Human-Robot Interaction Control Based on the RBF Neural Network for Upper Limb Rehabilitation Robot

Radical basis function (RBF) neural network is proposed for the upper limb rehabilitation robot driven by series elastic actuator (SEA). The human-in-charge (HIC) mode and the robot-in-charge (RIC) mode are performed via the devised interaction control algorithm, where in the different control modes can be switched smoothly to avoid human-robot confrontation and improve the rehabilitation effect. On the basis of the RBF neural network with nonlinear mapping property, the human-robot interaction control modes can be dynamically adjusted via utilizing human-robot interaction force, which realizes the human-robot collaboration and adaptive control. The simulation results prove that the advised control method has a excellent performance different control modes.