Adaptive Extended State Observer-Based Velocity-Free Servo Tracking Control with Friction Compensation

Weiyang Lin,Zhongjin Zhang,Xinghu Yu,Jianbin Qiu,Imre J. Rudas,Huijun Gao,Dongsheng Qu

DOI: https://doi.org/10.1109/tsmc.2023.3299953

2024-01-01

Abstract:In this article, a velocity-free adaptive controller is proposed for the tracking control of servo mechanisms with friction compensation. A continuously differentiable friction model is employed to compensate for the dominant friction nonlinearity of servo mechanisms. Besides, a projection-type adaptive law is applied to handle parameter uncertainties in the system model. Since only the output position signal is directly measurable, an adaptive extended state observer (AESO) is constructed to estimate the indeterminate velocity state, which can also provide an estimation of unmodeled dynamics. Moreover, the dynamic gain switching of AESO can effectively suppress the peaking phenomenon at the motion beginning. Specifically, the parameter adaptation law utilizes the desired velocity state instead of the estimated value, avoiding the coupling problem between parameter and state estimation. The proposed control strategy theoretically demonstrates the transient performance and boundedness of the error in output tracking. Asymptotic stabilization of the system can also be implemented when only parameter uncertainty exists. Comparative experiments are conducted on a linear motor platform to demonstrate the effectiveness of the proposed control scheme.