Torque Ripple Compensation For The Servo System Using Periodic Adaptive Learning Control

Permanent Magnet Synchronous Motor(PMSM)servo system is widely used in industrial fields due to its advantages of simple structure,low maintenance cost and fast speed response.With the advancement of technology,various application fields have put forward higher requirements for the performance of the servo system.However,torque ripple disturbance degrades the speed smoothness and position tracking performance.Therefore,it is of great theoretical and practical significance to use modern control algorithms to design a torque ripple disturbance compensation controller and improve the dynamic performance of the servo system.This paper considers a typical PMSM servo system with torque ripple disturbance and analyzes the influence of torque ripple on the performance.Combining Sliding Mode Control(SMC),Model Reference Adaptive Control(MRAC)and Periodic Adaptive Learning Control(PALC),a sliding mode periodic adaptive learning controller(SMC-MRAC-PALC)is proposed to effectively eliminate the influence of torque ripple on the performance of the servo system.The details are as follows:First of all,based on the characteristics of the PMSM,the mathematical model of the PMSM position servo system with torque ripple disturbance is established.By analyzing the mechanism of torque ripple disturbance,the dominant harmonic frequency is identified via Fast Fourier Transform(FFT),and the dominant harmonic model of torque ripple is established.Secondly,to eliminate the influence of torque ripple on the servo system,based on the torque ripple dominant harmonic model,a Model Reference Adaptive and Periodic Adaptive Learning Controller(MRAC-PALC)is designed,and the system stability is given through Lyapunov stability theorem.Considering the difference in the torque ripple of the servo system under high and low velocity operating conditions,some comparative simulations with MRAC-PALC,MRAC and PD controller under high and low velocity operating conditions are presented,respectively.The simulation results show that the proposed MRAC-PALC controller can effectively eliminate the torque ripple disturbance in the servo system and improve the position tracking performance.Thirdly,in order to improve the robustness of the closed-loop system,based on the above MRAC-PALC controller,a sliding mode periodic adaptive learning controller(SMC-MRAC-PALC)is designed.Using Particle Swarm Optimization(PSO)algorithm,the controller parameters are tuned offline.By comparing with the MRAC-PALC controller under high and low velocity operating conditions,respectively,it is verified that the proposed SMC-MRAC-PALC controller has a better anti-disturbance ability.Finally,with the PMSM position servo system experimental platform,some comparative experiments are performed under the SMC-MRAC-PALC controller,MRAC controller and PD controller.The experimental results further validate the effectiveness and superiority of the proposed SMC-MRAC-PALC.