Upper limb rehabilitation robots can substitute traditional practitioners in training programs, effectively alleviating the shortage of rehabilitation professionals in the market. However, human-robot interactions during rehabilitation tasks often introduce disturbances, degrading the performance of these robots. This paper proposes a linear active disturbance rejection control (LADRC-SMC) strategy, based on sliding mode control, to improve trajectory tracking performance of a desktop upper limb rehabilitation robot (DULRR).
Based on the basic mechanical structure and drive characteristics of DULRR, the dynamic model of the system is established. A linear expanded state observer (LESO) estimates the total perturbation within the LADRC framework, and a sliding mode control (SMC) based control law is designed to eliminate observer bandwidth limitations, thereby enhancing dynamic system performance.
Three controlled experiments are conducted to demonstrate the algorithm’s performance. Experimental results align with simulations, showing that the proposed LADRC-SMC overcomes bandwidth limitations, offering better disturbance rejection and tracking performance than PID and LADRC.
The LADRC-SMC method is proposed to enhance the performance of LADRC by integrating SMC. This method improves the disturbance rejection capability of DULRR in rehabilitation tasks, laying the foundation for its future applications.
