RPCA-SVM fault diagnosis strategy of cascaded H-bridge multilevel inverters

Abstract

In order to improve the accuracy of the fault diagnosis and accelerate the operation speed in a cascaded H-bridge multilevel inverter system (CHMLIS), a fault diagnosis strategy based on Relative Principle Component Analysis-Support Vector Machine (RPCA-SVM) is presented in this paper. In this strategy, the output voltage of CHMLIS, which is preprocessed through the fast Fourier transform (FFT), is used to identify the type and location of occurring fault through a SVM model. Then RPCA is utilized to reduce input sample's dimension. A lower dimensional input sample will reduce the time necessary to train the SVM model, and the reduced noise may improve the mapping performance. Compared with other traditional fault diagnosis methods, the proposed strategy has much higher computing efficiency and diagnosis accuracy in fault diagnosis. Simulation results and experimental results have validated the RPCA-SVM fault diagnosis strategy in CHMLIS.