Open-Circuit Fault Diagnosis Strategy for Novel DFPMM Drive Based on Fuzzy Logic

To diagnose different types of open-circuit (OC) faults in novel dual-winding fault-tolerant permanent magnet motor (DFPMM) drive, a fault diagnosis strategy based on fuzzy logic is proposed and investigated. The proposed DFPMM drive requires less independent power supply and power switches, its reliability, utilization, and power density are improved compared with some existing multiphase fault-tolerant drives. To locate power switch intermittent OC fault that is difficult to be diagnosed by traditional strategies, the fuzzy logic theory is used in the proposed strategy, the membership functions and fuzzy rules in the proposed fuzzy logic system are specially designed and optimized to meet the requirements of fault diagnosis. The proposed fault diagnosis strategy has strong robustness, due to the optimized dynamic average calculation cycle and normalization, its diagnosis results are not affected by speed or load mutation; Additionally, an error-proof processor is added to the system as a part of the proposed strategy, which prevents misdiagnosis by improving the shape of phase current waveform when DFPMM operates at no-load or light-load state. The proposed strategy can quickly and accurately locate winding OC, power switch OC, and power switch intermittent OC faults in DFPMM drive, furthermore, it can quantify the fault level of intermittent OC fault for staff to take different remedial measures according to different situations. Simulation and experimental results verify the correctness and efficiency of the proposed strategy.

Automated summary

This article proposes a fault diagnosis strategy based on fuzzy logic for diagnosing different types of open-circuit faults in a novel dual-winding fault-tolerant permanent magnet motor drive. The strategy demonstrates strong robustness and accuracy, allowing for quick identification and quantification of open-circuit faults in the DFPMM drive, as well as providing appropriate remedial measures.