Model Predictive Self-tuning Control For Permanent Magnet Synchronous Motor Based On Parameter Identification

During the operation of a permanent magnet synchronous motor(PMSM),it may cause the control system’s dynamic response capability to deteriorate and abnormal steady-state fluctuations,when the operating conditions of the motor change and cause problems such as changes in its own parameters and load torque fluctuations.Aiming at these problems that may exist in the process of permanent magnet synchronous motor control,this paper improves traditional model prediction algorithms and conducts self-tuning control related research starts to improve the dynamic response ability and steady-state anti-disturbance ability of the motor control system,First,the principle of the decoupling of the current realized by the coordinate change theory in the vector control is described in detail.Secondly,the derivation of SVPWM and the derivation of model predictive control are carried out.Then the four current control methods of the motor(i_d=0 etc.)are introduced.Finally,the response characteristics and dynamic and static capabilities of the vector control are analyzed through simulation to pave the way for the development of the project in this paper.Then,for the traditional finite set model predictive control(FCS-MPC),the discretization processing equation of the current prediction algorithm is analyzed,and it is found that the response is fast but the stability is not strong.The flux prediction algorithm has relatively good stability due to the influence of small inductance during the discretization process.In order to improve the steady-state anti-disturbance ability of the control system,this paper studies a deadbeat model predictive flux control method.This method combines incremental flux prediction and current prediction,and uses feedback switching vector to predict current,and then predicts the effect of incremental flux linkage to reduce control delay.In order to improve the accuracy of the mathematical model,the real-time motor parameter identification link is combined to make the motor’s speed response performance more superior,which is verified by simulation experiments.Next,optimize the anti-disturbance ability in motor control,starting from two aspects of load torque fluctuation and parameter change.By analyzing the structure block diagram of the improved model predictive control,the control block diagram could be decomposed into five small block diagrams:motor model,speed loop PI controller,current and voltage position signal sampling,model predictive control,and three-phase inverter.The transfer function of each part to get the transfer function of the entire control system is researched.In view of the fluctuation of load torque during motor operation,a frequency analysis method is designed.Making the motor still meet the expected amplitude-frequency and phase-frequency characteristics when the load torque fluctuates under the control of parameter auto-tuning,this ensured the stable operation of the motor.In view of the disturbance of motor parameters,an analysis method based on the typical type II system is designed.The system transfer function is calibrated into the form of a typical system.According to the dynamic anti-disturbance index of the system,the parameters of the PI controller are self-tuning so that the motor can still run stably under the condition of parameter disturbance.Finally,this paper designs a hardware experiment platform based on Infineon’s XE164FN control chip corresponding to the simulation analysis.Part of the research has been done on vector control and deadbeat flux linkage predictive control.Parameter identification and some self-tuning control experiments combined with identification theory are carried out.By comparing the results of simulation and experiment,the correctness of the control method proposed in this paper is verified.