An elbow joint movement control model with visual feedback.

Abstract

A motor program generator control model is proposed to simulate neuromuscular control. Three muscles (Biceps, Triceps, Brachialis) driving elbow joint flexion in a plane are simulated by integrating their nonlinear dynamic property and spinal neural circuitry. The motor descending commands are described by a visual feedback signal from the joint and an excitation signal for the motor neuron pool. The visual feedback signal mimics the gamma command whereas the excitation signal mimics another descending co-activation command. The gamma command is expressed as the output of a PID controller with the visual feedback error signal as the input. The gamma command and the motoneuron pool background activity are the inputs to the motoneuron pool model coupled with the Renshaw cell recurrent inhibitions. The output of the motoneuron pool model mimics the alpha command feeding directly to the muscle dynamics. A movement is produced by reducing the error signal between goal position and actual position and altering excitation signal properly. The simulation results show that a burst pattern of excitation signal and a PID controller can accurately trace the terminal goal and generate a smooth movement with a bell shaped velocity profile. The muscle activation signals have the characteristic similar to the smoothed EMG. Changing different parameters of the PID can cause the same effects as the stimulus pulse intensity or duration modulation.