基于改进随机森林算法的低压用户窃电行为自适应监测方法  

An Adaptive Monitoring Method Based on Improved Random Forest Algorithm for Power Stealing Behavior of Low-voltage Users

作　　者：王骏 梁东 高迪 董振祥 高晓婧 陈婧 WANG Jun;LIANG Dong;GAO Di;DONG Zhenxiang;GAO Xiaojing;CHEN Jing(State Grid Jibei Electric Power Co.,Ltd.,Beijing 100052,China;Zhangjiakou Power Supply Company of State Grid Jibei Electric Power Co.,Ltd.,Zhangjiakou 075001,China;State Grid Info-Telecom Great Power Science and Technology Co.,Ltd.,Fuzhou 350003,China)

机构地区：[1]国网冀北电力有限公司,北京100052 [2]国网冀北电力有限公司张家口供电公司,河北张家口075001 [3]国网信通亿力科技有限责任公司,福建福州350003

出　　处：《微型电脑应用》2024年第6期202-205,210,共5页Microcomputer Applications

摘　　要：用户窃电行为频发,针对窃电后的日用电量,电压、三相电流变化监测过程困难,精度低的问题,提出基于改进随机森林算法的低压用户窃电行为自适应监测方法。利用分布式结构采集低压用户的用电信息,通过控制振荡器(CCO)与单线程单元(STA)的相互配合,实现一天内96点电表曲线数据采集;从已采集到的用电信息中提取窃电用户的电量、电流与电压的用电曲线偏离度特征;将这些特征输入随机森林算法开始训练,经训练后获取自适应监测样本,并向随机森林算法中引入惩罚项因子,提升训练能力,使其具有自适应性,完成低压用户窃电行为的自适应监测。经实验验证,该方法能够精确采集用户的日用电量,在监测时可迅速发现用户窃电后的电压、三相电流变化,同时还可以实现多种窃电行为的监测。The electricity stealing behavior of users occurs frequently.Aiming at the problems of difficult monitoring process and low accuracy of daily power consumption,voltage and three-phase current changes after the electricity theft,an adaptive monitoring method of low-voltage users’electricity stealing behavior based on improved random forest algorithm is proposed.The distributed structure is used to collect the power consumption information of low-voltage users.Through the cooperation between the control oscillator(CCO)and the single threaded unit(STA),the 96 point meter curve data collection in one day is realized.It extracts the deviation degree characteristics of power consumption curve of electric quantity,current and voltage of power stealing users from the collected power consumption information.These features are input into the random forest algorithm to start training.After training,the adaptive monitoring samples are obtained,and the penalty factor is introduced into the random forest algorithm to improve the training ability and make it adaptive,so as to complete the adaptive monitoring of low-voltage users’electricity theft behavior.The experimental results show that this method can accurately collect the daily power consumption of users,quickly find the changes of voltage and three-phase current after the power theft,and realize the monitoring of a variety of power theft behaviors.

关 键 词：改进随机森林算法 低压用户 窃电行为监测 三相电流 

分 类 号：TM714[电气工程—电力系统及自动化]