Research On Demagnetization Fault Diagnosis Of PMSM Based On Magnetic Flux Density Component

In recent years,permanent magnet synchronous motor(PMSM)is widely used in various industries because of its high efficiency,low temperature rise and high power factor.At the same time,its safety and reliability has also become the primary consideration for people to choose the motor.Demagnetization fault is one of the faults of PMSM,which affects the operation efficiency of the motor and cause damage to the motor,resulting in unacceptable consequences.If the demagnetization fault can be diagnosed and repaired in time at the initial stage of PMSM operation,the demagnetization degree will be prevented from further expanding,so as to reduce the maintenance cost and ensure personal safety.Therefore,it is of great significance to diagnose demagnetization fault of PMSM.Among the various diagnosis methods of PMSM demagnetization fault,the air gap flux density is one of the most intuitive parameters to judge the demagnetization fault.Therefore,this paper proposes a demagnetization fault diagnosis method of PMSM based on magnetic flux density component.The main research contents are as follows:(1)Firstly,the magnetic flux density component of PMSM under no-load and load are calculated and analyzed,and the demagnetization fault diagnosis scheme based on magnetic flux density component is proposed.Secondly,the Hall effect method and the Hall element are selected as the magnetic field detection method and device of the diagnosis scheme.Then,the theory of EKF is described and the filtering effect is analyzed,and the EKF is selected as the filtering method.Finally,the RBF neural network is proposed to predict the demagnetization fault degree;(2)The finite element simulation model of ANSYS Maxwell is built.And the magnetic flux density component of normal motor,local demagnetization fault motor with 30%,50%,70%,100% demagnetization of single pole,with 50%demagnetization of two poles,with 50% demagnetization of three poles,and uniform demagnetization fault motor with 30%,50% demagnetization of all poles are simulated and analyzed.Then the prediction model of demagnetization fault based on RBF neural network is built to predict the demagnetization fault degree,and the feasibility of demagnetization fault diagnosis scheme is verified;(3)The software and hardware of the whole demagnetization fault diagnosis system are designed and implemented.The signal acquisition module and DSP drive module are designed in hardware;In the aspect of software,the signal acquisition,transmission and filtering are realized;(4)The demagnetization fault diagnosis platform is built,and the experimental signal acquisition under different demagnetization faults is realized on the platform.Then the prediction model of the trained demagnetization fault is used to predict the degree of demagnetization fault.Finally,the experimental results are compared with the simulation results.The simulation and experimental results show that once the demagnetization pole passes through the Hall element under the demagnetization fault of PMSM,the Hall voltage detected by the Hall element will change,that is,the magnetic flux density component will change.Then the demagnetization fault prediction model is used to predict the changed magnetic flux density component.The prediction result is basically consistent with the real data,and the prediction result curve has a certain trend,and is kept synchronized with the change of curve of the real data,achieving the expected effect.In addition,the changing trend under other demagnetization faults can be analogized through the changes of the magnetic flux density component under the above several demagnetization faults.