Research On High Precision Multi-step Model Predictive Position Control Method For Planar Switched Reluctance Motor

Planar Switched Reluctance Motors(PSRMs)have great application prospects in the fields of high precision two-dimensional driving devices including micro-machinery manufacturing,optical manufacturing,ultra-precision machining and so on,because of their simple structure,easy manufacturing,easy installation,low heat consumption,low cost and high reliability.However,the double salient structure,highly saturated magnetic circuit and switching mode of the PSRMs seriously restrict its high precision position control,thus affecting its motion accuracy.Model Predictive Control(MPC)theory has attracted extensive attention and research in the field of motor because of its excellent static and dynamic control performance and anti-jamming performance.In order to improve the motion accuracy of the PSRM,the high precision multi-step model predictive position control method of the PSRM is investigated in this paper.The mechanical structure and basic principle of the PSRM are described.The control system of the PSRM is established,and the experimental platform of the PSRM is built.Built on the mechanical motion equation of the PSRM,the dynamic model of the PSRM is constructed by Euler discrete method.According to the established dynamic model of the PSRM and model predictive control theory,a multi-step model predictive position controller of the PSRM based on a dynamic model is designed.Aiming at the problem of predictive deviation of motor prediction model based on the dynamic model,the traditional cost function of model predictive controller is modified,and the performance of multi-step model predictive position controller before and after the improvement is compared through simulation.The PD position control algorithm based on parameter optimization is designed.The simulation results are compared with the multi-step model predictive position control algorithm based on a dynamic model.The trajectory tracking motion control experiment of the PSRM is done on the experimental platform.A parameter regression model of the PSRM is proposed and established.In the actual operation of the motor,the parameters of the regression model are indicated with the recursive least square method with forgetting factor.A multi-step model predictive position controller for the PSRM based on parameter regression model is developed.The simulation study is conducted and the experimental study of trajectory tracking motion control is carried out based on the experimental platform.Aiming at the problem that the starting thrust of the PSRM may exceed the maximum allowable thrust when starting,the traditional method limits the maximum thrust output by adding a limiting module at the output end of the position controller.This method cannot guarantee the stable operation of the motor.A constraint characteristic method based on threshold function is therefore proposed to saturate the thrust of the motor,a multi-step model predictive position controller for constrained optimization of the PSRM based on threshold function is designed.The simulation study and experimental verification are performed.The results show that the proposed position controller based on model predictive control theory has better dynamic and steady-state performance than the PD position control method based on parameter optimization.The proposed parameter regression model based multi-step model predictive control method and constrained optimization multi-step model predictive position control method for the PSRM achieve high precision trajectory tracking motion control of the motor.The validity of the proposed multi-step model predictive position control method for the PSRM is verified.