摘要
针对传统遗传算法易于陷入局部最优解和随着配电网规模的扩大搜索效率降低的问题,借鉴多种群和自适应思想,提出了基于自适应多种群遗传算法的配电网规划算法。通过对目标函数进行处理,引入了多个物种,并采用自适应遗传算法和考虑进化稳定的改进多种群遗传算法分别对不同的物种进行操作,通过转移优秀个体,实现了物种之间的协同作用。同时为解决遗传算法应用于配电网规划时产生的大量不可行解的问题,借助图论知识和搜索技术给出了不可行解的修复方案,通过对孤岛,孤链和环进行修复,将非辐射状网络修复为辐射状网络。算例结果验证了该算法的实用性和有效性。
To solve the problem that traditional genetic algorithm（GA） often falls into local optimal solution and along with the expansion of distribution network scale the searching efficiency declines,distribution network planning based on adaptive multiple population genetic algorithm is proposed through drawing the idea of multiple population and adaptive thinking.By means of processing the objective function,multi-species are led in,the adaptive GA and the improved multiple population GA,in which the evolutionarily stability is considered,are applied to different species,and using the translocation of superior individuals the cooperation among species is implemented.Meanwhile,to solve the problem of large amount infeasible solutions caused by applying GA to distribution network planning,by use of graph theory and search technique a restoration scheme for infeasible solution is given,by means of restoring isolated nodes,isolated chain and closed loop,it is implemented to restore non-radial network to radial network.Calculation examples show that the proposed method is practical and effective.
出处
《南方电网技术》
2010年第5期87-91,共5页
Southern Power System Technology
关键词
配电网规划
多种群
遗传算法
自适应
进化稳定
不可行解修复
distribution network planning
multiple population
genetic algorithm（GA）
adaptive
revolutionarily stability
infeasible solution restoration
