ARTICLE IN PRESS

Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

Contents lists available at ScienceDirect

Engineering Science and Technology,

an International Journal Press: Karabuk University, Press Unit
ISSN (Printed) : 1302-0056
ISSN (Online) : 2215-0986
ISSN (E-Mail) : 1308-2043

j o u r n a l h o m e p a g e : h t t p : / / w w w. e l s e v i e r. c o m / l o c a t e / j e s t c h
H O S T E D BY

Available online at www.sciencedirect.com

ScienceDirect

Full Length Article

Design of optimal input–output scaling factors based fuzzy PSS using

bat algorithm
D.K. Sambariya a,*, R. Gupta a, R. Prasad b
a Rajasthan Technical University, Kota, Rajasthan, India
b
Indian Institute of Technology Roorkee, Roorkee, Uttaranchal, India

A R T I C L E I N F O A B S T R A C T

Article history: In this article, a fuzzy logic based power system stabilizer (FPSS) is designed by tuning its input–output
Received 14 September 2015 scaling factors. Two input signals to FPSS are considered as change of speed and change in power, and
Received in revised form the output signal is considered as a correcting voltage signal. The normalizing factors of these signals
10 January 2016
are considered as the optimization problem with minimization of integral of square error in single-
Accepted 10 January 2016
Available online
machine and multi-machine power systems. These factors are optimally determined with bat algorithm
(BA) and considered as scaling factors of FPSS. The performance of power system with such a designed
BA based FPSS (BA-FPSS) is compared to that of response with FPSS, Harmony Search Algorithm based
Keywords:
Bat algorithm (BA) FPSS (HSA-FPSS) and Particle Swarm Optimization based FPSS (PSO-FPSS). The systems considered are
Fuzzy logic controller (FLC) single-machine connected to infinite-bus, two-area 4-machine 10-bus and IEEE New England 10-
Fuzzy logic based power system stabilizer machine 39-bus power systems for evaluating the performance of BA-FPSS. The comparison is carried
(FPSS) out in terms of the integral of time-weighted absolute error (ITAE), integral of absolute error (IAE) and
Harmony search algorithm (HSA) integral of square error (ISE) of speed response for systems with FPSS, HSA-FPSS and BA-FPSS. The su-
Input–output scaling factors perior performance of systems with BA-FPSS is established considering eight plant conditions of each
Particle swarm optimization (PSO) system, which represents the wide range of operating conditions.
Performance indices (PIs)
Copyright © 2016, The Authors. Production and hosting by Elsevier B.V. on behalf of Karabuk
Power system stabilizer (PSS)
University. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction adaptive technique is to estimate the dynamic model with resting

uncertainties in the system online based on measured signals. The
Modern electric power systems (EPSs) are complex, intercon- erroneous estimation of states and the uncertainty may lead to
nected and susceptible to low frequency oscillations (LFOs) in the design of PSS with degraded performance. In online transient sta-
frequency range of 0.2 Hz–3.0 Hz. Power system stabilizers (PSSs) have bility assessment, some selected contingencies need to be evaluated
been commonly used to damp out these LFOs. The changes in loading as fast as possible before occurrence of a fault or a disturbance in
conditions, operating conditions and some sort of disturbance are the system. Therefore, the computational time is very critical. H∞
main causes to develop LFOs in EPSs. Conventional power system sta- optimization technique is used to design robust PSS, but it gives the
bilizers (CPSSs) consisting of lead-lag networks are generally used PSS order as high as that of the plant, which increases the com-
PSS for damping out these oscillations because of simple structure plexity to the system and reduces its applicability [3].
and easy installation. The design of CPSS is based on linear control In early phase of optimization, CPSS parameters have been tuned
theory and involves the linearized dynamic model with a specific op- using gradient based optimization technique. It requires the com-
eratic condition of EPS. These controllers give degraded performance putation of sensitivity and eigenvectors at the end iteration, which
with varying operating conditions and sometime unable to main- resulted with heavy computational burden and slow convergence rate.
tain stability of EPS on a higher degree of loading conditions [1,2]. The heuristic based optimization techniques are employed to tune
The wide range of operating conditions for real EPS has moti- the parameters of CPSS and proportional integral derivative (PID)
vated researchers to develop different methods to design PSS with based PSS. Among these are Tabu search algorithm [1], real coded
improved performances and this resulted to the application of adap- genetic algorithm (RCGA) [4], genetic algorithm [3], particle swarm
tive and robust control to design PSS. The basic idea behind the optimization (PSO) [4,5], and breeder genetic algorithm [6]. Bacte-
ria foraging algorithm [3], simulated annealing [7], differential
evolution [1] and strength pareto evolutionary algorithms [8] have
* Corresponding author. Tel.: +91 744 2472429, fax: +91 9982252205.
E-mail address: [email protected]; [email protected] (D.K.
been used successfully to tune the CPSS parameters. However, genetic
Sambariya). and simulated annealing algorithms have the tendency of revisit-
Peer review under responsibility of Karabuk University. ing the sub-optimal solutions and thus the designed CPSS may give

http://dx.doi.org/10.1016/j.jestch.2016.01.006
2215-0986/Copyright © 2016, The Authors. Production and hosting by Elsevier B.V. on behalf of Karabuk University. This is an open access article under the CC BY-NC-
ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
2 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

deteriorated performance. These optimization techniques fail with systems. In this paper, bat algorithm (BA) has been used to opti-
an epistatic objective function, which have closely related to multi- mize scaling factors of FPSSs for SMIB, 4-machine and IEEE New
model problems and the higher number of variables [9]. England 10-machine 39-bus power systems. The performance of the
To mitigate these limitations, an artificial intelligence based proposed BA-FPSS is to be compared to the PSO-FPSS [5] and HSA-
methods of PSS design, such as artificial neural networks (ANNs) FPSS [2] for the three-power systems. The performance evaluation
[10], fuzzy logic [11–14], adaptive fuzzy [15], neuro-fuzzy [16,17], is carried out in terms of ITAE, IAE and ISE in each case of a con-
and interval type-2 [18,19], have been reported in literature. In the troller as well as a power system under study.
case of ANN, the gradient algorithm is being used to learn its pa- In the organization of this paper, the problem formulation is con-
rameters using either input/output [20] parameters or online data sidered by introducing test power systems, and an objective function
from different operating points in a power system network. used for optimization of scaling factors in Section 2.2. The bat al-
Fuzzy logic controllers (FLCs) can cope with those that naturally gorithm used to determine optimal set of input–output scaling
have lots of vagueness or uncertainty in their behavior. These do not factors is mentioned in Section 3. Optimization of scaling factors
require a mathematical model of the controlled process. These have of FPSSs for all three-power systems using bat algorithm is carried
rigidity and robustness as their profound and interesting character- out in Section 4. The optimal set of scaling factors for SMIB power
istics in comparison to other methods. The properly designed fuzzy system using bat algorithm and performance comparison with BA-
logic based PSS works similar to PD or PID based PSS [21]. Develop- FPSS, with HSA-FPSS, PSO-FPSS [5] and with FPSS (without scaling
ment of an equivalence between the scaling factor of a fuzzy controller factors) is discussed in Section 4.1. It is repeated for 4-machine power
and linear PID controller coefficients is reported in [22]. The selec- system in Section 4.2. The process of optimal parameters for 10-
tion of scaling factors, appropriate membership function, number of machine power system is determined using harmony search, as well
linguistic variables and the corresponding rule table are the major as bat algorithm in Section 4.3. The detail on harmony search is not
requirement in designing PSS based on FLC. The detail on linguistic given and considered as in [2] with same initializing parameters.
variables and selection of membership function is well reported in The nonlinear time-domain simulation is carried out on this power
[23]. Based on an organized approach, a standardized rule table is system using BA-FPSS and HSA-FPSS and compared with FPSS
proposed in [24]. The optimization of scaling factors using particle (without scaling factors) in this section. Lastly, Section 5 concludes.
swarm optimization is reported in [5]. The harmony search algo-
rithm (HSA) is proposed by Geem et al in 2001 [25], and is inspired 2. Problem formulation
by the process of the improvisation used by musicians to achieve
harmony. The HS algorithm [26] is a meta-heuristic optimization al- The aim of this paper is to utilize the superior performance of
gorithm that is similar to the PSO [27] and GA [28]. It has been Bat algorithm for tuning input and output scaling factors of FPSS
implemented extensively in the fields of engineering optimization in connection with power systems; therefore, the EPS elements such
in [26]. It became an alternative to other heuristic algorithms like PSO as generators, excitation system and PSS must be modeled. To com-
[27] and simulated annealing (SA) [7]. It is a derivative free, meta- plete the tuning process, an objective function to obtain satisfactory
heuristic optimization (which does not use trial-and-error), inspired results is necessary and should be defined. Accordingly, the system
by the way musicians improvise new harmonies [29], and it uses model and an objective function used in PSS parameter tuning
higher-level techniques to solve problems efficiently [2]. process for SMIB, and multi-machine power systems, should be
In the field of optimization, much of algorithms are floating with elaborated.
unique properties. Some are useful to one application, while others
are not so. The bat algorithm reported by Yang (2010) is meta-
heuristic in nature [30]. It is based on the echolocation based 2.1. Test power systems
behavior of micro bats [31]. It was established by considering bench-
mark functions that the behavior is superior to PSO and GA [32]. 2.1.1. SMIB power system
It has also reported that the application of GA and PSO is inappro- The power system is a multi-component system. The equiva-
priate with multi-model problems. The frequency-tuning and lent of system can be represented by using differential equations.
automatic zooming are out of the main features of the bat algorithm. Assuming that the vector of states and the vector of inputs are rep-
Particle swarm optimization (PSO) and Firefly algorithm (FA) gen- resented by X and U, respectively, then the power system may be
erate an efficient codebook, but undergo instability in convergence represented as in Eqn. (1).
when particle velocity is high and with the non-availability of
X = f ( X , U ) (1)
brighter fireflies in the search space, respectively. The application
of Bat Algorithm (BA) on the initial solution of LindeBuzoGray (LBG)
A nonlinear power system can be linearized by considering small
is presented in [33]. It produces an efficient codebook with less com-
perturbation around an operating point. It is easy to design PSS to
putational time and results due to its automatic zooming feature
such linearized model of power system [9,36]. The EPS repre-
using adjustable pulse emission rate and loudness of bats [33]. The
sented by Eqn. (1) may be shown by state equations as in Eqn. (2).
design of fuzzy proportional derivative controller and fuzzy pro-
portional derivative integral controller for speed control of brushless ΔX = AΔX + BU (2)
direct current drive has been presented in [34]. The problem of con-
troller design is considered as an optimization using nature inspired The infinite-bus of the SMIB power system can be considered
optimization algorithms such as particle swarm, cuckoo search, and by Thevenin’s equavalent of the large and complex power system.
bat algorithms [34]. A Firefly Algorithm (FA) optimized fuzzy PID The components and inter-connections of the SMIB power system
controller is proposed for Automatic Generation Control (AGC) of are shown in Fig. 1. The inadequate damping of the generator is the
multi-area multi-source power system in [35]. main cause of small signal oscillations. The PSS may be connected
In [5], the scaling factors associated with two inputs and one to excitation system to add extra-damping of the generator as elabo-
output are optimized by PSO for single-machine infinite-bus (SMIB) rated in [9]. The pioneer work on the design of appropriate PSS is
and two-area 4-machine 10-bus power system. These scaling factors presented in [37].
(input–output) have been further optimized using harmony search In system representation by Eqn. (2), A is the system matrix of
algorithm in [2]. The performance of the HSA-FPSS has been com- an order as 4 × 4 and is given by δf/δX, while B is the input matrix
pared and was found better as compared to PSO-FPSS for both power with order 4 × 1 and is given by δf/δU. The order of state vector is

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■ 3

stabilizer connected in decentralized manner to the generators. The

state model can be represented as in Eqn. (2), where A is the system
matrix with order 4N × 4N (16 × 16) and is given by δf/δX, while B
is the input matrix with order 4N × N pss (16 × 4) and is given by δf
δU. The order of state vector ΔX is 4N × 1 (16 × 1), and the order of
ΔU is N pss × 1 (4 × 1). Here, the well-known Heffron-Phillip linear-
ized model is used to represent the large multimachine power system
as in [2,39] and the system dynamics is given in [38,40].
Fig. 1. Line diagram of single-machine infinite-bus power system.
2.1.3. IEEE New England 10-machine 39-bus power system
The state equations to the power system, consisting of N, the
4 × 1; the order is 1 × 1. Here, the well-known Heffron-Phillip lin- number of generators, and N pss , the number of power system sta-
earized model is considered for fabricating the model in MATLAB bilizers, can be written as in Eqn. (2). In this case, A is the system
2011b as in [2,9]. The SMIB power system dynamics in terms of dif- matrix of the order 4N × 4N (40 × 40) and B is the input matrix with
ferential equations are considered as in [38]. the order 4N × N pss (40 × 10). The order of state vector ΔX is 4N × 1
(40 × 1), and the order of ΔU is N pss × 1 (10 × 1). Here, the well-
known Heffron-Phillip linearized model is used to represent the large
2.1.2. Two-area 4-machine 10-bus power system
multi-machine power system, as in Fig. 3, and the single-line diagram
The 2nd system considered is two-area 4-machine 10-bus power
of IEEE 39-bus power system is shown in Fig. 4.
system [2]. The line-diagram of the system is shown in Fig. 2. The
development of the small-signal model of multi-machine power
system is well explained in [38]. It can be represented by a large 2.2. Objective function
number of differential and algebraic equations. The general repre-
sentation of Heffron-Philip model for multi-machine power systems The scheme of input–output scaling factors of FPSS is consid-
is shown in Fig. 3. Consider N as the number of generators of multi- ered as presented in [2]. The input signals to the FPSS are considered
machine power system, with N pss the number of power system as change in speed (Δω) and change in power (Δp) with associ-
ated scaling factors as Kω and Kp, respectively. The output signal of
FPSS is considered as change in correction voltage (Δu) and the
scaling factor as Ku [2]. In this paper, these scaling factors are de-
termined using bat algorithm. The problem of tuning scaling factors
is considered as an optimization with minimization of integral
squared error (ISE) of change in speed signal as a fitness function.
As an objective function, the ISE based cost function is repre-
sented for SMIB, four-machine and ten-machine power system by
Eqns. (3)–(5), respectively. The connections of scaling factors of FPSS
are shown in Fig. 5, where the change in speed is subjected to min-
imize using the bat algorithm to obtain optimal set of input–
output scaling factors.
Tsim

∫ Δω (t ) ⋅ dt
2
J= (3)
0

4 Tsim
J=∑ ∫ Δω i (t ) ⋅ dt
Fig. 2. Line diagram of two-area 4-machine 10-bus power system. 2
(4)
i =1 0

Fig. 3. General representation of Heffron-Philip model for multi-machine power

systems. Fig. 4. Line diagram of IEEE New England 10-machine 39-bus power systems.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
4 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

Fig. 6. Representation of tuning scheme for input–output scaling factors of FPSS using
bat algorithm.

• The bats are able to differentiate/detect the prey and back-

ground barriers in the search path using echolocation behavior.
• Assuming that r th bat is randomly moving with velocity, loca-
tion, frequency, wavelength and intensity represented by vr, xr,
fmin , λr and A0, respectively, the pulse frequency is regulated
and the pulse rate is adjusted in the range pr = [0,1] on the basis
of the distance of the prey.
• The loudness of the pulse is adjusted according to the distance
of the prey as A0 (maximum for large distance) to Amin (minimum
for lower distance) [42].

In optimization problems, an objective function is represented

by minimization of F(r) and subjected to xr ∈ Xr, r = 1, 2, … , n . In ini-
tialization step of the bat algorithm, the bat population is generated
Fig. 5. Representation of Heffron-Philip model for SMIB power system with input– with velocity vr and position xr for r = 1, 2, … , n . The pulse frequen-
output scaling factors of FPSS.
cy is selected in the range fr ∈[ fmin, fmax ] . Pulse rate and the loudness
are set as above, while the search loop is set to maximum itera-
tion counts as t ≤ Tmax [31,42].
In step 2, the new solutions are generated by considering the
following equations of frequency, velocity and position. For r th bat,
9 Tsim
J=∑ ∫ Δω i (t ) ⋅ dt the new position and velocity at time step t are represented by xrt
2
(5)
i =1 0 and vrt , respectively [43].

The parameter bounds for SMIB power system are as in Eqn. (6) fr = fmin + ( fmax − fmin ) β (8)
[2].
vrt = vrt −1 + ( xrt −1 − x ′ ) fr (9)
K min
p ≤ K p ≤ K max
p

K ω ≤ K ω ≤ K ωmax
min
xrt = xrt −1 + vrt (10)
K umin ≤ K u ≤ K umax (6)
where β represents the uniform distribution in the range β ∈ [0, 1].
K ≤ K pi ≤ K
min
pi
max
pi
The value represents the best location in the search step for n bats.
In step 3, the local search is applied for the generation of the new
i ≤ K ωi ≤ K ωi
K ωmin max
solutions using local random walk behavior as described by the fol-
K ui ≤ K ui ≤ K ui
min max
(7)
lowing Eqn. (11). The ε is selected in the range of [−1, 1] with average
value of loudness At at time t.
Eqn. (7) includes parameter bounds for both multi-machine
power systems [2]. The i stands for ith generator in the multi-
xnew = xold + ε At (11)
machine power system and Tsim refers to simulation time during
optimization process and specified as 100 seconds. In the case of In step 4, the loop operation for generation of the new solu-
IEEE 10-machine power system, the value of i is 09, because 10th tions is considered. On advancement of iterations, the loudness and
generator is considered as slack without controller at this genera- the rate of pulse emission have to be updated by Eqns. (9)–(10). The
tor. Considering one of the above objectives corresponding to the rate of pulse emission is increased when shortening the path to prey.
system under investigation, the proposed approach employs the bat
algorithm with parameter bounds to solve this optimization problem Art +1 = xold + α Art (12)
for an optimal set of input–output scaling factors of FPSS.
prrt +1 = prr0 [1 − e −γ t ] (13)

3. Review on bat algorithm where α and γ represent the constant values in the range of 0 ≤ α ≤ 1
and 0 < γ. The process behaves like the cooling factor of a cooling
This algorithm is based on the echolocation behavior produced schedule in the simulated annealing [44]. The generally selected value
by natural bats in locating their prey. The pulse generated by of these constants is 0.9 in the literature [45].
microbats lasts for 8–10 seconds, with frequency range of In the last step 5, the stopping criterion is checked as the
25–150 kHz and with associated wave length of 2–14 mm. Neces- maximum count of iterations is reached and termination of com-
sary assumptions are required to be considered during development putation is executed. Otherwise, go to steps 3–4 to repeat the process.
of the echolocation characteristics of microbats [9,41]. The tuning scheme of input–output scaling factors is shown in Fig. 6,

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■ 5

Table 1 −4
x 10
Plant configuration of SMIB power system [38]. 7.035
Fitness function
PS model Pg 0 Q g0 Xl
7.03
Plant-1 0.50 0.0251 0.20
Plant-2 0.50 0.0505 0.40 7.025
Plant-3 0.75 0.0566 0.20
Plant-4 0.75 0.1152 0.40 7.02
Plant-5 1.00 0.1010 0.20
Plant-6 1.00 0.2087 0.40 7.015
Plant-7 1.10 0.2550 0.40

min
Plant-8 1.20 0.3068 0.40 7.01

F
7.005

7
where the speed deviation is minimized using bat algorithm to
decide optimal set of parameters. As the connection of scaling factors 6.995
is already shown in Fig. 5, Δp is left open intentionally to save space.
6.99

4. Results and discussion 6.985

0 20 40 60 80 100
Iterations
4.1. SMIB power system
Fig. 7. Plot of fitness function using bat algorithm in tuning of input–output scaling
4.1.1. Plant creation for simulation factor for SMIB power system with nominal operating condition.
The line diagram and the small signal model of SMIB power
system are represented in Figs. 1 and 5, respectively. The operating
conditions of SMIB power system are represented by different considered for nonlinear simulations. Such obtained eight-plants
sets of active power Pg 0 and transmission line reactance Xl as (covering wide range of operating conditions) are examined for
mentioned in Table 1. The plants are designed to represent oper- the speed response with FPSS, HSA-FPSS and BA-FPSS in this
ating conditions and weak conditions through heavy loading section.
conditions. Plant-6 represents the nominal operating conditions as
in [40].
60
4.1.2. Optimal set of scaling factors (a)
The problem is formulated in MATLAB environment and ex- 40
w

ecuted on Intel (R) Core (TM) – 2 Duo CPU T6400 @ 2.00 GHz with
K

3 GB RAM, 32-bit operating system. The SMIB system is equipped 20

with FPSS along with input–output scaling factors. The scheme of
optimization is shown in Fig. 6. The problem of optimization of 0
(b)
scaling factors is considered with an ISE based objective function 10
as in Eqn. (3). The steps of the bat algorithm are shown in Section
p
K

3. In [32,42], the generally opted values of initializing parameters, 5

such as intensity (A) and pulse rate (r) are 0.5 and 0.5, respective-
ly. However, the proper initializing parameters for bat algorithm are 0
considered after long efforts and found as A = 0.9 and r = 0.1. The other (c)
constraint such as initializing population is selected as n = 25 and 4
the bandwidth are considered as fmin = 0 and fmax = 2.0 . The plant
u
K

(SMIB power system) operating at nominal operating condition 2

(where in Xl = 0.4pu and Pg 0 = 1.0 pu ) is considered for optimal tuning
of input–output scaling factors of FPSS. The scaling factors are 0
0 20 40 60 80 100
considered with lower and upper bounds as 0.001 ≤ K ω ≤ 50.0 , Iterations
0.001 ≤ K p ≤ 10.0 , and 0.001 ≤ K u ≤ 5.0. The optimization process
with bat algorithms is set to terminate with maximum iteration Fig. 8. Plot of input–output scaling factors with iteration count for SMIB power system.
counts as 100. The behavior of bat algorithm in terms of fitness func-
tion with iterations is shown in Fig. 7. The variation of the PID
parameters with iteration count is shown in Fig. 8. The optimal set Table 2
Optimal values of input–output scaling factors of FPSS with PSO [5], HSA [2] and
of parameters obtained using the bat algorithm is enlisted in Table 2.
proposed BA on SMIB power system.
The scaling factors using PSO in [5] and HSA in [2] for SMIB system
Controller Parameters Bounds
are also included in the table for the purpose of comparison.
Symbol Values Lower Upper

4.1.3. Speed response analysis BA-FPSS (Prop.) Kω 13.2238 0.001 50.0

A SIMULINK based block diagram, including all the nonlinear Kp 3.0358 0.001 10.0
Ku 2.0128 0.001 5.00
blocks, is generated in MATLAB software. The SMIB power system
HSA-FPSS [2] Kω 26.3928 0.001 50.0
performance under nonlinear mode is carried out by creating self- Kp 5.3353 0.001 10.0
clearing fault at time 5 seconds and persistent for 0.1 second with Ku 2.4531 0.001 5.00
the wide range of operating conditions. The system with unlike com- PSO-FPSS [5] Kω 59.80 0.0 70.0
binations of different active power and transmission line reactance Kp 4.0 0.0 10.0
Ku 1.0 0.0 10.0
as in Table 1 (eight different plants) and system data as in [9,38] is

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
6 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

The fuzzy logic based PSS (FPSS) reported in [14,46] is consid- Plant−7 FPSS (Sambariya, 2010)
ered for comparison purpose. The numbers of linguistic variables PSO−FPSS (El−Zonkoly, 2009)
are five as LN (large negative), MN (medium negative), Z (zero), MP 0.01 HSA−FPSS (Sambariya, 2015)
(medium positive) and LP (large positive). The input signals to FLC BA−FPSS (Proposed)
have been considered as change in speed (Δw) and change in power
0.005

Speed devn. (pu)

(Δp), while that of the output signal is considered as correction
voltage ( ΔVpss ). The corresponding 25 rules of the rule-base are con-
sidered as presented in [46]. The triangular type membership 0
function is considered for both input and output signals. The crisp
value is obtained using centroid type defuzzification method.
−0.005
The SIMULINK model of SMIB system is prepared in the MATLAB
software equipped with FPSS, BA-FPSS and HSA-FPSS controllers.
These systems are simulated for all eight plants as created in −0.01
Table 1. The comparison of speed response of SMIB system with
FPSS, with PSO-FPSS [5], with HSA-FPSS [2] and with BA-FPSS is
carried out for each plant configuration. The comparative re- −0.015
sponse is carried out for 8-plant conditions but shown only for
4 6 8 10 12 14
plant-3, plant-6 and plant-7 in Figs. 9–11, respectively. However, Time (s)

Fig. 11. Speed response for Plant-7 with FPSS [46], PSO-FPSS [5], HSA-FPSS [2] and
−3 proposed BA-FPSS for SMIB power system.
x 10
8
Plant−3 FPSS (Sambariya, 2010)
PSO−FPSS (El−Zonkoly, 2009)
6 HSA−FPSS (Sambariya, 2015)
the response with other plants are not shown because of space
BA−FPSS (Proposed) limitation. Clearly, the settling time with BA-FPSS is better as
compared to HSA-FPSS [2], PSO-FPSS [5] and greatly improved
4
Speed devn. (pu)

with respect to FPSS [14,46]. The response with HSA-FPSS [2] and
BA-FPSS is comparable but the response with FPSS [46] settles in
2 more than 25 seconds. The closely related responses with HSA-
FPSS [2] and BA-FPSS are to be differentiated by recording
performance indices.
0
To carry out the analysis with clear perceptiveness and com-
pleteness about the system response for all the system conditions,
−2 three performance indices that reflect the settling time and over-
shoot are introduced and evaluated as in [2,9]. These indices are
−4 defined as folowing in Eqns. (14)–(16).

• Integral of the Time-Weighted Absolute Error (ITAE)

−6
4 5 6 7 8 9 10 T =t sim
Time (s) ITAE = ∫ t Δω (t ) dt (14)
t =0

Fig. 9. Speed response for Plant-3 with FPSS [46], PSO-FPSS [5], HSA-FPSS [2] and
proposed BA-FPSS for SMIB power system.
• Integral Square Error (ISE)
T =t sim
ISE = ∫ Δω (t ) dt
2
(15)
t =0

0.01
Plant−6 FPSS (Sambariya, 2010) • Integral of the Absolute Error (IAE)
0.008 PSO−FPSS (El−Zonkoly, 2009)
T =t sim
HSA−FPSS (Sambariya, 2015) IAE = ∫ Δω (t ) dt (16)
0.006 BA−FPSS (Proposed) t =0
Speed devn. (pu)

0.004 where t sim is the simulation time of the system and Δω(t) repre-
0.002
sents the instantaneous speed change. To prove superiority of the
BA-FPSS, the SMIB system is simulated one by one with all four
0 controllers (FPSS [46], PSO-FPSS [5], HSA-FPSS [2] and BA-FPSS)
and the performance indices (ITAE, IAE and ISE) of speed response
−0.002
are recorded for the simulation time as 40 seconds and enlisted in
−0.004 Table 3. The closely related responses with HSA-FPSS and BA-FPSS
are well differentiated by distinct values of performance indices.
−0.006 The lower value of performance index (PI) represents the compar-
atively better performance of the system with reduced settling
−0.008
time and overshoot. In Table 3, the value of performance indices
−0.01 (PIs) with BA-FPSS is lesser as compared to others, resulting to
4 6 8 10 12 14
good performance. The value of PIs of system response with PSO-
Time (s)
FPSS [5] or plant-7 and plant-8 are higher as compared to that of
Fig. 10. Speed response for Plant-6 with FPSS [46], PSO-FPSS [5], HSA-FPSS [2] and with BA-FPSS. Therefore, the performance of system with PSO-
proposed BA-FPSS for SMIB power system. FPSS is degraded against the proposed BA-FPSS.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■ 7

Table 3 nonlinear simulation is considered by creating self-clearing fault at

Performance comparison of speed response with FPSS [46], PSO-FPSS [5], HSA- bus no. 3 and bus no. 4 in plant-1 and pant-2 configuration, respec-
FPSS [2] and proposed BA-FPSS for SMIB power system.
tively. The active power plant-1 associated to 4-generators is [7, 7,
PS model Controllers ITAE IAE ISE 7.2172] and changed to [7.2, 7.1, 7.0, 6.9] in plant-3. The load con-
Plant-1 FPSS [46] 0.0140 0.0025 5.1939E-06 nected to system are 2 as [11.59 + j2.12; 15.75 + j2.88] in plant-1
PSO-FPSS [5] 0.0072 0.0013 3.0359E-06 configuration but only real parts are enlisted in Table 4 because imag-
HSA-FPSS [2] 0.0073 0.0013 3.0437E-06 inary part remains the same for all plant conditions. In this way, the
BA-FPSS (Prop.) 0.0073 0.0013 2.9941E-06
Plant-2 FPSS [46] 0.0221 0.0036 7.0078E-06
eight different plants of the system are considered as shown in Table 4.
PSO-FPSS [5] 0.0119 0.0021 3.9812E-06
HSA-FPSS [2] 0.0117 0.0039 3.9599E-06 4.2.2. Optimal set of scaling factors
BA-FPSS (Prop.) 0.0115 0.0020 3.8997E-06
Plant-3 FPSS [46] 0.0259 0.0044 1.4628E-06
The system model referring to plant-1 configuration as in Table 4
PSO-FPSS [5] 0.0159 0.0029 9.5283E-06 is equipped with FPSS to all four-machines (named as Gen-1 to Gen-
HSA-FPSS [2] 0.0131 0.0024 8.6405E-06 4) and subjected to design using the bat algorithm (as described
BA-FPSS (Prop.) 0.0110 0.0020 6.7405E-06 in Section 3), with a simple time domain based minimization of ISE
Plant-4 FPSS [46] 0.0453 0.0071 2.2915E-05
as an objective function as in Eqn. (4) with bounds as defined in
PSO-FPSS [5] 0.0162 0.0029 1.0369E-05
HSA-FPSS [2] 0.0181 0.0032 1.0954E-05 Eqn. (7). The speed signal from each generator is sensed and the
BA-FPSS (Prop.) 0.0177 0.0032 1.0785E-05 minimum value of sum of ISE of the error signal is minimized to
Plant-5 FPSS [46] 0.0529 0.0086 3.9531E-05 tune input–output scaling factors of four FPSSs with parameter
PSO-FPSS [5] 0.0229 0.0041 1.6368E-05 bounds as 40 ≤ Kω ≤ 70, 1.0 ≤ K p ≤ 10 , and 1.0 ≤ K u ≤ 5.0 . The initial-
HSA-FPSS [2] 0.0245 0.0044 2.1180E-05
BA-FPSS (Prop.) 0.0150 0.0028 1.1477E-05
izing parameters for BA are considered the same as in the previous
Plant-6 FPSS [46] 0.1364 0.0182 8.1390E-05 section. The termination criterion of the tuning process is consid-
PSO-FPSS [5] 0.0411 0.0070 3.1181E-05 ered as the maximum number of iterations and set as 100. The
HSA-FPSS [2] 0.0328 0.0057 2.6481E-05 parameter bounds are selected by using the trial-and-error method;
BA-FPSS (Prop.) 0.0247 0.0044 1.9924E-05
therefore, several attempts are required. The optimized scaling factors
Plant-7 FPSS [46] 0.2791 0.0313 1.5210E-04
PSO-FPSS [5] 0.0821 0.0129 6.7567E-05 are shown in Table 5. The behavior of BA during optimization in
HSA-FPSS [2] 0.0448 0.0077 4.0846E-05 terms of fitness function is plotted in Fig. 12.
BA-FPSS (Prop.) 0.0398 0.0069 3.4579E-05
Plant-8 FPSS [46] 0.97088 0.07295 3.7782E-04
PSO-FPSS [5] 124.26 7.7840 6.6780
HSA-FPSS [2] 0.0733 0.0120 7.6007E-05 Table 5
BA-FPSS (Prop.) 0.0644 0.0107 6.7486E-05 Comparison of input–output scaling factors of FPSS using PSO-FPSS [5], HSA-FPSS
[2] and BA-FPSS for two-area 4-machine 10-bus power system.

Controllers Generators K ωi K pi K ui

4.2. Two-area 4-machine 10-bus power system PSO-FPSS [5] Gen-1 59.8000 4.0000 1.0000
Gen-2 59.8000 4.0000 1.0000
Gen-3 59.8000 4.0000 1.0000
4.2.1. Plant creation for simulation Gen-4 59.8000 4.0000 1.0000
The single-line diagram of the two-area four-machine ten-bus HSA-FPSS [2] Gen-1 61.1017 3.9703 0.7327
power system is shown in Fig. 2, which is a benchmark power system Gen-2 60.8977 4.7107 0.5536
Gen-3 57.0917 3.8375 0.6540
to study small signal oscillations [40]. The line data, load flow and
Gen-4 60.3711 3.6118 0.5258
machine data are considered as in [38,40]. The above multimachine BA-FPSS (Prop.) Gen-1 58.6538 4.0109 1.8991
system is modeled using SIMULINK Toolbox with machine model Gen-2 56.0157 4.0016 1.0021
1.0. The test system (four-machine system) is considered with the Gen-3 59.3950 6.4531 4.0501
wide range of operating conditions of power system and system con- Gen-4 40.0012 7.997 3.9996

nection configuration. Here, the different test models are created

by changing the active power of generation, distributed load, line
outage and fault at different bus location as mentioned in Table 4. 0.04
In Table 4, the configuration of the 4-machine power system is Fitness function
considered by varying active power, active load, bus structure and 0.0399
fault at a particular bus of Fig. 2. In plant-1, the bus structure is as
in Fig. 2 but a line between bus no. 9 and bus no. 10 is disconnect- 0.0398
ed in plant-2 configuration of the system. It can be observed that the
0.0397
Fmin

Table 4 0.0396
Plant configuration with different operating conditions for two-area 4-machine 10-
bus power system [38].
0.0395
PS model Active power Active load F/Ba L/Ob

Plant-1 7, 7, 7.2172, 7 11.59; 15.75 B/No. 3 As in Fig. 2 0.0394

Plant-2 7, 7, 7.2172, 7 11.59; 15.75 B/No. 4 B/No. 9–10
Plant-3 7.2, 7.1, 7.0, 6.9 11.59; 15.75 B/No. 5 As in Fig. 2 0.0393
Plant-4 7.2, 7.1, 7.0, 6.9 11.59; 15.75 B/No. 6 B/No. 7–10
Plant-5 7.2, 7.1, 7.0, 6.9 11.99; 15.45 B/No. 7 As in Fig. 2
Plant-6 7.1, 6.9, 7.5, 6.5 11.19; 15.95 B/No. 8 As in Fig. 2 0.0392
0 20 40 60 80 100
Plant-7 7.1, 6.9, 7.5, 6.5 11.19; 15.95 B/No. 9 B/No. 5–9 Iterations
Plant-8 5, 8, 6.2172, 8 11.59; 15.75 B/No. 10 As in Fig. 2
a
Fault location at a particular bus for non-linear study. Fig. 12. Fitness function plot for simultaneous tuning of input–output scaling factors
b System as in in Fig. 2 or with line outage between two buses. of FPSSs for 4-machine 10-bus power system using bat algorithm.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
8 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

4.2.3. Speed response analysis −3

x 10
The two-area four-machine ten-bus power system is described 2.5
Gen−3: Plant−3
and the creations of system models based on operating condi- FPSS (Sambariya, 2010)
2 PSO−FPSS (El−Zonkoly, 2009)
tions are elaborated in the previous section. The FPSS [46], PSO-
HSA−FPSS (Sambariya, 2015)
FPSS [5], HSA-FPSS [2] and proposed BA-FPSS are connected to the 1.5 BA−FPSS (Proposed)
system and simulations are carried out for the speed response. In
each plant condition as listed in Table 4 is considered with fault 1

Speed devn. (pu)

location. The disturbance is considered as self-clearing at differ-
0.5
ent buses at 1.0 second and cleared after 0.05 second. As a sample,
the speed response of Gen-1 to Gen-4 for plant-3 is compared with 0
FPSS [46], PSO-FPSS [5], HSA-FPSS [2] and BA-FPSS in Figs. 13–16.
These graphical representations of the simulation results reveal −0.5
that the performance of the system with PSO-FPSS [5], HSA-FPSS
−1
[2] and BA-FPSS is greatly improved as compared to FPSS [46]. The
−1.5

−2

−2.5
−3 1 2 3 4 5 6 7 8 9 10
x 10 Time (s)
3 Gen−1: Plant−3
FPSS (Sambariya, 2010)
PSO−FPSS (El−Zonkoly, 2009) Fig. 15. Speed response for Gen-3 of Plant-3 with FPSS [46], PSO-FPSS [5], HSA-
HSA−FPSS (Sambariya, 2015) FPSS [2] and BA-FPSS.
2 BA−FPSS (Proposed)

−3
Speed devn. (pu)

x 10
1 2
Gen−4: Plant−3
FPSS (Sambariya, 2010)
1.5 PSO−FPSS (El−Zonkoly, 2009)
0 HSA−FPSS (Sambariya, 2015)
BA−FPSS (Proposed)
1
Speed devn. (pu)

−1
0.5

−2 0

−0.5
1 2 3 4 5 6 7 8 9 10
Time (s) −1

Fig. 13. Speed response for Gen-1 of Plant-3 with FPSS [46], PSO-FPSS [5], HSA-
−1.5
FPSS [2] and BA-FPSS.

−2
1 2 3 4 5 6 7 8 9 10
−3
x 10 Time (s)
Gen−2: Plant−3
5 FPSS (Sambariya, 2010) Fig. 16. Speed response for Gen-4 of Plant-3 with FPSS [46], PSO-FPSS [5], HSA-
PSO−FPSS (El−Zonkoly, 2009) FPSS [2] and BA-FPSS.
HSA−FPSS (Sambariya, 2015)
4
BA−FPSS (Proposed)

3
Speed devn. (pu)

responses of the system with FPSS [46], PSO-FPSS [5], HSA-FPSS

2 [2] and BA-FPSS are closely related, therefore differentiating the
associated performance indices was to be carried out in the next
1 section.
To evaluate the robustness of the proposed BA-FPSS, simula-
0 tion is carried out for all eight plant configurations, which represent
the wide range of operating conditions and system configura-
−1
tions. The system is simulated FPSS [46], PSO-FPSS [5], HAS-FPSS
[2] and BA-FPSS for comparison purpose with eight plant condi-
tions. Each time the performance indices (ITAE, IAE and ISE) are
−2
recorded and enlisted in Table 6. Since the system possesses four
generators, the PI values in Table 6 are the sum of PIs of four
1 2 3 4 5 6 7 8 9 10
Time (s) generators. Comparatively lower value of PI refers to better perfor-
mance. It is clear from this table that the performance of the
Fig. 14. Speed response for Gen-2 of Plant-3 with FPSS [46], PSO-FPSS [5], HSA- system is enhanced by using proposed BA-FPSS as compared to
FPSS [2] and BA-FPSS. other controllers.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■ 9

Table 6 Table 7 refers to the self-clearing fault at a bus for nonlinear

Performance comparison of speed response with FPSS [46], PSO-FPSS [5], HSA- behavior of system.
FPSS [2] and BA-FPSS for 4-machine 10-bus power system.

PS model Controllers ITAE IAE ISE

4.3.2. Optimal set of scaling factors
Plant-1 FPSS [46] 0.0763 0.0231 4.5996E-05
The creation of experimental plants for IEEE New England ten-
PSO-FPSS [5] 0.0308 0.0128 4.0754E-05
HSA-FPSS [2] 0.0356 0.0135 4.2913E-05 machine thirty nine-bus power system is well explained in the
BA-FPSS (Prop.) 0.0285 0.0133 3.9055E-05 previous section. The required machine data, load flow data, trans-
Plant-2 FPSS [46] 0.1823 0.0473 1.4331E-04 former data and line data for the system configuration are considered
PSO-FPSS [5] 0.0828 0.0322 1.6105E-04 as presented in [38,40]. The system model referring to plant-1 con-
HSA-FPSS [2] 0.0846 0.0354 1.8268E-04
BA-FPSS (Prop.) 0.0775 0.0302 1.4252E-04
figuration as in Table 7 is equipped with FPSS at all nine-machines
Plant-3 FPSS [46] 0.0596 0.0169 1.4331E-04 (named as Gen-1 to Gen-9) except Gen-10, which is considered as
PSO-FPSS [5] 0.0191 0.0053 8.3550E-06 the slack and subjected to controller design using harmony search
HSA-FPSS [2] 0.0169 0.0050 7.4998E-06 algorithm (as described in [2]) and the bat algorithm in Section 3,
BA-FPSS (Prop.) 0.0155 0.0046 6.4506E-06
with parameter bounds as 0.001 ≤ K ω ≤ 60 , 0.001 ≤ K p ≤ 8.0 , and
Plant-4 FPSS [46] 0.0497 0.0132 1.1295E-05
PSO-FPSS [5] 0.0301 0.0062 9.0234E-06 0.001 ≤ K u ≤ 5.0 . With the initializing parameters as above for BA
HSA-FPSS [2] 0.0262 0.0059 8.2087E-06 and as in [2] for HSA; the systems are simulated for an iteration count
BA-FPSS (Prop.) 0.0183 0.0055 7.6245E-06 as 200. The fitness function variation for 200 iterations with HSA
Plant-5 FPSS [46] 0.0625 0.0167 1.7306E-05 and BA is shown in Fig. 17. The optimal values of scaling factors with
PSO-FPSS [5] 0.0224 0.0052 6.5374E-06
HSA-FPSS [2] 0.0280 0.0064 7.6419E-06
bat and harmony search algorithm are mentioned in Table 8.
BA-FPSS (Prop.) 0.0171 0.0048 5.7325E-06
Plant-6 FPSS [46] 0.0362 0.0111 1.0156E-05
PSO-FPSS [5] 0.0206 0.0050 6.1350E-06
HSA-FPSS [2] 0.0270 0.0058 6.7488E-06 −4
x 10
BA-FPSS (Prop.) 0.0132 0.0046 5.8600E-06 10
Plant-7 FPSS [46] 0.0613 0.0149 9.2326E-06 Fitness function
PSO-FPSS [5] 0.0320 0.0038 1.7054E-06
HSA-FPSS [2] 0.0309 0.0037 1.7049E-06
BA-FPSS (Prop.) 0.0195 0.0036 1.6846E-06 9.5
Plant-8 FPSS [46] 0.0428 0.0111 6.8407E-06
PSO-FPSS [5] 0.0082 0.0021 1.2035E-06
HSA-FPSS [2] 0.0168 0.0037 2.0258E-06
BA-FPSS (Prop.) 0.0094 0.0027 1.5931E-06 9
Fmin

8.5
4.3. IEEE New England 10-machine 39-bus power system

4.3.1. Plant creation for simulation

The IEEE 39-bus power system is configured with different sets 8
of active power and active load connected to the system shown in
Fig. 4. It has 10 generators and 19 loads connected as in [38]. The
active power assigned to plant-1 (base case) are as [5.519816,
0 50 100 150 200
10.0, 6.5, 5.08, 6.32, 6.5, 5.6, 5.4, 8.3, 2.5]. The load assigned to
Iterations
plant-1 (base-case) for bus nos. [1, 2, 13, 14, 17, 18, 21, 23, 24, 25,
26, 27, 28, 29, 30, 32, 35, 36, 38] = [(0.092 + j0.046), (11.04 + j2.5), Fig. 17. Fitness function plot for simultaneous tuning of input–output scaling factors
(3.22 + j0.024), (5.0 + j1.84), (2.3380 + j0.8400), (5.22 + j1.76), of FPSSs for 10-machine 39-bus power system using bat algorithm.
(2.74 + j1.15), (2.745 + j0.8466), (3.086 + j0.922), (2.24 + j0.472),
(1.39 + j0.17), (2.81 + j0.755), (2.06 + j0.276), (2.835 + j0.269),
(6.28 + j1.030), (.075 + j0.88), (3.20 + j1.53), (3.294 + j0.323), Table 8
(1.58 + j0.30)]. To generate 8-plant configurations, the different Comparison of input–output scaling factors of FPSS using HSA-FPSS and BA-FPSS
for IEEE 10-machine 39-bus power system.
sets of active power of generators and active load are considered.
These system plants are shown in Table 7. The last column of Controllers Generators K ωi K pi K ui

HSA-FPSS Gen-1 10.0000 9.7770 9.9991

Gen-2 56.8062 6.8892 9.9382
Gen-3 10.0899 9.9099 10.000
Table 7 Gen-4 59.8892 5.8811 10.000
Plant configuration with different operating conditions for IEEE New England 10- Gen-5 55.5804 8.9766 9.7362
machine 39-bus power system [38]. Gen-6 59.5663 9.8928 10.000
Gen-7 60.0000 6.5596 8.9733
Power system model Active powera Active loadb Fault at bus
Gen-8 52.9681 9.0131 8.7615
Plant-1 Base case Base case Bus-16 Gen-9 56.4434 7.9649 6.4976
Plant-2 3,5 2,13,27,28 Bus-13 BA-FPSS Gen-1 19.6546 4.7485 3.0247
Plant-3 1,2,3,4 17,24 Bus-11 Gen-2 59.7079 8.4225 4.8785
Plant-4 7,8 27,28,30,32 Bus-9 Gen-3 13.1155 8.4256 3.9920
Plant-5 2,7 30,35,36,38 Bus-7 Gen-4 59.4908 7.1621 3.1274
Plant-6 1,3,9,10 24–27,30,35,36 Bus-17 Gen-5 25.9536 2.9191 4.6584
Plant-7 1,4,5,6 13,25,30,35 Bus-19 Gen-6 17.4769 4.2649 4.8460
Plant-8 4,5,6,7 18,21,27,28,36,38 Bus-21 Gen-7 44.3857 9.0319 3.6861
a Gen-8 45.5560 4.6530 3.9345
Active power of the generators is changed w.r.t. base case.
b Gen-9 27.4008 7.8812 4.7202
The connected load to the buses is changed w.r.t. base case.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
10 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

4.3.3. Speed response analysis x 10

−4

The 9-generators of the system are equipped controllers, and sim- 6

Gen−5: Plant−5 FPSS
ulation is carried out for the speed response from the system. The
HSA−FPSS
system is equipped with FPSS [47,48], HSA-FPSS and BA-FPSS and 4 BA−FPSS
graphical comparison is recorded. The response from the system
with all controllers and for all generators is impossible because of 2
space constraint. The response of plant-5 for Gen-1, Gen-3, Gen-5,

Speed devn. (pu)

Gen-8 and Gen-9 is shown in Figs. 18–22. The improved perfor-
0
mance from the system with the bat algorithm can be observed with
reduced settling time and overshoot as compared to others. The
graphical representation of the response due to these controllers −2
is quite clear to interpret best performance with BA-FPSS and worst
as with FPSS [47,48]. It can be seen that the overshoot as well as −4
the settling time with BA-FPSS is greatly improved as compared to
that of FPSS [47,48]. However, the response with BA-FPSS is closely
−6

−8
−4
x 10 0 1 2 3 4 5 6 7 8
Gen−1: Plant−5 FPSS
Time (s)
3 HSA−FPSS
BA−FPSS Fig. 20. Speed response of Gen-5 for plant-5 of 10-machine 39-bus power system
with FPSS [47,48], HSA-FPSS and BA-FPSS.
2
Speed devn. (pu)

1
related to that of HSA-FPSS; therefore, the performance indices based
analysis is needed to differentiate the degree of performance.
0 To evaluate the robustness of the proposed BA-FPSS, simula-
tion is carried out for all eight plant configurations, which represent
the wide range of operating conditions and system configura-
−1 tions. The system is simulated with FPSS and with HSA-FPSS for
comparison purpose with eight plant conditions. Each time the per-
formance indices (ITAE, IAE and ISE) are recorded and enlisted in
−2
Table 9. Since the system possesses ten generators, the PI values in
Table 9 are the sum of PIs of ten generators. Comparatively lower
value of PI refers to better performance. It is clear from this table
0 1 2 3 4 5 6 7 8
Time (s) that the performance of the system is enhanced by using pro-
posed BA-FPSS as compared to performance with FPSS and with
Fig. 18. Speed response of Gen-1 for plant-5 of 10-machine 39-bus power system HAS-FPSS.
with FPSS [47,48], HSA-FPSS and BA-FPSS.

−4 −4
x 10 x 10
4 5
Gen−3: Plant−5 FPSS Gen−8: Plant−5 FPSS
HSA−FPSS 4 HSA−FPSS
3 BA−FPSS BA−FPSS
3
2
2
Speed devn. (pu)

Speed devn. (pu)

1 1

0
0
−1

−1 −2

−2 −3

−4
−3
−5
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
Time (s) Time (s)

Fig. 19. Speed response of Gen-3 for plant-5 of 10-machine 39-bus power system Fig. 21. Speed response of Gen-8 for plant-5 with of 10-machine 39-bus power system
with FPSS [47,48], HSA-FPSS and BA-FPSS. FPSS [47,48], HSA-FPSS and BA-FPSS.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■ 11

x 10
−4 that speed response with BA-FPSS is much better as compared to
4 Gen−9: Plant−5 others. The superiority of the proposed controller (BA-FPSS) proved
FPSS
in terms of performance indices.
HSA−FPSS
3 In case of IEEE ten-machine power system only FPSS [47,48] is
BA−FPSS
available; therefore, the harmony search and bat algorithms are con-
2
sidered to optimize the input–output scaling factors. The system
1 responses with FPSS [47,48], with HSA-FPSS and with BA-FPSS are
Speed devn. (pu)

compared and found that the BA-FPSS appeared with superior per-
0
formance. The speed response is compared graphically as a sample
−1 for plant-5 and superior performance with BA-FPSS is validated over
eight plant conditions using performance indices.
−2 The strong aspect of the bat algorithm is its quick start proper-
−3 ty and the strength to optimize in global space. The harmony search
is able to optimize system globally but after a prolonged number
−4 of iterations.
−5
Acknowledgments
−6

0 1 2 3 4 5 6 7 8 This research was supported by All India Council for Technical

Time (s) Education, New Delhi, India in the name of D. K. Sambariya. The
author is grateful to the University College of Engineering, Rajast-
Fig. 22. Speed response of Gen-9 for plant-5 with of 10-machine 39-bus power system han Technical University, Kota, for sponsoring him under Quality
FPSS [47,48], HSA-FPSS and BA-FPSS.
Improvement Programme. He also thanks his colleagues for sharing
the responsibility at the parent institute during the stay at Roorkee.
Table 9
Performance comparison of speed response with FPSS [47,48], HSA-FPSS and BA- Nomenclature
FPSS for IEEE New England 10-machine 39-bus power system.
A0 Loudness of sound
PS model Controllers ITAE IAE ISE
A System matrix
Plant-1 FPSS [47,48] 0.0236 0.0122 8.3233E-06
B Input matrix
HSA-FPSS 0.0104 0.0064 5.8288E-06
BA-FPSS 0.0099 0.0063 5.3339E-06 Di Damping coefficient for the ith generator
Plant-2 FPSS [47,48] 0.0339 0.0149 8.4324E-06 ε Elite percentage
HSA-FPSS 0.0097 0.0056 3.1842E-06 E fdi Equivalent excitation voltage of the ith generator
BA-FPSS 0.0080 0.0051 3.0877E-06 Eqi′ Internal voltage behind the d-axis transient reactance
Plant-3 FPSS [47,48] 0.0115 0.0063 1.5598E-06
fmax Maximum frequency
HSA-FPSS 0.0021 0.0022 5.5395E-07
BA-FPSS 0.0017 0.0019 5.0791E-07 fmin Minimum frequency
Plant-4 FPSS [47,48] 0.0185 0.0104 1.0332E-05 i ith generator
HSA-FPSS 0.0066 0.0048 6.3518E-06 K Ai AVR gain of the ith generator
BA-FPSS 0.0063 0.0048 5.6729E-06
K 1ij − K 6ij Heffron-Phillip constants
Plant-5 FPSS [47,48] 0.0132 0.0086 7.0048E-06
HSA-FPSS 0.0082 0.0056 6.3767E-06 Kp Scaling factor for Δp signal
BA-FPSS 0.0079 0.0054 5.7404E-06 Ku Scaling factor for Δu signal
Plant-6 FPSS [47,48] 0.0204 0.0107 5.8676E-06 Kω Scaling factor for Δω signal
HSA-FPSS 0.0081 0.0052 3.4451E-06 Mi Machine inertia coefficient for the ith generator
BA-FPSS 0.0076 0.0049 3.6366E-06
Plant-7 FPSS [47,48] 0.0120 0.0066 1.9293E-06
N Number of generators
HSA-FPSS 0.0030 0.0025 8.0286E-07 N pss Number of PSS
BA-FPSS 0.0027 0.0023 7.7544E-07 Pg 0 Active power
Plant-8 FPSS [47,48] 0.0291 0.0135 8.2817E-06 Δp Change in power
HSA-FPSS 0.0084 0.0056 4.3251E-06
r Pulse rate
BA-FPSS 0.0075 0.0052 4.3215E-06
TAi AVR time-constant of the ith generator
Td′0i d-axis open-circuit transient time-constant of the ith gen
Tmi Mechanical torque of the ith generator
5. Conclusion U Input variable vector
Δu Change in PSS output signal
In this paper, the application of the bat algorithm is used to tune Δω Change in speed
the input–output scaling factors of fuzzy logic based power system Δωi Change in speed for the ith generator
stabilizer for three systems such as single-machine infinite-bus power ω0 Synchronous speed of the generator
system (SMIB), two-area four-machine ten-bus power system and X State variable vector
IEEE New England ten-machine thirty nine-bus power systems. Xl Transmission line reactance
The SMIB power system is equipped with FPSS [46], PSO-FPSS Ui PSS output signal of the ith generator
[5], HSA-FPSS [2] and proposed BA-FPSS. The system is simulated
for eight plant conditions, and consequently the speed response from
the system for different plants is compared. The performance indices References
with BA-FPSS are greatly improved as compared to others.
The two-area four-machine ten-bus power system is simulated [1] M. Abido, Robust Design of Power System Stabilizers for Multimachine Power
Systems Using Differential Evolution, Vol. 302 of Studies in Computational
for speed response comparison with FPSS [46], PSO-FPSS [5], HSA- Intelligence, Springer, Berlin and Heidelberg, 2010, pp. 1–18, doi:10.1007/978-
FPSS [2] and proposed BA-FPSS. The simulation study is revealed 3-642-14013-6_1.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006
 ARTICLE IN PRESS
12 D.K. Sambariya et al./Engineering Science and Technology, an International Journal ■■ (2016) ■■–■■

[2] D.K. Sambariya, R. Prasad, Optimal tuning of fuzzy logic power system stabilizer [25] Z.W. Geem, J.H. Kim, G. Loganathan, A new heuristic optimization algorithm:
using harmony search algorithm, Int. J. Fuzzy Syst. 17 (3) (2015) 457–470, harmony search, SIMULATION 76 (2) (2001) 60–68, doi:10.1177/
doi:10.1007/s40815-015-0041-4. 003754970107600201.
[3] S. Abd-Elazim, E. Ali, Power system stability enhancement via bacteria foraging [26] Z. Geem, Harmony search applications in industry, in: B. Prasad (Ed.), Soft
optimization algorithm, Arabian J. Sci. Eng. 38 (3) (2013) 599–611, doi:10.1007/ Computing Applications in Industry, Vol. 226 of Studies in Fuzziness and Soft
s13369-012-0423-y. Computing, Springer, Berlin and Heidelberg, 2008, pp. 117–134, doi:10.1007/
[4] A.D. Falehi, Design and scrutiny of maiden PSS for alleviation of power system 978-3-540-77465-5_6.
oscillations using RCGA and PSO techniques, J. Electr. Eng. Technol. 8 (3) (2013) [27] J. Yu, P. Guo, Improved PSO algorithm with harmony search for complicated
402–410. function optimization problems (2012). doi:10.1007/978-3-642-31346-2_70.
[5] A.M. El-Zonkoly, A.A. Khalil, N.M. Ahmied, Optimal tuning of lead-lag and fuzzy [28] Y. Abdel-magid, M. Abido, Optimal multiobjective design of robust power system
logic power system stabilizers using particle swarm optimization, Expert Syst. stabilizers using genetic algorithms, IEEE Trans. Power Syst. 18 (3) (2003)
Appl. 36 (2) (2009) 2097–2106, doi:10.1016/j.eswa.2007.12.069. 1125–1132, doi:10.1109/TPWRS.2003.814848.
[6] M. Linda, N. Nair, A new-fangled adaptive mutation breeder genetic optimization [29] D.K. Sambariya, R. Prasad, Design of harmony search algorithm based tuned
of global multi-machine power system stabilizer, Int. J. Electr. Power Energy fuzzy logic power system stabilizer, Int. Rev. Electr. Eng. 8 (5) (2013) 1594–1607,
Syst. 44 (1) (2013) 249–258, doi:10.1016/j.ijepes.2012.06.005. doi:10.15866/iree.v8i5.2117.
[7] M. Abido, Simulated annealing based approach to PSS and FACTS based [30] X.S. Yang, A New Metaheuristic Bat-Inspired Algorithm, vol. 284, Springer, Berlin
stabilizer tuning, Int. J. Electr. Power Energy Syst. 22 (4) (2000) 247–258, and Heidelberg, 2010, pp. 65–74, doi:10.1007/978-3-642-12538-6_6.
http://dx.doi.org/10.1016/S0142-0615(99)00055-1. [31] S.K. Saha, R. Kar, D. Mandal, S.P. Ghoshal, V. Mukherjee, A new design method
[8] H. Yassami, A. Darabi, S. Rafiei, Power system stabilizer design using strength using opposition-based bat algorithm for IIR system identification problem, Int.
Pareto multi-objective optimization approach, Electr. Power Syst. Res. 80 (7) J. Bio-Inspired Comput. 5 (2) (2013) 99–132.
(2010) 838–846, doi:10.1016/j.epsr.2009.12.011. [32] X.S. Yang, A.H. Gandomi, Bat algorithm: a novel approach for global engineering
[9] D.K. Sambariya, R. Prasad, Robust tuning of power system stabilizer for small optimization, Engng. Comp. 29 (5) (2012) 464–483, doi:10.1108/
signal stability enhancement using metaheuristic bat algorithm, Int. J. Electr. 02644401211235834.
Power Energy Syst. 61 (0) (2014) 229–238, doi:10.1016/j.ijepes.2014.03.050. [33] C. Karri, U. Jena, Fast vector quantization using a bat algorithm for image
[10] H.N. Al-Duwaish, Z.M. Al-Hamouz, A neural network based adaptive sliding compression, Eng. Sci. Technol. Int. J. (2015), http://dx.doi.org/10.1016/j
mode controller: application to a power system stabilizer, Energy Convers. .jestch.2015.11.003.
Manag. 52 (2) (2011) 1533–1538, http://dx.doi.org/10.1016/j.enconman [34] K. Premkumar, B. Manikandan, Bat algorithm optimized fuzzy PD based speed
.2010.06.060. controller for brushless direct current motor, Eng. Sci. Technol. Int. J. (2015),
[11] D.K. Sambariya, Power system stabilizer design using compressed rule base of http://dx.doi.org/10.1016/j.jestch.2015.11.004.
fuzzy logic controller, J. Electr. Electron. Eng. 3 (3) (2015) 52–64, doi:10.11648/ [35] P.C. Pradhan, R.K. Sahu, S. Panda, Firefly algorithm optimized fuzzy PID
j.jeee.20150303.16. controller for AGC of multi-area multi-source power systems with UPFC and
[12] D.K. Sambariya, R. Prasad, Design of PSS for SMIB system using robust fast SMES, Eng. Sci. Technol. Int. J. (2015), http://dx.doi.org/10.1016/j.jestch
output sampling feedback technique, in: 7th International Conference on .2015.08.007.
Intelligent Systems and Control (ISCO-13), 2013, pp. 166–171. doi:10.1109/ [36] H. Alkhatib, J. Duveau, Dynamic genetic algorithms for robust design of
ISCO.2013.6481142. multimachine power system stabilizers, Int. J. Electr. Power Energy Syst. 45 (1)
[13] R. Gupta, D.K. Sambariya, R. Gunjan, Fuzzy logic based robust power system (2013) 242–251, http://dx.doi.org/10.1016/j.ijepes.2012.08.080.
stabilizer for multi-machine power system, in: IEEE International Conference [37] F. Demello, C. Concordia, Concepts of synchronous machine stability as affected
on Industrial Technology, ICIT-06, 2006, pp. 1037–1042. doi:10.1109/ by excitation control, IEEE Trans. Power Appar. Syst. 88 (4) (1969) 316–329,
ICIT.2006.372299. doi:10.1109/TPAS.1969.292452.
[14] D.K. Sambariya, R. Gupta, A. Sharma, Fuzzy applications to single machine power [38] D.K. Sambariya, Small signal stability enhancement using power system
system stabilizers, J. Theor. Appl. Inf. Technol. 5 (3) (2009) 317–324. stabilizer (Ph.D. Thesis), Department of Electrical Engineering, Indian Institute
[15] T. Hussein, M. Saad, A. Elshafei, A. Bahgat, Damping inter-area modes of of Technology Roorkee, Roorkee, 2015, pp. 1–287.
oscillation using an adaptive fuzzy power system stabilizer, Electr. Power Syst. [39] R. Gupta, B. Bandyopadhyay, A. Kulkarni, T. Manjunath, Design of decentralized
Res. 80 (12) (2010) 1428–1436, http://dx.doi.org/10.1016/j.epsr.2010.06.004. power system stabilizer for multi-machine power system using periodic output
[16] D.K. Chaturvedi, O.P. Malik, Neurofuzzy power system stabilizer, IEEE Trans. feedback technique, in: 7th International Conference on Control, Automation,
Energy Conv. 23 (3) (2008) 887–894, doi:10.1109/TEC.2008.918633. Robotics and Vision, ICARCV-02, Vol. 3, 2002, pp. 1676–1681, doi:10.1109/
[17] A. Albakkar, O. Malik, Adaptive neuro-fuzzy controller based on simplified ANFIS ICARCV.2002.1235027.
network, 2012, pp. 1–6. doi:10.1109/PESGM.2012.6344842. [40] K.R. Padiyar, Power System Dynamics Stability and Control, second ed., B. S.
[18] D.K. Sambariya, R. Prasad, Evaluation of interval type-2 fuzzy membership Publications, Hyderabad, India, 2008.
function & robust design of power system stabilizer for SMIB power system, [41] E. Ali, Optimization of power system stabilizers using BAT search algorithm,
Sylwan J. 158 (5) (2014) 289–307. Int. J. Electr. Power Energy Syst. 61 (2014) 683–690 http://dx.doi.org/10
[19] D.K. Sambariya, R. Prasad, Power system stabilizer design for multimachine .1016/j.ijepes.2014.04.007.
power system using interval type-2 fuzzy logic controller, Int. Rev. Electr. Eng. [42] X.S. Yang, Bat algorithm for multi-objective optimisation, Int. J. Bio-Inspired
8 (5) (2013) 1556–1565, doi:10.15866/iree.v8i5.2113. Comput. 3 (5) (2011) 267–274, doi:10.1504/ijbic.2011.042259.
[20] E. Abu-Al-Feilat, M. Bettayeb, H. Al-Duwaish, M. Abido, A. Mantawy, A neural [43] D.K. Sambariya, H. Manohar, Model order reduction by integral squared error
network-based approach for on-line dynamic stability assessment using minimization using bat algorithm, in: Proceedings of 2015 RAECS UIET Panjab
synchronizing and damping torque coefficients, Electr. Power Syst. Res. 39 (2) University Chandigarh, 21–22nd December 2015, 2015, pp. 1–7.
(1996) 103–110, http://dx.doi.org/10.1016/S0378-7796(96)01099-1. [44] S. Kirkpatrick, C. Gelatt, M. Vecchi, Optimization by simulated annealing, Science
[21] D.K. Sambariya, R. Prasad, Design of robust PID power system stabilizer for 220 (4598) (1983) 671–680, doi:10.1126/science.220.4598.671.
multimachine power system using HS algorithm, Am. J. Electr. Electron. Eng. [45] Z. Yang, A. Bose, Design of wide-area damping controllers for interarea
3 (3) (2015) 75–82, doi:10.12691/ajeee-3-3-3. oscillations, IEEE Trans. Power Syst. 23 (3) (2008) 1136–1143, doi:10.1109/
[22] T. Hiyama, Real time control of micro-machine system using micro-computer tpwrs.2008.926718.
based fuzzy logic power system stabilizer, IEEE Trans. Energy Conv. 9 (4) (1994) [46] D.K. Sambariya, R. Gupta, Fuzzy applications in a multi-machine power system
724–731, doi:10.1109/60.368335. stabilizer, J. Electr. Eng. Technol. 5 (3) (2010) 503–510.
[23] D.K. Sambariya, R. Prasad, Robust power system stabilizer design for single [47] M. Ramirez-Gonzalez, O.P. Malik, Self-tuned power system stabilizer based on
machine infinite bus system with different membership functions for fuzzy logic a simple fuzzy logic controller, Electr. Power Comp. Syst. 38 (4) (2010) 407–423,
controller, in: Intelligent Systems and Control (ISCO), 2013 7th International doi:10.1080/15325000903330591.
Conference on, 2013, pp. 13–19. doi:10.1109/ISCO.2013.6481115. [48] M. Ramirez-Gonzalez, O. Malik, Simplified fuzzy logic controller and its
[24] M. Sanaye-Pasand, O.P. Malik, A fuzzy logic based PSS using a standardized application as a power system stabilizer, in: 15th International Conference on
rule table, Electr. Mach. Power Syst. 27 (3) (1999) 295–310, doi:10.1080/ Intelligent System Applications to Power Systems (ISAP ’09), 2009, pp. 1–6,
073135699269316. doi:10.1109/ISAP.2009.5352817.

Please cite this article in press as: D.K. Sambariya, R. Gupta, R. Prasad, Design of optimal input–output scaling factors based fuzzy PSS using bat algorithm, Engineering Science
and Technology, an International Journal (2016), doi: 10.1016/j.jestch.2016.01.006