Design and Control of an Enhanced Self-Reconfigurable Parallel Ankle Rehabilitation Exoskeleton

This paper presents design, modeling, and experimental validation of a self-reconfigurable parallel ankle rehabilitation platform. The proposed system employs a three-limb parallel mechanism with mechanically reconfigurable joints, enabling seamless transition between static mobilization and dynamic gait-oriented operation while preserving alignment with the anatomical ankle joint center. A position-level inverse kinematics formulation is developed to directly compute actuator lengths from desired platform orientation, eliminating drift associated with velocity-integration approaches. The corresponding Jacobian matrix is derived to describe the velocity relationship and is evaluated over the rehabilitation workspace to ensure nonsingular operation. A closed-loop control framework based on Jacobian mapping and Lyapunov-guided design is implemented for stable motion tracking. Simulation results demonstrate accurate tracking of multi-axis trajectories with limited cross-axis coupling. Experimental validation on a physical prototype confirms the approach’s feasibility, demonstrating stable trajectory tracking and setpoint regulation under practical conditions, with minor deviations attributable to actuator and mechanical nonlinearities. The results highlight the potential of the proposed self-reconfigurable parallel architecture for anatomically consistent, adaptive ankle rehabilitation in both clinical and home-based settings.