配电网用户日用电量异常智能分析研究  

Intelligent Analysis Research on Abnormal Daily Electricity Consumption of Distribution Network Users

作　　者：冯建伟 王联智 谢敏 符艺超 周吉星 Feng Jianwei;Wang Lianzhi;Xie Min;Fu Yichao;Zhou Jixing(China Southern Power Grid Hainan Digital Power Grid Research Institute Co.,Ltd.,Haikou 570100,China)

机构地区：[1]南方电网海南数字电网研究院有限公司,海口570100

出　　处：《现代科学仪器》2021年第2期193-197,共5页Modern Scientific Instruments

摘　　要：当前我国经济发展的脚步逐渐加快,人民的物质生活水平不断提升,但与此同时,部分用户的异常用电行为导致电力损失的情况愈加严重。为了对配电网中所有用户的日用电量异常进行智能分析,解决异常用电的问题,减小电力损失,此次研究通过应用PCA对配电网用电用户数据进行预处理与特征构建,再重点应用LOF算法进行用电量异常分析。结果显示,此次研究中应用的LOF算法能够高效准确地对用户的日用电量进行分析,进而对该用户进行识别。这表明基于LOF算法的配电网用户日用电量异常智能分析,具有较强的实用性与准确性。At present,the pace of China's economic development is gradually accelerating,and people's material living standards are constantly improving,but at the same time,the abnormal electricity consumption behavior of some users leads to more and more serious power loss.In order to intelligently analyze the abnormal daily electricity consumption of all users in the distribution network,solve the problem of abnormal power consumption,and reduce the power loss,this study uses PCA to preprocess the data of power users in the distribution network and construct the characteristics,and then focuses on the analysis of abnormal electricity consumption by using LOF algorithm.The results show that the LOF algorithm used in this study can effectively and accurately analyze the daily power consumption of the user,and then identify the user.This shows that the intelligent analysis of abnormal daily electricity consumption of distribution network users based on LOF algorithm has strong practicability and accuracy.

关 键 词：配电网用户 LOF算法 日用电量 异常用电 

分 类 号：TM769[电气工程—电力系统及自动化]