Application of neural network-based controller for the knee-joint position control with quadriceps stimulation

A neuro-control system has been designed to control the knee joint to move in accordance with the desired trajectory of movement with stimulation of quadriceps muscle. This control system consists of a neural controller in feedforward loop and a fixed parameter PID feedback controller. The neural network is trained as an inverse model of quadriceps-lower leg system for directly feedforward control. With the aid of neural controller, the knee joint angle is controlled within small region around the desired trajectory, then is fine controlled by the PID controller. Experimental results showed that the neuro-PID controller is promising for the position control of knee joint with quadriceps stimulation