应用进化规划变异算子的配电网重构算法

针对进化规划算法普遍存在的进化过程缓慢和进化后期难以平稳收敛等问题,考虑到配电网重构功率损耗最小和负荷均衡2个目标,提出一种新的进化规划变异算子。为保证变异产生的新个体积极地向最优个体靠拢,一方面使变异算子的拓扑调整次数与进化代数成反比,另一方面在拓扑调整过程中增加待合开关选择算子和待分开关选择算子。同时,为提高整个配电网重构算法的计算速度,研究了功率损耗、二次负荷矩、待合开关选择算子和待分开关选择算子的简化计算方法。所提方法的可行性在河南省商丘供电局的10 kV配电网运行优化与辅助决策系统上得到了验证。

A Distribution Network Reconfiguration Algorithm Based on Evolutionary Programming Mutation Operator

In allusion to the problems of slow evolution and hard to reposefully converge which commonly exist in evolutionary programming algorithm and considering the two objects of minimum loss and load equalization,a new mutation operator for evolutionary programming is proposed.To ensure that the new individuals due to the mutation can draw close to optimal individual actively,on the one hand the topological restructure times of the mutation operator should be in inverse proportion to the generations of evolution,on the other hand the selection operator for switches to be switched on and that for switches to be switched off should be added during the topological restructure process.Meanwhile,to speed up the calculation of the whole distribution network reconfiguration algorithm,a simplified method to calculate power loss,quadratic load moment,selection operator for switches to be switched on and that for switches to be switched off is researched.The feasibility of the proposed method is verified by operation optimization and auxiliary decision-making system for an actual 10 kV distribution network in Henan province,China.