Advances in Engineering Research, volume 150

4th International Conference on Machinery, Materials and Computer (MACMC 2017)

The Approach of Genetic Algorithms Application on Reactive Power

Optimization of Electric Power Systems

RuiJin ZHU 1, a
1
XiZang agriculture and Animal Husbandry College, Linzhi City 860000,China
a
[email protected]

Keywords: Electric Power Systems; Genetic Algorithms; Reactive Power Optimization

Abstract: With the rapid growth of China's economy and the development of industry, the demand
of power quality in various departments of national economy is more and more strict. In the power
system, reactive power plays a special role. The research on reactive power optimization of power
system has significant practical significance for reducing the extra active power consumption and
improving the voltage operation level caused by the unreasonable allocation of reactive power.
Using the improved genetic algorithm proposed in this paper, the standard test system is used to
simulate the reactive power optimization, and the optimization results of the simple genetic
algorithm and the improved genetic algorithm are compared. The simulation results show that the
proposed algorithm is feasible and effective, and the improved genetic algorithm has lower active
network loss and better global convergence performance and convergence speed.

Introduction
With the rapid development of China's national economy and power industry, the total installed
capacity of power generation in China and the demand for electricity in all walks of life are
increasing at an unprecedented rate. At present, China grid scale has surpassed the United States
ranked first in the world, the installed power generating capacity continues to rank second in the
world, the power system security, reliability, economy and resource allocation capabilities improve,
basically meet the electricity needs of economic and social development [1-2]. With the continuous
growth of power generation capacity, power generation and social electricity consumption, higher
requirements are put forward for the safety of power grid and power quality. Therefore, the security,
economic and stable operation of power system is becoming more and more important. Reactive
power optimization is an effective means to ensure the safe, economic and stable operation of
power system, and is an important measure to improve the quality of system voltage [3].
In order to improve the quality of voltage and improve power system and reduce the power loss
of the whole system, ensure the economic benefit of power system, it needs the appropriate
transmission power, reactive power, power distribution network reactive power compensation
device and adjustable pressure control device, to ensure adequate system reactive power reserve
capacity, voltage level the stability of the entire system, protection of electrical equipment, to
maintain its normal state of operation and improve the economic benefits of equipment operation.
At the same time, in the process of transmission and distribution, it is necessary to avoid the
long-distance transmission of reactive power and redundant reactive power transmission in the
power grid, so as to reduce the active power loss of the system. This is of great significance to
realize the security, stability and economic operation of the whole power system. The optimization
of reactive power in power system is an optimization problem. It has many kinds of variables, more
than one goal, complex mathematical model, large processing scale, and high requirement for
real-time algorithm. Especially in recent years, the scale of power system is becoming larger and
larger, and the technology is becoming more and more complex. Due to the inherent limitations of
the traditional algorithm, it has been unable to adapt to the modern large-scale power system. In
recent years, artificial intelligence algorithm has been applied in the field of reactive power

Copyright © 2018, the Authors. Published by Atlantis Press.

This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/). 206
 Advances in Engineering Research, volume 150

optimization, among which genetic algorithm is more widely used than other intelligent algorithms.
In this paper, genetic algorithm is improved on the basis of the existing research, in order to further
improve the speed and accuracy of its solution.

Related work
Reactive power optimization of power system is a large-scale, dynamic, multi objective, multi
constrained nonlinear optimization problem. The optimization process is very complex, and the
operation variables include not only continuous variables, but also discrete variables [4]. The
reactive power optimization of power system up to now, many domestic and foreign experts and
scholars carried out a lot of research work, a series of effective algorithms and mathematical model
is proposed to solve the reactive power optimization problem, made a lot of achievements.
Especially in recent decades, with the development of modern artificial intelligence algorithms and
the rapid development of computer science and technology, people will use these advanced theory
to solve the problem of reactive power optimization, provides a new theoretical method for the
research and analysis of reactive power optimization [5-6]. So far, the solution methods of reactive
power optimization can be divided into two categories: conventional optimization algorithm and
modern artificial intelligence algorithm.

Genetic algorithm research for power system operation

The genetic algorithm is a random search method simulating the evolutionary mechanism proposed
in 1970s by American professor J.Holland of the University of Michigan. The algorithm can
globally optimal solution space in multi path search problem, and to solve the optimization problem
in the mathematical model of operating variables encoding processing, the object of genetic
manipulation by the operating variables are converted to digital string, which makes the selection,
crossover and mutation genetic operation becomes easy. GA searching direction is directly
determined by the objective function, does not need the derivation for the inverse of complex
mathematical operations, improves the efficiency of the algorithm, at the same time, GA is adopted
to group N the number of individuals on the unit to carry out optimization, can also handle a large
number of patterns, with strong processing line ability, which can fast global convergence. At the
same time, because the GA is able to handle the constraints of the optimization problem, the
introduction and not too restrictive, more conducive to the algorithm converges to global optimal
solution space solution of non-linear programming problem in the solution of reactive power
optimization of power system with multiple variables and constraints in this category are wide
application. In view of the above characteristics of genetic algorithm, many scholars at home and
abroad have done a lot of work in recent years. GA is applied to solve the problem of reactive
power optimization in power system, and has achieved some experience and results [7].
The solution steps of the IGA are shown in Figure 1. (1) Initialize the requirements and set
number of “catastrophe”. (2) Come into being out the initial race. (3) Operate the genetic
evolution. (4) Record the maximum and the minimum adaptation value individual in the population.
(5) Carry out the catastrophe and preserve the current optimal solution while the difference between
the maximum and the minimum adaptation value individual is smaller than a threshold, and return
to step (3). Otherwise, take the next step.

207
 Advances in Engineering Research, volume 150

Figure 1. Flow chart for Genetic Algorithm Optimization

Application process of Genetic algorithm for power system operation is shown in Figure 2.

Figure 2. Flow chart for power system operation

Simulation and discussion

The proposed GA is implemented using the MATLAB platform for RPO. And the calculations are
deployed on IEEE 14, 30, 57-bus system. All of the bus voltage ranges of these three benchmark
systems are 0.95 ~ 1.1 (per-unit value). The increment step size of the transformer is given as
1.25%, the lower and upper ratio limit are 0.9 and 1.1, respectively. And, the upper limit for the
account of stunt capacitor is 5, its step size is 0.1. Parameters used in the GA are set as the
following: popsize = 50, maximum generation loop t = 100, P C0 = 0.8, increment step size P cstep
0.002, P Cmin = 0.6, mutation rate P m0 = 0.1, the decrement step size P mstep = 0.001, P max
=0.2, penalty coefficient of the voltage V = 1.05, penalty coefficient of generator’s reactive power
Q = 1.025. Both of the simple genetic algorithm (SGA) and the improved genetic algorithm (IGA)
will be tested on all the three systems for ten times. Data from Table 1 manifest that the optimal
value of IEEE 14 system by the improved genetic algorithm (IGA) is 0.126935, and the reduction
ratio is 8.35%, better than the SGA result, 0.129240. The reduction ratio of IEEE 30 and IEEE57
are 11.44% and 17.43%, respectively, proving that the optimization performance effect is notable.
Figure 3 shows the convergence history of the objective function along the evolutionary generation
of the optimal optimization result out of the ten runs, regarding the IEEE 14 bus system.

208
 Advances in Engineering Research, volume 150

Figure 3. Comparison of the convergence performances of IEEE 14 system

The y-axis refers to the minimum evolutionary value of each generation. As revealed in Figure 3
and 4, the improved GA can more efficiently return smaller optima than the SGA. The comparison
indicates that the proposed IGA is feasible and effective in solving the RPO problems.

Figure 4. Comparison of the convergence performance of IEEE 30 system

Conclusion
Reactive power optimization of power system is not only an effective means to ensure the security,
stability and economic operation of power system, but also an important measure to effectively
reduce the power system network loss and improve the voltage quality. Therefore, the research on
reactive power optimization of power system has both theoretical significance and practical
application value. Reactive power optimization of power system is a multi-variable, multi
constrained and nonlinear mixed programming problem. Its operation variables include continuous
variables and discrete variables, and the optimization process is also very complicated. Genetic
algorithm is widely used in solving the problem of reactive power optimization because it can
search the global optimal solution by multi path, and can solve the discrete and nonlinear problems
naturally. Therefore, based on the establishment of the mathematical model of reactive power
optimization, an improved genetic algorithm is proposed and applied to solve the reactive power
optimization problem, and the effect of reactive power optimization is improved. Using the

209
 Advances in Engineering Research, volume 150

improved genetic algorithm proposed in this paper, the standard test system is used to simulate the
reactive power optimization, and the optimization results of the simple genetic algorithm and the
improved genetic algorithm are compared. The simulation results show that the proposed algorithm
is feasible and effective, and the improved genetic algorithm has lower active network loss and
better global convergence performance and convergence speed.

Acknowledgements
1.XiZang agriculture and Animal Husbandry College High-level Talents Introduction Project (2014DQ-ZN-01).
2.Key research projects of Tibet Autonomous Region for Innovation and Entrepreneur(Grant No.Z2016D01G01/01).

Reference
[1] Zhu M, Xu Z, Li X, et al. Dynamic Reactive Power Optimization Based on Genetic Algorithms
in Power Systems[C]// International Conference on Mechatronics, Control and Electronic
Engineering. 2014.
[2] Wang L Y. Research on Reactive Power Optimization Based on Genetic Algorithms in
Distribution Network[J]. Applied Mechanics & Materials, 2014, 644-650(3):2476-2478.
[3] Liu Z, Wang X, Li C. Reactive power optimization in power system based on chaos ant colony
algorithm[C]// International Conference on Sustainable Power Generation and Supply, 2009.
Supergen. IEEE, 2010:1-4.
[4] Liu B Y, Yang R G, Hui L I, et al. Reactive Power Optimization Based on Chaos Genetic
Algorithm in Electric Power Systems[J]. Proceedings of the CSU-EPSA, 2006, 18(5):49-52.
[5] Ma X, Liu R L. Reactive power optimization in power system based on improved niche genetic
algorithm[C]// International Conference on Computer Design and Applications. IEEE,
2010:V3-413-V3-416.
[6] Dai C, Chen W, Zhu Y, et al. Reactive power dispatch considering voltage stability with seeker
optimization algorithm[J]. Electric Power Systems Research, 2009, 79(10):1462-1471.
[7] Kim H M, Kim J Y, Yoon C D, et al. Optimal Voltage and Reactive Power Control of Local
Area Using Genetic Algorithm[C]// International Conference on Knowledge-Based and
Intelligent Information and Engineering Systems. Springer, Berlin, Heidelberg, 2006:900-906.

210