基于加权K-means聚类和遗传算法的变电站规划  

Substation Planning Based on Weighted K-means Cluster Algorithm and Genetic Algorithm

作　　者：成乐祥[1] 季丽 

机构地区：[1]国网南京供电公司,江苏南京210019 [2]国网江苏省电力公司电力科学研究院,江苏南京211103

出　　处：《江苏电机工程》2016年第6期9-12,共4页Jiangsu Electrical Engineering

摘　　要：针对变电站规划问题,提出了基于加权K-means聚类的变电站供电范围划分方法,并在此基础上提出了基于加权K-means聚类和遗传算法的变电站规划算法。该算法运用遗传算法的全局搜索能力确定变电站的座数、主变台数和容量的最优组合,解决了应用加权K-means聚类算法划分变电站供电范围时初始聚类数确定的问题。加权K-means聚类算法能够综合考虑变电站的负载率和供电半径的约束,并在迭代过程中自适应调节。算例结果表明所提算法能够较好地求解变电站优化规划问题。The paper proposed a partitioning method of substation service areas based on weighted K-means clustering algorithm, based on which substation planning algorithm was put forward further with genetic algorithm. The algorithm determines the optimal combination of the number of substations and main transformers and the capacity of main transformer by the global searching ability of genetic algorithm, which solves the problem of determining initial clustering number when partitioning substation service areas by weighted K-means clustering algorithm. The weighted K-means clustering algorithm can comprehensively meet the constraints of load ratio and supply radius of substation and take adaptive adjustment during iteration. The example shows that the proposed algorithm can serve for substation planning well.

关 键 词：变电站规划 加权K-means聚类算法 遗传算法 变电站站址 供区优化 

分 类 号：TM721[电气工程—电力系统及自动化]