模糊C均值聚类算法在用电异常稽查中的应用  

Application of Fuzzy C-means Clustering Algorithm in Inspection of Abnormal Electricity

作　　者：王鹏伍[1] 孙志杰[1] 傅军[1] 周国鹏[1] 

机构地区：[1]国网冀北电力有限公司电力科学研究院(华北电力科学研究院有限责任公司),北京100045

出　　处：《华北电力技术》2016年第4期14-18,共5页North China Electric Power

摘　　要：在电力系统电能损失中,非技术损失占有相当大的比例。为了有效稽查出电力系统中的用电异常用户,文中使用电力用户负荷历史曲线数据和模糊C均值算法进行聚类分析,得到电力用户负荷典型曲线,然后通过比较用户实际负荷曲线与相应行业负荷典型曲线的Camberra距离,分析可能发生异常用电的用户,从而对相应用户进行针对性用电异常稽查。In power losses of power system,nontechnical losses occupy a large proportion. To inspect abnormal elec- tricity customers in power system effectively, historical load curve data and fuzzy C-means algorithm are used for clus- tering analysis in this paper. Typical load curves of electricity customers are obtained. By comparing Camberra dis- tances of actual load curves and typical load curves of corresponding industry, potential abnormal electricity customers are analyzed. Corresponding customers are inspected specifically.

关 键 词：电力系统 负荷曲线 模糊C均值 用电异常 Camberra距离 

分 类 号：TM930.1[电气工程—电力电子与电力传动]