Computer Simulation of the Ontogeny of Bipedal Walking

Abstract

Computer simulations of human infant walking, changing from supported walking with weight support to independent bipedal walking, were conducted to identify the biomechanical factors of the neuro-muscular system necessary to acquire bipedal walking. We simulated the autonomous acquisition of independent walking by repeating walking trials and modifying neuronal structures and parameters so that walking efficiency is maximized, and forces supporting the upper body are minimized. The results of simulation showed that reinforced interaction with the system of body dynamics, such as feedback from somatic senses, especially information concerning angular velocities of joints, is essential to the development of walking. Other significant factors for bipedal walking are the development of neuronal feedback about the hip joint position, which determines the posture of the upper body, and improvements in the extensors' ability to produce joint moments.