A Novel RBF Neural Network-Based Iterative Learning Control For Lower Limb Rehabilitation Robot With Strong Robustness

Abstract

In this paper, a novel RBF neural network-iterative learning control (RBFILC) methodology is proposed, investigated and analyzed for low limb rehabilitation robot (LLRR) to assist patients in passive training. The asymptotic stability of the RBFILC approach is proposed and verified via a classical Lyapunov theory, which simulates the behavior of healthy human body by 3-degree of freedoms (DOFs) sagittal plane LLRR system. Furthermore, the MATLAB software is utilized to conduct simulation experiments on LLRR model. The simulation results show that the RBFILC controller is rapidly convergent in a satisfactory way, and the RBFILC controller with strong robustness of the LLRR model is feasible and effective. Therefore, the patients can safely and effectively complete repetitive movements during rehabilitation treatment, and save a lot of medical expenses.