A sliding walk method for humanoid robots using ZMP feedback control

In this paper, we propose two methods for a highly stable sliding walk for humanoid robots working in narrow areas under constrained postures. The first method uses a new displacement of the center of gravity with Zero Moment Point (ZMP) feedback control. ZMP is a force distribution on the soles of the feet. This method is faster and more stable than previous techniques. The second method utilizes sliding forward movements using the inverse kinematics of the leg mechanism, which are calculated from the positions variation of the sole. The sliding walk is experimentally demonstrated using the humanoid robot HOAP-2, and an evaluation of the lateral variations in ZMP and the lateral acceleration variations are shown.