基于协同进化算法的配电网络重构方案  

Distribution network reconfiguration based on co-evolution algorithm

作　　者：欧帝宏[1] 陈皓勇[1] 荆朝霞[1] 

机构地区：[1]华南理工大学电力学院,广东广州510640

出　　处：《电力自动化设备》2012年第7期108-113,共6页Electric Power Automation Equipment

摘　　要：提出了一种基于协同进化算法的配电网络重构方案,将配电网络重构分解为各个回路优化的子问题,采用分开进化、定期协同的思想来处理各个子问题,即生成与回路数目相等的种群,在确保全局最优解为优化方向的前提下,各种群独立进化对应各回路的优化过程。特殊的基因操作减少了不可行解的数量,提高了优化效率;使用"确定树"方法进行不可行解的修复,降低了陷入局部最优解的概率。算例结果表明所提算法具有较好的计算结果和较高的计算效率。A distribution network recorlfiguration scheme based on co-evolution algorithm is proposed,which breaks down the distribution network reconfiguration into the optimization of sub-loops and applies the concept of separate evolution under periodic coordination to realize the optimization. The populations same in quantity as the sub-loops are generated and independently evolved with the globally optimal solution as its optimization direction. A specific genetic operation is used to reduce the infeasible solutions for enhancing the optimization efficiency and the ＂certain tree＂ is adopted to restore the infeasible solution for avoiding the locally optimal solution. Case study proves its better effect and efficiency.

关 键 词：电力系统 配电网管理系统 配电网络重构 协同进化算法 确定树 进化算法 

分 类 号：TM727.2[电气工程—电力系统及自动化]