کنفرانس مهندسی برق ایران

صفحه اصلی / سی و چهارمین کنفرانس بین المللی مهندسی برق

Constrained Adaptive Control of a Lower-Limb Rehabilitation Exoskeleton using Barrier Lyapunov Functions to Ensure Patient Safety

نویسندگان :

Mohammad Olyai¹ Ashkan Jelodar² Kevin Babakhanloo³ Khalil Alipour⁴ Bahram Tarvirdizadeh⁵ Mohammad Ghamari⁶

1- University of Tehran 2- University of Tehran 3- University of Tehran 4- University of Tehran 5- University of Tehran 6- Department of Electrical Eng, California Polytechnic State University

کلمات کلیدی :

Rehabilitation Robotics،Barrier Lyapunov Function،Adaptive Control،Physical Human-Robot Interaction

چکیده :

Lower-limb rehabilitation exoskeletons operate in close physical human--robot interaction, where tracking performance alone is insufficient: the controller must also enforce hard safety limits so that joint motion remains within the wearer’s physiological range of motion, including during involuntary events such as spastic contractions. This paper presents a computationally light safety-critical control architecture for a 2-DOF lower-limb exoskeleton with uncertain coupled human--robot dynamics and actuator torque limits. The proposed method integrates (i) an adaptive model-compensation term with a projection operator to keep inertial parameter estimates physically meaningful, (ii) a robust continuous high-gain term to reject bounded disturbances, and (iii) a symmetric logarithmic Barrier Lyapunov Function (BLF) that renders the tracking error set invariant, thereby preventing joint-limit violations provided the reference trajectory respects a safety margin. Lyapunov analysis establishes boundedness of all closed-loop signals and constraint satisfaction for all time under standard robot structural properties and bounded disturbances. Comparative simulations under severe spasm-like torques show that baseline PID and unconstrained adaptive control can transiently violate the safety envelope, while the proposed controller maintains error trajectories strictly inside the prescribed bounds with feasible torque demands. These results support the use of BLF-based constrained adaptation as a principled safety layer for rehabilitation exoskeleton control loops.