Research On Key Technologies Of Fault Diagnosis For Permanent Magnet Synchronous Motor In High-speed Railway

While China’s high-speed railway is actively promoting the construction of the "eight vertical and eight horizontal" main channels,with the upgrade of the power system and the increase of railway electrification,the safety of traction motor is more significant.The traction motor is the key equipment of the high-speed railway traction system,which determines the safe operation of the high-speed railway train group.However,as a complex electromechanical coupling system,a series of defects such as electrical faults may occur in the production and assembly process.In recent years,permanent magnet synchronous motor(PMSM)has been developing continuously in the field of high-speed railway traction.But the current high-speed railway traction motor is the asynchronous motor,the research on fault diagnosis technology of PMSM is relatively backward,and there is a lack of targeted fault diagnosis technology.To solve this problem,this study with 600 k W permanent magnet synchronous traction motor as the research object,with eccentric rotor fault diagnosis,permanent magnet demagnetization fault diagnosis,the stator inter-turn short circuit fault diagnosis as the research target,the new three kinds fault of mathematical model is proposed.According to the motor parameters,the simulation model based on the finite element method is constructed respectively obtained the health motors and the fault simulation of motor stator three-phase current signals.Besides,the ShortTime Fourier transform of the Kaiser window and the six-layer wavelet packet decomposition based on Coif4 wavelet are adopted to process the fault current signal,and the fault feature vector is extracted successfully.Then,a motor test platform is built and a 20 k W permanent magnet synchronous motor is used to conduct the rotor eccentric fault experiment,which verified the feasibility of the signal processing method.Furthermore,this study proposes a fault diagnosis method based on least square support vector machine for permanent magnet synchronous motor(PMSM)eccentricity.After adjusting broadband parameters and penalty factors,the fault diagnosis accuracy based on the Gaussian kernel function reaches 85.76%.