Novel Predefined‐Time Observer‐Based Sliding Mode Controller for Lower‐Limb Exoskeletons: Design and Real‐Time Validation

Abstract

Control systems play a critical role in lower‐limb exoskeletons, directly influencing user safety, comfort, and adaptability to varying physiological conditions. However, achieving fast, robust, and predictable tracking performance remains a significant challenge due to model uncertainties, external disturbances, and human–robot interaction dynamics. In particular, conventional asymptotic, finite‐time, and fixed‐time control strategies do not provide explicit guarantees on convergence time, which is essential for ensuring reliable and user‐friendly assistance in rehabilitation and assistive applications. To address these limitations, this paper proposes a novel Predefined‐time Observer‐based Sliding Mode Control (POSMC) scheme for lower‐limb exoskeleton systems. The dynamics of the human limb and the exoskeleton are modeled and integrated to capture coupled behavior during operation, while interaction forces are explicitly incorporated to reflect realistic usage conditions. The proposed framework combines a state observer and a disturbance observer with predefined‐time stability theory to ensure robust and accurate trajectory tracking within a user‐specified convergence time. The effectiveness of the proposed method is validated through real‐time simulation on a Real‐time Digital Simulator (RTDS) platform under two practical scenarios. Comparative results with finite‐time and fixed‐time sliding mode controllers demonstrate that the proposed POSMC approach achieves faster convergence, improved tracking accuracy, and enhanced robustness against disturbances and modeling uncertainties. These results highlight the potential of the proposed control strategy to improve safety and performance in both rehabilitation and assistive exoskeleton applications.