Abstract: In order to improve the quasi-determination of the fault diagnosis of mine substation equipment, based on the analysis of the sources and hazards of the fault of high voltage electrical equipment and the causes of the fault of high voltage electrical equipment, this paper constructs the real-time temperature change monitoring system of mine substation electrical equipment with the help of advanced technology to monitor the temperature change parameters, and integrates BP neural network and D-S evidence theory to analyze and optimize the monitoring data. Diagnose the fault status of the electrical equipment.