PD Controller of a Lower Limb Exoskeleton Robot Based on Sliding Mode RBF Neural Network

Abstract

The lower limb exoskeleton robot (LLER) is a human-robot interaction device that combines human functions and mechanical characteristics. Due to the complexity and strong coupling of human gait, it is difficult for LLER to be worn comfortably and safely for training. In such a scenario, the paper proposes a kind of proportional-derivative(PD) controller of LLER based on sliding mode RBF neural network(SMRBF-nn). In order to verify the effectiveness of the proposed control scheme, pertinent experiments were carried out. The gait data of the subject was collected through the motion capture system. A simulating model was established, different control methods, like conventional SMRBF-nn controller and PD controller based on SMRBF-nn, have been tested on the LLER. The experimental results show that the control strategy proposed in this paper can not only make LLER track the human body’s gait trajectory, but also output appropriate torque when there is a disturbance.