Open Switch Fault Detection in Electric Vehicle Drives Using Support Vector Machine

This paper presents an open switch fault (OSF) detection and faulty switch identification technique in a three-phase Electric Vehicle (EV) drive using Support Vector Machine (SVM). Three-phase instantaneous current signals obtained from the inverter output have been given to SVM for OSF detection and faulty switch identification. The performance of SVM has been assessed as a binary-class and multi-class classifier for OSF detection and classification. Various cases have been simulated under healthy and OSF conditions with different load and speed variations. The performance of the proposed technique has been satisfactory for OSF diagnosis in inverter-fed EV drives. Moreover, it reduces the computation burden on the controller.