Demagnetization Fault Diagnosis And Fault-tolerant Control Of PMSM For Electrical Vehicle

With the limitation of mounting space,PMSM used for electrical vehicle has higher power density,but worse heat environment and more complex operation condition,which will result in a strong armature reaction,combining with the natural aging,the uniform demagnetization and local demagnetization fault usually appears in permanent magnet,furthermore these will cause the torque reduction and fluctuation of PMSM.Therefore,based on the in-depth and systematic researches of demagnetization fault diagnosis and fault-tolerant control of PMSM,the integrated solution of permanent magnet demagnetization fault of PMSM for electrical vehicle including demagnetization fault modeling,demagnetization fault diagnosis,fault mode recognition,fault degree evaluation and fault-tolerant control is presented,the research undoubtedly has significant theoretical and engineering application value for the safe and reliable operation of PMSM electrical drive system for electrical vehicle.First,the demagnetization mechanism of PMSM is intensively analyzed firstly in this dissertation,the mathematic model and simulation model of PMSM drive system considering permanent magnet demagnetization fault are proposed to realize qualitative and quantitative description of the electric characteristics of PMSM demagnetization fault.Then,the typical nonlinear filtering algorithms,such as kalman filter,extended kalman filter,particle filter and unscented particle are presented to implement permanent magnet flux linkage identification under the constraint of non-stationary operation condition of electrical vehicle and measurement noise,and their identification performances are compared comprehensively.In order to guarantee the uniqueness of permanent magnet flux identification results under the PMSM parameters variation,taking the extened kalman filter as an example,an unified full rank identification algorithm of PM flux linkage considering the PMSM parameters variation is studied and proposed on the basis of parameters sensitivity analysis.Based on the above researches,in order to conquer the key technology of eliminating the negative influences on permanent magnet flux linkage identification accuracy induced by PMSM parameters variation and the deficiency rank of identification model with less computation time,an algebraic identification method is studied and proposed to realize simultaneous,online and accurate identification of all of the PMSM electromagnetic parameters and the accurate diagnosis of permanent magnet uniform demagnetization fault.Furthermore,the weak fault characteristic signal extraction capability of HHT is studied,in view of the fact that this method has some technical shortcomings,such as weak fault characteristic signal around fundamental frequency is easily merged and the representation of local demagnetization fault is rather ambiguous because of the influnences from inverter harmonic wave,the method based on fractal dimension,integrated with the adaptive fundamental wave extraction and EMD is presented to eliminate the negative influences on box dimension value of fault characteristic signal induced by the fundamental wave of stator current,high frequency harmonic of PMSM drive system,and the low trend signal of EMD,to realize accurate diagnosis of PMSM local demagnetization fault.At last,the reliable recognition method of different fault modes,demagnetization fault degree evaluation method and fault-tolerant control strategy are also proposed based on the accurate diagnosis of permanent magnet uniform demagnetization and local demagnetization fault.On the basis of theoretical analysis and simulation researches,the hardware and software of PMSM drive system is designed and the experimental platform of PMSM drive system is developed with the dSPACE modular system PX10,the availability and feasibility of the proposed method including permanent magnet demagnetization fault diagnosis,fault mode recognition,fault degree evaluation method and fault-tolerant control strategy are verified by the experimental test.