Output Feedback Learning Control of Atomic Force Microscopes

In this study, an observer-based output feedback learning controller is proposed to eliminate the needs for accurate dynamic model information of the atomic force microscope and the measurements of cantilever tip velocity. Provided that the desired cantilever tip displacement is periodic with a known period, the proposed learning-based controller, together with a model-free observer formulation, ensures asymptotic tracking of the cantilever tip displacement despite the absence of velocity measurements and the presence of uncertainties in the system dynamics. Numerical simulations are provided to illustrate the feasibility and effectiveness of the designed observer--controller couple.