摘要
针对窃电形式多元化的趋势,综合现有窃电检测技术的不足,提出了基于关联度分析的窃电模型。设立失压断相预警、电量差动预警、电量波动预警、功率差动预警、电能表停走预警、电流异常预警、恒定磁场干扰预警、线损异常预警,并通过关联度分析对线损值进行判定;基于python搭建线损检测模型,可快速实现线损异常用户识别,最终验证窃电模型的准确性,降损会取得较好的治理效果。
Aiming at the trend of diversified forms of electricity theft,combined with the shortcomings of existing electricity theft detection technologies,a electricity theft model based on correlation analysis is proposed.It establishes early warning for loss of voltage and phase,differential power warning,power fluctuation warning,power differential warning,electric energy meter stop and running warning,current abnormal warning,constant magnetic field interference warning,line loss abnormal warning,and line loss value analysis through correlation analysis to make a judgment.It builds a line loss detection model based on python,and tests the test set to quickly realize user identification of abnormal line loss.The actual line verifies the accuracy of the electricity theft mode lultimately,and the line loss has a better control effect.
作者
何立强
隋政澔
杜峰
HE Liqiang;SUI Zhenghao;DU Feng(State Grid Dandong Power Supply Company,Dandong,Liaoning 118000,China)
出处
《东北电力技术》
2021年第4期29-31,45,共4页
Northeast Electric Power Technology
关键词
反窃电
台区线损
大数据分析
电能计量
anti-stealing power
station line loss
big data analysis
electric energy measurement
