Research on Online Fault Diagnosis Method for Single-Phase Two-Level Pulse Rectifier

Traction rectifier is the most front-end power electronic equipment in the traction drive system of high-speed train. Its safety and reliability directly determine whether other equipment in the system can work normally. The research object of this paper is the single-phase two-level pulse rectifier of traction converter. For the sensor fault and open circuit fault diagnosis of power equipment, the online fault diagnosis method of single-phase two-level pulse rectifier based on data-driven learning is studied. Firstly, the fault model is built, and the fault data set is formed by collecting the operation data. The fault diagnosis model is trained and tested. In this paper, Random Forest, Extreme Learning Machine and Support Vector Machine are used to train the model, and the off-line diagnosis results are obtained. Based on the off-line diagnosis model, the on-line fault diagnosis model is established by using Matlab/Simulink, and the fault diagnosis of single-phase two-level pulse rectifier is verified online, and good off-line and on-line diagnosis results are obtained.