基于结合均匀变异的粒子群算法的配电网重构研究

摘要

Distribution network reconfiguration is one of methods to maintain safety,stability and re-liability of electrical power system. To make network active loss lower after reconfiguration,an binary particle swarm optimization( BPSO) combining uniform mutation and improved inertia weight is pres-ented. Firstly,a linear decreasing inertia weight is used. Secondly uniform mutation operator of ge-netic algorithm( GA) is introduced which makes the both global and local search ability of BPSO and overcomes the drawback of prematurity. Finally,the improved BPSO is used to reconstruct the IEEE 33 buses system. The simulation results show that compared with traditional particle swarm optimiza-tion( PSO) ,by using the improved BPSO to reconstruct electrical power system can reach better opti-mization. And the network active power loss reduce to 32. 42% compared with the original after reconfiguration.%配电网重构是确保电力系统安全、稳定、可靠运行的有效方法之一。为使重构后的配电网有功网损更低，文中提出了一种结合均匀变异与惯性权重的改进的二进制粒子群优化算法( BPSO)。首先，采用一种线性递减的惯性权重，然后引入遗传算法的均匀变异算子，使BPSO兼具良好的全局和局部搜索能力，克服了容易早熟的缺点。最后，将算法用于IEEE33节点系统进行重构。仿真结果表明，与传统粒子群算法相比，采用改进后的算法进行重构寻优效果更佳，且重构后的网损较初始状态降低了32.42%。