Abstract: This article addresses the intelligent operation and maintenance requirements of the traction power supply system for urban rail transit, and proposes an intelligent diagnosis method based on the Mallat algorithm of wavelet transformation. The contact network fault features are extracted through multi-resolution analysis, enabling reliable and rapid fault location. This method has been applied in the No. 1 Line of Shijiazhuang Metro and has formed a complete and feasible intelligent operation and maintenance plan. The operation results show that the accuracy and response speed of fault diagnosis using the Mallat algorithm have been significantly improved. It can detect fault signs earlier and accurately identify the fault points, promoting the operation and maintenance mode from "post-event handling" to "pre-event prevention", and becoming the most powerful support for the intelligent and digital development of rail transit operation and maintenance.