Improved Genetic Algorithm for Reactive Power Optimization of Electric Network

任震,钟红梅,张勇军,唐卓尧
DOI: https://doi.org/10.3969/j.issn.1006-6047.2002.08.005
2002-01-01
Abstract:Regarding to the low speed of traditional genetic algorithm (TGA) used in real time reactive power optimization in power systems,an improved algorithm,period evolution genetic algorithm (PEGA), is presented.The population size of the group and the maximum evolution generation are no longer fixed.The evolution process is divided into several periods, the group maximum population is increasing gradually and the maximum generation of each period is set accordingly.It can improve the direction in searching optimal solution,prevent the premature convergence of algorithm,and ensure the efficiency of each alternation during latter evolution periods for increasing the calculation speed.The proposed algorithm is evaluated in an IEEE 30 bus power system,the result shows that PEGA has better global searching ability with faster convergence rate and better performance in real time reactive power control.
What problem does this paper attempt to address?