Fault Diagnosis Using Artificial Neural Network for Two-Level VSI in PMSM Drive System

A permanent magnet synchronous motor's (PMSM) drive system depends heavily on the voltage source inverter (VSI). However, the VSI shows vulnerability to faults during operation, especially when it comes to open circuit switch faults, which are the most frequent and have the most significant negative impact on the drive system. In order to diagnose Open Circuit Faults (OCF) in the Voltage Source Inverter (VSI) of a three-phase permanent magnet synchronous motor drive system, this work employs the vector average current technique in Clark's transformation and an artificial neural network. Diagnostic variables are computed based on the current distortion to gather details about the fault. An artificial neural network (ANN) is put to use to process the fault information in order to determine and locate the specific location of the faulty switch. The study's simulation results show that the methodology employed effectively identifies and localises open circuit faults in a two-level, three-phase inverter.