A Self-Adaptive Genetic Algorithm for Multi-Objective Reactive Power Optimization
XIA Ke-qing1,ZHAO Ming-qi2,LI Yang1 (1.Department of Electrical Engineering,Southeast University,Nanjing 210096,Jiangsu Province,China; 2.Yangzhou Power Supply Company,Yangzhou 225009,Jiangsu Province,China)
Based on the factor of self-adaptive weight sum and self-adaptive penalty function, a self-adaptive genetic algorithm is proposed and applied to solve multi-objective reactive power optimization. By use of the proposed method the multidirectional property of searching can be ensured and the defect of fuzzy membership algorithm, namely the overlong computing time, can be avoided. During searching process the self-adaptive penalty function can effectively utilize the available information in infeasible solution and appropriately punish the infeasible solution. Results of IEEE 14-bus testing system show that the proposed algorithm is an effective method for multi-objective reactive power optimization.