Periodic adaptive learning control of PMSM servo system with LuGre model-based friction compensation

Abstract

In this paper, based on LuGre friction model, a compensation method by combining of model reference adaptive control and periodic adaptive learning control (MRAC-PALC) is proposed to eliminate the adverse influence of nonlinear friction disturbance on tracking performance in permanent magnet synchronous motor (PMSM) servo systems under periodic task conditions. Specifically, the controller consists of a linear algorithm with PD, feedforward and velocity feedback components and a nonlinear friction compensator which is divided into two parts as follows: in the first time period, an MRAC algorithm is developed to ensure the boundedness of tracking errors; from the second time period, a PALC compensation algorithm, which learns from past information and updates the controller parameters in real time, is deployed to accurately capture the friction dynamics and guarantee the tracking performance. The stability of the proposed MRAC-PALC approach is guaranteed through Lyapunov stability theorem. Some comparative simulations and experiments are conducted to illustrate the superiority of the proposed MRAC-PALC strategy.