Fault Diagnosis of Switched Reluctance Motor Power Converter Based on VMD-MPE

Aiming at the characteristics of power converter fault signals that are nonlinear and unstable, and useful information is easily covered by noise, this paper proposed a new fault feature extraction method. This method introduced variational modal decomposition, and obtained several intrinsic mode functions through the variational modal decomposition of DC bus current, used mutual information analysis to select practical modal components, calculated the average value of multi-scale effective component permutation entropy as the feature vector, and used support vector machine classifiers for fault identification. To verify the above algorithm's feasibility, the author established a simulation model, compared it with traditional wavelet transform and other diagnostic algorithms, and established an experimental platform for the switched reluctance motor system to test the open circuit and short circuit fault states. Simulation and experimental results show that the method proposed in this paper can reduce noise interference and improve fault identification accuracy.