Output Feedback Control of a Robotic Exoskeleton with Input Deadzone via Neural Networks

In this paper, adaptive output feedback control via neural networks is designed for a robotic exoskeleton with unknown dynamics. Neural networks are used to compensate for the unknown deadzone effect induced by the actuators and the unknown dynamics of the robot. High-gain observer is employed to estimate the velocity information and then integrated in the design of output feedback controller. The deadzone effect is approximated by a Radial Basis Function Neural Network (RBFNN) and the tracking error for the deadzone effect is bounded and converging. The unknown dynamics of the robotic exoskeleton are estimated with another RBFNN. The proposed control is able to compensate for the estimated deadzone effect and track the desired trajectory. Finally, numerical simulation and experiment on a two-joint rigid exoskeleton demonstrate the effectiveness of the proposed method.