Predictive Antiwindup Compensation for Repetitive Control Supporting Nanoscanning

Ultrahigh precision scanning has emerged as a promising and challenging technology in nanoscale measurement and manipulation. Besides the requirement of dynamical tracking of periodical references, the existence of control saturation, due to the limited stroke of microactuators, poses additional challenges for the control of such systems. To address this problem, we propose a repetitive control structure with predictive antiwindup compensation to support nanoscale repetitive tracking/scanning with a minimized impact of actuator saturation. In particular, the proposed antiwindup compensator is activated by the time-lead control output ahead of saturation due to the time-delay block in the repetitive control structure. According to the input/output (I/O)-based equivalent representation and sector bound criterion, stability conditions for the antiwindup compensator are derived, which are further formulated as an H-infinity optimization problem with the robustness against model uncertainties, where the optimal compensator is developed. The effectiveness of the proposed tracking control architecture is further verified in real-time experiments on a piezoelectric-actuator-driven nanostage, where significant improvements are demonstrated comparing with the existing results.