Equilibrium-point-based synergies that encode coordinates in task space: A practical method for translating functional synergies from human to musculoskeletal robot arm

Abstract

A practical method for exploiting the neuromuscular coordination in a human arm posture to control a musculoskeletal robotic arm is proposed. The method enables the low-dimensional control of multiple muscles in the arm robot without considering the differences in the size and force balance between the human and artificial muscles. The central idea is based on the assumption that the equilibrium point (EP) of an arm is always located at its endpoint position while maintaining the posture on the horizontal plane. The mathematical formulation obtained using this physical constraint leads to the EP-based synergies that encode the polar coordinates in the task space. We tested this theory by extracting EP-based synergies from electromyography activities during human posture maintenance and implementing them for a musculoskeletal robotic arm. The results suggest EP-based synergies could be functional modules of muscle mechanical impedance which is an essential primitive for human motor control.