基于基本环矩阵及改进并行遗传算法的配电网重构  

Distribution Network Reconfiguration Based on Basic Ring Matrix and Improved Parallel Genetic Algorithm

作　　者：王宁[1] 王淳[1] 熊云[1] 黄辉[1] 孔令飞[1] 

机构地区：[1]南昌大学电气与自动化系,江西南昌330031

出　　处：《水电能源科学》2016年第10期198-201,共4页Water Resources and Power

基　　金：国家自然科学基金项目(51167012)

摘　　要：为了快速准确地去除配电网重构过程中的不可行解,给出基本环信息矩阵以包含所有配电网含环路情况,从而在一定条件下保证非基本环信息矩阵内的方案可行,同时避免小生境遗传算法收敛于局部最优解,提出小生境—粗粒度混合遗传算法求解配电网重构模型,即将小生境遗传算法中的子父代共同竞争淘汰机制与粗粒度并行遗传算法中的迁移操作相结合,再通过试验获取最优参数配置。算例结果表明,该方案较传统小生境遗传算法求解配电网重构模型具有更高的寻优成功率和更快的收敛速度。In order to remove infeasible solution in the process of distribution network reconfiguration quickly and accurately,this paper put forward basic ring information matrix,which includes all loop distribution network.So,the plan in the non-basic ring information matrix is feasible under certain conditions.At the same time,for avoiding the niche genetic algorithm converging to local optimal solution,this paper put forward niche coarse-grained genetic algorithm（NCGA）to solve the distribution network reconfiguration model,which is composed by combination child-parents＇ mutual competition and elimination mechanism of niche genetic algorithm（NGA）and migration of coarse-grained genetic algorithm（CGA）.The optimal parameter configuration was obtained by the test.Numerical example results show that the NCGA has higher success rate of searching optimization and faster convergence speed than traditional NGA for solving distribution network reconfiguration model.

关 键 词：小生境 粗粒度 基本环信息矩阵 遗传算法 迁移 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TM76[自动化与计算机技术—控制科学与工程]