Multi-switch Fault Diagnosis for the Voltage Source Inverter fed Multi-phase Motor Drives Based on Machine Learning

Existing fault diagnosis methods cannot be used to various types of multi-phase machines due to different types of stator phase arrangements and its faults. With the advent of big data and artificial intelligence, a machine-learning-based fault diagnosis approach is proposed for multi-phase drive for an electric transportation applications. In this paper, the possible ML techniques and their performance have been studied and implemented for the five-phase voltage source inverter (VSI) fed induction motor drive. For this, the possible switch faults (open and short) has been considered to collect the data for training and testing. The testing comparative results have been presented for fault detection and classification.