Investigation of Extracting a Transfer Function from Transient Signals During Switching Operations of Inverters for On-Line Monitoring of Electrical Machines

This paper investigates the potential of using electric terminal signals of induction machines for online fault monitoring through experimental studies. The results reveal that by analyzing the transient voltage signals at electrical machine terminals, including the neutral point, it is possible to detect inter-turn short-circuit electrical faults online in real-time. The detection mechanism is based on analyzing signals and their correlation in high-frequency domains. These transient signals naturally exist due to the inverter-fed operations. Therefore, this work invokes the potential of realizing a non-invasive and convenient online monitoring of insulation fault detection.