Detection of Electric Power Theft Behavior Based on BP Neural Network Optimized by L-M Algorithm

The traditional electric power theft technology is difficult to identify the existing behavior of electric power theft. This paper proposes a neural network model optimized by L-M (Levenberg-Marquardt) algorithm. This model is based on the historical power consumption data of power grid users, and reveals the normal and abnormal power consumption data of power grid users from the perspective of data the difference.