Adaptive controller design for upper-limb rehabilitation robot

This paper studies the problem of robust output feedback control for a class of upper-limb rehabilitation robot, which can assist with shoulder, elbow and wrist joint movements. The proposed controller can guarantee the robustness of the robotic system in the presence of model uncertainties, as well as the ultimate boundedness of the presented adaptive observer-based controller. An obvious advantage of the proposed method is that the implemented information for feedback is only the measurement of position. Without additional sensors, the information of velocity and acceleration is provided by an adaptive observer. By utilizing a novel adaptive logic-based switching control scheme, a trajectory tracking performance is well achieved with a tunable error bound. Finally, a simulation example verif es the effectiveness of the proposed control scheme.