15 with a lower terminal of is used since we are dealing with a
stationary problem where the initial conditions are irrelevant. Rewriting 3.6 with 3.3 as
the stationary solution expi
t yields
_
i
_
q p,
,
1
_
i
_
p
. 3.7
We now equate the real and imaginary parts of the above equation to arrive at
q
_
p,
,
_
tan
1
p
sin
_
/2p
_
_
1
2
_
p
cos
_
/2p
_
_
p,
,
_
q
_
_
1
2
_
2
_
2
2
_
1
2
_
p
cos/2p.
3.8
10 International Journal of Dierential Equations
The stationary solution for 3.5 can be written as x
t Acos
F
0
_
_
1
2
_
2
_
2
2
_
1
2
_
p
cos/2p
,
tan
1
p
sin
_
/2p
_
_
1
2
_
p
cos
_
/2p
_
.
3.9
The procedure used to obtain the functional form for q and is easily extendable to systems
that consist of multiple viscoelastic terms. For n terms, the expression for q is
q tan
1
n
k1
p
k
sin/2p
k
_
1
2
_
n
k1
p
k
cos/2p
k
. 3.10
Similarly, the expression for is
1
_
1
k1
p
k
cos
2
p
k
_
2
_
n
k1
p
k
sin
2
p
k
_
2
.
3.11
Equations 3.83.9 clearly show that q represents a scaled phase shift and is
the scaled ratio of the amplitude of the forcing to the response. Thus, for the stationary
solution of the oscillator the order of the derivative is connected to a physical quantity
that is characteristic of the system. The multiterm equation in frequency that represents the
stationary motion of a viscoelastic operator can be replaced by a single term parametric
operator in frequency, where both the order of the derivative and the scaling function
have physical meaning. A phase shift of /2 implies that the response is proportional to
the velocity and hence to the 1storder derivative, while a phase shift of implies that it is
proportional to the acceleration or to the 2ndorder derivative. The order of the VO operator
that captures the whole dynamics of the systems is thus naturally connected with the phase
shift between the response and the forcing.
Plots of q and versus
for damping orders of p 0.25, 0.50, 0.75, and 1 and various
values of the damping ratio are shown in Figures 3, 4, 5, and 6. Also shown are the maps of q
and .
The expression for q reveals regions in which the three terms of the original equation of
motion are dominant. For example, regions in which the order of the derivative is near 2 such
as systems with lower damping ratio and higher forcing frequency are primarily dominated
by the inertial term. Lower values of damping ratio and orders of viscoelastic damping reach
the asymptotic value more quickly, suggesting that the change in the dynamics and also
the phase shift is more sensitive to changes in frequency in those cases. The dependence of
International Journal of Dierential Equations 11
0
0.5
1
1.5
2
q
0 1 2 3 4 5
a
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
b
0
0.5
1
1.5
2
2.5
0 1 2 3 4 5
c
0.5
1
.5
.5
1.5
2
2.5
3 3.5
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
d
Figure 3: Plots of q and for p 0.25. a q
, versus
for 0.1, 0.5, 1, 1.5, and 2 going from light to
dark. Note that q aysmptotes to 2 as
increases. b Map of q
, versus
for 0.1, 0.5, 1, 1.5, and
2 going from light to dark. d Map of
> 1 then the behavior is switched with the systems with higher damping ratio having lower
values for the order. For all cases, when
1, then q p for any damping ratio.
A scaled behavior identical to q is shown in the plots of . The normalized amplitude
ratio reaches an asymptotic value of 1 as
increases. Thus, at higher frequencies the
12 International Journal of Dierential Equations
0
0.5
1
1.5
2
q
0 1 2 3 4 5
a
0.2
0.4
0.8
1
1.2
0.6
1.4
1.6
1.8
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
b
0
0.5
1
1.5
2
2.5
0 1 2 3 4 5
c
0.5
1
2
2.5
3
3.5
1.5
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
d
Figure 4: Plots of q and for p 0.50. a q
, versus
for 0.1, 0.5, 1, 1.5, and 2 going from light
to dark. The asymptotic behavior is similar to the case when p 0.25. b Map of q
, where lower
values are represented by darker colors. When
1, q 0.50 for any value of . c
, versus
for
0.1, 0.5, 1, 1.5, and 2 going from light to dark. d Map of
, .
amplitude ratio is proportional to
p
, and the order of the damping and the damping ratio
do not have any eect on the amplitude of the stationary motion. Similar to the case for q,
approaches the asymptotic value more quickly for small viscoelastic damping orders p and
lower damping ratios. Also for the cases with damping order p < 1, the peak amplitude
response shifts to higher frequencies for increasing damping ratio. For increasing p and
damping ratio, the peak response begins to atten out. The maps of show that as the order
of the viscoelastic damping increases, the regions where the amplitude of the response is
damped increase.
International Journal of Dierential Equations 13
0
0.5
1
1.5
2
q
0 1 2 3 4 5
a
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
b
0
0.5
1
1.5
2
0 1 2 3 4 5
c
0.5
1
1.5 2
2.5
3
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
d
Figure 5: Plots of q and for p 0.75. a q
, versus
for 0.1, 0.5, 1, 1.5, and 2 going from light
to dark. b Map of q
, versus
for 0.1, 0.5, 1, 1.5, and 2 going from light to dark. d Map of
, .
4. Conclusions
This work advances our understanding of the use of variable order VO dierential
operators in dynamics in two substantial ways. First, we compare several denitions
proposed in literature, and select the most suitable denition based on a few criteria: 1 the
VO operator must be able to return all intermediate values between 0 and 1 that correspond
to the argument of the order of dierentiation, 2 the VO operator must be eectively
evaluated numerically, and 3 all derivatives of a true constant a function that is constant
from to must be zero. The operator dened in 2.15 satises these criteria for
modeling dynamic systems. We then proceed to illustrate the meaning of a variable order
14 International Journal of Dierential Equations
0
0.5
1
1.5
2
q
0 1 2 3 4 5
a
0.2
0.4 1.4
1.6
0.6
0.8
1
1.2
1.8
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
b
0
0.5
1
1.5
2
0 1 2 3 4 5
c
0.5
1
2
2.5
1.5
0
0.5
1
1.5
2
2.5
3
n
0 0.5 1 1.5 2 2.5 3
d
Figure 6: Plots of q and for p 1, signifying the oscillator with viscous damping. a q
, versus
for
0.1, 0.5, 1, 1.5, and 2 going from light to dark. b Map of q
, versus
for 0.1, 0.5, 1, 1.5, and 2
going from light to dark. d Map of
, .
of dierentiation by analyzing a familiar dynamical problem the stationary analysis of a
viscoelastic oscillator. We determine that the order of dierentiation for a single operator
describing all dynamic elements in the stationary equation of motion mass, damping and
spring is equal to the normalized phase shift. The normalization is easily understood as the
quantity that transforms the maximum phase shift between acceleration and position
into the maximum dierence between the order of acceleration 2 and the order of position
0. The normalization constant is thus 2/, and the variable order dierentiation is seen
as being just the normalized phase shift in this problem, which gives us a straightforward
interpretation of the meaning of variable orders of dierentiation in dynamic systems.
International Journal of Dierential Equations 15
References
1 C. F. M. Coimbra, Mechanics with variableorder dierential operators, Annalen der Physik, vol. 12,
no. 1112, pp. 692703, 2003.
2 D. Ingman and J. Suzdalnitsky, Control of damping oscillations by fractional dierential operator
with timedependent order, Computer Methods in Applied Mechanics and Engineering, vol. 193, no. 52,
pp. 55855595, 2004.
3 C. M. Soon, C. F. M. Coimbra, and M. H. Kobayashi, The variable viscoelasticity oscillator, Annalen
der Physik, vol. 14, no. 6, pp. 378389, 2005.
4 I. Podlubny, Fractional Dierential Equations, vol. 198 of Mathematics in Science and Engineering,
Academic Press, San Diego, Calif, USA, 1999.
5 K. B. Oldham and J. Spanier, The Fractional Calculus: Theory and Applications of Dierentiation and
Integration to Arbitrary Order, vol. 11, Academic Press, London, UK, 1974.
6 S. G. Samko, A. A. Kilbas, and O. I. Marichev, Integraly i proizvodnye drobnogo poryadka i nekotorye ikh
prilozheniya, Nauka i Tekhnika, Minsk, Belarus, 1987.
7 K. S. Miller and B. Ross, An Introduction to the Fractional Calculus and Fractional Dierential Equations,
John Wiley & Sons, New York, NY, USA, 1993.
8 A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo, Theory and Applications of Fractional Dierential
Equations, vol. 204 of NorthHolland Mathematics Studies, Elsevier, Amsterdam, The Netherlands, 2006.
9 C. F. M. Coimbra and R. H. Rangel, General solution of the particle momentumequation in unsteady
Stokes ows, Journal of Fluid Mechanics, vol. 370, pp. 5372, 1998.
10 C. F. M. Coimbra and R. H. Rangel, Spherical particle motion in harmonic Stokes ows, AIAA
Journal, vol. 39, no. 9, pp. 16731682, 2001.
11 C. F. M. Coimbra and M. H. Kobayashi, On the viscous motion of a small particle in a rotating
cylinder, Journal of Fluid Mechanics, vol. 469, pp. 257286, 2002.
12 L. E. S. Ramirez, E. A. Lim, C. F. M. Coimbra, and M. H. Kobayashi, On the dynamics of a spherical
scaold in rotating bioreactors, Biotechnology and Bioengineering, vol. 84, no. 3, pp. 382389, 2003.
13 R. L. Bagley and P. J. Torvik, A theoretical basis for the application of fractional calculus to
viscoelasticity, Journal of Rheology, vol. 27, no. 3, pp. 201210, 1983.
14 R. L. Bagley and P. J. Torvik, On the fractional calculus model of viscoelastic behavior, Journal of
Rheology, vol. 30, no. 1, pp. 133155, 1986.
15 N. Heymans and J.C. Bauwens, Fractal rheological models and fractional dierential equations for
viscoelastic behavior, Rheologica Acta, vol. 33, no. 3, pp. 210219, 1994.
16 T. Pritz, Fiveparameter fractional derivative model for polymeric damping materials, Journal of
Sound and Vibration, vol. 265, no. 5, pp. 935952, 2003.
17 R. Metzler and J. Klafter, The random walks guide to anomalous diusion: a fractional dynamics
approach, Physics Reports, vol. 339, no. 1, pp. 177, 2000.
18 I. Podlubny, Fractionalorder systems and PI
Гораздо больше, чем просто документы.
Откройте для себя все, что может предложить Scribd, включая книги и аудиокниги от крупных издательств.
Отменить можно в любой момент.