Abstract: Accurate diagnosis of cable faults is crucial for the reliable operation of power systems. Traditional traveling wave signal recognition methods are affected by noise interference and waveform complexity, resulting in insufficient accuracy and low efficiency. This research aims to develop an artificial intelligence-based method for identifying cable fault traveling wave signals, improving the accuracy and intelligence level of fault location and type judgment. The method utilizes deep learning technology to construct a recognition model, preprocesses and extracts features from the original traveling wave signal, and trains the model to distinguish various typical fault signal features under complex background noise. The experiment is validated using a simulation dataset, and the results show that the proposed artificial intelligence model effectively overcomes the limitations of traditional methods, significantly improves the recognition accuracy and robustness of cable fault traveling wave signals under different working conditions, and provides reliable technical support for intelligent cable fault diagnosis.