Electrical Power and Energy Systems 155 (2024) 109615

Contents lists available at ScienceDirect

International Journal of Electrical Power and Energy Systems

journal homepage: www.elsevier.com/locate/ijepes

Robust multi-machine power system stabilizer design using bio-inspired

optimization techniques and their comparison
Dhanraj Chitara a, P.K. Singhal a, S.L. Surana a, Gulshan Sharma b, R.C. Bansal c, d, *
a
Department of Electrical Engineering, Swami Keshvanand Institute of Technology (SKIT), Management & Gramothan, Jaipur (Raj.), India
b
Department of Electrical Engineering Technology, University of Johannesburg, Johannesburg, 2006, South Africa
c
Department of Electrical Engineering, University of Sharjah, Sharjah, United Arab Emirates
d
Department of Electric, Electronic and Computer Engineering, University of Pretoria, Pretoria, South Africa

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

Keywords: This paper reports a comparative study among four bio-inspired meta-heuristic techniques i.e. Sooty-Tern
Low-frequency Oscillations Optimization (STO), Grey Wolf Optimization (GWO), Genetic Algorithm (GA), and Particle Swarm Optimiza­
Power System Stabilizer tion (PSO) to tune the robust Power System Stabilizer (PSS) parameters of the multi-machine power system.
Grey Wolf Optimization
These approaches are successfully tested on two bench-mark systems: sixteen-machine, sixty-eight-bus New
Sooty-Tern Optimization
England Extended Power Grid (NEEPG) and three-machine, nine-bus Western System Coordinating Council
(WSCC). The efficacy of planned PSS via STO and GWO is validated by extensive non-linear simulations,
eigenvalue analysis, and performance indices for numerous operating conditions under decisive perturbations,
and outcomes are matched with those of GA and PSO techniques. In addition, the robustness is also tested for
these algorithms. The results indicate that the PSS design using STO and GWO improves the small-signal stability
and damping performance for mitigating inter-area and local area modes of low-frequency oscillations compared
to GA and PSO.

1. Introduction based on the linearized theory of control system that helps to mitigate
the low-frequency oscillations efficiently only for a specific operating
In the recent past, the Small-Signal Stability (SSS) of Multi-Machine point. The CPSS designs are unsuccessful for a variation in the extensive
Power Systems (MMPS) has become a bigger challenge for engineers. range of operating settings of non-linear power systems. Although,
The SSS concerns low-frequency electromechanical oscillations that classical control techniques like root-locus [4], frequency response [5],
arise due to unbalance between mechanical and electrical torques at digital control [6], pole-placement [7], non-linear & adaptive control
synchronous generators after small perturbations [1]. These distur­ methods [8,9], etc. perform satisfactorily but are not appropriate for
bances cause system separation, endangering system security, and non-convex and non-differentiable problem functions.
power transfer capability, creates stress on the mechanical shaft, and With the development of the power system, complication in the
decrease the overall operating efficiency of the power system if tolerable systems has increased enormously. As a result of this analysis of power
damping is not introduced. As per the literature, these oscillations in systems by conventional methods has become very complicated, diffi­
MMPS are inter-area and local areas [1,2]. To boost the damping per­ cult, and time-consuming. Now, Artificial Intelligence (AI) based
formance and SSS, Power System Stabilizers (PSSs) are adopted to damp methods like fuzzy logic, artificial neural network, and combined neuro-
such oscillations via excitation control of synchronous generator [3]. fuzzy [10–12] have emerged as active tools to resolve the issue of low-
However, in [3], it is mentioned that the proper tuning of PSS param­ frequency oscillation. The advantages of AI methods are fast processing
eters plays a vital role in performance during the system perturbation. speed, more robustness, fault-tolerant capability, and can handle con­
Literature reveals that the Conventional PSSs (CPSS) have been ditions of incomplete information about data and corrupt data. In the
employed as effective damping controllers in power systems due to their literature, it is seen that the optimization methods are categorized into
simplicity and satisfactory performance [2,3]. The CPSS designs are two major groups namely, mathematics-based classical techniques and

* Corresponding author at: Department of Electrical Engineering, University of Sharjah, Sharjah, United Arab Emirates.
E-mail addresses: [email protected] (D. Chitara), [email protected] (P.K. Singhal), [email protected] (S.L. Surana), [email protected] (G. Sharma),
[email protected] (R.C. Bansal).

https://doi.org/10.1016/j.ijepes.2023.109615
Received 23 June 2023; Received in revised form 2 October 2023; Accepted 23 October 2023
Available online 16 November 2023
0142-0615/© 2023 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Table 1 Table 1 (continued )

Performance of optimization techniques in PSS design. Year Optimization Technique Problem Solutions
Year Optimization Technique Problem Solutions
of speed deviation based objective
1996 Genetic Algorithm (GA) [13] PSS designed for Single Machine function and results compared with
Infinite Bus (SMIB) system for some GA and CPSS.
operating cases 2020 Slime Mould Algorithm [30] PSS designed for specific operating
1999 Tabu Search (TS) Algorithm [14] PSS designed for specific operating cases of SMIB system using
cases of SMIB system and MMPS minimization of squared eigenvalue
using minimization of eigenvalue based multi-objective function and
based single objective function. results compared with GA.
2000 Simulated Annealing (SA) PSS designed for specific operating 2021 Henry Gas Solubility Algorithm PSS designed for specific operating
Algorithm [15] cases of MMPS using minimization [31] cases of SMIB system using
of eigenvalue based single objective minimization of speed deviation
function. based objective function and results
2002 Particle Swarm Optimization (PSO) PSS designed for specific operating compared with Atomic Search
Algorithm [16] cases of MMPS using minimization Optimization (ASO).
of eigenvalue based multi-objective 2021 Harris Hawk Optimization PSS designed for specific operating
function. Algorithm [32] cases of MMPS using maximization
2002 Evolutionary Programming PSS designed for specific operating of damping ratio based objective
Algorithm [17] cases of MMPS using minimization function and results compared with
of squared eigenvalue based single- differential evolution.
objective function. 2022 Reptile Search Algorithm (RSA) PSS designed for specific operating
2007 Bacteria Forging Algorithm (BFA) PSS designed for specific operating [33] cases of SMIB system and MMPS
[18] cases of MMPS using minimization using minimization of speed
of squared eigenvalue based single deviation based objective function
and multi-objective functions. and results compared with GA and
2010 Chaotic Algorithm (CA) [19] PSS designed for specific operating BA.
cases of MMPS using minimization 2022 Improved Atomic Search PSS designed for specific operating
of squared eigenvalue based multi- Optimization (IASO) Algorithm cases of SMIB system using
objective function. [34] minimization of speed deviation
2012 Ant Colony Algorithm [20] PSS designed for specific operating based objective function and results
cases of MMPS using minimization compared with GA, PSO, SA and
of squared eigenvalue based multi- ASO.
objective function and results 2022 Hybrid Water Cycle Moth-Flame PSS designed for specific operating
compared with GA, PSO and CA. Optimization Algorithm [35] cases of SMIB system using
2013 Cultural Algorithm [21] PSS designed for specific operating minimization of speed deviation
cases of MMPS using minimization based objective function and results
of squared eigenvalue based multi- compared with water cycle
objective function and results algorithm, moth-flame optimization
compared with GA. and artificial ecosystem
2014 Bat Algorithm (BA) [22] PSS designed for specific operating optimization.
cases of MMPS using minimization 2022 Rat Swarm Optimization Algorithm PSS designed for specific operating
of squared eigenvalue based multi- [36] cases of MMPS using minimization
objective function and results of squared eigenvalue based multi-
compared with GA and CPSS. objective function and results
2015 Orthogonal Learning Artificial Bee PSS designed for specific operating compared with CPSS.
Colony (ABC) Algorithm [23] cases of SMIB system using 2022 Revamped Sine Cosine Algorithm PSS and POD designed for specific
minimization of speed deviation [37] operating cases of MMPS with DFIG
based objective function and results using minimization of squared
compared with ABC and CPSS. eigenvalue based multi-objective
2016 Chaotic Teaching Learning PSS designed for specific operating function and results compared with
Algorithm (TLA) [24] cases of MMPS using minimization whale optimization algorithm.
of speed deviation based objective 2023 Weighted Mean of Vectors PSS designed for specific operating
function and results compared with Algorithm [38] cases of SMIB system using
GA and TLA. minimization of speed deviation
2016 Cuckoo Search Algorithm [25] PSS designed for specific operating based objective function and results
cases of MMPS using minimization compared with grey wolf
of squared eigenvalue based multi- optimization, BA, RSA and IASO.
objective function and results 2023 Mayfly Algorithm [39] PSS designed for specific operating
compared with GA and CPSS. cases of MMPS using maximization
2017 Back Tracking Search Algorithm PSS designed for specific operating of damping ratio based objective
[26] cases of MMPS using minimization function and results compared with
of squared eigenvalue based multi- firefly algorithm and PSO.
objective function and results 2023 Hybrid Gorilla Troops Optimization PSS designed for specific operating
compared with PSO and BFA. (GTO) and Gradient-Based cases of MMPS using maximization
2018 Salap Swarm Algorithm (SSA) [27] PSS designed for specific operating Optimization (GBO) Algorithm of damping ratio based objective
cases of MMPS using minimization [40] function and results compared with
of squared eigenvalue based multi- AEO, GTO and GBO.
objective function and results
compared with TS.
2018 Ant Lion Algorithm [28] PSS designed for specific operating meta-heuristic techniques. Performance of optimization techniques
cases of MMPS using minimization employed in literature for PSS designing are depicted in Table 1.
of speed deviation based objective
However, the meta-heuristic techniques suffer from the issue of high
function and results compared with
PSO and SSA. run time, and slow convergence depending on the system size under
2019 Hyper-Spherical Search Algorithm PSS designed for specific operating study, but these are off-line techniques and free from mathematical
[29] cases of MMPS using minimization modelling. The main advantage of meta-heuristic techniques is that they
are based on evolutionary schemes and information exchange between

2
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 1. Flow Chart of STO.

individuals. 2. Formulation of problem

In this paper, four bio-inspired optimization techniques: Sooty-Tern
Optimization (STO) [41], GWO [42], GA [43], and PSO [44] techniques 2.1. Model of power system
have been explored for designing the PSS of MMPS. The tuning pro­
cedure is modeled as a multi-objective based on eigenvalue modes In this work, every generator is modelled as a two-axis, fourth-order
optimization problem for relocating the unstable right-half of s-plane model. For all selected operating settings, the power system can be
eigenvalues to a definite stable D-shape sector in the left-half of the s- modelled as a set of non-linear differential equations given by
plane. To depict the performance of the designed PSS, these algorithms
Ẋ = f (X, U) (1)
are successfully tested on two benchmark test systems: sixteen-machine,
sixty-eight-bus New England Extended Power Grid (NEEPG) and three- T
machine, nine-bus Western System Coordinating Council (WSCC) for where state vector X = [δ, w, E′q , Efd ] is a set of variables: rotor angle δ,
various operating cases under decisive perturbations. The effectiveness rotor speed w, internal voltage E′q and the field voltage Efd respectively
of tuned PSS is checked by non-linear simulations, performance indices: and input variables vector U. Moreover, the system is assumed to be a
Integral of Absolute Error (IAE) & Integral of Time-Absolute Error linearized incremental model around a particular operating point which
(ITAE), and eigenvalue analysis using the Power System Analysis is commonly utilized for designing of PSS [17–22].
Toolbox (PSAT) [45] and matched with each other. The results depict
the hat designed STO-stabilizer guarantees to produce robust damping 2.2. Modelling of PSS
performance for an extensive variety of operating settings as well as
hidden operating cases also under critical disturbance as compared to The elementary task of PSS is to utilize an auxiliary stabilizing signal
GA-stabilizer, PSO-stabilizer and GWO-stabilizer. The key feature of this that helps in providing damping for generator rotor oscillations by
paper is that the designed parameters of PSS using all algorithms for regulating its field excitation. To offer sufficient damping, PSS produces
selected operating cases show the robustness performance by testing a constituent of electrical torque in phase with rotor speed deviations
them under critical disturbance for other severe operating cases also. [17–40]. The basic transfer function of PSS includes (i) damping

3
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 2. Flow Chart of GWO.

Table 2 Table 3
Three Generator and Load Operating Cases of WSCC Power System [25]. Six Operating Conditions of NEEPG [49].
Operating Cases Generator Load Cases Operating Conditions

G1 G2 G3 A B C Load at Case- Nominal generation and load

G1 1
Case- Transmission line 1–2 out of service
Normal P 0.71 1.63 0.06 1.25 0.90 1.00 1.00
2
Loading Q 0.62 0.85 –0.10 0.50 0.30 0.35 0.35
Case- Transmission line 1–27 out of service
3
Light P 0.96 1.00 0.45 0.70 0.50 0.60 0.60 Case- Transmission line 8–9 out of service
Loading Q 0.22 – 0.19 – 0.26 0.35 0.30 0.20 0.20 4
Case- Enhance 20 % load at bus-17
5
Heavy P 3.57 2.20 0.71 2.00 1.80 1.60 1.60 Case- Transmission line 46–49 out of service with load enhanced by 25 % at bus-
Loading Q 1.81 1.35 0.43 0.90 0.60 0.65 0.65 6 20, 21 and generation raised by 20 % at G9

controller gain (ii) washout time constant (iii) two stages lag-lead between excitation input and electrical torque output for a wide range of
compensator. The standard transfer function of PSS is specified as: inter-area and local area modes of oscillations. The main design problem
[
sTw
][
(1 + sT1i ) (1 + sT3i )
] of PSS involves the proper selection of dynamic gain Ki and time co­
ΔUi = Ki Δwi (2) efficients Tw, T1i, T2i, T3i, and T4i for ith machine. For the sake of
1 + sTw (1 + sT2i ) (1 + sT4i )
simplicity, the numerical values of Tw, T2i, and T4i are preferred as fixed
The phase lead block provide necessary phase lag compensation constant value while other parameters Ki and T1i and T3i values are to be

4
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Table 4
Eigenvalues Analysis without PSS for Three Loading Cases of WSCC System.
Cases Eigenvalues Damping Ratio (%) Frequency (p. u.) Participation Factor Participation
Modes

Normal Loading – 0.110 ± j 8.588 12.0 1.366 0.290 w2, δ 2

– 0.653 ± j 13.023 5.02 2.072 0.374 w3, δ 3

Light Loading – 0.637 ± j 8.515 7.40 1.355 0.278 w2, δ 2

– 1.274 ± j 12.752 9.90 2.029 0.355 w3, δ 3

Heavy Loading 0.158 ± j 8.372 – 1.80 1.332 0.288 w2, δ 2

– 0.308 ± j 12.896 2.40 2.052 0.384 w3, δ 3

np ∑
∑ np ∑
∑
Table 5 ( )2 ( )2
J= σ0 − σ i,j + ξ0 − ξi,j (3)
Eigenvalues Analysis without PSS for Six Operating Cases of NEEPG System. j=1 σi,j ≥σ0 j=1 ξi,j ≤ξ0
Cases Eigenvalues & Frequency Participation Participation
Percentage (p. u.) Factor Modes where np, ξ0, σ0, ξi,j and σ i,j are the number of operating cases to be
Damping Ratio
selected, chosen damping ratio, chosen damping factor, the damping
Case- 0.388 ± j 6.439, 1.024 0.350 w9, δ9 ratio and the real-part of the ith eigenvalue mode of the jth operating
1 –6.02 % case respectively.
0.030 ± j 6.662, – 1.060 0.189 w2, δ2
0.45 %
Minimize J subject to:
0.009 ± j 11.306, 1.799 0.440 w11, δ11
Kimin ≤ Ki ≤ Kimax (4)
– 0.085 %

T1imin ≤ T1i ≤ T1imax (5)

Case- 0.358 ± j 6.411, ¡ 1.020 0.325 w9, δ9
2 5.58 %
0.032 ± j 6.658, ¡ 1.059 0.195 w2, δ2 T3imin ≤ T3i ≤ T3imax (6)
0.48 %
0.023 ± j 11.248, 1.790 0.442 w11, δ11 where and Tjimax and Tjimin the upper and lower bounds of time-coefficients
¡ 0.20 %
of PSS design, Kmax
i and Kmin
i are upper and lower bounds of PSS design
gain.
Case- 0.382 ± j 6.428, – 1.023 0.348 w9, δ9
3 5.94 %
0.030 ± j 6.661, – 1.060 0.195 w2, δ2 3. Meta-heuristic optimization techniques
0.45 %
0.012 ± j 11.306, 1.799 0.440 w11, δ11 3.1. Sooty-tern optimization algorithm
– 0.11 %
Case- 0.410 ± j 6.400, – 1.018 0.363 w9, δ9
4 6.39 % Sooty terns named as onychoprion fucatus are the sea birds which
0.016 ± j 6.450, – 1.026 0.192 w2, δ2 are found on banks of the oceans proposed by Gaurav Dhiman et al [41].
0.26 % These sooty terns are found in broad range and groups, with various
0.017 ± j 11.285, 1.796 0.439 w11, δ11 sizes and mass. These terns are omnivores which eat fish, spiders,
– 0.15 %
Case- 0.387 ± j 6.437, – 1.024 0.350 w9, δ9
earthworms, reptiles, amphibians and likewise little insects of ocean.
5 6.00 % Sooty tern is a very shrewd bird that lures earthworms which are deep
0.031 ± j 6.663, – 1.060 0.196 w2, δ2 beneath the soil by making a rain-like noise with the feet and use bread
0.47 % crumbs to lure fish around. Mostly these birds spend their lives in the
0.014 ± j 11.315, 1.800 0.441 w11, δ11
groups. Knowledge is shared by ranging and assaulting the prey. The
– 0.12 %
most important thing of sooty terns is to migrate and assaulting actions.
Immigration is known as seasonal shifting of sooty terns from one
Case- 0.585 ± j 6.312, – 1.004 0.342 w9, δ9
6 9.23 %
location to another for exploring the wealthiest and most plentiful areas
0.017 ± j 6.666, – 1.061 0.192 w2, δ2 of resources which will supply enough food and power [41]. Hunting
0.26 % process of sooty bird can be understood as follows:
0.010 ± j 11.284, 1.795 0.437 w11, δ11
– 0.08 %
1. Sooty terns fly in a flock throughout immigration. The primary
location of sooty terns goes varying to discourage collision in be­
evaluated [17–40]. tween them.
2. Sooty terns fly as a team towards the best sooty terns direction, so
2.3. Optimization problem that the sooty tern having lowest fitness level can also travel to the
best solution.
For simultaneous regulation of both damping ratio and real-part of 3. Remaining sooty terns will change their places according to the
eigenvalues, the tuned parameters of PSS may be planned to minimize fittest one.
the following multi-objective function in such a way that the eigen­
values of MMPS with planned PSS are significantly relocated in a D- The execution of the STO algorithm for tuning the PSS parameters
shape sector of the stable zone in the s-plane [17–22,25–27,30,36,37]. are described as shown in Fig. 1 through the flow chart representation:

5
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 3. Convergence Characteristics of various optimization techniques for (a) WSCC System (b) NEEPG System.

Table 6 Table 7
Statistical Results of GA, PSO, GWO and STO Algorithm for WSCC and NEEPG Optimized Parameters of GA-stabilizer, PSO-stabilizer,GWO-stabilizer and STO-
System. stabilizer for WSCC system.
Desired Convergence Best Average Generators GA- PSO- GWO- STO-
Solution Time (sec) Fitness Fitness stabilizer stabilizer stabilizer stabilizer
Iteration [47] [47] [47]

For GA 62 629 0 0 K G2 1.000 1.000 5.372 3.370

WSCC PSO 42 458 0 0 G3 1.000 1.000 1.000 1.000
System GWO 28 321 0 0
STO 18 258 0 0
T1 G2 0.464 1.000 0.355 0.212
G3 0.610 0.400 0.155 0.180
For GA 93 2758 0 0
NEEPG PSO 65 2146 0 0
T3 G2 0.060 0.156 1.000 0.060
System GWO 42 1863 0 0
G3 0.679 0.06 0.060 0.102
STO 32 1582 0 0

3.2. Grey Wolf Optimization algorithm shown in Fig. 2 through the flow chart representation:

The Grey Wolf Optimization (GWO) technique was created by usual 4. Design and simulation results of PSS
behavior of wolf pack and developed by Mirjalili et al [42]. Grey wolf is
the apex predator in food chain and devote life in the groups. They have 4.1. Three-machine, nine-bus WSCC power system
preserved a harsh community behavior and work in a group. They are
characterized into four brands as: Alpha, Beta, Delta, and Omega based The layout of a well-known 3-machine, 9-bus WSCC system and its
on their role played in hunting. The three main stages of hunting are data are specified in [46]. All three generators of WSCC power system
look for prey, surrounding prey and aggressive prey [42]. The key merits are modelled as: fourth order with static exciter and constant impedance
of GWO are that it has very less parameters to be tuned and also do not loads. Table 2 shows three different operating cases for which the study
require any derivative information during initialization. The execution is carried out [25].
of the GWO algorithm for tuning the PSS parameters are described as

6
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Table 8
Optimized Parameters of GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for NEEPG System.
With GA-stabilizer With PSO-stabilizer With GWO-stabilizer With STO-stabilizer

K1 T1 T3 K1 K1 K1 K1 T1 T3 K1 T1 T3

G1 76.425 0.162 0.888 31.388 0.785 0.501 76.499 0.105 0.906 68.815 0.254 1.000
G2 26.315 0.551 0.510 59.808 0.100 0.345 42.591 0.261 0.563 40.920 0.092 0.352
G3 47.143 0.323 0.572 30.629 0.763 0.010 22.364 0.863 0.642 37.496 0.209 0.076
G4 17.558 0.994 0.509 23.942 0.100 0.100 27.451 0.511 0.267 18.839 0.235 0.072
G5 59.035 0.281 0.329 63.755 0.100 0.281 23.563 0.653 0.823 8.043 0.201 0.699
G7 10.756 0.814 0.677 13.813 0.913 0.693 100.00 0.229 0.435 96.216 0.085 0.306
G8 52.683 0.431 0.109 44.771 0.772 0.163 18.225 0.418 0.767 15.662 1.000 0.301
G9 26.648 0.520 0.489 44.431 0.709 0.329 50.159 0.430 0.213 19.437 0.684 0.219
G10 75.727 0.549 0.245 29.253 0.826 0.813 8.279 0.242 1.000 60.817 0.150 0.855
G11 15.281 0.310 0.249 12.186 0.564 0.615 35.328 0.162 0.124 7.897 0.064 1.000
G12 4.496 0.766 0.687 88.476 0.100 0.594 67.328 0.071 0.382 8.664 0.632 0.158
G13 37.969 0.512 0.235 52.810 0.100 0.325 23.146 0.116 0.502 42.864 0.3847 0.305
G15 38.367 0.388 0.263 13.955 0.766 0.297 31.542 0.580 0.152 100.000 0.397 0.266
G16 42.826 0.478 0.878 26.569 0.100 0.100 74.559 0.329 0.269 60.784 1.000 0.373

Table 9 Table 10
Eigenvalues Analysis with Designed GA-stabilizer, PSO-stabilizer, GWO- Eigenvalues Analysis with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and
stabilizer, STO-stabilizer for Three Loading Cases of WSCC System. STO-stabilizer, for six-operating cases of NEEPG System.
Normal loading Light Loading Heavy Loading Cases With GA- With PSO- With GWO- With STO-
stabilizer stabilizer stabilizer stabilizer
With GA- – 1.778 ± j 8.323, – 1.659 ± j 7.724, – 0.961 ± j 7.148,
stabilizer 20.9 % 21.0 % 13.3 % Case- – 0.940 ± j – 1.109 ± j – 0.894 ± j – 0.586 ± j
[47,28] – 1.887 ± j 7.160, – 2.811 ± j 7.480, – 1.930 ± j 8.508, 1 7.086, 3.1 % 10.400, 10.6 % 7.047, 12.5 % 3.670, 15.7 %
25.4 % 35.1 % 22.1 % – 0.744 ± j – 1.059 ± j – 1.772 ± j – 1.824 ± j
3.996, 18.3 % 6.923, 15.1 % 12.969, 13.5 % 11.300, 15.9 %
– 0.601 ± j – 0.641 ± j – 0.827 ± j – 2.073 ± j
With PSO- – 1.212 ± j 7.549, – 1.614 ± j 7.563, – 0.768 ± j 7.381,
2.214, 26.2 % 2.628, 23.6 % 3.998, 20.2 % 10.421, 19.5 %
stabilizer 15.8 % 20.8 % 10.3 %
[47,28] – 2.007 ± j – 2.669 ± j – 1.570 ± j
14.393, 13.8 % 14.041, 35.1 % 14.157, 11.0 % Case- – 0.969 ± j – 1.127 ± j – 0.920 ± j – 0.589 ± j
2 7.081, 13.5 % 10.395, 10.7 % 7.039, 12.9 % 3.669, 15.91 %
– 0.750 ± j – 1.106 ± j – 0.830 ± j – 1.837 ± j
With GWO- – 2.008 ± j 7.363, – 2.235 ± j 7.598, – 1.561 ± j 7.244,
3.996, 18.4 % 6.927, 15.7 % 3.980, 20.4 % 11.290, 16.0 %
stabilizer[47] 26.3 % 28.2 % 21.0 %
– 0.655 ± j – 0.673 ± j – 0.694 ± j – 1.248 ± j
– 2.619 ± j – 3.351 ± j – 1.944 ± j
2.214, 28.3 % 2.595, 25.1 % 2.112, 31.2 % 7.014, 17.5 %
17.189, 15.0 % 17.010, 19.3 % 17.526, 11.0 %

Case- – 0.946 ± j – 1.113 ± j – 0.899 ± j – 0.585 ± j

With STO- – 1.945 ± j – 2.719 ± j – 2.284 ± j
3 7.081, 13.2 % 10.393, 10.6 % 7.040, 12.6 % 3.668, 15.7 %
stabilizer 12.189, 15.7 % 11.675, 22.6 % 17.348, 13.0 %
– 0.746 ± j – 1.069 ± j – 1.759 ± j – 1.828 ± j
– 2.135 ± j – 2.658 ± j – 1.518 ± j
3.992, 18.3 % 6.915, 15.2 % 13.055, 13.3 % 11.295, 15.9 %
17.161, 12.3 % 16.729, 15.6 % 12.520, 12.0 %
– 0.645 ± j – 0.640 ± j – 0.825 ± j – 1.224 ± j
2.218, 27.9 % 2.624, 23.7 % 3.986, 20.2 % 6.987, 17.2 %

4.2. New England Extended power Grid (NEEPG) system comprising of

16-Machine, 68-Bus Case- – 0.959 ± j – 1.124 ± j – 0.913 ± j – 0.585 ± j
4 7.089, 13.4 % 10.407, 10.7 % 7.048, 12.8 % 3.672, 15.7 %
– 0.746 ± j – 1.103 ± j – 1.765 ± j – 1.809 ± j
Layout of standard sixteen-machine, sixty-eight-bus New England 3.994, 18.3 % 6.948, 15.6 % 13.085, 13.3 % 11.282, 15.8 %
Extended Power Grid (NEEPG) and its data are given in [48], [49]. – 0.652 ± j – 0.679 ± j – 0.824 ± j – 1.261 ± j
Table 3 presents six different operating cases including under brutal and 2.225, 28.1 % 2.595, 25.3 % 3.989, 20.2 % 7.070, 17.5 %
decisive line outage condition for which the study is carried out [24].
Case- – 0.941 ± j – 1.110 ± j – 0.895 ± j – 0.586 ± j
5 7.086, 13.1 % 10.398, 10.6 % 7.047, 12.6 % 3.671, 15.7 %
4.3. Eigenvalue analysis without PSS – 0.744 ± j – 1.060 ± j – 1.760 ± j – 1.825 ± j
3.997, 18.3 % 6.923, 15.1 % 13.064, 13.3 % 11.298, 15.9 %
The participation factor [50] can be used to identify the various – 0.601 ± j – 0.640 ± j – 0.827 ± j – 1.213 ± j
2.215, 26.2 % 2.629, 23.6 % 3.999, 20.2 % 6.989, 17.1 %
modes of oscillations and their detail data for unstable and/or margin­
ally stable modes of three loading cases of WSCC system and six oper­
ating cases of NEEPG system are depicted in Tables 4 and 5 respectively. Case- – 0.939 ± j – 1.111 ± j – 0.893 ± j – 0.594 ± j
6 7.105, 13.1 % 10.483, 10.5 % 7.068, 12.5 % 3.608, 16.2 %
On analysing Table 4, it is observed that WSCC system is charac­
– 0.761 ± j – 1.047 ± j – 1.768 ± j – 1.776 ± j
terized by two local-area modes and is unstable due to large value of 3.953, 18.9 % 6.959, 14.8 % 13.042, 13.4 % 11.359, 15.4 %
negative percentage damping ratio under heavy loading case as – 0.623 ± j – 0.639 ± j – 0.847 ± j – 1.179 ± j
compared to other operating cases. Thus, according to high participation 2.186, 27.4 % 2.635, 23.5 % 3.948, 20.9 % 7.000, 16.6 %
factor [50] of two-generators G2, G3, they were equipped with PSS for
inserting damping to local area modes.
owing to large value of negative percentage damping ratio in compari­
On analysing Table 5, it is observed that NEEPG system has three
son to other operating cases. Thus, according to high participation factor
inter-area modes, eleven local-area modes and is unstable due to three-
[50], fourteen-generators except G6 and G14 were equipped with PSS to
pairs of eigenvalues modes with negative damping for the chosen six-
insert damping for both inter and local area modes.
operating cases. Moreover, the operating Case-6 is extremely unstable

7
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 4. Eigenvalue Map analysis with No-PSS and Designed PSS for three loading cases of WSCC System.

4.4. Eigenvalue analysis with designed PSS depicted in the Appendix. Typical convergence of GA-stabilizer, PSO-
stabilizer, GWO-stabilizer and STO-stabilizer for WSCC system and
To stabilize the systems, six-PSS parameters of WSCC system and NEEPG system are illustrated in Fig. 3 (a) and (b) respectively.
forty-two parameters of NEEPG system are designed considering by Fig. 3 shows that all algorithms are capable to discover the desired
minimization of objective function J depicted in equation (3) using GA, solution for which fitness function J is zero. The figure shows that the
PSO, GWO and STO techniques. For both case studies, real-part of ei­ STO algorithm discover the best solution at a faster rate compared with
genvalues σ0 is set as − 0.5 and damping ratio ξ0 is set as 10 %. The target that for GWO, PSO, and GA for WSCC system and NEEPG system. The
value of J = 0 causes the system either unstable or marginally stable. statistical results of GA, PSO, GWO and STO Algorithm for WSCC and
Eigenvalue modes of the system are relocated to a desired D-shape stable NEEPG System are depicted in Table 6.
sector in left-half of s-plane to assure the system stability. Hence, for The final optimized 6-parameters of WSCC and 42 parameters of
optimizing six-PSS parameters of WSCC system and forty-two parame­ NEEPG systems using GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and
ters of NEEPG system, the gain Ki of PSS is varied from 1 to 100 and, the STO-stabilizer are listed in Tables 7 and 8 respectively. The assessment
lower and upper bounds of T1i and T3i are set at 0.01 and 1.0 respectively of eigenvalues and percentage damping ratio with planned GA-
[25]. To reduce the computation burden, the washout time constants Tw, stabilizer, PSO-stabilizer, GWO-stabilizer and STO-stabilizer for three-
T2 and T4 are kept constant at 5 sec, 0.05 sec and 0.05 sec respectively. operating cases of WSCC system and six-operating cases of NEEPG sys­
The control parameters of GA, PSO, GWO and STO algorithm are tem are listed in Tables 9 and 10 respectively. Eigenvalue map analysis

8
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 5. Eigenvalue Map analysis with No-PSS and Designed PSS for six operating cases of NEEPG System.

of WSCC system and NEEPG system with No-PSS and GA-stabilizer, PSO- using the STO-stabilizer is better than the stabilizers planned using GA,
stabilizer, GWO-stabilizer and STO-stabilizer are depicted in Figs. 4 and PSO, and GWO techniques.
5 for selected operating cases respectively.
It is clear from Tables 9 and 10, Figs. 4 and 5 that eigenvalues of the
designed GWO-stabilizer and STO-stabilizer are not only repositioned 4.5. Simulation results
from the unstable and/or lightly damped oscillations zone to signifi­
cantly far away from a selected D-shape stable zone in the s-plane but To analyse the performance of the planned STO-stabilizer described
also shift other oscillation modes to the left-half of s-plane as compared in section 4.4 for WSCC and NEEPG systems, decisive perturbations are
to same obtained by GA-stabilizer and PSO-stabilizer for selected oper­ selected, and their performance is matched with the performance of GA-
ating cases of WSCC and NEEPG systems respectively. From the out­ stabilizer, PSO-stabilizer, and GWO-stabilizer. The results are shown in
comes, it is clear that the damping performance of the stabilizer planned Table 11.
The responses of the generator speed deviations Δw12, Δw23, and

9
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 5. (continued).

Δw31 without employing PSS for Strategy-1 of heavy loading case of

Table 11
WSCC system and all sixteen generator speed deviations for Strategy-2
Critical Perturbation for Testing the Performance of Planned PSS.
of critical operating Case-6 of NEEPG system are illustrated in Figs. 6
Strategies Most Critical Disturbances (a) and 7 (a) respectively whereas for the same strategies with planned
Strategy- A 6-cycle, 3-phase short circuit fault occur at t = 1 sec on bus 7 by GA-stabilizer, PSO-stabilizer, GWO-stabilizer and STO-stabilizer,
1 tripping the lines 5–7 of the WSCC system generator speed deviations Δw12, Δw23 and Δw31 of WSCC system and
Strategy- A 6-cycle, 3-phase short circuit fault occur at t = 1 sec on bus 21 critical generator speed deviations Δw5, Δw6 and Δw9 of NEEPG system
2 without tripping the lines 21–22 of the NEEPG system
are illustrated in Figs. 6 (b)-(d) and 7 (b)-(d) respectively.

10
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 6. Generator speed deviations (a) without-PSS and (b) Δw12 (c) Δw23 (d) Δw31 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for Strategy-
1 of heavy loading case for WSCC system.

Fig. 7. Generator speed deviations (a) without PSS and (b) Δw5 (c) Δw6, (d) Δw9 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for Strategy-2
of operating Case-6 of NEEPG system.

11
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 8. Comparison of performance indices (a) IAE (b) ITAE for three loading cases of WSCC system with Designed GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and
STO-stabilizer.

Fig. 9. Comparison of performance indices (a) IAE (b) ITAE for six operating cases of NEEPG system with Designed GA-stabilizer, PSO-stabilizer, GWO-stabilizer,
STO-stabilizer.

On comparison of Figs. 6 (a) and 7 (a), it is clear that with the stabilizer for Strategy-1 of WSCC system and Strategy-2 of NEEPG sys­
increment of load, the size and duration of oscillations of all three tem are depicted in Figs. 8 (a)-(b) and 9 (a)-(b) respectively.
generators increase in the same direction and are oscillatory in nature, These bar-charts indicate that the performance of the STO-stabilizer
and finally all generators lose synchronism. Furthermore, it is noticed is better than the other three stabilizers for the selected decisive
from Figs. 6 (b)-(d) and 7 (b)-(d) that speed deviations for all generators perturbation for both strategies. The performance of the GWO-stabilizer
with planned GWO-stabilizer and STO-stabilizer show fast decaying of and STO-stabilizer is almost similar for time domain specifications.
oscillations, settle fast and improve the relative stability as compared to Furthermore, it may be concluded that local-area and inter-area modes
other techniques. Furthermore, the performance of the planned STO- have been well stabilized with less overshoot, peak values, and settling
stabilizer is comparable with GWO-stabilizer and superior to GA- time using all planned PSS under selected decisive perturbations [47].
stabilizer and PSO-stabilizer.
A comparison of performance indices IAE and ITAE as bar-charts
with designed GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-

12
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Table 12 marginally stable mechanical modes of WSCC [51] and NEEPG systems
Three Hidden Operating Conditions of WSCC Power System [47]. are illustrated in Tables 14 and 15 respectively. Eigenvalue map analysis
Operating Cases Generator Load of WSCC system and NEEPG system with No-PSS and GA-stabilizer, PSO-
stabilizer, GWO-stabilizer, and STO-stabilizer are depicted in Figs. 10
G1 G2 G3 A B C
(a)-(c) and 11 (a)-(c) for selected hidden operating cases respectively.
Hidden Case-I P 0.33 2.00 1.50 1.50 1.20 1.00 From Table 14, it is seen that one-pair of eigenvalues without-PSS fall
Q 1.12 0.57 0.38 0.90 0.80 0.50
in the unstable sector of the s-plane and have a negative damping ratio
for all chosen hidden operating cases of the WSCC system. Moreover, the
Hidden Case-II P 1.09 2.45 1.27 1.90 1.30 1.50
hidden operating Case-3 is extremely unstable due to a high negative
Q 0.79 0.57 0.21 0.75 0.45 0.50
percentage damping ratio compared to other hidden cases as well as the
previous three operating cases for which the PSS are planned. From
Hidden Case-III P 1.41 2.60 1.20 2.00 1.50 1.60
Table 15, it is seen that the NEEPG system is unstable with two, two, and
Q 0.59 0.38 0.02 0.60 0.30 0.20
five pairs of eigenvalues without-PSS for hidden operating cases 1, 2,
and 3 respectively and they lie in the unstable zone of the s-plane with
Table 13 negative percentage damping ratio. The hidden operating Case-3 is
Three Hidden Operating Conditions of the NEEPG System. highly unstable due to a larger negative percentage damping ratio as
Cases Operating Conditions compared to other hidden operating cases as well as for six-operating
cases studied earlier.
Hidden Case- Total real and reactive power enhanced by 25 %
1 Eigenvalues and percentage damping ratio with earlier planned PSS
Hidden Case- Total real and reactive power reduced by 15 % using GA, PSO, GWO, and STO techniques for three hidden operating
2 cases of WSCC and NEEPG systems are evaluated using PSAT [45] and
Hidden Case- Transmission lines 1–31,10–11, 30–32, and 33–34 are out of are listed in Tables 16 and 17 respectively.
3 service
On analysing Tables 16 and 17, Figs. 10 and 11, it is revealed that
with planned GWO-stabilizer and STO-stabilizer, the eigenvalues are
5. Robustness performance analysis relocated far away from the selected D-shape stable sector in the s-plane
with superior damping performance as compared to PSS planned using
In sub-section 4.5, it is depicted that the planned GWO-stabilizer and GA and PSO techniques and ensure dynamic stability for all selected
STO-stabilizer are more effective than the PSS design using GA and PSO hidden operating cases also.
techniques. Therefore, to examine the impact of robustness performance To illustrate the impact of the robustness performance of STO-
of GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer on stabilizer for selected hidden operating cases for WSCC and NEEPG
WSCC and NEEPG systems; three hidden cases of WSCC [51] and NEEPG systems, decisive perturbations listed in Table 11 are selected for testing
systems are selected, and the operating conditions chosen are illustrated their simulation performance and compared with without stabilizer, GA-
in Tables 12 and 13 respectively. stabilizer, PSO-stabilizer, and GWO-stabilizer. Fig. 12 (a) illustrates the
In this section comparison of planned GWO-stabilizer, STO-stabilizer responses of generator speed deviations Δw12, Δw23 and Δw31 with No-
with GA-stabilizer, and PSO-stabilizer is done based on eigenvalue PSS for Strategy-1 of decisive hidden operating Case-3 of WSCC and
analysis, performance indices, and nonlinear simulation results. The Fig. 13 (a) shows all generator speed deviations with No-PSS for
eigenvalue analysis detail data without PSS for unstable and/or Strategy-2 of decisive hidden operating Case-3 of NEEPG system

Table 14
Eigenvalues Analysis without PSS for Three Hidden Operating Cases of WSCC System.
Eigenvalues Damping Ratio (%) Frequency (p. u.) Participation Factor Participation Modes

Hidden Case-1 0.341 ± j 8.339 – 4.00 1.327 0.269 w2, δ 2

– 0.109 ± j 12.803 0.85 2.037 0.363 w3, δ 3

Hidden Case-2 0.465 ± j 8.357 –5.50 1.330 0.272 w 2, δ 2

– 0.250 ± j 12.931 1.90 2.058 0.382 w3, δ 3

Hidden Case-3 0.604 ± j 8.375 – 7.20 1.333 0.270 w 2, δ 2

– 0.233 ± j 12.981 8.00 2.065 0.383 w3, δ 3

Table 15
Eigenvalues Analysis without PSS for Three Hidden Operating Cases of NEEPG System.
Eigenvalues Damping Ratio (%) Frequency (p.u.) Participation Factor Participation
Modes

Hidden Case-1 0.310 ± j 6.418 – 4.83 1.021 0.345 w9, δ 9

0.0005 ± j 6.628 – 0.008 1.055 0.185 w2, δ 2

Hidden Case-2 0.406 ± j 6.425 – 6.31 1.022 0.347 w9, δ 9

0.035 ± j 6.610 – 0.54 1.052 0.188 w2, δ 2

Hidden Case-3 0.386 ± j 6.420 – 5.99 1.021 0.352 w9, δ 9

0.329 ± j 5.485 – 5.99 0.873 0.350 w11, δ11
0.038 ± j 6.661 – 0.57 1.060 0.235 w2, δ 2
0.013 ± j 7.368 – 0.18 1.172 0.165 w2, δ 2
0.003 ± j 8.090 – 0.03 1.287 0.374 w10, δ10

13
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 10. Eigenvalue Map analysis with No-PSS and Designed PSS for hidden operating cases of (a)-(c) WSCC system.

whereas, for the same hidden operating cases, speed deviations Δw12, and 13 (b)-(d), it is noticed that system provided with planned GWO-
Δw23, and Δw31 with planned GA-stabilizer, PSO-stabilizer, GWO- stabilizer and STO-stabilizer, oscillations die down quickly thus
stabilizer, and STO-stabilizer provided with WSCC system are illus­ improving the relative stability as compared to GA-stabilizer and PSO-
trated in Fig. 12 (b)-(d) and critical generator speed deviations Δw5, Δw6 stabilizer.
and Δw9 of NEEPG system are depicted in Fig. 13 (b)-(d) respectively. To test the robust performance of the planned STO-stabilizer, a
From Figs. 12 (a) and 13 (a), it is clear that with a decisive line comparison of performance indices: IAE and ITAE are evaluated and
outage, the amplitude and duration of oscillations for all three genera­ plotted as bar-charts for Strategy-1 of the WSCC system and Strategy-2
tors increase indefinitely and finally lose synchronism. Fig. 12 (b)-(d) of NEEPG system in Figs. 14 (a)-(b) and 15 (a)-(b) respectively. It is

14
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 11. Eigenvalue Map analysis with No-PSS and Designed PSS for hidden operating cases of (a)-(c) NEEPG system.

15
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Table 16
Eigenvalues Analysis with Planned PSS for Three Hidden Operating Cases of WSCC System.
Hidden Case-1 Hidden Case-2 Hidden Case-3

With GA-stabilizer [47] – 0.766 ± j 7.225, 10.5 % – 1.228 ± j 8.052, 15.0 % – 0.746 ± j 8.283, 8.9 %
– 1.829 ± j 8.273, 21.5 % – 1.327 ± j 7.440, 17.5 % – 2.587 ± j 26.412, 9.7 %

With PSO-stabilizer [47] – 0.664 ± j 7.530, 8.7 % – 0.557 ± j 7.442, 7.4 % – 0.465 ± j 7.442, 6.2 %
– 1.565 ± j 13.977, 11.1 % – 1.587 ± j 14.234, 11.0 % – 1.495 ± j 14.387, 10.3 %

With GWO-stabilizer [47] – 1.311 ± j 7.403, 17.4 % – 1.516 ± j 7.229, 20.5 % – 1.547 ± j 7.106, 21.2 %
– 2.538 ± j 17.182, 14.6 % – 2.017 ± j 17.487, 11.4 % – 1.825 ± j 17.495, 10.3 %

With STO-stabilizer – 1.413 ± j 11.853, 11.8 % – 1.553 ± j 12.297, 12.5 % – 1.527 ± j 12.467, 12.1 %
– 2.220 ± j 16.966, 12.9 % – 2.065 ± j 17.316, 11.8 % – 1.913 ± j 17.369, 10.9 %

Table 17
Eigenvalues Analysis with Planned PSS for Three Hidden Operating Cases of NEEPG System.
Cases With GA-stabilizer With PSO-stabilizer With GWO-stabilizer With STO-stabilizer

Hidden Case-1 – 0.688 ± j 2.229, 29.5 % – 0.640 ± j 2.649, 23.5 % – 0.681 ± j 2.208, 29.5 % – 0.557 ± j 3.771, 29.5 %
– 0.706 ± j 4.035, 17.2 % – 0.768 ± j 4.029, 18.7 % – 0.773 ± j 4.028, 18.8 % – 1.153 ± j 7.027, 16.1 %

Hidden Case-2 – 0.953 ± j 7.048, 13.4 % – 1.121 ± j 10.319, 10.8 % – 1.556 ± j 12.160, 12.6 % – 1.280 ± j 6.965, 18.0 %
– 0.810 ± j 3.862, 20.5 % – 0.818 ± j 4.148, 19.3 % – 0.910 ± j 7.001, 12.8 % – 0.651 ± j 3.415, 18.7 %
– 0.539 ± j 2.032, 25.6 % – 0.633 ± j 2.642, 23.3 % – 0.696 ± j 2.174, 30.4 % – 0.754 ± j 1.702, 40.5 %

Hidden Case-3 – 0.956 ± j 7.077, 13.3 % – 1.123 ± j 10.385, 10.7 % – 1.761 ± j 12.765, 13.6 % – 1.868 ± j 11.366, 16.2 %
– 0.754 ± j 7.084, 18.6 % – 1.085 ± j 6.919, 15.5 % – 0.905 ± j 7.035, 12.7 % – 1.256 ± j 6.988, 17.6 %
– 0.831 ± j 3.084, 26.0 % – 0.891 ± j 4.181, 20.8 % – 0.842 ± j 3.995, 20.8 % – 1.777 ± j 9.915, 17.6 %
– 0.750 ± j 2.434, 29.4 % – 0.512 ± j 2.937, 17.1 % – 0.855 ± j 2.226, 35.8 % – 0.842 ± j 1.753, 43.2 %
– 0.624 ± j 2.188, 27.4 % – 0.472 ± j 2.416, 19.1 % – 0.606 ± j 2.177, 26.8 % – 0.593 ± j 3.623, 16.1 %

clear from Figs. 14 and 15, that the numerical values of both indices for different shunt and series FACTS damping controllers for improving the
systems provided with planned STO-stabilizer are lowest as compared to small-signal stability of power systems.
GA-stabilizer, PSO-stabilizer, and GWO-stabilizer for decisive pertur­
bations of three hidden operating cases. CRediT authorship contribution statement
It can also be concluded that local-area and inter-area modes have
been well stabilized with less overshoot, peak values, and settling time Dhanraj Chitara: Conceptualization, Formal analysis, Investigation,
using all planned PSSs for both chosen loading cases as well as hidden Visualization, Writing – original draft, Writing – review & editing. P.K.
operating cases of WSCC and NEEPG systems. Singhal: Investigation, Visualization, Writing – original draft, Writing –
review & editing. S.L. Surana: Investigation, Visualization, Writing –
6. Conclusion and future scope original draft, Writing – review & editing. Gulshan Sharma: Data
curation, Methodology, Supervision, Writing – review & editing. R.C.
This paper offering a comparative analysis of bio-inspired meta- Bansal: Conceptualization, Data curation, Methodology, Project
heuristic optimization techniques: STO, GWO, PSO, and GA for administration, Supervision, Validation.
designing robust PSS for the MMPS. The tuning method is considered a
multi-objective optimization problem for relocating unstable right-half
Declaration of Competing Interest
of s-plane eigenvalues to a definite stable D-shape sector in the left-
half of the s-plane. To check the performance of the designed PSS,
The authors declare that they have no known competing financial
these techniques are effectively tested on two benchmark test systems:
interests or personal relationships that could have appeared to influence
sixteen-machine, sixty-eight-bus New England Extended Power Grid
the work reported in this paper.
(NEEPG) and three-machine, nine-bus Western System Coordinating
Council (WSCC) for an extensive variety of operating conditions under
Data availability
critical perturbations. The superiority of the STO-stabilizer is revealed
by analyzing its performance using non-linear simulations, performance
Data will be made available on request.
indices, and eigenvalue analysis and by comparing it with GA-stabilizer,
PSO-stabilizer, and GWO-stabilizer. The results prove that the designed
Appendix
STO-stabilizer produces better damping performance for a wide range of
operating conditions as well as for hidden operating cases under critical
Static exciter for WSCC system: Ka = 100, Ta = 0.05 sec, for NEEPG
disturbance as compared to others. This research work can also be
system: Ka = 50, Ta = 0.001 sec
extended to advance coordinated tuning of PSS parameters with
The control parameters of GA, PSO, GWO and STO algorithm are

16
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 12. Generator speed deviations (a) without-PSS and (b) Δw12 (c) Δw23 (d) Δw31 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for
Streatgy-1 of hidden operating Case-3 of WSCC system.

Fig. 13. Generator speed deviations (a) without-PSS and (b) Δw5 (c) Δw6 (d) Δw9 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer and STO-stabilizer for Strategie-2
of hidden operating Case-3 of NEEPG system.

17
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

Fig. 14. Comparison of performance indices (a) IAE and (b) ITAE for Strategy-1 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for hidden
operating cases 1–3 of WSCC system.

Fig. 15. Comparison of performance indices (a) IAE and (b) ITAE for Strategy-2 with GA-stabilizer, PSO-stabilizer, GWO-stabilizer, and STO-stabilizer for hidden
operating cases 1–3 of NEEPG system.

given by: [12] Radaideh SM, Nejdawi IM, Mushtaha MH. Design Of Power System Stabilizers
Using Two-Level Fuzzy And Adaptive Neuro-Fuzzy Inference Systems. Int J Electr
Population = 70, Number of iteration = 70, Number of individuals =
Power Energy Syst 2012;35(1):47–56.
06 (WSCC System) [13] Abdel-Magid YL, Dawoud MM. Tuning of power system stabilizers using genetic
Population = 100, Number of iteration = 100, Number of in­ algorithms. Electr Pow Syst Res 1996;39(2):137–43.
dividuals = 42 (NEEPG System). [14] Abido MA. A Novel Approach to Conventional Power System Stabilizer Design
Using Tabu Search. Int J Electr Power Energy Syst 1999;21(6):443–54.
GA: Crossover rate = 0.75 (WSCC System), 0.80 (NEEPG System), [15] Abido MA. Robust Design Of Multi-Machine Power System Stabilizers Using
Mutation rate = 0.01 Simulated Annealing. IEEE Trans Energy Convers 2000;15(3):297–304.
PSO: c1, c2 = 2, wmin = 0.4, wmax = 0.9 [16] Abido MA. Optimal Design of Power-System Stabilizers Using Particle Swarm
Optimization. IEEE Trans Energy Convers 2002;17(3):406–13.
GWO: a = [2, 0], A = [-2a, 2a], C = [0, 2] [17] Abido MA, Abdel-Magid YL. Optimal Design of Power System Stabilizer Using
STO: Sa = [2, 0], CB ¼ 0.5R, where R = [0, 1] Evolutionary Programming. IEEE Transactions Energy Conversions Sep. 2002;17
(4):429–36.
[18] Mishra S, Tripathy M, Nanda J. Multi-Machine Power System Stabilizer Design By
References Rule-Based Bacteria Foraging. Electr Pow Syst Res 2007;77(12):1595–607.
[19] Shayeghi H, Shayanfar HA, Jalilzadeh S, Safari A. Multi-Machine Power System
[1] Kundur P. Power System Stability and Control. Edited by Neal J. Balu, and Mark G. Stabilizers Design Using Chaotic Optimization Algorithm. Energ Conver Manage
Lauby 4.2 1994. March 2010;51:1572–80.
[2] Kundur P, et al. Definition and Classification of Power System Stability IEEE/ [20] Linda MM, Nair NK. Optimal Design of Multi-Machine Power System Stabilizer
CIGRE Joint Task Force on Stability Terms and Definitions. IEEE Trans Power Syst Using Robust Ant Colony Optimization Technique. Trans Inst Meas Control 2012;
2004;19(3):1387–401. 34(7):829–40.
[3] Demello FP, Concordia C. Concepts of Synchronous Machine Stability as Affected [21] Khodabakhshian A, Hemmati R. Multi-Machine Power System Stabilizers Design
By Excitation Control. IEEE Trans Power Syst 1969;88(4):316–29. By Using Cultural Algorithms. Int J Electr Power Energy Syst Sep. 2013;44:571–80.
[4] Bollinger K, et al. Power Stabilizer Design Using Root Locus Methods. IEEE Trans [22] Ali ES. Optimization of Power System Stabilizers Using BAT Search Algorithm. Int J
Power Syst 1975;94(5):1484–8. Electr Power Energy Syst 2014;61:683–90.
[5] Bollinger KE, Winsor R, Campbell A. Frequency Response Methods for Tuning [23] Eke Ī, Taplamacıoğlu MC, Lee KY. Robust Tuning Of Power System Stabilizer by
Stabilizers to Damp Out Tie-Line Power Oscillations: Theory and Field-Test Results. Using Orthogonal Learning Artificial Bee Colony. IFAC-Papers online 2015;48(30):
IEEE Trans Power Syst 1979;PAS-98(5):1509–15. 149–54.
[6] DeMello FP, et al. A Power System Stabilizer Design Using Digital Control. IEEE [24] Farah A, Guesmi T, Abdallah HH, Ouali A. A Novel Chaotic Teaching–Learning-
Trans Power Syst 1982;8:2860–8. Based Optimization Algorithm for Multi-Machine Power System Stabilizers Design
[7] Shrikant Rao P, Sen I. Robust Pole Placement Stabilizer Design Using Linear Matrix Problem. Int J Electr Power Energy Syst 2016;75:197–209.
Inequalities. IEEE Trans Power Syst 2000;15(1):313–9. [25] Abd Elazim SM, Ali ES. Optimal Power System Stabilizers Design Via Cuckoo
[8] Chen S, Malik OP. H∞ Based Power System Stabilizer Design. IEE Proc Gener Search Algorithm. Int J Electr Power Energy Syst 2016;75:99–107.
Transm Distr March 1995;142(2):179–84. [26] Islam NN, Hannan MA, Shareef H, Mohamed A. Azah Mohamed, “An Application of
[9] Shah R, Mithulananthan N, Lee KY, Bansal RC. Wide-Area Measurement Signal- Backtracking Search Algorithm in Designing Power System Stabilizers for Large
Based Stabilizer For Large-Scale Photovoltaic Plants With High Variability And Multi-Machine System. Neurocomputing 2017;237:175–84.
Uncertainty. IET-Renew Power Gener 2013;7(6):614–22. [27] Serdar E, Hekimoglu B. Parameter Optimization of Power System Stabilizer via
[10] Kamalasadan S, Swann GD, Yousefian R. A Novel System-Centric Intelligent Salp Swarm Algorithm. 5th IEEE International Conference on Electrical and
Adaptive Control Architecture for Power System Stabilizer Based On Adaptive Electronic Engineering (ICEEE). 2018.
Neural Networks. IEEE Syst J Dec. 2014;8(4):1074–85. [28] Špoljarić T, Pavić I. Performance analysis of an ant lion optimizer in tuning
[11] Chopra R, Joshi D, Bansal RC. Analysis Delta-Omega and Fuzzy Logic Power generators’ excitation controls in multi machine power system. 41st IEEE
System Stabilizer Performances under Several Operating Conditions. J Renew International Convention on Information and Communication Technology,
Sustain Energy 2009;1(3):1–11. Electronics and Microelectronics (MIPRO). 2018.

18
D. Chitara et al. International Journal of Electrical Power and Energy Systems 155 (2024) 109615

[29] Meysam R, Seyedtabaii S. Multi-machine optimal power system stabilizers design [40] El-Dabah MA, Hassan MH, Kamel S, Zawbaa HM. Robust Parameters Tuning Of
based on system stability and nonlinearity indices using Hyper-Spherical Search Different Power System Stabilizers Using A Quantum Artificial Gorilla Troops
method. Int J Electr Power Energy Syst 2019;105:729–40. Optimizer. IEEE Access 2022;10:82560–79.
[30] Serdar E, et al. An Application of Slime Mould Algorithm for Optimizing [41] Dhiman G, Kaur A. STOA: A Bio-Inspired Based Optimization Algorithm for
Parameters of Power System Stabilizer. 4th IEEE International Symposium on Industrial Engineering Problems. Eng Appl Artif Intel Jun 2019;82:148–74.
Multidisciplinary Studies and Innovative Technologies (ISMSIT). 2020. [42] Mirjalili S, Mirjalili SM, Lewis A. Grey Wolf Optimizer. Adv Eng Softw March 2014;
[31] Ekinci S, Izci D, Hekimoglu B. Implementing the Henry Gas Solubility Optimization 69:46–61.
Algorithm for Optimal Power System Stabilizer Design. Electrica 2021;21(2): [43] Goldberg DE. Genetic algorithms in search, optimization and machine learning.
250–8. Addison-Wesley; 1989.
[32] Chaib L, Choucha A, Arif S, Zaini HG, El-Fergany A, Ghoneim SSM. Robust Design [44] Kennedy J, Eberhart RC. Particle Swarm Optimization. In: Proc. IEEE International
Of Power System Stabilizers Using Improved Harris Hawk Optimizer For Conference on Neural Networks, Piscataway, NJ, vol. IV, pp. 1942-1948, 1995.
Interconnected Power System. Sustainability Oct. 2021;13(21):11776. [45] Milano F. Power System Analysis Toolbox Manual-Documentation for PSAT
[33] Serdar E, et al. Development of Lévy Flight-Based Reptile Search Algorithm with Version 2.1. 6. Univ Coll Dublin, Dublin Irel Tech Rep 2010.
Local Search Ability for Power Systems Engineering Design Problems. Neural [46] Anderson PM, Fouad AA. Power System Control and Stability, ISBN-
Comput & Applic 2022;34(22):20263–83. 9780470545577, Piscataway, NJ: IEEE Press, 2nd Edition, 2003.
[34] Davut I. A novel improved atom search optimization algorithm for designing power [47] Sharma RK, Chitara D, Raj S, Niazi KR, Swarnkar A. Multi-machine Power System
system stabilizer. Evol Intel 2022;15(3):2089–103. Stabilizer Design Using Grey Wolf Optimization. In: Proceedings of International
[35] Boucetta I, Naimi D, Salhi A, Abujarad S, Zellouma L. Power System Stability Conference on Computational Intelligence and Emerging Power System, pp. 331-
Enhancement Using A Novel Hybrid Algorithm Based On The Water Cycle Moth- 343, Springer, Singapore, 2022.
Flame Optimization. Energies 2022;15(14):1–17. [48] Rogers G. Power system oscillations. Springer; 1999.
[36] Moghadam AT, Aghahadi M, Eslami M, Rashidi S, Arandian B, Nikolovski S. [49] Shayeghi H, Ghasemi A. A multi objective vector evaluated improved honey bee
Adaptive Rat Swarm Optimization For Optimum Tuning Of SVC And PSS In A mating optimization for optimal and robust design of power system stabilizers.
Power System. Int Trans Electr Energy Syst Jan. 2022;2022:1–13. International Journal of Electrical Power & Energy Systems May 2014;62:630–45.
[37] Grace SS, Kumaravel S, Ashok S. Revamped Sine Cosine Algorithm Centered [50] Hsu YY, Chen CL. Identification of optimum location for stabilizer applications
Optimization of System Stabilizers and Oscillation Dampers for Wind Penetrated using participation factors. IEE Proc C Gener Transm and Distr May 1987;134(3):
Power System. IEEE Access 2022;11:1890–905. 238–44.
[38] Snášel V, et al. Weighted Mean of Vectors Optimization Algorithm and its [51] Chitara D, Meena NK, Yang J, Niazi KR, Swarnkar A, Gupta N, Vega-Fuentes E.
Application in Designing the Power System Stabilizer. Appl Soft Comput 2023;136 Small-Signal Stability Enhancement of Multi-Machine Power System Using Cuckoo
(110085):pp. and Harmony Search Optimization Techniques, In: Vol. 05: Proceedings of 11th
[39] Ruswandi DM, Robandi I, Prakasa MA. Stability Enhancement of Sulselrabar International Conference on Applied Energy, Part 4, Sweden, pp. 1-6, Aug. 2019.
Electricity System Using Mayfly Algorithm Based on Static Var Compensator and
Multi-Band Power System Stabilizer PSS2B. IEEE Access 2023;11:57319–40.

19