The Substation Fault Prediction and Health Management System is based on design data, online monitoring data, offline test data, and equipment maintenance data. It utilizes signal processing technology, data processing technology, artificial intelligence technology, and advanced reasoning models to perform prediction and assessment. The system achieves pre-fault early warning and prediction, in-fault diagnosis and rapid handling, and post-fault evaluation and analysis for traction substation equipment. Ultimately, it realizes the goal of full-lifecycle management for all equipment in traction substations.