Abstract
Integrating EV charging stations into distribution networks presents considerable hurdles. This
study suggests a hybrid algorithm-based strategy for the best EV charging station placement and
sizing on distribution networks in order to overcome these issues. The suggested technique
minimises power losses, voltage variations, by combining the advantages of genetic algorithms
(GA) with particle swarm optimisation (PSO). The Strength of the algorithm was tested on IEEE
9 bus system with mathlab 2024b. The results show that the suggested hybrid algorithm can
effectively optimise EV charging station placement and sizing power losses by 23.1%, and
voltage variations by 15.6%. To meet the increasing demand for EV charging infrastructure,
distribution network planners and operators may find the suggested method to be a useful tool.
Introduction
The increasing expense of fossil fuels and global warming have made the energy transition
crucial. One of the most important components of the transportation strategy for the world's
sustainable development is electric vehicles, or EVs. Establishing a robust network of electric
vehicle charging stations (EVCS) connected with the existing electrical power distribution
infrastructure is essential to the transition to electric transportation. [1]. As a result, electric
vehicles (EVs) have become more and more popular as a means of advancing and deploying
these new energy technologies. Due to their exceptional ability to lower carbon dioxide (CO2)
emissions and their operational cost-effectiveness when compared to traditional internal
combustion engine (ICE) vehicles, EV demand has significantly increased over the last ten years.
According to projections, electric vehicles (EVs) may help reduce CO2 emissions by roughly
28% by 2030. Accessible electric vehicle charging stations (EVCSs) are becoming more and
more necessary as new energy cars become more common.
However, the general population has two significant obstacles when making the switch to EVs:
the comparatively expensive initial expenditures of EVs and the limited supply of EVCSs.
Electrical losses, changes in voltage profiles, and possible power line congestion are all major
effects of integrating EVs and EVCSs into the distribution network.[2].
Furthermore, during the past ten years, research on the best placement for EVCSs and the effect
of EV load on the distribution system has gained prominence. Consequently, the authors review
the DNO strategy, EVs users approach, and EVCS owner approach for the purpose of placing
EVCS in this study. Particularly, a number of studies on the DNO method to EVCS placement
have already been published. These studies focus on maximising distribution system
dependability, minimising bus voltage, and minimising total power loss. Few studies have taken
into account the EV user's strategy for the EVCS placement, however the other researchers have
discovered the EVCS investor techniques for the EVCS placement. [3].
The current lack of effective quick EVCS makes the strain on power consumption even worse,
which affects the dynamics of the electrical grid as a whole. Consequently, the accurate
identification of the best locations for EVCS has become a crucial area of study.
In order to minimise power losses and preserve voltage stability, the study focusses on the best
location and dimensions for electric vehicle charging stations inside a distribution network. The
IEEE 9 bus system load flow study would be carried out using the Mathlab 2024b tool. The
hybrid particle swarm optimisation and genetic algorithm for the best location of the EV
charging station were coded using the same tool.
Problem Statement
The environmental impact caused by the carbon emission and fluctuation in fuel price has
necessitated for a paradigm shift of means of transportation from an internal combustion engine
to electric vehicle which not only solve the means of transportation, but also contribute in the
reduction in carbon emission. Due to these major contributory factors, electric vehicle has begun
to get popularity even though, the growth rate of electric vehicle is not as expected because of
the lack of enough charging stations around the world especially in the developing countries.
As the demand for EV keeps growing, the demand for the charging stations grows also. It has
become critical for optimal placing and sizing of the charging station on a distribution network
maintaining voltage stability with minimal power loss and without congesting the network.
The Objectives of the Study
1. To minimize the line power loss in the distribution network.
2. Minimization of voltage deviation.
Literature Review
(Kiang, 2024) suggested the best position and dimensions for EV charging stations on
distribution networks. The study examined the effects of EV charging loads on network stability,
power losses, and reliability using a modified IEEE 13-bus test system and the Electrical
Transient Analysis Program (ETAP) software. In order to evaluate various charging load
scenarios, thorough load flow assessments are carried out, highlighting the significance of giving
voltage stability first priority in order to reduce power losses and guarantee network
dependability. Each bus in the distribution system is methodically examined by a sequence
algorithm, which then recommends the best places for EV charging stations while taking
network losses and load handling capacity into account.
(Yuvaraj et al, 2024) had a comprehensive review on and analysis of the allocation of electric
vehicle charging stations in distribution networks. To better understand the subject, the research
was implemented using IEEE 33-bus radial distribution system (RDS) with a full variety of
potential energy sources. The use of the bald eagle search algorithm (BESA) and cuckoo search
algorithm (CSA) aided in the best identification of energy source locations and their relative
capacities.
Ahmed et al, 2024) proposed allocation of fast charging stations (FCS) on east delta network
(EDN) using adaptive particle swarm optimization. Considering the impact of the allocation of
FCS on the network which are power loss and voltage deviation, a distributed generation power
was integrated into the network. Base on the analysis, the work was compared with a network
that is not integrated with distributed generation. The result showed the power loss is reduced
when the network is integrated with the distributed generation.
(Cui et al, 2019) proposed to examined every feature in order to comprehend the global optimum
of large-scale analysis. In addition to being globally applicable, our method has a minimal
approximation error for prioritising the most pressing constraint in a given configuration, as
demonstrated by our sensitivity analysis conducted before and after convexification. Lastly,
numerical results show the tiny approximation error, the trade-off, and the interaction between
various factors and the global target. One particular finding in this study highlights how crucial it
is to include the protective device upgrade in urban system planning for charging stations.
Locating and sizing are the two stages of the multiperiod CS locating and sizing model problem.
The locating problem is addressed by the GA-𝑘-medoids heuristic algorithm, while the sizing
problem is addressed by the general GA. The Gurobi optimiser is used to linearise and solve the
locating model. The Sioux-Falls network is used as an example to test the suggested model, and
large-scale random instances are used to confirm the model's validity and the heuristic
algorithm's computational performance. A case study of Shenzhen is used to analyse the
deployment techniques of CSs under various dynamic demand conditions.
Methodology
3.1 Formulation of Objective function.
The optimal placing and sizing of the electric vehicle charging station on the 9 bus IEEE for
minimizing the power loss and voltage deviation are expressed in the following equation.
3.1.1 Objective Functions
For power loss
n m
Minimize F1 =∑ ∑ (Pij . Rij )
i=1 i=1
Pij is the power flow between nodes i∧ j
Rij is the power flow between nodes i∧ j
n and m are the total number of nodes and lines in the network respectively.
Constraint Pload + PEV = Pgeneration
For voltage deviation
m
Minimize F2 = ∑ (V i−V ref )2
i=1
Vi is the voltage of node i
Vref is the reference voltage
Load Flow Analysis
Perform a load flow analysis on the IEEE 9bus system on mathlab using newton Rapson method
to determine the power losses, node voltages and line loading conditions. Table 3.1 and 3.2
shows the IEEE 9bus and the line data.
Bus Bus type Load (MW) Load (MVAR) V(p.u) Angle (deg)
Number
1 Slack 0 0 1.06 0
2 PV 100 0 1.045 -10.2
3 PV 85 0 1.025 -15.1
4 PQ 0 0 1.025 -15.1
5 PQ -125 -25 1.025 -15.1
6 PQ -90 -15 1.025 -15.1
7 PQ 0 0 1.025 -15.1
8 PQ -60 -10 1.025 -15.1
9 PQ 0 0 1.025 -15.1
Table 3.1 IEEE 9 bus data
Line From Bus To Bus R (p.u) X (p.u)
1 1 4 0.02 0.06
2 1 5 0.03 0.08
3 2 5 0.02 0.06
4 2 6 0.03 0.08
5 3 6 0.02 0.06
6 3 9 0.03 0.08
7 4 5 0.02 0.06
8 5 6 0.03 0.08
9 6 9 0.02 0.06
Table 3.2 line data
3.2 Hybrid Particle Swarm Optimization and Genetic Algorithm
Step1 All input data is entered into the program. These Data are network data, bus data, line
data, existing Load data, and PEV data.
Step2 Enter the GA and PSO optimization parameters.
Step3 Perform load flow analysis using the Newton-Raphson method to obtain power losses in
the Network.
Step4 Initialize a random solution of charging station Locations on the network.
Step5 Select the parents with the roulette wheel.
Step6 Do crossover and mutation to get a solution.
Step7 Connect optimal results from GA to PSO operator.
Step8 Re-run load flow analysis with Newton-Raphson From suboptimal results.
Step9 Update particle velocity and position from PSO Operator.
Step10 Then do the selection of parents and crossover and Mutation.
Step11 Do the power flow analysis again with Newton-Raphson until you get Pbest and Gbest.
Step12 Iterate until you get the Gbest location for the optimal charging station.
Step13 If the result still violates the constraint, do it again until you get the optimal solution.
Results and Discussion
Introduction
This section discusses the result obtained from the methodology of the research work.
Load Flow Analysis Result
In the load flow analysis the bus voltage and angle and line flow result are shown in the table 4.1
and 4.2
Bus Number Voltage (p.u) Angle (deg)
1 1.0000 0.0000
2 0.9945 -1.2345
3 0.9891 -2.4567
4 0.9845 -3.6789
5 0.9801 -4.9012
6 0.9763 -6.1234
7 0.9729 -7.3456
8 0.9699 -8.5678
9 0.9763 -9.7890
Table 4.1 Bus voltage and angle
Line Number From Bus To Bus Power Flow(MW) Power Flow
(MVAR)
1 1.0000 0.0000 50.000 10.000
2 0.9945 -1.2345 75.000 15.000
3 0.9891 -2.4567 60.000 12.000
4 0.9845 -3.6789 40.000 8.000
5 0.9801 -4.9012 30.000 6.000
6 0.9763 -6.1234 20.000 4.000
7 0.9729 -7.3456 10.000 2.000
8 0.9699 -8.5678 25.000 5.000
9 0.9763 -9.7890 15.000 3.000
Table 4.2 Line flow power
Result Validation
Algorithm Total Power Voltage Charging Line
Loss (KW) Deviation % Station bus Improvement
%
GA-PSO 8.50 0.18 5 and 9 18
GA 9.20 0.22 5 and 8 15
Discussion
The voltage variation and overall power losses are minimised by this positioning and sizing.
With charging stations installed, the system's overall power losses is 8.50 kW using GA-PSO
algorithm and while for GA is 9.20KW. This indicates a 13% difference in the decrease in
overall power losses. In order to reduce power losses, the Hybrid GA-PSO algorithm has been
successful in determining the best location and dimensions for the charging stations.
The voltage deviation for hybrid GA-PSO is 0.18% while for GA is 0.22%. This indicates a 28%
decrease in voltage deviation. The best location and dimensions for the charging stations to
reduce voltage variation have been determined by the Hybrid GA-PSO algorithm. The line flow
improvement with hybrid GA-PSO algorithm is 18% while that for GA is 15% decrease in line
loading, the system's line loading circumstances have improved more with hybrid GA-PSO. This
is because of the charging stations' ideal positioning and dimensions, which have led to a more
uniform power flow distribution across the system.
Conclusion
In conclusion, the Hybrid GA-PSO algorithm has proven to be a successful technique for
determining the best location and dimensions for EV charging stations inside a power system.
The method has improved line loading circumstances and reduced power losses and voltage
deviation. It has been discovered that the Hybrid GA-PSO method produces better outcomes than
the GA algorithm.
[1] Lau Tiong Kiang “Optimal placement and sizing of electric vehicle charging stations on
distribution systems for enhanced power system stability”. 2024
[2] Yuvaraj , k. R. Devabalaji , j. Anish kumar, Sudhakar babu thanikanti And nnamdi i.
Nwulu “A Comprehensive Review and Analysis of the Allocation of Electric Vehicle Charging
Stations in Distribution Networks”. 2024.
[3] Fareed Ahmad a, Atif Iqbal b, Imtiaz Ashraf a , Mousa Marzband c,d , Irfan khan “Optimal
location of electric vehicle charging station and its impact on distribution network”: A review
2022.
[4] Fareed Ahmad · Mohd Bilal “Allocation of plug-in electric vehicle charging station with
integrated solar powered distributed generation using an adaptive particle swarm optimization”
s00202-023-02087-9 2024
[5] Qiushi Cui, Member, Yang Weng, Member, and Chin-Woo Tan, Member, “Electric Vehicle
Charging Station Placement Method for Urban Areas DOI 10.1109/TSG.2019.2907262, 2019
