Вы находитесь на странице: 1из 9

International Journal of Advanced Robotic Systems

Dynamic Modelling and Trajectory


Tracking of Parallel Manipulator
with Flexible Link

Regular Paper



Chen Zhengsheng
1,*
, Kong Minxiu
1
, Liu Ming
1
and You Wei
1


1 School of Mechatronics Engineering, Harbin Institute of Technology, Harbin, China
* Corresponding author E-mail: zschen88200@gmail.com

Received 23 Dec 2012; Accepted 14 Jun 2013



DOI: 10.5772/56765

2013 Zhengsheng et al.; licensee InTech. This is an open access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.


Abstract This paper mainly focuses on dynamic
modelling and realtime control for a parallel
manipulator with flexible link. The Lagrange principle
and assumed modes method (AMM) substructure
technique is presented to formulate the dynamic
modelling of a twodegreesoffreedom (DOF) parallel
manipulator with flexible links. Then, the singular
perturbation technique (SPT) is used to decompose the
nonlinear dynamic system into slow timescale and fast
timescalesubsystems.Furthermore,theSPTisemployed
to transform the differential algebraic equations (DAEs)
for kinematic constraints into explicit ordinary
differential equations (ODEs), which makes realtime
control possible. In addition, a novel composite control
scheme is presented; the computed torque control is
applied for a slow subsystem and the H

technique for
the fast subsystem, taking account of the model
uncertainty and outside disturbance. The simulation
resultsshowthecompositecontrolcaneffectivelyachieve
fastandaccuratetrackingcontrol.

Keywords Dynamic Modelling, Composite Control, Two


TimeScale,FlexibleLink
1.Introduction

In modern industry there are increasing demands for


high speed, high acceleration and high accuracy
manipulators. Compared to conventional serial
mechanisms, parallel manipulators offer better
performance in terms of accuracy, rigidity, and payload
capacity, and show more potential to deal with these
tasks. Lightweight designofmodern mechanical systems
can bring a higher operation speed, greater payloadto
arm weight ratio, smaller actuators, and lower energy
consumptionduetocomparativelylowinertia.However,
lightweight mechanisms will cause structural flexibility
under conditions of high speed and high acceleration.
Unfortunately,conventionalPIDcontroldoesnottakethe
dynamic effects into account, which causes deterioration
in tracking accuracy for the flexible structure. Therefore,
flexible dynamic modelling and advanced modelbased
control of parallel manipulators are essential for high
accuracytracking[1][2].

The dynamic modelling of flexible link manipulators has


been being investigated for more than 20 years; the
1 Chen Zhengsheng, Kong Minxiu, Liu Ming and You Wei: Dynamic Modelling
and Trajectory Tracking of Parallel Manipulator with Flexible Link
www.intechopen.com
ARTICLE
www.intechopen.com
Int. j. adv. robot. syst., 2013, Vol. 10, 328:2013
modellingoftheflexiblelinkcanbeobtainedbythelump
massmethod(LMM),theassumedmodesmethod(AMM)
and the finite element method (FEA), while the system
dynamicscanbeintegratedwiththehelpoftheLagrange
principle, theNewtonEulermethod or the Kane method
[3]. For serial manipulators, flexible dynamic modelling
has been the subject of intense research [4], whereas few
papers have been published on parallel mechanisms
because of the closed kinematic chains constraints
(CKCCs). CKCCs are usually expressed in the form of
nonlinear DAEs, which usually cannot be solved
explicitly, making flexible modelling and control very
complex[5].Theassumedmodesmethodwasappliedto
derivetheflexibledynamicmodellingofa3DOFparallel
manipulator in [6]. Dynamic modelling using FEM is
formulated only in [7] with consideration of the flexural
deformation; however, the obtained equations are too
complex for realtime control. Mills used various
methods to formulate a planar parallel manipulator and
carriedoutactivevibrationcontrolusingsmartmaterials
[810]. To date, the challenge still exists as no previous
study has given a definite description of the kinematic
chains.Moreover,researchoncontrolschemesforflexible
parallel manipulators is onlyin its first phase. Therefore,
effective dynamic modelling is necessary for realtime
control and control scheme design of a flexible parallel
manipulator.

Inthisstudy,theAMMbasedsubstructuretechniqueand
the Lagrange principle are applied to formulate the
dynamic modelling of a 2DOF flexiblelink parallel
manipulator towards the end of achieving realtime
control; then, an SPT form offlexible dynamic modelling
isdevelopedtodividethenonlineardynamicsysteminto
two subsystems for the design of a reducedorder
controller. Subsequently, the SPT is proposed to
transform the nonlinear DAE constraints into explicit
ODEs for the flexible parallel manipulator, in order to
reduce computation cost. In order to obtain accurate
trajectory tracking performance, novel composite control
concepts are developed for the two respective
subsystems; the slow subsystem controller is designed
usingthewellknowncomputedtorquecontroltechnique
for rigid motion control, while the fast subsystem
controller is designed using the H

scheme for vibration


suppression. At the end of this paper, numerical
simulationsareperformedtoinvestigatetheeffectiveness
ofthenovelcontrolscheme.

2.Flexibledynamicmodelling
2.1Structuredescription
The structure shown in Figure 1 has two degrees of
freedom, and is an example of the PRR (Prismatic
Revolute Revolute) type parallel mechanism. Two
lineardirectdrivingmotors,representedas
i
A ,canmove
along the guide rails, which are the actuated joints with
coordinates { }
i
y with respect to fixed coordinate OXY .
This manipulator can realize highspeed positioning and
manipulating in the predefined region of the XY plane.
{ }
i
w and { }
i
u represent the flexural and axial
displacementoftheflexiblelinkiwithrespecttofloating
coordinate
i i i
A X Y ,respectively; ' C C

represents the
deflection of the end effector, and { }
i
represents the
passive joint angle of link i with respect to the fixed
coordinate. It is assumed that
i
m represents the lump
mass of the driving part and
0
m the end effectors
payload.

1
2
A
F
B
F
2
u
1
w
2
w
1
y
2
y
1 1
( ) A m
2 2
( ) A m
X
Y
O
1
l
2
l
1
u
2
u
0
( ) C m
' C
1
u

(1) Payload;(2)flexiblelink;(3)and(4)linemotors
Figure1.Twodegreesoffreedomparallelmanipulator
2.2Dynamicmodellingofclosedchainflexiblemechanism
For an open chain mechanism, the flexible link dynamic
modelling of the serial manipulators can be written as
follows[11]:

1
2
f ( ) 0
M( ) + + =
f ( ) K 0
( ( ( (
( ( ( (

y y, y, q, q
y, q q
q y, y, q, q


(1)

where y is the joint position vector, q is the generalized


coordinate of deformation, ( ) M y, q represents the n n
inertial matrix,
i
f ( ) y, y, q represents the Coriolis and
centrifugal terms, K representsthe stiffness coefficient of
flexiblelinks,and istheinputforceortorque.
2 Int. j. adv. robot. syst., 2013, Vol. 10, 328:2013 www.intechopen.com
However, for the closedchain mechanisms, the parallel
manipulators haveatleast onekinematic loop.Toderive
the dynamic equation for the serial manipulators, the
twoDOF flexible parallel mechanism is cut open, as
shown in Figure 2, resulting in two separate open
mechanisms; this brings two extra dependent degrees of
freedom and constraint equations. The equations of
motionofthesystemcanbeobtainedasfollows:

0
M ( ) + + 0 = 0 +A
K 0
f( )=0
' ( ' ( ( (
(
( ( (
' '
(
( ( (
`
(
( ( (

)
1
2
2
y
f
y , q f q
q f
y, , q

(2)

where

istheconstrainedforceand ( ) 0 = f y, , q is
theconstraintfunctions.

1
F
2
F
1
y
2
y
1
O 2
O
2
A
1
Y
1
X
2
X
2
Y
1 l
2
l
1
A
1
u
2
u

Figure2.Freesystemoftheparallelmanipulator

Unlike the realopenmanipulators, the axialdeformation


consists of two parts: one comes from the bending
deformation and the other from the coupling kinematics
of the closed chains; although the former could be
ignored in the case of small deformation, the latter must
be considered because of coupling kinematics. We
assume that the flexible link is an EulerBernoulli beam
with hybrid bending and axial deformation. Since the
manipulator is constrained in the horizontal plane, the
influence of gravity on the movement is ignored. The
position vector
i
r

relative to the fix frame at an arbitrary


locationxoftheflexiblelinkcanbeexpressedas:

i i i i
i i i i i
(x+u )cos w sin
=
y +(x+u )sin +w cos
(
(

i
r , i=1,2 (3)

where w
i
and u
i
are the bending displacement and the
axialdisplacement, respectively. Then,the kineticenergy
ofthesystemcanbeobtainedas:


}

li
2 n
T T T
i i i i i0 i0 0 1m 1m
i=1 i=1
0
1 1 1
T= dx+ m + m
2 2 2
r r r r r r (4)

whereadotdenotesthederivativeinrelationtotime;
i
,
i
m ,
0
m are the linear density of the two flexible link, the
lump mass of the two driving ends, and mass of the
payload,respectively.Velocityvector
i
r isexpressedas:

i i i i i i i i i
i
i i i i i i i i i i
(u cos w sin )((x+u )sin +w cos )
=
(y +u sin +w cos )+((x+u )cos w sin )
(
(
(

r

(5)

Importantly,
0
| x=0
i i
= r r ,
1
|
1m 1
x=l = r r .Itcanbeseenfrom
theaboveequationthatnotonlytheaxialdisplacement
i
u ,
but also the axial displacement velocity
i
u is taken into
accountinthedynamicmodelling.

Thepotentialenergyofthesystemcanbeexpressedas:

2
2 l l 2 2 2
i i
2
i=1 i=1
0 0
1 w 1 u
V= EI dx+ EA dx
2 x 2 x
| | c c | |
| |
c c
\ .
\ .

} }
(6)

For realtime control, the AMMbased substructure


method was employed to discretize the flexible link. The
structural dynamic behaviour is dominated by the first
several vibration modes, and the axial natural vibration
frequency is much higher than the bending natural
frequency;thefirstvibrationmodeforaxialvibrationand
first two vibration modes for the bending vibration were
chosen.

Based on the AMM and geometry boundary,


i
w and
i
u
canbederivedasin[12].

( ) ( )
( ) ( )

2
i ij i ij
j=1
i i i i
w = x q t i=1,2
u = x t i=1,2
(7)

| |

|

\ .
`

)
j
ij i
i
j
i
i i
i
x
(x )= i=1,2,j=1,2
L
x
(x )= i=1
L
(8)

Forsimplicity,define | |
11 12 21 22 1 2
q q q q
T
= q .

Hence,theequationsofmotionforthetwoDOFflexible
manipulatorsareobtainedbyusingthefollowinggeneral
Lagrangeapproach:

T
d T (TV)
( ) =F+A
dt
c c
c c

X X

(9)

where
T
=[ ] X y, , q is the generalized coordinate vector, F
is the generalized force vector, and
T
A is the
generalizedconstraintforcevector.

Two holonomic constraints can be derived from the


geometryoftheclosedloopchain:
3 Chen Zhengsheng, Kong Minxiu, Liu Ming and You Wei: Dynamic Modelling
and Trajectory Tracking of Parallel Manipulator with Flexible Link
www.intechopen.com
| |
1 2
1 1 1 2 2
2 1 1 1 2 2 2
(y,)= h h =0
h =l cos Ll cos
h =l sin +y l sin y

)
(10)

This shows that the constraint equation can be at least


differentiated twice with respect to time; according to
ImplicitFunctionTheory(IFT),foragivenconfigurationof
the parallel manipulator, passive joint parameter can be
uniquely determined. Hence, if equation (10) is twice
differentiatedwithrespecttotime,then

canbeobtained:

2 2
2 2

y yy
1
y yy
y+ y + + =0
= ( y+ y + )
(11)

where
y
and

representpartialderivativesof y and ,
and
yy
and

represent second partial derivatives of


y and , respectively. As shown in Figure 1, the axial
displacementoftheflexiblelinkcannotbeignoredbecause
ofthedeformationcoordinationprinciple.Thedeformation
coordinationcanbederivedasbelow:

1m 1 1 1m 1
2m 2 2 2m 2
1m 1 1 1m 1
2m 2 2m 2 2
u sin( + )+w cos
=u sin( + )+w cos
u cos( + )w sin
=w sin u cos( + )

)
(12)

where
i
im i x=l
u =u| ,
i
im i x=l
w =w| ,
i
i
i
x=l
w
=
x
c
c
aretheaxial
displacement,flexuraldisplacementandflexuralanglein
payloadposition,respectively.
Because
im
u ,
im
w and
i
are all small variables
comparedtorigidmotion,fromtheTaylorserialequation
wecanobtain:

1 1 1 1 1
1 1 1 1 1
sin( + ) sin + cos
cos( + ) cos sin
~
`
~
)
(13)

Substitute Equations (7), (8) and (13) into Equation (12).


Thedeformationcoordinationisexpressed:

| || |
31 31 u c
B B =0 q q (14)

where
1 1 2 2
31
1 1 2 2
cos cos cos cos
B =
sin sin sin sin
(
(

,
1 2
32
1 2
sin sin
B =
cos cos
(
(

, | |
T
u 1 2 3 4
= q q q q q
| |
T
c 5 6
= q q q

ThroughEquation(14),theaxialmodes
c
q canbeexpressed
bytheflexuralmodes
u
q ,whereby q canberewrittenas

u
=B q q (15)

where
1
32 31
I
B=
B B
(
(

Substitute Equations (5), (6), (7), and (8) into Equation (9).
Theequationsofmotionofthesystemcanbeobtainedthus:

11 12 13
21 22 23
31 32 33
1
2
T
3
3
m m m
m m m + +
m m m
B 0
0 = 0 + B
K 0
B
B q=0
(y,)=0
(
( (
(
( (
(
( (
(
( (


(
( (
`
(
( (

(
( (

(
( (

)
1
2
3
1
1
2
y
f (y, y, q, q)
f (y, y, q, q)
q f (y, y, q, q)

q



(16)

From Equations (15) and (16), the full dynamic model of


theflexibleparallelmanipulatorcanbeexpressedthus:

11 12 13
21 22 23
T T T
31 32 33
1 1
2 2
T T
3
m m m
m m m +
B m B B m B B m B
f ( B B ) B 0
f ( B B ) + 0 = 0 + B
0 0 B KB B f ( B B )
(y,)=0
(
(
(
(
(
(
(
(


( ( ( (
`
( ( ( (
( ( ( (
( ( ( (

)
u
u u 1
u u u 1
u u
y

q
y, y, q , q
y, y, q , q q
y, y, q , q

(17)

FromEquation(17),Lagrangemultipliervector
1
canbe
expressedthus:

1
2 21 22 23
=B (m +m +m + )
1 u 2
y q f

(18)

BasedonEquation(18),thedynamicmodellingofEq.(17)
canbereducedto:

11 12
21 22
T
M M
+ +
M M
0
=
0 B KB
' ( ( (
( ( (
'

`
( (

( (

)
1 u u
u 2 u u
y f (y, y, q , q )
q f (y, y, q , q )

q


(19)

where

1 1 1 T
11 11 1 2 21 12 1 2 22 1 2
M =m B B m (m B B m )(B B )
1
12 13 1 2 23
M =m B B m
T T 1 T
21 31 32 1 2
M =B m BB m B(B B )
T
22 33
M =B m B
( )( )
( )
T
1 1 1
1 2 12 1 2 22 2 2
T
T T 1
2
32 2 2
B B m B B m B B
=
B B m B B B
(
'
(
(
(
(
'

(

1 2
1
3
f f
f
f
f

4 Int. j. adv. robot. syst., 2013, Vol. 10, 328:2013 www.intechopen.com


Therefore, flexible dynamic modelling of the parallel
manipulator can be achieved without the Lagrange
multiplier,whichwouldmakethedesignofthecontroller
process easier. The only problem is that, with the un
actuatedjointvariables
i
,and
i

,directintegrationmay
cause error accumulation. The SPTbased technique will
beintroducedinSection3.
2.3Singularperturbationdynamicmodel
When the flexure modes of each link introduce
independent variables and the degrees of freedom of
flexible manipulators are more than the driving forces,
the obvious difficulty is that the inverse dynamic and
control of the flexible manipulators is quite complex. To
solve this problem effectively, the singular perturbation
technique has been successfully used in flexible link or
flexible joint serial manipulators, dividing the dynamic
system into several subsystems in different time scales.
Based on previous work [13], this method will be
extendedtoparallelmanipulators.

Since the inertial matrix M is positive definite and


symmetrical,itsinversematrix Hcanbepartitioned,and
thedynamicEquation(19)canbewrittenasfollows:

11 12

21 22
0 H H
=
H H 0 K
' | | ( ( ( ( (
|
( ( ( ( (
|
'
\ .
1 u u
u
u 2 u u
y f (y, y, Bq , Bq )
q
q f (y, y, Bq , Bq )

(20)

where
T
K =B KB

Toobtainsingularperturbationequations,aperturbation
parameter should be introduced first. Usually the
stiffness parameter

K is comparatively large. Defining
the minimum stiffness

c
k =min(K ) as the scale factor, the
stiffness parameters can be scaled by the scale factor;
substituting

c
K=K /k

,
c
=1/K and
u
=q into Equation
(20)gives:

11 11
12 12
21 21
22 22
=H H
H H K
=H H
H H K
1
2
1
2
y f (y, y, , )
f (y, y, , )
f (y, y, , )
f (y, y, , )
(21)

To design the control system, Equation (21) should be


transformed into the state equation, = ; the state
variablesareasfollows:

1 2 1 3 4 3
x =y,x =x ,x =,x =x (22)

Putting Equation (22) into Equation (21), the state


equationscanbeobtained:

11 11
12 12
21 21
22 22
=
=H H
H H K
=
=H H
H H K
1 2
2 1
2
3 4
4 1
2
x x
x f (y, y, , )
f (y, y, , )
x x
x f (y, y, , )
f (y, y, , )
(23)

Assuming K , issettozero,theslowsubsystem
canbeobtained:

22
1 1
21 21 22
( K H H H H

=

22
11 11 12
1
12 21 21 22
1
11 12 22 21
1
11
H H H
H H (H H H )
=(H H H H )( )
=M ( )
s 1 2
s 1 2
s 1 2
s 1
s 1
f f
y f f
f f
f
f
(24)

Thefasttimescalecanbeobtainedby
f
t =t/ , =
f1
x and
=
f2
x

. In the boundary layer, =0 , , 0 =


1 2
x y C x
andthefastsubsystemcanbegiventhus:

21 21
22 22
=
d
=H H
dt
H H K

)
f1
f2
f2
f 1 1 2
2 1 2 f1
dx
x
dt
x
f (x , x , 0, 0)
f (x , x , 0, 0) x

(25)

Because the terms of centrifugal force and Coriolis force


are the quadratic form of the joint velocity, so
= =0
1 1 2 2 1 2
f (x , x , 0, 0) f (x , x , 0, 0) ,andthefastsubsystemcan
begiven:

f f
=A +B
f f f
x x (26)

where
T
=[ ]
f f1 f2
x x x
(
(

f
22
0 E
A =
H K 0
, | |
T
f 21
B = 0 H

3.SingularPerturbationBasedKinematic
ConstraintsEquationTransformation

FromEquations(10),(24)and(26),thestatefunctionof
flexibleparallelmanipulatorscanbewrittenas:

=
=
0=
1 s
f 2 f f
y g (y, , )
x g (x , , )
(y, )
(27)
These are a form of nonlinear algebraic differential
equations, which are timeconsuming for realtime
control if a conventional NewtonRaphson numerical
interactive technique is applied. In [14], the author
successfully employed SPT in the rigid parallel
5 Chen Zhengsheng, Kong Minxiu, Liu Ming and You Wei: Dynamic Modelling
and Trajectory Tracking of Parallel Manipulator with Flexible Link
www.intechopen.com
manipulator, and the result showed high accuracy and
robustness to the initial value. Based on previous work,
the SPTbased flexible parallel manipulator will be
analysed. According to the reasoning of [5], an auxiliary
variablewiswrittenas:

w= (y,) (28)

w was supposed to asymptotically approach zero and


wasdefinedasfollows:

-
1
1
w w

= (29)

where
1
isasmallpreselectedsmallpositivenumber.

For any parallel manipulator, the kinematic constraint


functions reveal the relationship between geometry and
kinematics. Except for the redundant manipulators, the
constraint equations are of the index one type, which
means that if the general coordinates are given, the un
actuating variable can be determined according to the
ImplicitFunctionTheory.DifferentiateEquation(28)and
substitute with Equation (29), and anexplicit form ODEs
canbeobtained:

-

= =

=

)

y
1
1
y
1
1
w y+ w

1
( w y)



(30)

Thus, the nonlinear DAEs have been transformed into


explicit ODEs, which greatly reduce the computational
burden and make realtime control for flexible parallel
manipulators possible. An analytical explanation of
choosing
1
willnotbediscussedhere.

4.Compositecontroldesign

Figure3presentsthestructureofthecompositecontroller
based on twotime scale of the parallel manipulator with
the two flexible links given in Equations (24) and (26).
Following the composite control strategy, the composite
controltorque canbedeterminedasin[14]:

= +
s f
(31)

where
s
is the slow subsystem control torque and
f
is
thefastsubsystemcontroltorque.

The controller for the slow subsystem can be designed


accordingtocomputedtorquecontroltechnique:

11 p v
=M [ +K ( )+K ( )]+
s d d d 1
y y y y y f (y, y, 0, 0)

(32)
where
p
K and
v
K are diagonal position and velocity gain
matrices of the slow subsystem control, respectively, and
d
y representsthedesiredtrajectoriesofthetwolinks.

SubstituteEquation(32)intoEquation(24),andweobtain:

v p
+K +K =0 e e e (33)

where =
d
e y y

Byselectingappropriate
p
K and
v
K ,therigidmotionof
thesystemcanbeexponentiallystable.

d
y
,
d d
y y
+

s
t +

f
t
t
, y y
, y y
, q q

Figure3.Compositecontrolscheme

Considering the unmodelled highorder modes of the


flexible links and external disturbance, an H

based
controllerwillbedesignedforthefastsubsystem.

The design of the fast subsystem aims to suppress the


flexible vibration effectively during trajectory tracking in
the presence of unmodelled highorder modes and
disturbance.Inthispaper,arobustcontrollerisproposed
for the fast subsystem based on the H

strategy. Defining
thestrainrateY oftheflexiblelinkas:

2 2
2 2
T
1 2
1 2
w w
Y
x x
( c c
=
(
c c

(34)

SubstitutingEquation(7)intoEquation(34),wecanobtain:

=C
f
Y x (35)

where C is the constant matrix related to the position of


thestraingaugeandtheshapefunctions.

From Equation (26) and Equation (35), the transfer


functionofthefastsystemcanbeexpressedas:

( ) ( )
1
2
22 21
G s =C s IH K H

(36)

Definingtransferfunctionofthelimittorquemodestobe
( )
0
G s ,then:

( ) ( ) ( )
m 0
G s = I+ G s (37)

6 Int. j. adv. robot. syst., 2013, Vol. 10, 328:2013 www.intechopen.com


where
m
isthemultiplicativeperturbation,whichsatisfies
thefollowingequation:

( ) ( ) | ) 0,
m 2
j < W j ( e

(38)

where denotesthemaximumsingularvalue.

The fast subsystem controller is robust to the un


modelled uncertainty and the external disturbances,
whichmeansthat ( ) K s satisfiesthefollowingequation:

( ) ( )
( ) ( )
1
2
W s S s
1
W s T s

(
< (
(

(39)

where ( )
1
W s , ( )
2
W s , ( ) S s and ( ) T s are the low pass
filter, high pass filter, transfer function between
disturbance and output, and transfer function between
disturbanceandcontrolinput,respectively.

Equation(39)canbeappliedtothestandard H

problem,
andthecontroller ( ) K s canthenbeobtained[15].

5.SimulationsandDiscussions

Simulations will be carried out for the twoflexiblelink


parallel manipulator to compare the performances of the
proposedcompositecontrolandthetraditionalcomputed
torquecontrol(CTC)techniques.Relevantparametersare
given in Table 1. The manipulator was commanded to
trackthesmoothdesiredtrajectory,givenbelow:

d 0
d 0
3 2
3 2
X =X +Rsin( )
Y =
4t +6t
4t +6t Y +Rcos( )

)
(40)

where ( )
d d
X ,Y and ( )
0 0
X ,Y are the desired tip point
position and initial position relative to the fixed
coordinate frame, respectively, and R is the radius of the
desired circle. The workspace of a twoDOF parallel
manipulator can be decomposed into four parts [16], but
hereweonlyconsiderthepartwhen
i

0< <
2
.Thedesired
jointposition
id
y canbeobtainedasfollows:

d
1d d 1
1
d
2d d 1
2
X
y =Y l sin(arccos )
l
LX
y =Y l sin(arccos )
l

)
(41)

Parameter Symbol Value


FlexuralRigidity EI
3 2
2.67 10 N m
Length
i
L 140mm
Payload
o
m 0.5kg
Lumpmass
i
m 1.9637kg

Table1.Parametersoftheflexiblemanipulator

Figure4.Trackingperformanceofthedesiredcircle

Figure5.TrackingperformanceofXdirection

Figure6.TrackingperformanceofYdirection

The trajectories of the proposed composite control


scheme and the computed torque control are shown in
Figure4,andthetrackingperformancesoftheXdirection
and Y direction are given in Figure 5 and Figure 6,
respectively. The blue curves refer to the desired
trajectory,theredcurvesrefertotheCTCcontrol,andthe
black curves refer to composite control. Figure 7 and
Figure 9 show the tracking errors of the two control
7 Chen Zhengsheng, Kong Minxiu, Liu Ming and You Wei: Dynamic Modelling
and Trajectory Tracking of Parallel Manipulator with Flexible Link
www.intechopen.com
schemes in X direction and Y direction, respectively. The
simulations show that the proposed composite controller
can improve the tracking accuracy during the whole
motion,andcanclearlydecreasethemaximumerror.

Figure7.TrackingerrorofXdirection

Figure8.TrackingerrorofYdirection

To evaluate the performances of the two controllers, the


roots square mean error (RSME) of the endeffector
position is introduced as the performance index, and the
RSMEisdefinedby:



N
2 2
d d
i=1
1
RSME= x i x i + y i y i
N

(42)

where
d
x i and
d
y i represent the X direction and Y
directionpositioncoordinatesatthe ith samplingtimeof
the desired trajectory, respectively; x i and y i
represent the X direction and Y direction position
coordinates at the ith sampling time of the actual
trajectory,respectively.

Controllers RSME
Compositecontroller 0.0158mm
CTC 0.0715mm

Table2.RSMEofcompositecontrollerandtheCTC
The RSME of the trajectory tracking simulation using
composite controller and CTC are given in Table 2.
AccordingtoTable2,theproposedcompositecontrolcan
decreasetheRSMEsignificantlycomparedwiththeCTC.

6.Conclusion

Generalized dynamic modelling has been established for


the twoflexiblelink parallel manipulator by using the
LagrangeEuler principle, the assumed mode method
based substructure technique. In contrast to previous
research on flexible parallel manipulators modelling, the
proposed model found a successful tradeoff between
model accuracy and computation cost for the purpose of
realtime control. Axial deformation was taken into
account, which was very important for the parallel
manipulator in coupling deformation. The singular
perturbation technique was applied to decompose the
model into a slow subsystem and a fast subsystem; in
doing so, the design of the controller becomes much
easier. The proposed singular perturbation technique for
the transformation of the DAEs constraint into explicit
ODEs for the parallel manipulator greatly reduced the
computation cost of kinematics and made realtime
control possible. Moreover, a novel composite control
scheme for the two subsystems was presented, and
simulation showed that the proposed composite control
scheme demonstrates better trajectory tracking
performance in comparison to traditional computed
torquecontrol.

7.Acknowledgement

ThisworkwassupportedbytheNationalNaturalScience
FoundationofChina(GrantNo.51075086).

8.Reference

[1] MerletJP.ParallelRobots.Springer,Dordrecht,2006.
[2] Pietsch IT, Krefft M, Becker OT. How to reach the
dynamic limits of parallel robots? An autonomous
control approach. IEEE Transactions on Automation
ScienceandEngineering,2005,2(4):369380.
[3] Theodore RJ, Ghosal A. Comparison of the assumed
modes and the finite element models for flexible
multilink manipulator. International Journal of
RoboticsResearch,1995,14(2):91111.
[4] Dwivedy SK, Eberhard P. Dynamic Analysis of
Flexible Manipulators: A Literature Review.
Mechanism and Machine Theory, 2006, 41(7): 749
777.
[5] Gordon BW and Liu S. A singular perturbation
approach for modelling differentialalgebraic
systems. Journal of dynamic systems measurement
and control. Transactions of the ASME, 1998, 120(4):
541545.
8 Int. j. adv. robot. syst., 2013, Vol. 10, 328:2013 www.intechopen.com
[6] Luo L, Wang S. On the modeling and composite
control of flexible parallel mechanism. International
Journal Advanced Manufacturing Technology, 2006,
29:78793.
[7] DuZ,YuY.Dynamicmodelingandinversedynamic
analysis of flexible parallel robots. Int. J. Adv. Rob.
Syst.,2008,5(1):115122.
[8] Wang X, Mills JK. Dynamic modeling of a flexible
link planar parallel platform using a substructuring
approach. Mechanism and Machine Theory, 2006,
41(6):671687.
[9] Piras G, Cleghorn WL, Mills JK. Dynamic finite
element analysis of a planar highspeed, high
precision parallel manipulator with flexible links.
Mechanism and Machine Theory, 2005, 40(7): 849
862.
[10] ZhangX,MillsJK,CleghornWL.DynamicModeling
and Experimental Validation of a 3PRR Parallel
Manipulator with Flexible Intermediate Links.
Journal of Intelligent and Robotic Systems, 2007,
50(4):323340.

[11] Subudhi B, Morris AS. Dynamic modeling,


simulationandcontrolofamanipulatorwithflexible
links and joints. Robotics and Autonomous Systems,
2002,41(4):257270.
[12] Ginsberg JH. Mechanical and Structural Vibrations:
TheoryandApplications.Wiley&SonsInc.,2001
[13] Siciliano B, Khatib O. Handbook of Robotics.
Springer,VerlagBerlinHeidelberg,2008.
[14] Dabney JB, Ghorbel FH. Modeling closed kinematic
chainsviasingularperturbations.Proceedingsofthe
AmericanControlConference,2002,5:41044110.
[15] Dorato P, Fotuna L, Muscato G. Robust Control for
Unstructured Perturbations: An Introduction.
SpringerVerlag,USA,1992.
[16] Wang Z, Ghorbel FH, Dabney JB. On the Domain
and Error Characterization in the Singular
Perturbation Modeling of Closed Kinematic Chains.
Proceedings of American Control Conference, 2004,
1:493498.

9 Chen Zhengsheng, Kong Minxiu, Liu Ming and You Wei: Dynamic Modelling
and Trajectory Tracking of Parallel Manipulator with Flexible Link
www.intechopen.com