A Human-Centered Prototype of a 3-DOF Upper Rehabilitation Exoskeleton with Multi-Objective Optimization

Exoskeleton robots have demonstrated significant potential in rehabilitation therapy. However, existing upper-limb exoskeleton designs struggle to balance moment of inertia (MOI), motor transmission efficiency, and range of motion (ROM). To address this, a human-centered 3-DOF upper-limb rehabilitation exoskeleton prototype was proposed, incorporating a optimization framework based on the Non-Dominated Sorting Genetic Algorithm II. With predetermined elbow flexion/extension motor placement, this framework identified optimal configurations for wrist flexion/extension and wrist pronation/supination (WPS) motors that achieved minimized MOI, maximized transmission efficiency, and enhanced ROM. Furthermore, a parallelogram-based remote-center-of-motion mechanism was implemented for the WPS motor to simultaneously preserve torque capacity and ROM. OpenSim-based simulations under a feeding paradigm revealed a trajectory relative mean error of 1.04% relative to the entire path length, demonstrating the feasibility of the prototype. The proposed methodology establishes a systematic framework for optimizing wearable robotic systems, facilitating the development of more efficient and adaptable solutions for rehabilitation applications.