Rfac Based Task-Oriented Active Sharing Control For A Class Of Robotic Rehabilitation Training Systems

This paper proposes a robust fuzzy adaptive critic controller (RFAC) for a class of robotic rehabilitation systems. Based on the RFAC control approach, the robotic rehabilitation system can utilize a robust fuzzy actor to generate an optimal task-oriented active sharing control behavior under the guidance of a fuzzy critic. Specifically, if the control behavior generated by the fuzzy actor is good for rehabilitation training, then the critic will give a reward, otherwise, the critic will punish the fuzzy actor to modify its control behavior. Finally, the RFAC can find the optimal control policy for the real robotic rehabilitation training through the reward-punishment mechanism. In addition, to attenuate the effects of the approximation error and various uncertainties in the system, a recently developed robust integral of the sign of the error feedback technique is also integrated into the developed RFAC agent. Lyapunov stability analysis shows the RFAC can yield a semi-global asymptotic result. Also, simulation experiments verify that the RFAC has good performance.