Robust Adaptive Control of Manipulators

This paper develops an adaptive controller for robot manipulators. The design decouples the system’s adaptation and control loops to allow for fast estimation rates, while guaranteeing robustness. The control scheme is tested in different manipulation scenarios, namely, (i) trajectory tracking where the desired joint motions are predefined and (ii) command following where the desired motions are not known a priori and instead inferred using measurements and inverse kinematics. We consider, in addition, an operating modality in which the control scheme switches between command following and static positioning, for which the predefined desired joint motions are constant in time and identical to the manipulator angles at the time of switching. The simulation results illustrate the performance of the proposed control algorithm and its ability to deal with unmodeled dynamics, measurement noise, and time delay, while maintaining smooth control signals.