Rotor Fault Monitoring And Diagnosis Of Permanent Magnet Synchronous Motor With Fewer Flux Search Coils

With the application of permanent magnet synchronous motors(PMSM)expanding to large-capacity and key area,its reliability and safety problems are becoming more and more prominent,and online diagnosis and monitoring of various faults are also becoming more and more important.The demagnetization and eccentricity are unique and common faults in PMSM,therefore,it is of great engineering significance and practical value to diagnose these kinds of faults accurately and in real time and take effective preventive measures.Among the existing diagnosis methods,various non-invasive methods based on motor current,voltage,torque,vibration and other signals are simple and easy to implement,but are easily affected by motor operating conditions,resulting in low diagnosis accuracy and inaccurate fault location.Comparatively speaking,the invasive diagnosis method based on the internal magnetic field monitoring of the motor with search coils has higher diagnosis accuracy and stronger anti-interference ability,which has higher practicability and research value.Therefore,this paper investigates the rotor faults of PMSM,proposes an online monitoring method that can better realize rotor faults diagnosis with fewer search coils(called fewer search coils method),and carries out theoretical research,simulation analysis and experimental verification on the number of search coils and the problems caused by fewer search coils on the fault location accuracy and fault degree diagnosis accuracy.Firstly,the feasibility of fewer search coils for rotor faults is theoretically analyzed based on the alternating-current winding theory.In order to adapt to the spatial resolution of the fault magnetic field information obtained from fewer search coils,a strategy of setting the minimum number of search coils with winding pole-phase group as unit is proposed.Then,based on the analysis and derivation of the motor air-gap magnetic flux expression under various rotor faults,this paper proposes to extract the fundamental wave component of the magnetic flux containing the rotor fault information by short-time Fourier transform and quantify it into the fault indicator which is more conducive to improving the accuracy of the demagnetization degree under fewer search coils and non-stationary operation conditions.On the basis of fault indicator,the fault characterization quantity is proposed to reflect the rotor fault degree and position more objectively,and the corresponding rotor fault diagnosis process is given.In the rotor static eccentricity diagnosis process,the power angle reflecting the motor load degree is introduced to correct the influence of the different distortion degrees of the air gap magnetic field under different loads on location accuracy of the static eccentricity fault.In this paper,an 8-pole 24-slot surface permanent magnet synchronous motor was selected as a prototype,the finite element simulation model of the motor under normal and fault conditions and the rotor fault experimental platform with fewer search coils were built.Through simulation analysis and experiments under different operation conditions and different fault settings,the feasibility and effectiveness of the rotor fault diagnosis method with fewer search coils were verified.