Determining an optimal multiarticular muscle arrangement of a musculoskeletal robot for a specific motion using human motion data

Abstract

In this paper, we propose a method to determine an optimal multiarticular muscle arrangement for a specific robot motion. Multiarticular muscles contribute to the realization of musculoskeletal robots with dynamic motion performance similar to humans and animals. However, the dynamic motion performance is affected by arrangement of the multiarticular muscles. Therefore, the arrangement is very important for the robots. In previous researches, muscle arrangements of robots were based on ones of humans and animals. However, we suppose that a more optimal muscle arrangement than any other one of humans and animals can be found by limiting robot motions. Therefore, in this research, we focus on a specific robot motion, and searched for an optimal multiarticular muscle arrangement for the specific robot motion. Specifically, we estimated force properties in the specific robot motion. The force properties were estimated by using human motion data corresponding to the specific motion. Then, on the basis of the evaluation of the force properties, we searched for the optimal muscle arrangement from among 736,281 of possible arrangements. Furthermore, we demonstrate the effectiveness of our method using simulation experiments. As a result, the search results indicate that there is a more optimal muscle arrangement than any other one of humans and animals on condition that we focus on the specific motion.