Walking Stability of a Humanoid Robot Based on Fictitious Zero-Moment Point

Abstract

A model of walking stability analysis of a humanoid robot is given, in which the different environments are considered. By applying the concept of fictitious zero-moment point (FZMP), a method to maintain the walking stability of humanoid robot under external disturbance is presented. The support polygon and the rotation edge in the case of losing balance in single and double foot support phase are determined in computerized form. A new control strategy is proposed to maintain stability in different environments. The measures to keep stable walking are the adjustment of the support polygon, the push or pull support of hand with environment and the movement modification of upper robot body. The relative position between the FZMP and the rotation edge represents the strength and direction of losing the stability. The optimization gait of humanoid robot was determined by using genetic algorithms. The feasibility of the proposed method is demonstrated by the simulation