Adaptive Control of Series Elastic Actuator Based on RBF Neural Network

For the interaction and safety between the exoskeleton and the human body, the Series Eelastic Actuator (SEA) is used as the torque source of the exoskeleton. It is difficult to accurately construct the interaction model between the load end of the driver and the human body, in order to improve the accuracy of the SEA torque control, an adaptive control method of the RBF neural network is proposed. The RBF neural network is used to approximate the muscle model that is difficult to accurately model to achieve the SEA stable torque control. The simulation results show that under the condition that the estimated human-computer interaction model and the actual human-computer interaction model are quite different, the control scheme composed of RBF network can effectively cope with the uncertainty of the model and keep the controller itself good performance.