Friction Model Identification Based on Robust Estimation of the Linear Motor

Linear motor is one of the most important drive units for high-speed machining centers. Nonlinear friction is an important factor affecting the positioning accuracy of linear motor drive system. One of the most important topics for controlling linear motor is how to accurately estimate the parameters of the friction and effectively compensate it. In this paper, to solve the problem existing in traditional estimation methods that the estimation accuracy is not high enough, a novel estimation method based on robust estimation algorithm is proposed. The proposed method can effectively void the modeling error of the least squares algorithm, which will greatly improve the estimation accuracy of the nonlinear LuGre friction model. First, LuGre friction model is introduced in linear motor servo system; then, the robust estimator is design to identify the static and dynamic friction parameters in the LuGre model. Finally, simulation results verify the effectiveness of the proposed estimation scheme.