Abstract:Proceedings of the Institution of Mechanical Engineers, Part I: Journal of Systems and Control Engineering, Ahead of Print. This article focus on the problems of trajectory tracking and motion constraint for physical human–robot interaction, and a compliant adaptive control method is proposed for stable and safe physical human–robot interaction during the interaction. First, a fuzzy variable impedance control strategy is given to make the robot to use suitable impedance parameters in different motion states, which can improve positioning accuracy and reduce the interaction force. Second, by transforming the safety interaction constraint into output constraint, a tan-type barrier Lyapunov function is presented to guarantee the safety of human partner in physical human–robot interaction. Third, an adaptive neural network is employed to design the adaptive controller to handle with the dynamic uncertainties and improve the robustness of the system. Finally, simulation results of a 2-degree-of-freedom manipulator are presented to show the effectiveness of the proposed method.

Fuzzy variable impedance-based adaptive neural network control in physical human–robot interaction