Fault Detection Method Of Modular Multilevel Converter Based On Current Mean Value And Svm

In this paper, the IGBT open-circuit fault characteristics in sub-module(SM) for Modular multilevel converter(MMC) are studied, and a fault detection method based on mean value of the three-phase ac current to extract fault characteristics and support vector machine(SVM) classification is proposed. When a fault occurs, the mean value of three-phase ac current is no longer zero, which can detect the occurrence of the fault. When the open-circuit fault occurs in the SMs of different bridge arms, the mean value of three-phase ac current produces different dc bias components. According to the change rule of the mean value of three-phase current, the fault characteristics are extracted, and then faults are classified with SVM classifier. Finally, experimental results show that the proposed method can detect the fault quickly and classify the faults accurately.