Force-Position Control for Upper Limb Exoskeleton Based on Muscle Model with Compensation

This paper investigates a force-position controller using neural network model free control with time-delay estimation(TDE-MFNNC) for the motion of the upper limb exoskeleton. The system establishes the muscle force prediction model based on the scaling principle of Hill model to estimate the torque provided by the upper limb. Then the upper limb exoskeleton is built in Solidworks as a virtual prototype which is used to instead the real exoskeleton model. We performed control stability to verify the effectiveness of the controller by comparing with TDE-iPD controller and with compensation the simulation result has a more satisfactory tracking performance.