Distribution Network Fault Diagnosis Method Based on Optimized Wavelet Transform and Convolutional Neural Network

Abstract

The distribution network connects the power system to power users. Fast and accurate fault diagnosis technology of distribution networks is an inevitable requirement to increase the security of electricity supply. A fault diagnosis method based on optimized wavelet transform with CNN is proposed for fast diagnosis of distribution network faults. Firstly, the original data obtained from the distribution network is subjected to a continuous wavelet transform to obtain the time-frequency matrix, and the fault feature data is constructed. The convolutional neural network model is used to judge and identify the types of fault to realize the diagnosis of a distribution network fault. Comparison with common fault diagnosis methods of neural networks, the optimization method can significantly enhance the accuracy of fault identification and has fast speed and good stability.