Accurate identification partial discharge of cable termination for high-speed trains based on wavelet transform and convolutional neural network

•WT-CNN effectively identifies PD in cable termination of high-speed train.•No need to manually extract features to achieve high-precision recognition.•Improved signal recognition accuracy of CNN by WT noise reduction. The cable termination is an important part of the energy transmission of high-speed trains and is also a weak link in the insulation. It is important to determine the insulation status of the cable termination by the state of the partial discharge signal, but the partial discharge signal in the field test circuit is mixed with a large amount of external corona interference, which affects the detection accuracy. This paper proposes a wavelet transform and convolutional neural networks (WT-CNN) based classification model for the accurate and fast identification of corona disturbances in the partial discharge signals of cable termination for high-speed train. The method uses wavelet transform noise reduction and adaptive moment estimation (Adam) as the optimizer. The effect of WT on the classification effect of CNN is analyzed, and it is shown that the method can still achieve over 94% accuracy with small training samples.