Generation of humanoid walking pattern based on human walking measurement

Abstract

Walking pattern generation has been a crucial issue for humanoid robot. To generate more natural walking patterns, and determine parameters more reasonably, this paper presents a method for humanoid walking pattern generation based on human walking characteristics. The characteristics of human walking parameters were obtained by capturing and analyzing human walking data. Walking pattern parameters are determined by using these characteristics, and then stable and harmonious walking patterns are generated. The effectiveness of the method was verified by experiment result on humanoid robot BHR-2.