Research On Fault Diagnosis Of Permanent Magnet Synchronous Motor

Permanent magnet synchronous motors(PMSMs)are widely used due to the characteristics of high-power density and high torque density.PMSM generally is the core of the transmission control system.The entire system would be failure if the motor were shut down accidentally,which may cause huge economic and property losses.Therefore,it is necessary to perform real-time fault diagnosis for PMSMs to ensure the safe and reliable operation of the motor.Inter-turn short-circuit faults and bearing faults are the two most common faults in PMSMs.This paper studies the diagnostic methods of these two faults.In this thesis,firstly,the fault characteristics of the inter-turn short-circuit fault and the bearing fault are studied.For the inter-turn short-circuit fault,this thesis firstly analyzes the cause of the inter-turn short-circuits and the equivalent circuit with the fault,and establish the mathematical model of PMSM with inter-turn short-circuit fault.Through the finite element simulation model of the inter-turn short-circuit fault permanent magnet synchronous motor,the amplitude,phase and characteristic frequency components of the motor current are analyzed;The effect of the _di=0 vector control strategy,speed and load change on the fault characteristics in voltage and current of the motor are analyzed by the MATLAB simulation model of the fault motor.A method of extracting the negative sequence component in voltage and current using coordinate transformation is proposed.According to MATLAB simulation results,the negative sequence voltage of the stator winding is proposed as the fault feature of the inter-turn short-circuit fault.For bearing fault,the vibration frequency caused by bearing failure is analyzed firstly.According to the increased complexity of the vibration signal after bearing fault arising,this thesis proposed that the sample entropy is used as the feature of bearing fault;according to the high-frequency resonance of the motor modulated by the feature frequency of bearing fault which caused by bearing fault,this thesis proposed that the signal energy distribution of vibration signal can be used as feature of bearing fault.Because directly using the sample entropy to process the vibration signal cannot effectively determine the type of fault,the vibration signal is decomposed by wavelet packet,and the sample entropy and relative wavelet packet energy of the reconstructed signal after the wavelet packet is combined as the fault feature.Secondly,according to the extracted fault features,this thesis proses two method for diagnosing inter-turn short-circuit:the first method based on negative sequence voltage;the second method based on back EMF estimated by sliding mode state observer.And a method for classifying bearing fault based on Support Vector machine.Thirdly,an experimental platform for fault diagnosis of PMSM that compatible with multiple load conditions is designed in order to collect the voltage,current,speed,vibration,torque and other data of the PMSM at different speeds,different load torques,different load inertia,etc.,which used to train support vector regression and support vector machine,and to validate the proposed fault diagnosis algorithm,.Finally,the simulation model of the experimental platform is established to verify the effectiveness of the inter-turn short-circuit diagnosis method.There are 76 figures,15 tables and 79 references in this thesis.