Research On Disturbance Rejection Method Of Precision Motor System With Permanent Magnet Linear Synchronous Machine

In order to meet the demanding requirements of the linear motor systems with highspeed,high-acceleration and high-precision for the industrial manufacturing equipment,they must have the capabilities of fast,precise and smooth thrust output.The current regulation with high dynamic response and high steady-state precision are the basic guarantee of thrust output quality,and fast and accurate force ripple suppression is the key to achieving smooth thrust output,which puts harsh requirements on the design of the control system.However,the effect of the high-bandwidth feedback control on the suppression performance for system parameter changes and various complex disturbances is limited,so a strong disturbance suppression strategy must be constructed at the same time.Aiming at the problem of disturbance suppression of the permanent magnet linear synchronous motor(PMLSM)system,this paper develops disturbance suppression methods from the perspective of electrical and mechanical subsystems with the consistent purpose of achieving high-quality thrust output.System modeling and model parameter identification are the basis for subsequent controller design.This paper first establishes a mathematical model of the electrical and mechanical subsystems when considering disturbances and then analysis the characteristics of the disturbances.The method of parameter identification of the dynamic model of the mechanical subsystem is studied and the parameter identification experiments are conducted.In the position control mode,the estimate and analysis the thrust ripple is done.The common characteristics of the disturbances of the two subsystems are studied,then the disturbance and its various derivatives are modeled as the extended states,and finally the extended state model of the entire motor system is established.The current controller design is very critical and the quality of current regulation determines the quality of thrust output and the stability of the outer loop and its achievable performance directly.The dynamic response of the deadbeat predictive current control(DPCC)is fast,but its parameter robustness is poor.To this end,under the extended state modeling of the disturbance and its derivatives,the design method of Kalman filter(KF)and general proportional integral observer(GPIO)as considering the current sampling noise are studied respectively in this paper.Then the estimation of the disturbance and the current at the next moment is realized,and the DPCC is improved by using the estimatedvalue.At the same time,the relationship between the performance of KF and GPIO with their parameters is analyzed,and parameter tuning strategies that are easy to implement are studied.Furthermore,a time-varying experiment of DPCC parameters is designed.The effectiveness of the above algorithm is verified experimentally and the performance is compared under single current closed loop.The results show that the above algorithms improve the DPCC parameter robustness effectively.Force ripple such as detent force of the linear motor and the cable force affect the smoothness of the thrust output directly,thereby reducing the position control performance.Observer-based force ripple estimation and compensation methods have strong trajectory adaptability,but their performance on fast time-varying disturbance estimation is limited.Therefore,this paper conducts in-depth research on this problem.Aiming at the shortcomings that the KF parameters are too mang and too difficult to be tuned as the extended state modeling is considered,the KF design method under incremental modeling is studied.In order to further reduce the difficulty of parameter-tuning and algorithm implementation,a GPIO design method considering position measurement noise is studied to achieve accurate estimation of the mechanical subsystem disturbances.And also,the relationship between the performance of KF and GPIO with their parameters is analyzed,and parameter tuning strategies that are easy to implement are studied.The disturbance estimation value under the above methods is fed forward,and a three-degree-of-freedom position control system is designed.The effectiveness of the above algorithm is experimentally verified and its performance is also compared at different speeds.The results show that the proposed observer-based methods can effectively improve the trajectory tracking accuracy.Although the above observer-based methods have strong trajectory adaptability,it is difficult to achieve high-precision disturbance estimation and compensation at high speeds.In order to solve the problem of high-precision position control under special occasions such as repetitive motion of PMLSM,this paper studies the application of a class of fractional-order iterative learning control(ILC)methods and the parameter tuning strategies in the frequency-domain so as to make indirect suppression of force ripple from the perspective of improving position control performance.First,the parameter tuning method in the frequency-domain of the P-plus-pure-phase-lead ILC is studied,and the design method of proportional gain is further analyzed.The basic principles and shortcomings of the fractional order D~?and PD~?-type ILC are analyzed and then thefractional order D~?and PD~?-type ILC with pure-phase-lead compensation are studied and its parameter tuning methods in the frequency-domain are given.Aiming at the shortcomings of the above ILC algorithms and the difficulty of parameter-tuning,a ILC with fractional-order phase-lead compensation is studied and its parameter tuning method in the frequency-domain is given.The proposed ILC algorithm has the advantage of high flexibility of phase compensation and high rationalization accuracy of fractionalorder operators.The effectiveness of the proposed methods and its parameter-tuning are verified experimentally at different trajectory speeds.The results show that the proposed ILC algorithms can effectively improve the trajectory tracking accuracy at high speed.The current inner-loop design with high dynamic response,high steady-state accuracy and strong parameter robustness and high-performance force ripple suppression under variable or repeated trajectories,and also high-precision motion control can be achieved with the disturbance suppression methods proposed in this paper.The research results are of great significance for the design of linear motor systems with high performance requirements.