Distribution cable arc-earth fault identification based on ensemble empirical mode decomposition and one-dimensional convolutional neural network

Distribution cables are an important part of the power grid, and arcing ground faults in cables can pose a major safety hazard to the grid. Due to fault signals' complex and non-smooth characteristics, extracting features and identifying them accurately remains a challenge. In this paper, a distribution cable arc-earth identification method based on ensemble empirical mode decomposition (EEMD) and a one-dimensional convolutional neural network (CNN) is proposed. Based on the relevant principles of EEMD, a more detailed decomposition of the signal is realized to extract its effective features; then a one-dimensional convolutional neural network is constructed to realize fault feature extraction and classification; finally, the PSCAD/EMTDC simulation software is used to generate a cable fault signal dataset to verify the effectiveness of the algorithm in this paper. The simulation results show that EEMD-CNN achieves higher accuracy in cable fault diagnosis compared with EMD decomposition and traditional classification models (99.4%), and is an effective cable fault diagnosis method that can improve the reliability and efficiency of grid operation.