Automatic generation control application with craziness based particle swarm

optimization in a thermal power system

Haluk Gozde

, M. Cengiz Taplamacioglu
Gazi University, Faculty of Engineering, Department of Electrical & Electronics Engineering, 06750 Maltepe, Ankara, Turkey
a r t i c l e i n f o
Article history:
Received 22 March 2009
Received in revised form 13 June 2010
Accepted 13 August 2010
Keywords:
Automatic generation control
CRAZYPSO
Thermal power system
PI-controller
a b s t r a c t
In this study, a novel gain scheduling Proportional-plus-Integral (PI) control strategy is suggested for
automatic generation control (AGC) of the two area thermal power system with governor dead-band non-
linearity. In this strategy, the control is evaluated as an optimization problem, and two different cost
functions with tuned weight coefcients are derived in order to increase the performance of convergence
to the global optima. One of the cost functions is derived through the frequency deviations of the control
areas and tie-line power changes. On the other hand, the other one includes the rate of changes which can
be variable depends on the time in these deviations. These weight coefcients of the cost functions are
also optimized as the controller gains have been done. The craziness based particle swarm optimization
(CRAZYPSO) algorithm is preferred to optimize the parameters, because of convergence superiority. At
the end of the study, the performance of the control system is compared with the performance which
is obtained with classical integral of the squared error (ISE) and the integral of time weighted squared
error (ITSE) cost functions through transient response analysis method. The results show that the
obtained optimal PI-controller improves the dynamic performance of the power system as expected as
mentioned in literature.
2010 Elsevier Ltd. All rights reserved.
1. Introduction
An interconnected electric power system generates, transports
and distributes electric energy. The aim of such these systems is
to supply electric energy with nominal system frequency and ter-
minal voltage, values and tolerances of those are dened by some
power quality standards. According to power system control the-
ory, a nominal system frequency depends on the balance between
generated and consumed real powers [1]. The difference between
generated power and instant load demand causes changing of
nominal system frequency at the normal state. If the amount of
generated power is less than the demanded amount, speed and fre-
quency of the generator units begin to decrease, and vice versa.
Hence, the amount of production of the synchronous generators
is made sense for frequency deviations occurred in the power sys-
tem in order to maintain that balance. For this purpose, an auto-
matic generation control concept is used. The aim of automatic
generation control is that the steady state error of the system fre-
quency deviations following a step load demand is made zero
error.
When the literature is investigated, it can be seen that early
works on AGC was initiated by Cohn [2]. However, a modern opti-
mal control concept for AGC designs of interconnected systems is
put forward by Elgerd and Fosha for the rst time [3]. They sug-
gested a proportional controller and different feedback form to de-
velop optimal controller. Until the present day, lots of different
control strategies such as conventional, adaptive, variable struc-
ture, robust and some based on articial intelligence have been re-
ported [4]. However, gain scheduling adaptive control can be
distinguished from the other control techniques because it makes
the process which is under control less sensitive to changes in pro-
cess parameters and in particular, it is also simpler to implement
than the other modern control techniques. For these reasons, it is
carried out to AGC system, frequently.
The rst gain scheduling control method for AGC of intercon-
nected power system was proposed by Lee and coworkers in
1991 [5]. Their controller provided better control performance
for a wide range of operating conditions than the performances ob-
tained so far. Later on, Rubaai and Udo presented a multi-variable
gain scheduling controller by dening a cost function with a term
representing the constraints on the control effort and then mini-
mizing that with respect to the control vector [6]. Since the con-
ventional gain scheduling methods may be unsuitable in some
operating conditions due to the complexity of the power systems
such as nonlinear load characteristics and variable operating
0142-0615/$ - see front matter 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijepes.2010.08.010

Corresponding author. Tel.: +90 312 2311340; fax: +90 312 2308434.
E-mail addresses: halukgozde@gmail.com (H. Gozde), taplam@gazi.edu.tr (M.C.
Taplamacioglu).
Electrical Power and Energy Systems 33 (2011) 816
Contents lists available at ScienceDirect
Electrical Power and Energy Systems
j our nal homepage: www. el sevi er . com/ l ocat e/ i j epes
points, a usage of articial intelligence based methods were pre-
ferred by researchers from the beginning of these dates. In 1997,
Chang reported the usage of fuzzy logic based gain scheduling
method for power system AGC [7]. As different from the usage of
two fuzzy rules for integral and proportional gains in PI-controller
by Chang, am and Kocaarslan who improved the performance of
this approach in 2005 by which the rules for the gains are chosen
to be identical [8]. Talaq suggested an adaptive fuzzy gain schedul-
ing method for conventional PI-controller in 1999 [9]. Then, Pingk-
ang optimized the gains of PI and PID controllers through real
coded genetic algorithm in a two area power system in 2002
[10]. After 1 year from Pinkangs study, Abdel-Magid and Abido
proposed a usage of PSO for the same purpose [11]. In 2004, Yes il
suggested the self tuning fuzzy PID type controller for AGC [12].
One year after, Juang put forward the genetic algorithm based fuz-
zy gain scheduling method for PI-controller [13]. In this study, to
reduce both the fuzzy system design effort and the number of fuz-
zy rules, the fuzzy system used for gain scheduling is automatically
designed by genetic algorithms. In 2006, Massiala used two layered
fuzzy gain scheduling controller in order to improve the dynamic
performance of AGC in a two area reheat thermal power system
with generation rate constraints [14]. Taher composed a hybrid
PSO algorithm for the gain scheduling of PI-controller in a two area
thermal power system in 2008 [15]. The hybrid PSO algorithm con-
tained evolutionary operators like selection, crossover and muta-
tion as in genetic algorithms or DE algorithm. In the year 2009,
Nanda suggested a maiden application of bacterial foraging optimi-
zation based gain scheduling PI-controller in a multi area AGC [16].
Rao and coworkers studied an automatic generation control on
TCPS based hydrothermal system [17].
The aim of this study which is different from the above litera-
ture is that a novel gain scheduling PI-control strategy is proposed
for automatic generation control (AGC) of a two area thermal
power system with governor dead-band nonlinearity. In this strat-
egy, the control is evaluated as an optimization problem, and two
different cost functions with tuned weight coefcients are derived
to increase the performance of convergence to the global optima.
The weight coefcients of the cost functions are also optimized
as the controller gains have been done. Then, the craziness based
particle swarm optimization (CRAZYPSO) algorithm is preferred
in order to optimize the parameters, because of its convergence
superiority and also its relatively simple codes. At the result of this
application, it is seen that the response of AGC of the power system
is improved.
2. Materials and methods
2.1. Power system model
A two area interconnected thermal power system is considered
for application of optimal automatic generation control, because of
being the simplest model of interconnected power system. A trans-
fer function model of this system is depicted in Fig. 1, and system
parameters are also given in Table 1. At the simulation study, it is
assumed that there is a step load changing in the control area-1.
In the above model, u
1
and u
2
are the control inputs from pro-
posed gain scheduling PI-controllers. DP
L1
is 1% of step load pertur-
bations of nominal loading in the control area-1. Df
1
and Df
2
are
frequency deviations of each control area, and DP
tie
is the changing
of tie-line power between control areas.
In the model, a governor dead-band effect is also added to all
control areas to simulate nonlinearity. A governor dead-band is de-
ned as the total magnitude of a sustained speed change where
there is no change in valve position of the turbine. Describing func-
tion approach is used to represent the governor dead-band in the
areas. The governor dead-band nonlinearity tends to produce a
continuous sinusoidal oscillation of natural period of about
T
0
= 2 s. This approach is being used to linearize the governor
dead-band in terms of change and rate of change in the speed
[18]. The nonlinearity of the hysteresis is dened as,
y Fx; _ x 1
In this function, x is taken as a sinusoidal oscillation with
f
0
= 0.5 Hz.
x Asinw
0
t 2
Since the dead-band nonlinearity tends to give continuously
sinusoidal oscillation, such an assumption is quite realistic. Then,
the F function can be evaluated as a Fourier series as follows,
Fx; _ x F
0
N
1
x
N
2
w
0
dx=dt 3
For an approximation, it is enough to consider the rst three
terms in (3). As the dead-band nonlinearity is symmetrical about
the origin, and then F
0
is equal to zero,
Fx; _ x N
1
x
N
2
w
0
dx=dt DBx 4
where DB denotes the dead-band. In this work, the backlash of
approximately 0.5% is chosen and the Fourier coefcients are ob-
tained as N
1
= 0.8 and N
2
= 0.2. At the result of the analysis, the
transfer function of the governor with dead-band nonlinearity can
be expressed in (5) [19]. It is used as this form in the power system
model.
G
g

0:8
0:2
p
s
1 T
g
s
5
2.2. Gain scheduling control
A gain scheduling control is an adaptive control technique that
changes some control parameters of the controller according to
scheduling variables related to different operating regions which
the plant works. This control technique deals with particularly
nonlinear processes, processes with time variations or situations
where the requirements on the control that change with the oper-
ating conditions. The main advantage of this control is that the
Nomenclature
K
i
integral gain constant
K
p
proportional gain constant
w
i
cost function weight coefcient
R
i
regulation constant
T
g
speed governor time constant
T
t
turbine time constant
T
p
power system time constant
Df
i
frequency deviation from nominal value in area-i
DP
tie
change in tie-line power between two area
DP
Li
load demand increment
B
i
frequency bias factor
T
12
synchronization coefcient
H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816 9
controller parameters can be adjusted very quickly in response to
changes in the plant dynamics. It is also simpler to implement than
the other adaptive control techniques. A typically block diagram of
gain scheduling control system is depicted in Fig. 2.
In this gure, the scheduling variables can be the measured sig-
nal, the control signal or an external signal. The control parameters
of the controller are determined with a gain scheduling algorithm.
The algorithm is run for all different operating conditions by auto-
matic tuning. However the classical tuning methods have been
used as a gain scheduling algorithm, nowadays the swarm intelli-
gence based methods are carried out increasingly due to their sim-
pler implementing and better performance of converging and less
run times [20]. As an optimization algorithm based on swarm
intelligence, craziness based particle swarm optimization method
is preferred in order to its superiority in respect of the standard
PSO and the other algorithms, and also its short and simple codes.
From this point of view, a microprocessor program of this algo-
rithm can be easily prepared in reality. Nowadays, this algorithm
is applied to the lots of power system control applications in order
to tune the parameters of the controller or the system such as eco-
nomic dispatch control [21].
2.3. Craziness based particle swarm optimization
Particle swarm optimization is a population based optimization
algorithm which is rst introduced by Kennedy and Eberhart in
1995 [22]. It can be obtained high quality solutions within shorter
calculation time and stable convergence characteristics by PSO
than other stochastic methods such as genetic algorithm.
PSO uses particles which represent potential solutions of the
problem. Each particles y in search space at a certain velocity
which can be adjusted in light of preceding ight experiences.
The projected position of ith particle of the swarm x
i
, and the veloc-
ity of this particle v
i
at (t + 1)th iteration are dened and updated
as the following two equations:
v
t1
i
v
t
i
c
1
r
1
p
t
i
x
t
i
c
2
r
2
g
t
i
x
t
i
6
x
t1
i
x
t
i
v
t1
i
7
where i = 1, , n and n is the size of the swarm, c
1
and c
2
are posi-
tive constants, r
1
and r
2
are random numbers which are uniformly
distributed in [0, 1], t determines the iteration number, p
i
repre-
sents the best previous position (the position giving the best tness
value) of the ith particle, and g represents the best particle among
all the particles in the swarm. At the end of the iterations, the best
position of the swarm will be the solution of the problem. It cannot
be always possible to get an optimum result of the problem, but the
obtained solution will be an optimal one. The owchart of PSO is
depicted in Fig. 3.
Since the standard PSO algorithm can fall into premature con-
vergence especially for complex problems with many local optima
and optimization parameters, the craziness based PSO algorithm
which is particularly effective in nding out the global optimum
in very complex search spaces is developed. The main difference
between PSO and CRAZYPSO is the propagation mechanism to
determine new velocity for a particle as follows:
v
t1
i
r
2
signr
3
v
t
i
1 r
3
c
1
r
1
p
t
i
x
t
i

1 r
2
c
2
1 r
1
g
t
x
t
i
8
x
t1
i
x
t
i
v
t1
i
Pr
4
signr
4
V
cr
9
where p
i
is the local best position of particle i, and g
i
is the global
best position of the whole swarm. r
1
, r
2
, r
3
and r
4
are random
Fig. 1. Block diagram of the two area interconnected thermal power system with governor dead-band nonlinearity.
Table 1
System parameters.
Parameter Quantity
T
g1,2
0.2 s
T
t1,2
0.3 s
K
p1,2
120 Hz/puMW
T
p1,2
20 s
T
12
0.0707 MW/rad
B
1,2
0.425 puMW/Hz
R
1,2
2.4 Hz/puMW
Fig. 2. Block diagram of a gain scheduling control system.
10 H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816
parameters distributed uniformly in [0, 1], and c
1
, c
2
are named step
constants and are taken 2.05 generally. The sign is a function de-
ned as follows for r
3
and r
4
,
signr
3

1 (r
3
6 0:05
1 (r
3
> 0:05

10
signr
4

1 (r
4
6 0:5
1 (r
4
> 0:5

11
In birds ocking or sh schooling, since a bird or a sh often
changes directions suddenly, in the position updating formula, a
craziness factor, V
cr
, is used to describing this behavior. In this
study, it is decreased linearly from 10 to 1. P(r
4
) is dened as
Pr
4

1 (r
4
6 P
cr
0 (r
4
> P
cr

12
where P
cr
is a predened probability of craziness and is introduced
to maintain the diversity of the particles. It is taken 0.3 in this study.
The CRAZYPSO algorithm can prevent the swarm from being
trapped in local minimum, which would cause a premature conver-
gence and lead to fail in nding the global optimum [23].
3. Control strategy
As a control strategy, the new control conguration which is de-
picted in Fig. 4 is suggested. In this conguration, to achieve the
control inputs, the optimal PI-controllers are used together with
area control errors, ACE
1
and ACE
2
, in (13) and (14) respectively.
ACE
1
B
1
Df
1
DP
tie1
13
ACE
2
B
2
Df
2
DP
tie2
14
In the control strategy, control inputs of the power system, u
1
and u
2
, are obtained with PI-controllers as below.
u
1
K
p1
ACE
1
K
i1
Z
ACE
1
dt 15
u
2
K
p2
ACE
2
K
i2
Z
ACE
2
dt 16
The object of the obtaining optimal solutions of control inputs is
taken as an optimization problem, and CRAZYPSO algorithm is
being used to tune the gains of the controllers and cost function
weights w
1
, w
2
and w
3
as a novel control approach. In AGC system,
in order to convergence to optimal solution, two different cost
functions in (17) and (18) are derived. While one of the cost func-
tions is derived through the frequency deviations of the control
areas and tie-line power changes, the rates of changes in these
deviations according to time are used in the other one.
J
Z
t
0
t w
1
dDf
1
dt

2
w
2
dDf
2
dt

2
w
3
dDP
tie
dt

2
" #
dt 17
J
Z
t
0
t w
1
Df
1

2
w
2
Df
2

2
w
3
DP
tie

2
h i
dt 18
Initialize the random
velocities and
positions of particles
Evaluate the fitness values
for each particles
Compare particles
fitness with its best
previous one
Best previous position
equal to the current value
Best previous position
equal to itself
Current
value is
better
Previous
value is
better
Best of the best position is
appointed to the global
best
Change velocities and positions
according to (6) and (7)
Is the criterion met
or end of iterations?
Yes No
Stop
Global best position is the
solution
Fig. 3. The owchart of the standard PSO algorithm.
Fig. 4. The novel control strategy of AGC.
Table 2
Tuned parameters with ISE, ITSE and proposed cost functions.
Cost function K
p
K
i
w
1
w
2
w
3
Proposed 1 0.5762 0.1962 0.7300 0.6848 0.7879
Proposed 2 0.4000 0.3000 0.5000 0.5000 0.5001
ITSE 0.2647 0.3317
ISE 0.2514 0.2491
Table 3
Settling times with ISE, ITSE and proposed cost functions.
Cost function Df
1
(s) Df
2
(s) DP
tie
(s)
Proposed 1 9.65 10.98 13.16
Proposed 2 11.64 12.42 17.55
ITSE 21.66 21.67 30.80
ISE 20.45 20.47 29.45
H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816 11
4. Results and discussion
The results are obtained by MATLAB 6.5 software run on Core2
of 2 GHz, and RAM of 1 GB. Forty particles are used, and 100
iterations are chosen for converging to solution in the craziness
based PSO algorithm. The simulations are realized in case of
DP
L1
= 0.01 puMW. At the end of the simulations, the tuned para-
meters of the control system are shown in Table 2, and the settling
5 10 15 20 25
-0.035
-0.03
-0.025
-0.02
-0.015
-0.01
-0.005
0
0.005
Time (s)
d
e
l
t
a

f
1

(
H
z
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
Fig. 5. The deviations of Df
1
with ISE, ITSE and proposed cost functions in case of DP
L1
= 0.01 puMW.
6 8 10 12 14 16 18 20 22 24
-1.5
-1
-0.5
0
0.5
1
1.5
x 10
-3
Time (s)
d
e
l
t
a

f
1

(
H
z
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
9.65 s 11.64 s 20.45 s 21.66 s
Fig. 6. The zoom of deviations of Df
1
with ISE, ITSE and proposed cost functions.
5 10 15 20 25
-0.04
-0.035
-0.03
-0.025
-0.02
-0.015
-0.01
-0.005
0
Time (s)
d
e
l
t
a

f
2

(
H
z
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
Fig. 7. The deviations of Df
2
with ISE, ITSE and proposed cost functions in case of DP
L1
= 0.01 puMW.
12 H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816
5 10 15 20 25
-1.5
-1
-0.5
0
0.5
1
x 10
-3
Time (s)
d
e
l
t
a

f
2

(
H
z
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
10.98 s 12.42 s 20.47 s 21.67 s
Fig. 8. The zoom of deviations of Df
2
with ISE, ITSE and proposed cost functions.
5 10 15 20 25 30
-10
-8
-6
-4
-2
0
x 10
-3
Time (s)
d
e
l
t
a

P
t
i
e

(
p
u
M
W
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
Fig. 9. The deviations of DP
tie
with ISE, ITSE and proposed cost functions in case of DP
L1
= 0.01 puMW.
15 20 25 30 35 40
-1.5
-1
-0.5
0
0.5
1
1.5
x 10
-4
Time (s)
d
e
l
t
a

P
t
i
e

(
p
u
M
W
)
Proposed cost function 1
Proposed cost function 2
ITSE cost function
ISE cost function
13.16 s 17.55 s 29.45 s 30.80 s
Fig. 10. The zoom of deviations of DP
tie
with ISE, ITSE and proposed cost functions.
H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816 13
times of the frequency and tie-line power deviations are repre-
sented in Table 3. These results are compared to the results ob-
tained with the integral of the squared error (ISE) and the
integral of time weighted squared error (ITSE) cost functions in
(19) and (20) respectively.
ISE
Z
t
0
ACE
i

2
dt 19
ITSE
Z
t
0
t ACE
i

2
dt 20
The results show that the proposed control conguration with
tuned cost function achieves good dynamic performance for the
power system. Especially, the proposed cost functions expose bet-
ter solutions than the standard cost functions. In addition to this,
the proposed cost function-1 depicted in (17) with the rates of
changes in the deviations shows the best solutions for the power
system. It can be clearly seen that the CRAZYPSO based gain sched-
uling PI-controller with this cost function improved to control of
the power system in order to minimize the frequency and tie-line
power deviations. These deviations and their zooms determining
the settling times are also depicted in Figs. 510.
These gures and Table 3 show that the settling time of Df
1
obtaining with proposed cost function-1 is better than that of Df
1
obtaining with proposed cost function-2, ISE and ITSE of 17.09%,
0
5
10
15
20
25
30
35
f1 f2 Ptie
Proposed 1
Proposed 2
ITSE
ISE
Fig. 11. The comparison of settling times according to cost functions.
0 5 10 15 20 25
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
Time (s)
d
e
l
t
a

f
1

(
H
z
)
nominal load
+25% load
+50% load
- 25% load
- 50% load
Fig. 12. Curves of Df
1
for different changes of DP
L1
.
0 5 10 15 20 25
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
Time (s)
d
e
l
t
a

f
2

(
H
z
)
nominal load
+25% load
+50% load
- 25% load
- 50% load
Fig. 13. Curves of Df
2
for different changes of DP
L1
.
14 H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816
55.44% and 52.81%, respectively. And then, the settling time of Df
2
obtaining with proposed cost function-1 is better than that of Df
2
obtaining with proposed cost function-2, ISE and ITSE of 11.59%,
49.33% and 46.36%, respectively. In addition to these, the settling
time of DP
tie
obtaining with proposed cost function-1 is better than
that of DP
tie
obtaining with proposed cost function-2, ISE and ITSE
of 25.01%, 57.27% and 55.31%, respectively. Furthermore, the com-
parison of settling times according to cost functions is depicted in
Fig. 11. According to this gure, the power change between control
areas is minimized by the proposed control strategy faster than the
other control strategies.
On the other hand, some different values of DP
L1
are applied to
the power system in order to evidence the robustness of the con-
trol strategy optimized by CRAZYPSO algorithm. For this purpose,
a step load change of the control area-1 is decreased the step of
25% and 50%, and increased the step of 25% and 50%, respectively.
In this case, the proposed cost function-1 is used due to its conver-
gence superiority. As a result, the obtained frequency deviation
curves of these cases are depicted in Figs. 1214. The tuned
parameters are shown in Table 4, and the maximum overshoots
of these curves are shown in Table 5, respectively. It can be seen
from these gures and tables that the deviation of the frequencies
and the tie-line power change are obtained less than half according
to nominal values. Then, the proposed control strategy provides a
robust control in the range of 50% of the step load change,
sufciently.
5. Conclusion
In this article, the new gain scheduling PI-controller strategy is
proposed for automatic generation control (AGC). In this strategy,
the control is evaluated as an optimization problem, and the
weight coefcients of the cost function are also optimized as the
controller gains have been done. CRAZYPSO algorithm which is
one of the recent population based optimization algorithms is used
because of its convergence superiority in order to optimize the
parameters and also its short and simple codes. The performance
of the proposed controller is compared with the performances
which are obtained with standard ISE and ITSE cost functions.
The results obtained from the simulations show that the proposed
control strategy optimized with new cost functions achieves better
dynamic performances than the standard cost functions as sum-
marized in Fig. 11. On the other hand, the robustness of the control
strategy with the proposed cost function-1 is also investigated.
Then, it can be said that this control approach is the effective
and the relatively robust strategy in order to provide optimal auto-
matic generation control to the power system, and the choosing
suitable cost function is also quite important for performance of
the convergence to the best solution. Finally, due to its superiori-
ties, the proposed control strategy can be applied to the different
control system applications, successfully.
Acknowledgment
The authors wish to thank to Prof.
_
Ilhan KOCAARSLAN for his
contributions to this paper.
References
[1] Elgard OI. Electric energy systems theory. New York: McGraw-Hill; 1982. p.
299362.
[2] Cohn N. Some aspects of tie-line bias control on interconnected power
systems. Am Inst Electr Eng Trans 1957;75:141536.
[3] Elgerd OI, Fosha CE. Optimum megawatt-frequency control of multi-area
electric energy systems. IEEE Trans Power Appar Syst 1970;89(4).
[4] Kumar IP, Kothari DP. Recent philosophies of automatic generation control
strategies in power systems. IEEE Trans Power Syst 2005;20(1).
[5] Lee KA, Yee H, Teo CY. Self-tuning algorithm for automatic generation control
in an interconnected power system. Electr Power Syst Res 1991;20(2):15765.
0 5 10 15 20 25
-16
-14
-12
-10
-8
-6
-4
-2
0
2
x 10
-3
Time (s)
d
e
l
t
a

P
t
i
e

(
p
u
M
W
)
nominal load
+25% load
+50% load
- 25% load
- 50% load
Fig. 14. Curves of DP
tie
for different changes of DP
L1
.
Table 4
Tuned parameters for different loads.
+50% Load +25% Load Nominal load 25% Load 50% Load
K
p
0.5389 0.4000 0.5762 0.4428 0.5478
K
i
0.2098 0.2464 0.1962 0.2000 0.2162
w
1
0.5473 0.8603 0.7300 0.7259 0.9882
w
2
1.2365 0.6533 0.6848 0.8438 0.9454
w
3
0.9877 1.1000 0.7879 0.6694 0.9781
Table 5
Maximum overshoots for different loads.
+50% Load +25% Load Nominal load 25% Load 50% Load
Df
1
0.04726 0.03764 0.03194 0.02293 0.01580
Df
2
0.05429 0.04171 0.03710 0.02570 0.01817
DP
tie
0.01411 0.01098 0.00958 0.00673 0.00472
H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816 15
[6] Rubaai A, Udo V. Self tuning LFC: multilevel adaptive approach. Proc Inst Electr
Eng Generat Transmiss Distribut 1994;141(4):28590.
[7] Chang CS, Fu W. Area load frequency control using fuzzy gain scheduling of PI
controllers. Electr Power Syst Res 1997;42:14552.
[8] am E, Kocaarslan I. A fuzzy gain scheduling PI controller application for an
interconnected electrical power system. Electr Power Syst Res
2005;73(3):26774.
[9] Talaq J, Al-Basri F. Adaptive fuzzy gain scheduling for load frequency control.
IEEE Trans Power Syst 1999;14(1):14550.
[10] Pinkang L, Hengjun Z, Yuyun L. Genetic algorithm optimization for AGC of
multi-area power systems. In: Proceedings of IEEE region 10 conference on
computers, communications, control and power engineering-TENCON02;
2002. p. 181821.
[11] Abdel-Magid YL, Abido MA. AGC tuning of interconnected reheat thermal
systems with particle swarm optimization. In: Proceedings of the 2003 10th
IEEE international conference on electronics, circuits and systems, vol. 1; 2003.
p. 376379.
[12] Yesil E, Gzelkaya M, Eks in I. Self-tuning fuzzy PID type load and frequency
controller. Energy Convers Manage 2004;45:37790.
[13] Juang C-F, Lu C-F. Power system load frequency control by evolutionary fuzzy
PI controller. In: Proceedings of IEEE international conference on fuzzy
systems, Budapest, Hungary; 2004. p.7159.
[14] Massiala M, Ghribi M, Kaddouri A. A two-layered self-tuning fuzzy controller
for interconnected power systems. In: Proceedings of the canadian conference
on electrical and computer engineering-CCECE 06; 2006. p. 10869.
[15] Taher SA, Hemati R, Abdolalipour A, Tabie SH. Optimal decentralized load
frequency control using HPSO algorithms in deregulated power systems. Am J
Appl Sci 2008;5(9):116774.
[16] Nanda J, Mishra S, Saikia LC. Maiden application of bacterial foraging based
optimization technique in multiarea automatic generation control. IEEE Trans
Power Syst 2009;24(2).
[17] Rao CS, Nagaraju SS, Sangameswara PR. Automatic generation control of TCPS
based hydrothermal system under open market scenario: a fuzzy logic
approach. Int J Electr Power Energy Syst 2009;31(78):31522.
[18] Lu C-F, Liu C-C, Wu C. Effect of battery energy storage system on load
frequency control considering governor dead-band and generation rate
constraints. IEEE Trans Power Syst 1995;10(3):55561.
[19] Pothiya S, Ngamroo I. Optimal fuzzy logic-based PID controller for load
frequency control including superconducting magnetic energy storage units.
Energy Convers Manage 2008;49:28338.
[20] Astrm KJ, Hagglund T. PID controllers. USA: Instrument Society of America;
1995. p. 2345.
[21] Roy R, Ghoshal SP. A novel crazy swarm optimized economic load dispatch for
various types of cost functions. Int J Electr Power Energy Syst
2008;30(4):24253.
[22] Kennedy J, Eberhart RC. Particles swarm optimization. In: Proceedings of IEEE
international conference on neural networks, Perth Australia, vol. 4; 1995. p.
19428.
[23] Ho SL, Yang S, Ni G, Lo EWC, Wong HC. A particle swarm optimization based
method for multi objective design optimizations. IEEE Trans Magnet
2005;41(5):17569.
16 H. Gozde, M.C. Taplamacioglu/ Electrical Power and Energy Systems 33 (2011) 816