Neural Networks Iterative Learning Impedance Control of Lower Limb Exoskeleton for the Later Stage of Rehabilitation

In the later stage of lower limb rehabilitation training, impedance control is used to realize human-machine force interaction. An expected impedance model-based iterative learning controller is proposed in this paper. Iterative learning control is used to realize the target impedance model, which not only meets the repeatability of rehabilitation training, but also ensures the transient performance of lower limb rehabilitation exoskeleton. In order to solve the model uncertainty, neural networks (NN) is used to estimate the unknown dynamics of the exoskeleton. Simulation results further verify the effectiveness of the proposed control scheme.