Improving stabilization of passive walking using chaos

Abstract

Passive walking is defined as walking down a shallow slope without using any muscular contraction as an active controller. Based on this definition, which was made to present the simple possible models of human gait, some knee-less models have been proposed. These simple passive models have the ability to show several different dynamics such as periodic and chaotic behaviors. Previously, the periodic behavior has been studied and considered as the only stable response, but recent dynamical analysis on human locomotion characteristics such as stride length exhibit that this process has chaotic behavior. The purpose of this paper is to demonstrate that the knee-less passive-dynamic models has the ability to show chaotic behavior as a stable and acceptable answer using a chaotic function in heel strike condition. Represented chaotic model's parameters have been examined and compared concluding that the novel model is robust to trivial changes of slope or initial conditions.