仿生粒子群算法在电力系统无功优化中的应用

摘要

Based on the particle sward optimization ( PSO) , bionic particle sward optimization ( BPSO) algorithm is presented in this paper which follows the biological feature of group optimizing. In the early period, group is dy⁃namically divided into several subgroups, each of which is relatively independent and evolves towards one target. Members change rapidly as the procedure of evolution proceeds. The latter increases the exchange of information between subgroups, and so the algorithm converges faster. Not only can this algorithm enrich the variety of popula⁃tion and avoid converging the local optimal solution, but it can also attain a fairly high rate of convergence. In this paper, the algorithm is applied in the reactive power optimization of power system. Compared with the standard PSO, the algorithm is proved to be feasible and practicable through simulation of IEEE30 bus system and IEEE118 bus system.%本文在标准粒子群算法的基础上，遵循群体寻优的生物特性，提出了仿生粒子群算法。初期将群体动态地分成多个子群，每个子群相对独立地向一个目标进化，子群的成员随着进化过程不断地更迭。后期增加子群间的信息交流，使算法更快收敛。该算法不仅丰富了种群的多样性，避免过早收敛于局部最优解，而且有较快的收敛速度。文中将该算法应用于电力系统无功优化中并与标准粒子群算法进行了比较，通过对IEEE30节点和IEEE118节点的算例仿真，证明了该算法的可行性和有效性。