Application of Improved Differential Evolution Algorithm in Reactive Power Optimization

In allusion to such features as nonlinearity,multi control variables,multi constraints and coexistence of continuous variables and discrete variables in power system reactive power optimization,an improved differential evolution algorithm is proposed.According to the accumulated experiences in evolutionary learning process and utilizing excellent group to lead the direction of mutation,the proposed algorithm simultaneously extracts information of each dimensional element of excellent group and guides crossover operation of each dimensional variable of individual by excellent group information.Calculation results of IEEE 30-bus system show that the proposed algorithm can effectively solve power system reactive power optimization problem and in aspects of convergence speed,calculation accuracy and stability the proposed algorithm is better than particle swarm optimization algorithm and the standard differential evolution.