Optimal Control of a Modular Robot Manipulator for Human-Robot Collaboration Based on Nonzero-Sum Differential Game

Abstract

This paper proposes a nonzero-sum differential game-based optimal control method for Modular Robot Manipulator (MRM) in Human-Robot Collaboration (HRC) tasks. The MRM dynamics are modeled using Joint Torque Feedback (JTF) techniques. A nonzero-sum differential game framework is established among subsystems within the MRM to determine optimal control strategies. Adaptive Dynamic Programming (ADP) based on fuzzy logic approximates solutions to the associated Hamilton-Jacobi equations. Lyapunov analysis proves that the tracking error of the MRM system remains Ultimately Uniformly Bounded (UUB). Experimental results validate the effectiveness and advantages of the proposed approach, demonstrating superior performance in trajectory tracking and energy efficiency.