High Precision Robust Control for Periodic Tasks of Linear Motor via B-Spline Wavelet Neural Network Observer

As one of the core components of high-end manufacturing equipment, the linear motor (LM) plays an important role in high efficiency and high quality production of the equipment. In this article, a high-precision robust control method with periodic dynamic compensation is proposed for the most common repetitive trajectory tracking task in the manufacturing industry. To overcome the influence of unknown periodic dynamics on the control accuracy, a B-spline wavelet neural network observer is designed to estimate the uncertainties. Moreover, the control framework in the form of robust control is employed to ensure the stability of the closed-loop system under other unknown aperiodic disturbance. Finally, the high-precision control for LM-driven stage in high-speed environment is realized, and the comparative experiments show that the control accuracy of the proposed method is significantly improved at least 28.9% compared with the other three existing control methods.