Abstract: At present, the traction power supply equipment has the problems of lack of accurate means of operation management and insufficient maintenance level. At the same time, the traditional monitoring scheme is more and more difficult to meet the new demand of high-speed railway equipment management. In order to strengthen the operation and maintenance management and control ability of high-speed railway traction power supply equipment, this paper presents an architecture of high-speed railway traction power supply equipment fault prediction and health management platform based on multi-source data fusion. With equipment digital maintenance as the core, intelligent data fusion and cloud decision-making as the main line, and supported by new technologies such as big data drive, this paper deeply excavates the professional data of operation, maintenance and repair of traction power supply equipment, builds relevant platforms to comprehensively manage the relevant system information of traction power supply network, and provides decision-making support for the intelligent operation, maintenance and repair of traction power supply equipment and the digitization of production management, so as to realize the orderly planning, fine maintenance and health management evaluation of traction power supply network.