Design of backstepping command-filtered adaptive sliding mode disturbance observer for finite-time tracking of wheelchair upper-limb exoskeleton robots

Abstract

In this paper, the main objective is to design a high-performance trajectory tracking control method for the wheelchair upper-limb exoskeleton robot system operating under significant uncertainty and external disturbances utilizing integrated command-filtered backstepping sliding mode control with an adaptive sliding mode disturbance observer. To this end, a general form of the dynamical model of the wheelchair upper-limb exoskeleton system under uncertainties and disturbances is first developed. The proposed approach overcomes chattering, computational complexity, or lack of robustness to uncertainties through adaptive compensation and command filtering techniques. Specifically, the command-filtered backstepping control reduces the computational burden associated with traditional backstepping, while the adaptive sliding mode disturbance observer effectively compensates for unknown dynamics and external perturbations in real time, enabling smooth and accurate control without chattering. Additionally, an asymmetric Lyapunov function is incorporated into the control design to ensure finite-time convergence of the tracking error and enhance system stability. The effectiveness of the proposed method is validated through both numerical and real-time simulations. Quantitative results from numerical simulations demonstrate that the proposed controller achieves less than 1 % steady-state tracking error and a 35 % improvement in convergence time compared to conventional sliding mode controllers. Moreover, real-time experiments using the Speedgoat real-time target system—interfaced with MATLAB/ Simulink for hardware-in-the-loop testing—confirm the practical implementation viability of the controller.