Exploiting invariant structure for controlling multiple muscles in anthropomorphic legs: II. Experimental evidence for three equilibrium-point-based synergies during human pedaling

摘要

Developing a musculoskeletal robot with multiple muscles is important for not only establishing a novel robot that achieves coordination with a human but also in understanding the framework of motor control in the human body. Our pioneering work proposed a biologically inspired control framework for multiple redundant muscles; however, very few practical results have been reported on controlling a musculoskeletal robot. This study focused on confirming the usability of the proposed biologically inspired control framework, which is referred to as equilibrium-point (EP)-based synergyies and is expressed by the activation balance of agonist-antagonist muscle pairs. Electromyography data obtained from the pedaling task of five subjects were analyzed based on the concept of the EP-based synergies, and were then used to control the musculoskeletal lower limb robot. Three EP-based synergies obtained from all the five subjects indicated that the musculoskeletal robot achieved forward pedaling. This indicates the utility of the EP-based synergies as well as the modularity of motor control in humans.