Sampled-data repetitive learning control of electro-hydraulic actuator used in lower limb exoskeleton for rehabilitation based on Euler approximation

In this paper, sampled-data repetitive learning control of electro-hydraulic actuator used in lower limb exoskeleton for rehabilitation is addressed. The controller is proposed entirely in the discrete-time domain. First, the discrete-time model of the electro-hydraulic actuator is derived by direct discretization using Euler approximation, where the equivalent mass and force acted on the electro-hydraulic actuator are considered as periodic due to the repetitive motion of the exoskeleton leg. Second, a repetitive learning controller is developed to learn the periodic ones by utilizing the backstepping technique. The boundedness and tracking convergence of the closed-loop system are obtained via rigorous analysis. Finally, the control performance of the proposed controller is verified by simulation.