• ZHENG Yingjun ¹

  ZHENG Yingjun

  College of Electrical and Information Engineering, Changsha University of Science and Technology, Changsha 410004, China
  在知网中查找
  在百度中查找
  在本站中查找
  
• YANG Yining ²

  YANG Yining

  China Electric Power Research Institute, Beijing 100192, China
  在知网中查找
  在百度中查找
  在本站中查找
  
• SHU Yifei ³

  SHU Yifei

  Marketing Service Center of State Grid Ningxia Electric Power Co., Ltd. (Metrology Center), Yinchuan 750001, China
  在知网中查找
  在百度中查找
  在本站中查找
  
• XUE Yang ²

  XUE Yang

  China Electric Power Research Institute, Beijing 100192, China
  在知网中查找
  在百度中查找
  在本站中查找
  
• LIU Aiguo ³

  LIU Aiguo

  Marketing Service Center of State Grid Ningxia Electric Power Co., Ltd. (Metrology Center), Yinchuan 750001, China
  在知网中查找
  在百度中查找
  在本站中查找
  
• SU Sheng ¹

  SU Sheng

  College of Electrical and Information Engineering, Changsha University of Science and Technology, Changsha 410004, China
  在知网中查找
  在百度中查找
  在本站中查找
  

1.College of Electrical and Information Engineering, Changsha University of Science and Technology, Changsha 410004, China;2.China Electric Power Research Institute, Beijing 100192, China;3.Marketing Service Center of State Grid Ningxia Electric Power Co., Ltd. (Metrology Center), Yinchuan 750001, China

Electricity theft of low-voltage users increases the line loss power. It is an effective way to identify the electricity theft users by the attribution analysis of line loss variation in distribution area. The frequent communication anomaly of low-voltage users may lead to the distortion of the data in the main station of the electricity information acquisition system, which tends to mislead the electricity theft detection. Using the characteristic that the distribution transformer terminal can collect the distribution area data completely and accurately, an electricity theft detection method for low-voltage users based on edge computing is proposed. Firstly, the correlation between electricity consumption and line loss power of electricity theft users in distribution area is analyzed under normal and abnormal communication conditions. Then, the attribution analysis is carried out on the line loss power and electricity consumption in the real distribution of the electricity theft detection module in distribution transformer terminal to identify electricity theft users. Finally, based on the simulation analysis of the actual data in the high-loss distribution area, the advantages of the proposed method compared with electricity theft detection using the abnormal data on the main station side and the data restored by different missing data filling algorithms are verified.

electricity theft detection;communication anomaly;distribution transformer terminal;edge computing;missing data completion