13 December 2021 in IEEE Transactions on Instrumentation and Measurement
Demagnetization fault in a permanent magnet synchronous motor (PMSM) degrades its performance and may cause serious catastrophes. Fault localization technique can improve the reliability and safety of PMSM and provide important information for developing operation and maintenance strategies. Thus, a fault localization method is so important. Mainstream is a common method that uses search coil installed on every stator tooth to diagnose fault location. Due to the use of a large number of search coils, the method is limited in terms of invasiveness, implementation cost, and installation. Moreover, the method can only diagnose the fault location in PMSM with one demagnetized permanent magnet (PM). This article proposes a novel method which can detect the specific location of demagnetized PMs in PMSM with any types of demagnetization faults. The method only needs three additional toroidal yoke coils wound around the stator yoke. Toroidal yoke coils are used to measure the back electromotive force (EMF). The demagnetization fault localization signals (DFLSs) are produced by basic calculation of the output signals of three search coils. The DFLSs are partitioned according to the electrical period. The location of demagnetization fault can be diagnosed through analyzing each electric period waveform of DFLSs. First, the arrangement of toroidal-yoke-type search coil and the construction method of DFLSs are introduced. Second, a partitioned analytical model of DFLSs is developed. Third, waveforms in one electric period of DFLSs under preset fault types are investigated. Finally, the method for diagnosing demagnetization fault in PMSM is proposed. The simulation and experimental results justify the precision of the proposed search coil which shows the maximum error of back EMF to be 4.1%. It also shows the ability of the method to locate demagnetized PMs when the demagnetization severity (the ratio of the change in coercive force of PM to the normal coercive force) is 10%. This study provides a new method for diagnosing demagnetization fault, as well as provides a significant reference for fault post-processing and fault-tolerant control of PMSM.
22 February 2022 in IET Electric Power Applications
The mass production of the motors causes tolerance of shape and dimension, deviation of permanent magnet remanence and rotor eccentricity error, which affect the cogging torque and torque ripple amplitude and performance consistency. In order to reduce the torque ripple of the motor in the actual condition, the robustness optimization is performed in the paper. Firstly, an improving convex arc type permanent magnet structure is adopted to improve air gap flux density and suppress the cogging torque. Secondly, the structure parameters of the magnetic pole are selected as optimization variables, and the magnetization angle, remanence and position of the permanent magnet, static and dynamic rotor eccentricity are considered as noise factors. To improve the overall robustness of the motor under different operating conditions, the dynamic Taguchi method is used to optimize the robustness of the motor, and the test data is processed through the relation analysis method to obtain the optimal combination of control factors. Finally, the prototype is manufactured for the experiment. Compared with the single‐operating conditions robust optimization results, the robust optimization method improves the robustness of the motor. The cogging torque amplitude is reduced by 39.2%. The torque ripple is 38.4% lower than that before optimization.