Use of a Central Pattern Generator for Control of a Muscle-Actuated Simulation of Pedaling

Muscle actuated forward dynamic simulations have provided tremendous insights into the mechanics of locomotion. However, the controllers used for large scale simulations have often been open-loop, with the muscle excitations prescribed as a function of time [1]. Due to the inherently unstable nature of bipedal movement, this means that perturbation-type analyses are often limited to short time frames after the perturbation is introduced [2]. However for many clinical problems, it would be desirable to predict how periodic locomotion reestablishes following a change to the system or perturbation from the environment.