Passive walking towards running

Passive walking emerges autonomously on a slight slope without an external input of energy. It is known that the walking motion on a steep slope evolves into a chaotic motion. In this paper a biped model for walking and running is presented, and a strategy is proposed to expand the range of stable passive walking by using a chaos-control technique based on the Ott-Grebogi-Yorke method. The resultant controller is a discrete type so that the input value changes at every step, and the generated walking motion is kept non-chaotic. Fast walking on a steep slope is achieved, and pseudorunning has also been realized in simulations. By adding an input to the biped model, in which the input corresponds to the effect of the artificial gravity field, it has been verified that pseudorunning can be realized on level ground.