PWM Voltage-Based Modeling for PM Machines With Interturn Short Circuit Fault Considering the Effect of Drives

• Published in: IEEE transactions on industrial electronics (1982) Vol. 70; no. 11; pp. 10981 - 10991
• Main Authors: Qin, Ying, Li, Guang-Jin, Jia, Chunjiang, Mckeever, Paul
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Analytical models; Cables; Circuit faults; Coils; Electric potential; Fault model; Field effect transistors; Harmonics; Integrated circuit modeling; interturn short circuit (ITSC) fault; Inverters; Mathematical models; Metal oxide semiconductors; Modelling; permanent magnet (PM) machines; Permanent magnets; Pulse duration modulation; Pulse width modulation; pulsewidth modulation (PWM) voltage; Semiconductor devices; Short circuits; Voltage; Windings
• ISSN: 0278-0046; 1557-9948
• DOI: 10.1109/TIE.2022.3229357

This article presents a novel analytical fault model based on pulsewidth modulation (PWM) voltage rather than ideal sinewave phase voltage for permanent magnet (PM) machines with interturn short circuit fault. This model is important because PM machines are usually driven by the PWM inverter. In addition, main contributors to the accuracy of the fault model such as the cables between the inverter and the machine, the precision of the rotor position estimation, and the inverter, in particular the metal-oxide semiconductor field-effect transistors have been considered. Therefore, this developed model can fully account for the effects of PWM harmonics and accurately predicts the machine's behaviors (particularly the fault current) under fault conditions considering the effect of the drives. This has rarely been studied in the literature as most existing fault modeling methods adopt sinewave phase voltage as inputs, which cannot account for the PWM harmonics in the fault currents. However, the investigations in this article showed that the PWM harmonic component could be close to 25% of the fundamental component in the fault current, and hence cannot be neglected. The accuracy of the proposed fault model has been validated by a series of experimental tests.