基于不平衡电能数据的反窃电智能监测系统研究  

Imbalanced-electricity-data-based Intelligent Monitoring System against Electricity Theft

作　　者：张爱梅 ZHANG Aimei(Guiyang Power Supply Bureau of Guizhou Power Grid Co.,Ltd.,Guiyang 550002,China)

机构地区：[1]贵州电网有限责任公司贵阳供电局,贵州贵阳550002

出　　处：《电工技术》2024年第4期134-136,共3页Electric Engineering

摘　　要：面对电网中的窃电行为,设计一个基于不平衡电能数据的反窃电智能监测系统,完成用电负荷数据的预处理后,运用随机过采样—迭代决策树算法建立计量装置的异常监测和用户异常用电行为的智能分析数学模型,快速检测出异常用电客户,并在此基础上开发嵌入式专家分析系统,实现用户的电能数据在线监测分析,自动展示异常用电数据,定位窃电用户。The present study designed an anti-electricity-theft intelligent monitoring system based on imbalanced energy data.After the preprocessing of electricity load data,the system realizes rapid determination of users with abnormal behaviors by using the combined algorithm of random oversampling examples-gradient boosting decision tree to construct an intelligent mathematical model for monitoring anomalies of metering devices and electric consumption anomalies of users.Based on this,an embedded expert analysis system is developed to achieve online monitoring and analysis of electricity data,automatically display abnormal electric consumption data,and locate electricity thefts.

关 键 词：不平衡数据 随机过采样 迭代决策树算法 反窃电系统 

分 类 号：TM72[电气工程—电力系统及自动化]