基于支持向量机的电力用户异常用电行为智能检测方法  

Intelligent Detection Method for Abnormal Electricity Consumption Behavior of Power Users Based on Support Vector Machine

作　　者：王奎 刘昊 WANG Kui;LIU Hao(State Grid Tianjin Jinghai Power Supply Co.,Ltd.,Tianjin 301600,China)

机构地区：[1]国网天津静海供电有限公司,天津301600

出　　处：《通信电源技术》2023年第20期20-22,共3页Telecom Power Technology

摘　　要：针对现有检测方法在检测电力用户异常用电行为时存在检测到的异常用电行为用户数与实际不符、检测精度受到严重影响等问题,引入支持向量机,设计电力用户异常用电行为智能检测方法。通过缺失值处理和归一化处理,实现对电力用户用电行为数据的预处理。利用支持向量机检测异常用电数据,结合得到的异常用电数据检测结果识别异常用电行为,并完成智能预警。通过对比实验证明,新的检测方法检测到的异常用电行为用户数与实际相接近,差距较小,具备极高的检测精度。In order to solve the problems that the number of users with abnormal power consumption behavior detected by the existing detection methods is inconsistent with the actual situation and the detection accuracy is seriously affected,support vector machine is introduced to design and study an intelligent detection method for abnormal power consumption behavior of power users.Through the missing value processing and normalization processing,the pretreatment of electricity consumption behavior data of power users is realized.Using support vector machine to detect abnormal electricity consumption data,combined with the detection results of abnormal electricity consumption data,abnormal electricity consumption behavior is identified and intelligent early warning is completed.Through comparative experiments,it has been proven that the new detection method detects a similar number of users with abnormal electricity consumption behavior to the actual situation,with a small difference and extremely high detection accuracy.

关 键 词：支持向量机 电力用户 异常用电行为 智能检测 

分 类 号：TM933.4[电气工程—电力电子与电力传动]