The problem of diagnosing faults in high speed railway traction network is addressed in this study. A fault diagnosis algorithm based on a multi-layer perceptron is proposed as a solution. The algorithm utilizes voltage and current data from 14 measurement points. Ten time-domain features are extracted from the data, including maximum value, minimum value, peak-to-peak value, mean value, root mean square value, waveform factor, rectified mean value, pulse factor, skewness, and kurtosis. The Relief-F algorithm is employed to rank the importance of these features, followed by a forward search process for optimization. The results demonstrate that the proposed method achieves a high level of accuracy in fault detection and classification. This approach provides valuable insights for further research in the field of fault diagnosis for overhead contact systems.

High speed railway traction network,Relief-F algorithm,multi-layer perceptron,fault diagnosis