45 25 Optimal Tunning

ARTICLE IN PRESS
Available online at www.sciencedirect.com
Expert Systems
with Applications
Expert Systems with Applications xxx (2008) xxxxxx
www.elsevier.com/locate/eswa
Optimal tunning of lead-lag and fuzzy logic power system

stabilizers using particle swarm optimization
A.M. El-Zonkoly *, A.A. Khalil, N.M. Ahmied
Arab Academy for Science and Technology, Faculty of Engineering and Technology, Egypt Miami, Alexandria, P.O. 1029, Egypt
Abstract
In this paper, the problem of simultaneous and coordinated tuning of brushless exciter and lead-lag power system stabilizer parameters of a single innite bus power system is considered. This problem is formulated as an optimization problem, which is solved using
particle swarm optimization technique. Also in this paper, the optimal tuning of a fuzzy logic power system stabilizer using particle
swarm optimization method is carried out.
Simulation results show the eectiveness of the proposed particle swarm optimization-based lead-lag power system stabilizer and particle swarm optimization-based fuzzy logic power system stabilizer to damp the oscillation of multimachine system and work eectively
under variable loading and fault conditions.
2007 Elsevier Ltd. All rights reserved.
Keywords: Particle swarm optimization; Power system control; Lead-lag power system stabilizer; Fuzzy logic power system stabilizer
1. Introduction
Power systems are inherently nonlinear and undergo a
wide range of transient conditions, which results in under
damped low frequency speed as well as power oscillations
that are dicult to control. The generator excitation system maintains generator voltage and controls the reactive
power ow using an automatic voltage regulator (AVR)
(Cheng, 1997). The role of an AVR is to hold the terminal
voltage magnitude of a synchronous generator at a specied level. Hence, the stability of the AVR system would
seriously aect the security of the power system. AVR
helps to improve the steady-state stability of power systems, but transient stability became a concern for the
power system operators. In transient stability the machine
is subjected to large impact, usually a fault, which is maintained for a short time and cause a signicant reduction in
the machine terminal voltage and ability to transfer synchronizing power.
*
Corresponding author. Tel.: +20 356 22586; fax: +20 356 22366.
E-mail address: amanyelz@hotmail.com (A.M. El-Zonkoly).
To enhance system damping, the generator is equipped

with power system stabilizer (PSS) that provide supplementary feedback stabilizing signal in the excitation system
(Anderson & Fouad, 1977; deMello & Concordia, 1969;
Sauer & Pai, 1998). The problem of the PSS design is a multimodal optimization problem (i.e., there exists more than
one local optimum). Hence, conventional optimization
techniques are not suitable for such a problem. Moreover,
there is no criterion to decide whether a local solution is also
the global solution. Therefore, conventional optimization
methods that make use of derivatives and gradients, in general, not able to locate or identify the global optimum.
An overview of the research eort developed in the last
decades and also on trends of small-signal studies in power
system dynamic analysis are presented in Gibbard et al.
(2001), Kundur (1994), EL-Zonkoly (2006), Urdaneta
et al. (1991) which discuss the modeling, control techniques
and analysis tools available.
Many random search methods, such as genetic algorithm (GA) and simulated annealing (SA) (Feliachi et al.,
1988; Xia & Heydt, 1983). have recently received much
interest for achieving high eciency and search global opti-
0957-4174/$ - see front matter 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eswa.2007.12.069
Please cite this article in press as: El-Zonkoly, A. M. et al., Optimal tunning of lead-lag and fuzzy logic power system ..., Expert Systems with Applications (2008), doi:10.1016/j.eswa.2007.12.069
ARTICLE IN PRESS
2
A.M. El-Zonkoly et al. / Expert Systems with Applications xxx (2008) xxxxxx
List of symbols
x (rad/s) the angular speed of the rotor
Wfd
the eld circuit ux linkage
xo (rad/s) the rated rotor electrical speed
d (elec rad) the rotor angle
Et (pu) the generator terminal voltage
H (MW. S/MVA) inertia constant
KD (pu) damping torque coecient
Te (NM) electrical torque
TA (s) amplier time constant
KA
the exciter gain
mal solution in problem space and they have been applied
to the problem of PSS design (Abe & DOI, 1983; Arredondo, 1997; Do Bmm et al., 2000).
In Do Bmm et al. (2000), authors propose a method that
simultaneously optimize both phase compensations and
gain settings for the stabilizer using Genetic Algorithms.
Although GA is very sucient in nding global or near global optimal solution of the problem, it requires a very long
run time that may be several minutes or even several hours
depending on the size of the system under study. Particle
swarm optimization (PSO), rst introduced by Kennedy
and Eberhart, is one of the modern heuristic algorithms. It
was developed through simulation of a simplied social system, and has been found to be robust in solving continuous
nonlinear optimization problems (Kennedy & Eberhart,
1995; Shi & Eberhart, 1998). The PSO technique can generate a high quality solution within shorter calculation time
and stable convergence characteristic than other stochastic
methods (Yoshida, Kawata, & Fukuyama, 2000). PSO has
been motivated by the behavior of organisms, such as sh
schooling and bird ocking. Generally, PSO is characterized
as a simple concept, easy to implement, and computationally ecient. Unlike the other heuristic techniques, PSO
has a exible and well-balanced mechanism to enhance the
global and local exploration abilities.
In this paper, the problem of simultaneous and coordinated tuning of brushless exciter and lead-lag power system
stabilizer parameters of a single machine innite bus power
system is considered. Also the design of a fuzzy logic power
system stabilizer using particle swarm optimization is carried out. Simulation results show the eectiveness of the proposed particle swarm optimization-based lead-lag power
system stabilizer (PSOLLPSS) and particle swarm optimization-based fuzzy logic power system stabilizer (PSOFLPSS)
to damp oscillation of multimachine system and work eectively under several loading and fault conditions.
2. Overview of particle swarm optimization
PSO is one of the optimization techniques and a kind of
evolutionary computation technique. The method has been
found to be robust in solving problems featuring nonlinear-
TR (s) the terminal voltage transducer time constant

KSTAB the power system stabilizer gain
TW (s) the time constant of the signal washout block
T1, T2 (s) the phase compensator time constants
Tc (s) lag time constant
TB (s) lead time constant
TE
integration time of the forward loop
Vs (pu) the output voltage of power system stabilizer
ity and non-dierentiability, multiple optima, and high

dimensionality through adaptation, which is derived from
the social-psychological theory. The features of the method
are as follows:
1. The method is developed from research on swarm such
as sh schooling and bird ocking.
2. It is based on a simple concept. Therefore, the computation time is short and requires few memories (EL-Zonkoly, 2006).
3. It was originally developed for nonlinear optimization
problems with continuous variables. It is easily
expanded to treat a problem with discrete variables.
According to the research results for birds ocking are
nding food by ocking.
PSO is basically developed through simulation of bird
ocking in two-dimension space. The position of each
agent is represented by XY axis position and also the velocity is expressed by vx (the velocity of X axis) and vy (the
velocity of Y axis). Modication of the agent position is
realized by the position and velocity information. Bird
ocking optimizes a certain objective function. Each agent
knows its best value so far (pbest) and its XY position. This
information is analogy of personal experiences of each
agent. Moreover, each agent knows the best value so far
in the group (gbest) among pbest. This information is analogy of knowledge of how the other agents around them
have performed. Namely, each agent tries to modify its
position using the following information:
The
The
The
The
current positions (x, y),

current velocities (vx, vy),
distance between the current position and pbest
distance between the current position and gbest
This modication can be represented by the concept of

velocity. Velocity of each agent can be modied by the following equation:
vk1
wvki c1 rand1 pbesti ski c2 rand2
i
gbest ski
ARTICLE IN PRESS
where
vki
w
ci
rand
S ki
pbesti
gbest
Governor
velocity of agent i at iteration k

weighting function
weighting factor
random number between 0 and 1
current position of agent i at iteration k
pbest of agent i
gbest of the group
Fig. 1 shows a concept of modication of searching point

by PSO.
Where
sk
current searching point
sk + 1 modied searching point
current velocity
vk
vk + 1 modied velocity
vpbest velocity based on pbest
velocity based on gbest
vgbest
3. Optimal tuning of lead-lag PSS parameters
The system considered for small-signal performance
study is shown in Fig. 2. The synchronous generator considered is equipped with brushless exciter (AC1A) and
lead-lag power system stabilizer (LLPSS). The details of
the power system with type AC1A excitation system block
diagram can be found in Kundur (1994).
The control parameters to be tuned through the optimization algorithm are exciter parameters (KA, TA, TB, TC),
sk+1
vk+1
vk
T.L
switch
Infinite
bus bar
Fig. 2. Schematic diagram for single machine-innite bus.
Using Eqs. (1) and (2) a certain velocity, which gradually

gets close to pbest and gbest can be calculated. The current
position can be modied by the following equation:
AVR
Vt
initial weight
nal weight
maximum iteration number
current iteration number
sk1
ski vk1
i
i
PSOFLPSS
Tm
Efd
PSOLLPS
The following weighting function is usually utilized in

(1).
wmax wmin
iter
2
w
itermax
where
wmax
wmin
itermax
iter
the lead-lag power system stabilizer parameters (T1, T2,

TW, KSTAB) and terminal voltage transducer parameter
(TR, as shown in Fig. 3. The inputs to the excitation system
are the terminal voltage, reference voltage VR and the signal from the PSS output VS. The input to the PSS is the
speed deviation Dx. We will investigate the performance
of particle swarm optimization (PSO) technique proposed
for tuning excitation system (AC1A) and lead-lag power
system stabilizer parameters.
For the linearized system model presented the eigenvalues of the system are evaluated. The proposed method
is aiming to search for the optimal parameters set of the
AC1A exciter and the lead-lag power system stabilizer so
that a comprehensive damping index (CDI) (objective
function) can be minimized (Cai & Erlich, 2005).
n
X
OF
1 fi
4
i1
where
OF
ni
n
objective function
the damping ratio
the total number of the dominant eigenvalues
The objective of the optimization problem is to maximize the damping ratio as much as possible. The computational ow chart of PSO algorithm is shown in Fig. 4.
The lower and upper limits of the exciter and lead-lag
power system stabilizer parameters are given in Table 1.
According to the simulation, the following PSO parameters are used for searching for the LLPSS and exciter
parameters:
The members of each individual are (KA, TA, TB, TC, T1,
T2, TW, KSTAB, TR).
Population size = 50.
Inertia weight factor w is set by Eq. (2), where
wmax = 0.9 and wmin = 0.4.
Acceleration constant c1 = 2 and c2 = 2.
The LLPSS and exciter parameters that showed the best
solution are summarized in Table 2.
vgbest
vpbest
4. Fuzzy logic-based power system stabilizers
sk
X
Fig. 1. Concept of modication of searching point by PSO.
The main components of a typical fuzzy logic controller

(FLC) include fuzzication, fuzzy inference engine (decision logic), and defuzzication stages. In this paper, power
ARTICLE IN PRESS
4
Lead-lag power system stabilizer
vs
1+sT1
sTw
1+sT2
1+sTw
KSTAB
K4
+
_
AC1A Efd +
Vref +
_
v1
fd K _Te
2
K3
1+sT3
Ief
0
s
Field circuit
Exciter
1
2*Hs+KD
K7
K6
k1
k8
Voltage transducer
Et
1
1+sTR
k5
Fig. 3. Block diagram representation a synchronous machine with a brushless exciter and LLPSS.
Table 2
Optimal values of control parameters
Start
Specify the system parameters for PSO
Initialize with random position (S)
and random velocity (v)
KA
TA
TB
TC
KSTAB
TW
T1
T2
TR
Optimal value
200
0.1
0.1
50
0.2
.005
0.1
(FLPSS) is injected as a supplementary stabilizing signal to

the AVR system.
The goal of this application is to stabilize and improve
the damping of the synchronous machine. Speed deviation
Dx and active power deviation Dp have been selected as the
controller inputs. The controller output is then injected
into the AVR system.
This conguration implies that the FLC has two input
parameters; KW and KP and one output parameter KU as
seen in Fig. 5. The selection of these parameters is usually
subjective and requires previous knowledge of the fuzzy
control variables (input and output signals).
The active power (Dp), speed deviation (Dx) and the
fuzzy output signal (U) are divided into seven triangular
shape membership functions with full range [1 to 1].
The membership function are shown in Fig. 6.
The membership functions related to the speed deviation
(Dx), active power deviation (Dp) are described by [Negative Big (NB), Negative Medium (NM), Negative small
(NS), Zero (ZE), Positive Small (PS), Positive Medium
(PM), Positive Big (PB)]. and output signal (U) is described
by [Output Negative Big (ONB), Output Negative Medium
evaluate O.F for each agent

If O.F < local best Then local
best =O.F
If local < global best Then
global best= local best
Update velocity
Update position
Reach
maximum
Iteration?
no
Parameter
yes
global best=parameters
of best solution
Stop
Fig. 4. Flowchart of particle swarm optimization process.
Table 1
Lower and upper limits of control parameters
TA
TB
TC
KSTAB
TW
T1
T2
TR
Lower limit
Upper limit
200
400
0.01
0.1
0.05
0.1
1
2
3
50
1
2
0.01
0.2
.001
.005
0.01
0.1
system model consisting of a synchronous machine connected to innite bus through a transmission line is used
in the simulation studies as shown in Fig. 2. The control
signal generated by the fuzzy logic power system stabilizer
KW
KP
Fuzzy
mapping
(Rules)
Defuzzification
KA
Fuzzification
Parameter
KU
Fig. 5. Schematic diagram of the FLPSS.
ARTICLE IN PRESS
a large number of iterations in searching for a proper minimum. Using some practical information about signal levels,
it is easy to set an operating range to the FLPSS parameters.
The PSO method is an excellent optimization methodology and a promising approach for solving the optimization
problem of FLPSS parameters. Therefore, this study develops the particle swarm optimization-based fuzzy logic
power system stabilizer (PSOFLPSS).
The PSO tuning algorithm tries to minimize three system
performance indices (PIs). These indices are the system
overshoot (OS) and the performance indices J1, J2, given as:
ry
100%
5
OS
Xr
J1
e2
6
X
2
J2
te
7
(p)
NB
NM NS ZE PS
PM
PB
-1 -.65 -.33
0 .33 .65 1
Input 1: Normalized membership function for active power
()
NB NM NS ZE PS PM PB
-1 -.65 -.33 0 .33 .65 1

Input 2: Normalized membership function for speed deviation
(U)
ONB ONM ONS OZE OPS OPM OPB
-1 -.65 -.33 0 .33 .65

1
Output: Normalized membership function for a consequence
Fig. 6. Membership function of (7*7) fuzzy controller.
Table 3
Rule base table for the FLPSSS
Speed
deviation
Active power
NB
NM
NS
ZE
PS
PM
PB
NB
NM
NS
ZE
PS
PM
PB
ONB
ONB
ONB
ONM
ONM
ONS
OZE
ONB
ONB
ONM
ONM
ONS
OZE
OPS
ONB
ONM
ONM
ONS
OZE
OPS
OPM
ONB
ONM
ONS
OZE
OPS
OPM
OPB
ONS
ONS
OZE
OPS
OPM
OPM
OPM
ONS
OZE
OPS
OPM
OPM
OPB
OPB
OZE
OPS
OPM
OPM
OPB
OPB
OPB
(ONM), output Negative Small (ONS), Output Zero

(OZE), Output Positive Small (OPS), Output Positive
Medium (OPM), Output Positive Big (OPB)]. Previous
experience of the controlled system dynamics is commonly
used in the design of the fuzzy control rule base as shown in
Table 3.
5. Optimal tuning of fuzzy logic based PSS parameters
Parameter tuning for the FLPSS plays an important role
in achieving the controller goals. Previous experience of the
controlled system is helpful in selection of the initial values
of the FLPSS parameters.
If sucient information is not available about the controlled system, the selection of suitable FLPSS parameters
can become a tedious trial-and-error process. Some eorts
have been reported in the literature to automate the tuning
of the FLPSS parameters at the design stage to get an optimal or near optimal system performance (EL-Hawary,
1998).
In case of a complete lack of information about the
parameters, the search for the best parameters may require
where r is the system reference, y is the system output, e is

the system error, and t is the time.
A FLPSS has two input parameters KW, KP and one
output parameter KU. These parameters can be tuned using
PSO algorithm which changes the three parameters in overlap loops, and calculates the performance index. The algorithm also detects if one of the parameters degrades the
performance indices or leads to instability. The tuning
algorithm tries to minimize three system PIs by varying
the FLPSS parameter (EL-Hawary, 1998).
The PSO algorithm was mainly utilized to determine the
optimal values of three optimal parameters KW, KP and
KU, such that the controlled system could obtain a good
output response. We dened the three FLPSS parameters
KW, KP and KU, to compose an individual K.
It is worth mentioning that the FLPSS is designed to
minimize the power system oscillation after a disturbance
so as to improve the stability.
These oscillations are expressed by adding the three system performance indices PIs.
W K J 1 J 2 OS
where
W(K)
K
performance criterion
[KW, KP, KU]
In the present study our aim is to minimize the objective

function (OF), which is formulated as
min OF absW K
With the variation of the parameter vector K the OF will

also be changed. It aimed to minimize this objective function in order to improve the system response in terms of
the settling time and overshoots.
Table 4
Lower and upper limits of control parameters
Parameter
KW
KP
KU
Lower limit
Upper limit
40
60
4
8
1
4
ARTICLE IN PRESS
6
Fuzzy Logic Power system Stabilizer

K
KU
KP
vs
K4
+
_
Vref +
_
v1
E fd
AC1A
K3
1+sT3
fd K P
2
1
2*Hs+KD
Field circuit
Exciter
Ief
0
s
K7
K6
k1
k8
Voltage transducer
1
1+sTR
E t
k5
Fig. 7. Block diagram representation a synchronous machine with a brushless exciter and FLPSS.
Table 5
Optimal values of control parameters
Parameter
KW
KP
KU
Optimal value
59.8
The change in the terminal voltage (Vt) with PSOFLPSS

is better than that with PSOLLPSS with less overshoot and
settling time as shown in Fig. 9.
The computational ow chart of PSO algorithm is

shown in Fig. 4.
The lower and upper limits of the three controller
parameters are given in Table 4.
According to the simulation, the following PSO parameters
are used for searching for the FLPSS controller parameters:
The member of each individual are KW, KP and KU.
Population size = 50.
Inertia weight factor w is set by Eq. (2) here wmax = 0.9
and wmin = 0.4.
Acceleration constant c1 = 2 and c2 = 2.
The system with FLPSS and AC1A exciter is shown in
Fig. 7. The optimal parameters that showed the best solution are summarized in Table 5.
Fig. 8. Speed deviation of a single machine system with dierent PSSs.
6. Simulation results
In this section a single-machine-innite bus system and a
multi-machine innite bus system will be tested after
equipped by the proposed PSOLLPSS and PSOFLPSS.
A comparison of the system responses will be done.
6.1. Single machine innite bus system
In the simulation of the single machine innite bus system the synchronous machine considered is equipped with
brushless excitation (AC1A) system and tested by a unitstep input signal.
Fig. 8 shows the speed deviation of a single machine system with dierent PSSs. The speed deviation (Dx) with the
PSOFLPSS and PSOLLPSS is nearly the same.
Fig. 9. Terminal voltage step response of a single machine power system

with dierent (PSOPSS).
ARTICLE IN PRESS
G1
110km
8 110km 9
25km 10km
C7
L7
G2
C9
10
11
10km 25km
L9
G3
G4
Fig. 10. Multi-machine power system.
Fig. 13. Speed deviation of machine 1 with double phase-to-ground fault

at bus 8.
Fig. 11. Speed deviation of machine 1 with single phase-to-ground fault at

bus 8.
Fig. 14. The terminal voltage of machine 1 with double phase-to-ground

fault at bus 8.
Fig. 12. The terminal voltage of machine 1 with single phase-to-ground

fault at bus 8.
6.2. Multi machine innite bus system

In this study, a two area interconnected four machine
power system shown in Fig. 10 is considered. The system
consists of four machines arranged in two areas inter-connected by a weak tie line. The parameters for all generating
units, transmission lines, load, and operating conditions
are given in Kundur (1994). The system is operating with
Fig. 15. Speed deviation of machine 1 with three phase-to-ground fault at

bus 8.
ARTICLE IN PRESS
8
Fig. 16. The terminal voltage of machine 1 with three phase-to-ground

fault at bus 8.
Fig. 19. Speed deviation of machine 1 with a load change of +20%.
Fig. 17. Speed deviation of machine 1 with one line out of service.
Fig. 20. The terminal voltage of machine 1 with a load change of +20%.

Single phase-to-ground fault at bus 8

Double phase-to-ground fault at bus 8
Three phase-to-ground at bus 8
One line out of service
Load change (+20%).
6.2.1. Single phase-to-ground fault

The speed deviation (Dx) of the system with PSOLLPSS
is better than that with PSOFLPSS with less under shoot
and settling time as shown in Fig. 11.
The terminal voltage (Vt) of machine 1 with PSOFLPSS
is better than that with PSOLLPSS with less overshoot but
the setting time is nearly the same as shown in Fig. 12.
Fig. 18. The terminal voltage of machine 1 with one line out of service.
area 1 exporting 400 MW to area 2. The simulation is carried out with the AC1Aexciter and PSOLLPSS or PSOFLPSS inserted in the model of generator 1 (G1). The
system is then tested in the following cases:
6.2.2. Double phase-to-ground fault

and less settling time as shown in Fig. 13.
Also in Fig. 14 the terminal voltage (Vt) of machine 1
with PSOFLPSS is better than that with PSOFLPSS with
less under shoot and settling time nearly the same.
ARTICLE IN PRESS
Table 6
Time response parameters for dierent test conditions using PSOFLPSS and PSOFLPSS
Test case
Type of response
PSOPSS
Types
PSOLLPSS
PSOFLPSS
PSOLLPSS
PSOFLPSS
Max. over shoot in

(pu)
1.60*e4
2.38*e4
1.4* e3
1.3*e3
Max.under shoot in
(pu)
6.99*e4
7.81*e4
1.1* e3
1 *e3
1-Single phase to ground

fault
Speed deviation
Fig. 11
Terminal voltage
Fig. 12
2-Double phase to
ground fault
0.42
0.45
0.1
.09
3
5
4
4
Speed deviation
Fig. 13
Terminal voltage
Fig. 14
PSOLLPSS
PSOFLPSS
PSOLLPSS
PSOFLPSS
6 *e4
4.3 *e3
4.1 *e3
6.45*e3
7.25*e3
6.1*e3
5.7*e3
0.87
0.91
1.33
1.33
4.5
5.8
6.0
6.0
3-Three phase to ground

fault
Speed deviation
Fig. 15
Terminal voltage
Fig. 16
PSOLLPSS
PSOFLPSS
PSOLLPSS
PSOFLPSS
5.5*e3
5*e3
4*e3
6*e3
3.2 *e2
3.4 *e2
6.6*e36.7*e3
2.2
2.3
0.1
0.12
8.5
9.0
10
11
4-Line out of Service
Speed deviation
Fig. 17
Terminal voltage
Fig. 18
PSOLLPSS
PSOFLPSS
PSOLLPSS
PSOFLPSS
4.1*e4
5.5*e4
2*e3
6*e3
8.5 *e3
9.8 *e4
7* e3
8 *e2
0.6
0.6
0.5
0.5
3
7
4.5
0.5
5-Load change (+20%)
Speed deviation
Fig. 19
Terminal voltage
Fig. 20
PSOLLPSS
PSOFLPSS
PSOLLPSS
PSOFLPSS
1.6 *e3
1.85 *e3
1.8*e3
1.9*e3
1.9*e3
1.9*e3
.60
0.7
0.15
0.15
4.5
6.0
5.5
4.5
6.2.3. Three phase-to-ground fault

is better than that with PSOFLPSS with less over shoot
6.2.4. Line out of service
and with PSOFLPSS is nearly the same where the settling
time and under shoot are nearly the same as shown in
Fig. 18.
6.2.5. Load change (+20%)
is better than that with PSOFLPSS with less over shoot
and time settling as shown in Fig. 19.
is better than that with PSOLLPSS with less settling time
and with over shoot nearly the same as shown in Fig. 20.
A comparative study between the time response parameters for dierent test conditions using PSOFLPSS and
PSOFLPSS is presented in Table 6.
7. Conclusion
This paper presented a particle swarm optimization
method (PSO) as a design method for determining the opti-
Peak time in
(s)
Settling time in
(s)
mal values of control parameters of the brushless exciter

(AC1A), lead-lag power system stabilizer (LLPSS) and
fuzzy logic power system stabilizer (FLPSS). Simulation
results showed the eectiveness of the proposed PSOLLPSS and PSOFLPSS to damp out the multimachine
(interarea) modes of oscillations and work eectively over
a wide range of loading conditions and fault conditions.
Furthermore, in multimachine system, the PSOLLPSS
has given a good response at dierent types of test conditions considering the speed deviation and small dierence
between the PSOLLPSS and PSOFLPSS in terminal voltage was observed.
References
Abe, S., & Doi, A. (1983). A new power system stabilizer synthesis in
multi-machine power systems. IEEE Transactions on Power Apparatus
and Systems, PAS-102, 39103918.
Anderson, P. M., & Fouad, A. A. (1977). Power system dynamics and
stability. Ames, IA: Iowa State Univ. Press.
Arredondo, J. M. (1997). Results of a study on location and tuning of
power system stabilizers. International Journal of Electrical Power and
Energy Systems, 19, 563567.
Cai, L. J., & Erlich, I. (2005). Simultaneous coordinated tuning of PSS and
FACTS damping controllers in LARGE power system. IEEE Transactions on Power Systems, 20(1), 294300.
Cheng, G. (1997). Genetic algorithms & engineering design. New York:
Wiley.
deMello, F. P., & Concordia, C. (1969). Concepts of synchronous machine
stability as aected by excitation control. IEEE Transactions on Power
Apparatus and Systems, PAS-88, 316329.
Do Bmm, A. L. B., Taranto, G. N., et al. (2000). Simultaneous tuning of
power system damping controllers using genetic algorithms. IEEE
Transactions on Power Systems, 15(1), 163169.
EL-Hawary, M. E. (1998). Electric power applications of fuzzy systems.
IEEE press.
ARTICLE IN PRESS
10
EL-Zonkoly, A. M. (2006). Optimal tuning of power systems stabilizers and

AVR gains using particle swarm optimization. International Journal of
Expert Systems with Applications (Vol. 31(939)). pp. 551557.
Feliachi, A., Zhang, X., et al. (1988). Power system stabilizers design using
optimal reduced order models Part II: design. IEEE Transactions on
Power Systems, 3(4), 16761684.
Gibbard, M. J., Martins, N., et al. (2001). Recent applications of linear
analysis techniques. IEEE Transactions on Power Systems, 16(1),
154162.
Kennedy, J., & Eberhart, R. (1995). Particle swarm optimization. In Proc.
IEEE Int. Conf. Neural Networks (Vol. 4) (pp. 19421948) Perth:
Australia.
Kundur, P. (1994). Power system control and stability. New York:
McGraw-Hill, Inc.
Sauer, P. W., & Pai, M. A. (1998). Power system dynamics and stability.
Englewood Clis, NJ: Prentice-Hall.
Shi, Y., & Eberhart, R. (1998). Amodied particle swarm optimizer. In
Proc. IEEE Int. Conf. Evol. Comput. Anchorage, UK (pp. 6973).
Urdaneta, A. J., Bacalao, N. J., et al. (1991). Tuning of power system
stabilizers using optimization techniques. IEEE Transactions on Power
Systems, 6(1), 127134.
Xia, D., & Heydt, G. T. (1983). Self-tuning controller for generator
excitation control. IEEE Transactions on Power Appararus and
Systems, 18771885.
Yoshida, H., Kawata, K., & Fukuyama, Y. (2000). A particle swarm
optimization for reactive power and voltage control considering
voltage security assessment. IEEE Transactions on Power Systems,
15, 12321239.

45 25 Optimal Tunning

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

45 25 Optimal Tunning

Hochgeladen von

Copyright:

Verfügbare Formate

ARTICLE IN PRESS

Available online at www.sciencedirect.com

Optimal tunning of lead-lag and fuzzy logic power system

To enhance system damping, the generator is equipped

TR (s) the terminal voltage transducer time constant

ity and non-dierentiability, multiple optima, and high

current positions (x, y),

This modication can be represented by the concept of

velocity of agent i at iteration k

Fig. 1 shows a concept of modication of searching point

Fig. 2. Schematic diagram for single machine-innite bus.

Using Eqs. (1) and (2) a certain velocity, which gradually

The following weighting function is usually utilized in

the lead-lag power system stabilizer parameters (T1, T2,

4. Fuzzy logic-based power system stabilizers

The main components of a typical fuzzy logic controller

(FLPSS) is injected as a supplementary stabilizing signal to

evaluate O.F for each agent

Fig. 5. Schematic diagram of the FLPSS.

-1 -.65 -.33 0 .33 .65 1

-1 -.65 -.33 0 .33 .65

(ONM), output Negative Small (ONS), Output Zero

where r is the system reference, y is the system output, e is

In the present study our aim is to minimize the objective

With the variation of the parameter vector K the OF will

Fuzzy Logic Power system Stabilizer

The change in the terminal voltage (Vt) with PSOFLPSS

The computational ow chart of PSO algorithm is

Fig. 8. Speed deviation of a single machine system with dierent PSSs.

Fig. 9. Terminal voltage step response of a single machine power system

Fig. 10. Multi-machine power system.

Fig. 13. Speed deviation of machine 1 with double phase-to-ground fault

Fig. 11. Speed deviation of machine 1 with single phase-to-ground fault at

Fig. 14. The terminal voltage of machine 1 with double phase-to-ground

Fig. 12. The terminal voltage of machine 1 with single phase-to-ground

6.2. Multi machine innite bus system

Fig. 15. Speed deviation of machine 1 with three phase-to-ground fault at

Fig. 16. The terminal voltage of machine 1 with three phase-to-ground

Fig. 19. Speed deviation of machine 1 with a load change of +20%.

Single phase-to-ground fault at bus 8

6.2.1. Single phase-to-ground fault

6.2.2. Double phase-to-ground fault

Max. over shoot in

1-Single phase to ground

3-Three phase to ground

4-Line out of Service

5-Load change (+20%)

6.2.3. Three phase-to-ground fault

mal values of control parameters of the brushless exciter

EL-Zonkoly, A. M. (2006). Optimal tuning of power systems stabilizers and

Das könnte Ihnen auch gefallen