Neuromuscular-Like Control for an Artificial Finger with SMA Actuators

Abstract

In comparison with robot manipulators, primate limbs excel robots in facile movements requiring compliance control. Based on this fact, a neuromuscular-like model that can emulate different involuntary and voluntary movements within constraints from neurophysiology is proposed in this paper. The neural controller is an intelligent system applied for multi-joint opponent muscle control of a robot finger. Artificial muscles are electric pistons with shape memory alloy (SMA) springs. The neural model proposes functional roles for pre-central cortical cell types in the computation of a descending command to spinal alpha and gamma motor neurons, and allows controlling desired joint movement trajectories by smoothly interpolating between initial and final muscle length commands for the antagonist muscles involved in the movement. Through experimental results, we showed that the neural controller exhibits key kinematic properties of human joint movements, dynamics compensation and including asymmetric bell-shaped velocity profiles. The neural controller allows varying the balance of static and dynamic feedback information to guide the joints’ movement command and to compensate external forces.