Available online at www.sciencedirect.

com
ScienceDirect

Energy Reports 9 (2023) 738–746

www.elsevier.com/locate/egyr

2022 9th International Conference on Power and Energy Systems Engineering (CPESE 2022),
Doshisha University, Kyoto, Japan, 9–11 September 2022

Implementation of genetic and particle swarm optimization algorithm

for voltage profile improvement and loss reduction using capacitors in
132 kV Manipur transmission system
Khaidem Bidyanath ∗, Sanasam Dhanabanta Singh, Shuma Adhikari
NIT Manipur, Langol, Imphal, 795004, India
Received 22 October 2022; accepted 7 November 2022
Available online xxxx

Abstract
Complex power plays a key role in maintaining and sustaining the fields i.e. magnetic and electric fields. The efficiency
of the power system depends entirely on the loss and voltage profile level of the system. The lower the level of loss and the
higher the voltage profile, the higher the system efficiency. Installation of the capacitors on a system greatly helps in controlling
and reducing levels of the reactive energy of the system. The location and size of the capacitors that need to be installed are
of great importance in the installation of the bank of capacitors. The idea of a Genetic Algorithm and also a Particle Swarm
Algorithm are deployed in this paper for calculating the optimal location and size of the capacitor. In this paper, the system
of the 132 KV Manipur Transmission System is designed in ETAP considering a radial distribution part and the optimization
method of a genetic method, and also the particle swarm algorithm is utilized for finding the optimal location and size of
the capacitor. The system is further optimized by reducing the loss and improving the voltage profile of the system. And the
comparison of both the optimization techniques is presented. The results show that PSO provides optimal solutions while GA
is simpler in the optimization process.
© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
Peer-review under responsibility of the scientific committee of the 9th International Conference on Power and Energy Systems Engineering, CPESE,
2022.

Keywords: Etap; Matlab; Reactive power; Genetic algorithm; Particle swarm optimization algorithm; Optimal capacitor placement; Power loss

1. Introduction
The efficient operation of any power system depends entirely on the controlling and monitoring of the power that
is being transferred. The quality of the transferred power is of great importance from the consumer’s perspective.
Power quality can be maintained by strict supervision of the voltage level of the system. The flow of the voltage
must be under strict control to avoid obstruction of the power from the sending end to the receiving end [1].
Reactive power compensating procedures is performed by the use of capacitors. Utilizing the required capacitors
∗ Corresponding author.
E-mail address: [email protected] (K. Bidyanath).

https://doi.org/10.1016/j.egyr.2022.11.098
2352-4847/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.
org/licenses/by/4.0/).
Peer-review under responsibility of the scientific committee of the 9th International Conference on Power and Energy Systems Engineering,
CPESE, 2022.
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

helps in maintaining the efficient system level of the system. Reducing power loss and improving the voltage level
of the system can be accomplished by the installation of capacitors on the system. Finding the optimal points of
the capacitors i.e. size and location of the capacitor is of great importance and one way to obtain this is with the
help of the GA [2] as well as PSO [3]. In the field of finding the optimal information of the capacitor, developing a
perfect problem or the required objective function is of enormous value and many approaches are being developed
over the past years.
In most situations, the optimal data is calculated by considering the objective function as the recovered energy
due to the energy loss reduction and also the reduction of the loss due to peak power in contrast to the cost of
the installed capacitors. Objective functions without considering the cost of the capacitors are also considered [4].
Optimization techniques are deployed in the fields of power systems with some distinctive purposes. One of the
optimization methods is the ant colony method used for the calculation of the feeder’s optimal data and also the
placement of the capacitor in the different systems [5]. The problem of finding the distinct site and size of the
capacitors is done by a specific term known as the loss sensitivity analysis and it includes a factor called the loss
sensitivity index [6]. Particle Swarm Algorithm (PSO) is also used for calculating the required optimal points for
voltage control, and loss reduction and also improves the efficiency of the system [7]. The absence of the operators
in PSO makes it easier for the implementation process. The particles inside the search space update their velocity
according to the algorithm without any specific operators. This search algorithm looks for the best solution and
only the best particle gives that information. This makes it suitable for the selection of the PSO for this research.
Unlike the methods of traditional algorithms, the solutions of GA are not trapped in the local optimum point and
it is based upon the rules of probabilistic selection. Influenced by the idea of Darwin’s theory of natural evolution,
the Genetic Algorithm (GA) is used for calculating the optimal solutions of distinct objective functions [8]. Besides
the capacitors, UPFCs can also improve the capacity of the power that is being transmitted [9]. AC system can
also combine with dc system at a high level of voltage for improving the loss level of the system [10]. GA can
be implemented in the fields of capacitor placement for solving various complex functions required for finding
the location, size, and also value of the capacitor. In this paper, the 132 kV transmission network of the Manipur
Transmission system is considered for research purposes. Manipur is a state located in the northeast region of India.
The entire system is designed with the help of Etap. For the optimization purpose, both GA and PSO are utilized
and the optimal capacitor’s location and size are determined. The use of a GA and PSO shows effective results
thereby reducing the loss and improving the voltage profile level of the system.

2. ETAP
The electrical transient analysis program is a great software used for the analysis and computing of the power in
electrical systems. Developed in the year 1983, this software includes the calculation of the power flow problems,
the study of the transient state analysis, and also the short circuit analysis of the system. ETAP also offers a real-time
simulation of distribution as well as the transmission system. The use of ETAP for the design and simulation of
various complex power system networks is easy as well as the user interface of the software environment is user-
friendly which makes the design and implementation process simple as well as effective. It is of utmost importance
for the power system engineers.

3. Genetic algorithm (GA)

A genetic algorithm is derived from the concept of natural selection or the theory of evolution. It is one of the
optimization techniques used for the calculation of the local and global data from a given objective function. The
entire process of this search algorithm s based on the concept of survival of the fittest. GA consists of different stages
and these stages are monitored by a specific operator called the genetic operators: Reproduction operator, crossover
operator, and mutation operator. The character of this algorithm is distinct as compared with other algorithms as
this has a diverse search space and this improves the ease of the computational calculations (see Fig. 1). .
Steps for the genetic algorithm:
1. Initialization of the size of the population as well as the generation.
2. Selection of the parents and crossover
3. Mutation of the offspring
4. Combination of the main population and the offspring
739
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Fig. 1. Stages of genetic algorithm.

5. Evaluate, sorting, and selection

6. Whenever required, go to step 2
Selection method of the genetic algorithm:
1. Random Selection
2. Tournament Selection
3. Roulette wheel selection

4. Particle swarm algorithm

PSO algorithm is the dominant technique used because of its ease of the implementation process and this
algorithm is developed by Kennedy et al. to calculate the solutions for complex mathematical problems [11]. The
main usage of this algorithm includes the optimal location and size for improving the voltage profile of the system
and also reducing the loss in the case of the radial distribution systems [12]. PSO, which is a search algorithm can
be deployed to different complex problems. It depends on the experience of the particle which includes the position
and velocity of the particle and finding the best or the global optimal points of the particles. Using this algorithm
for the optimization of the power system is very effective and it shows promising results.

5. Problem formulation
The main aims of this research are to improve the voltage profile and also to reduce the loss of the designed
system. This process can be achieved by the installation of the capacitors at the desired location. The location and
also the size as well as the number of capacitors that need to be installed entirely depends on the two algorithms
i.e. genetic algorithm and particle swarm algorithm.

5.1. Problem formulation for genetic algorithm

The expression for the current is as follows:

Pik − j Q ik
Iik = (1)
Vi
Where,
Iik = the current through the branch (i, k) where i indicates sending bus and k indicates receiving bus
Pik = the total power flow through (i, k)
Qik = reactive power through (i, k)
Vi = the voltage at the bus i
The total power loss Ploss as well as the objective function of the system, is as follows:
P 2 ik + Q 2 ik
Ploss = Iik2 ∗ Rik = ∗ Rik (2)
V 2i
740
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Where,
Rik = resistance through the branch (i, k)
Q ik = reactive power through the branch (i, k)

5.2. Problem formulation for particle swarm algorithm (PSO)

The objective function for reducing the reactive power loss of the system and also for improvement of the voltage
profile is as follows:
Min S = K p Plosstotal + K c Qctotal (3)
where
K p is the per year cost per capacitor
K c is the total capacitor purchase and installation cost
Qc is the total capacitor reactive power
Ploss is the total real power loss of the system
The objective function is considered concerning the following constraints:
(1) Bus Voltage Constraint
Each bus voltage must be within the given limits for satisfactory results:
Vimin ≤ Vi ≤ Vimax (4)
Where Vi Represents the voltages at each bus of the system.
(2) Number of Capacitor Constraints
This specific constraint plays a key role in reducing the total number of capacitors that need to be installed on
the system.
NC ≤ NCmax (5)
This indicates that the total number of the capacitors NC must be less as compared with the number of locations
that require the installation of the capacitors Nmax
C .

6. System design

The 132 kV Manipur Transmission System is designed in Etap. The considered system consists of 10 buses. Out
of the 10 buses, 1 bus is considered the slack bus i.e. IMPHAL PG bus is considered the slack bus. And the bus
designated as the LOKTAK is considered the voltage-controlled bus of the system (see Figs. 2–4).

7. Proposed algorithm
7.1. Genetic algorithm

The steps for the genetic algorithm are as follows:

1. the line data and bus data are being initialized
2. load flow study is performed through forward and backward sweep
3. the value of the capacitor is initialized as 2
4. the population and generation size of the system is considered as 50 and 500 respectively.
5. the objective function of the system is calculated and the parent is selected for the first iteration.
6. optimal point is calculated i.e. size of the capacitor and its location.
7. according to the optimal point of the system, crossover, and mutation is performed and it moves to the next
stage.
8. the process is iterated and the required point is calculated.
9. stop
741
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Fig. 2. 132 kV Manipur Transmission System.

Fig. 3. List of the buses for 132 kV Manipur Transmission System.

7.2. Particle swarm optimization

For the PSO algorithm, the selection procedure for the load buses and the size of the capacitor are explained
below:

7.2.1. Loss sensitivity index

This index is mainly used for the selection of the load buses. Selection of the bus is done by ranking the load
buses corresponding to the voltage’s active power loss and also the demand of the reactive power at the receiving
site. Two loss sensitivity indices are considered for finding the load buses location and they are represented by LS1
and LS2 respectively. LS1 is a negative value but LS2 gives a positive value. A higher negative value of LS1 indicates
a huge level of the variation of the voltage and this kind of bus is suitable for the placement of the capacitor. But
in the case of LS2 , the more positive value is suitable for the installation of the capacitor.
δPloss
LS1 = (6)
δV
δPloss
LS2 = (7)
δQ

7.2.2. Steps for the PSO algorithm are as follows

1. The line data, bus data and specify the base parameters of S, V, Z, and I are being initialized
2. the sensitivity index or factors are calculated using the above equations
742
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Fig. 4. Flowchart for genetic algorithm.

3. the total numbers of capacitors are being initialized (10%–15%)

4. the various parameters of PSO are computed i.e. population size, maximum iteration, weight factor, and
acceleration constants.
5. the particle’s random position and velocity are being generated and for each iteration, the position and velocity
of the particle are updated using the following equation.
vi (t + 1) = wi (t) vi (t) + c1r1 (pbest (t) − xi (t)) + c2r2 (gbest (t) − xi (t)) (8)
xi (t + 1) = xi (t) + vi (t + 1) (9)
6. location of the particle is updated using the following conditions
if
particle(i). cost < bparticle(i). cost (local best)
than bparticle(i) (local best) = particle(i);
else
bparticle(i). cost < gparticle.cost
gparticle(i) (local best) = bparticle(i);
7. the location of the capacitor is selected using the global best position.
gparticle.pos(i)
8. besides the above calculation, the global best position helps in the selection of the capacitor location.
9. algorithm ends if the solution is optimal otherwise the process is iterated.

8. Implementation and results of the system

The 132 kV Manipur Transmission System is implemented with the proposed algorithm presented in this paper.
Both the algorithm i.e. genetic algorithm and the particle swarm algorithm are being used for optimization purposes.
743
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

The outcoming values of both the optimization methods show improved and effective results as compared with the
original system. The implementation process is performed in MATLAB® version R2021b. The peak value of the
designed system is considered for simulation purposes. The PSO and GA algorithm is implemented for the proposed
system. The optimized buses and also the size is being calculated by the two optimization methods. Losses of the
system are improved up to a certain extent through the two algorithms.
Table 1 shows the results of the PSO algorithm for the system of 132 kV Manipur Transmission System.

Table 1. Capacitor placement using PSO algorithm.

Buses that require Size of the capacitor
a capacitor (KVAr)
7 1194.17
9 1200

After various iterations, the buses that require a capacitor to make an optimized system are shown in the above
table i.e. buses 7 and 9, and also the size of the bus is presented. Max number of 2 capacitors is being considered
for this specific research as the lower number of the capacitor can reduce the cost as well as the complexity of the
network (see Table 2).
Table 2. Loss comparison before and after optimization using PSO
algorithm.
Losses before optimization 31625.91 KW
Losses after optimization 18139.30 kW

Again Table 3 gives the result for the GA.

Table 3. Capacitor placement using GA.

Buses that require a capacitor Size of the capacitor (KVAr)
6 2000
8 2000

Besides the consideration of the 2 buses in PSO, GA also considers the same number of capacitors for the same
reason. Buses 6 and 8 are the buses that require the installation of the capacitors (see Table 4).

Table 4. Loss comparison before and after optimization using GA.

Losses before optimization 18314 kW
Losses after optimization 8382 kW

The improved voltage profile of the system through the PSO algorithm can be observed from Fig. 5.
From the result of the PSO, the minimum voltage level of the system before compensation of the system is
0.6571, and after compensation is 0.68332. The minimum voltage level for both of the cases is at bus number 7
(i.e. KAKCHING). The maximum level of the voltage for the system is 2.13 before compensation and the bus is at
5 (KONGBA). After compensation, the level of the voltage which is maxed is at 2.44 at bus number 5 (KONGBA).
Some voltage level of the system is improved after the optimization process i.e. the voltage level of bus 5 is improved
from 2.13 to 2.44 and for bus 4, the level of the voltage is improved from 1.29 to 1.94. The level of the voltage is
improved from 1.46 to 1.61.
The improved voltage profile of GA is as given in Fig. 6.
From the result of the GA, the minimum voltage level of the system before compensation the system is
0.17, and after compensation is 0.55. The minimum voltage level for both of the cases is at bus number 10
(i.e. NINGTHOUKHONG). The maximum level of the voltage for the system is 0.69 before compensation and
the bus is at 3 (IMPHAL). After compensation, the level of the voltage which is maxed is at 0.84 at bus number 3
(IMPHAL). All the voltage level of the buses is improved with the help of the genetic algorithm method.
744
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Fig. 5. Voltage Profile of the system by using PSO algorithm.

Fig. 6. Voltage Profile of the system by using GA algorithm.

9. Conclusion

In this research paper, the main purpose is to find the optimal parameters for improving the 132 kV Manipur
Transmission System. The 132 kV Manipur Transmission System is designed with the help of ETAP. Designing
through ETAP is very effective as the interface of the system is user-friendly and complex power system analysis can
be performed with effective results. To increase the voltage profile and reduce the losses, the location, size, and the
buses that require capacitors are calculated with the help of PARTICLE SWARM ALGORITHM and GENETIC
ALGORITHM. PSO and GA provide promising results but there is a difference between them in terms of the
implementation process and the results that they produce. PSO provides the most effective results in terms of
reducing the loss of the system but GA is easier in terms of the implementation process and well as the equation
is less.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could
have appeared to influence the work reported in this paper.

Data availability

The data that has been used is confidential.

745
K. Bidyanath, S.D. Singh and S. Adhikari Energy Reports 9 (2023) 738–746

Acknowledgment
The authors would like to thank the Director NIT Manipur for his constant support morally and technically
by providing the necessary research facilities. Also we would like to thank CPDA provided by the Institute for
supporting in materialization of this paper.

References
[1] Watpade Dnyaneshvar Y, Sonwane PM. Optimal capacitor placement for IEEE 14 bus system using genetic algorithm. Int J Innov
Res Adv Eng 2016;3(09).
[2] Chopade P, Bikdash M. Minimizing cost and power loss by optimal placement of capacitor using ETAP. In: 2011 IEEE 43rd southeastern
symposium on system theory. IEEE; 2011.
[3] Masoum Mohammad AS, Ladjevardi Marjan, Jafarian Akbar, Fuchs Ewald F. Optimal placement, replacement and sizing of capacitor
banks in distorted distribution networks by genetic algorithms. IEEE Trans Power Deliv 2004;19(4).
[4] Maxwell M. The economic application of capacitors to distribution feeders. AIEE Trans 1960;01(79):353–9.
[5] Shahgholian G, Azimi Z. Analysis and design of a DSTATCOM based on sliding mode control strategy for improvement of voltage
sag in distribution systems. Electronics 2016;5(3):1–12.
[6] Hashemi-Zadeh1 SA, Zeidabadi-Nejad O, Hasani S, Gharaveisi AA, Shahgholian G. Optimal DG placement for power loss reduction
and improve voltage profile using smart methods. Int J Smart Electr Eng 2012;1(3):141–7.
[7] Yoshida H, Kawata K, Fukuyama Y, Takayama S, Nakanishi Y. A particle swarm optimization for reactive power and voltage control
considering voltage security assessment. IEEE Trans Power Syst 2000;15(4):1232–9. http://dx.doi.org/10.1109/59.898095.
[8] Shahgholian G, Haghjoo E, Seifi A, Hassanzadeh I. The improvement DSTATCOM to enhance the quality of power using fuzzy neural
controller. J Intell Procedures Electr Technol 2011;2(6):3–16, Summer.
[9] Oo Moe Moe, Thuzar Myint. Transmission capacity improvement using unified power flow controller with new control strategy. Int J
Electr Electron Eng Telecommun 2021;10(3):225–32. http://dx.doi.org/10.18178/ijeetc.10.3.225-232.
[10] Majid Zulfiana S, Arief Ardiaty, Akil Yusri S. Minimization of transmission loss in application of HVDC networks under load increase
scenario. Int J Electr Electron Eng Telecommun 2021;10(5):333–40. http://dx.doi.org/10.18178/ijeetc.10.5.333-340.
[11] Kennedy J, Eberhart R. Particle swarm optimization. In: Proceedings of ICNN’95 - international conference on neural networks. 1995,
p. 1942–8. http://dx.doi.org/10.1109/ICNN.1995.488968, 4.
[12] Rani BJ, Reddy AS. Optimal allocation and sizing of multiple DG in radial distribution system using binary particle swarm optimization.
Int J Intell Eng Syst 2019;12:290–9.

746