Self-adaptive Observer Sensorless Control Of Permanent Magnet Linear Synchronous Motor Based On Auto-disturbance Rejection

With the rapid development of urban rail transit,the development of urban rail locomotives is also diverse,the traditional rotary motor as a traction device,but its drawbacks are increasingly reflected,and linear motor because of the advantages of load,energy saving and service life,etc.,by the development of urban rail locomotives.In this thesis,permanent magnet linear synchronous motor(PMLSM)is the research object,and sensorless control system is the research direction.Firstly,the principle structure of the linear motor is analyzed and modeled with reference to the permanent magnet synchronous motor,and the mathematical models of ABC coordinate system,DQ coordinate system and coordinate system are established respectively,and then the control simulation of the motor is simulated by field-oriented control,and the mathematical model of the motor is verified.On this basis,the structure of Model Reference Adaptive System(MRAS)sensorless observation control system is carried out on the motor,and the current and flux are used as state variables respectively,and the observation system is mathematically modeled and simulated,which verifies the effectiveness of the MRAS observation system for motor control.Secondly,aiming at the problems of large tracking error and poor system stability of traditional MRAS adaptive law control,the MRAS observation and control system of sliding mode control(SMC)adaptive law is proposed,which is insensitive to nonlinear systems and external disturbances,and can quickly switch the system by using the switching function,and the simulation results show that its tracking and stability are stronger than that of the traditional adaptive law.However,the discontinuity of the switching function and the existence of dead zone make the system have the problem of jitter,so in view of this problem,an MRAS observation and control system based on the adaptation law of artificial neural networks(ANN)is proposed,and the weight of the error signal is corrected in real time by the gradient descent method by the reverse propagation of the error signal,and the jitter reduction amplitude,control accuracy and tracking effect are improved through the comparative analysis of simulation results.Finally,aiming at the weak immunity of the traditional PI velocity loop ANNMRAS observation control system,an ANN-MRAS observation control system based on Active Disturbance Rejection Control(ADRC)is proposed.The tracking differential in ADRC is used to transition and track the signal in real time,then the feedback error is nonlinear combined by the nonlinear feedback control rate,and finally the total disturbance of the system is observed by the extended state observer.Through the comprehensive comparison of simulation results,it can be seen that the control accuracy,stability and immunity of the ADRC-ANN-MRAS observation and control system are all good.