ONLINE WALKING PATTERN GENERATION AND CONTROL FOR BIPED ROBOT WITH CONTACT CONSISTENCY

Wenqi Hou, Jian Wang, Jianwen Wang, and Hongxu Ma

Keywords

Walking pattern generation, biped robot, point foot, contact consistency, orbit energy, task space controller

Abstract

In this paper, a novel online biped walking pattern generation strategy with contact consistency is proposed. Generally, to realize biped walking without foot–ground slipping there are three prime questions need to be answered: When a step should be taken? Where the foot should be placed? How to take the desired step? By utilizing the linear inverted pendulum model (LIPM), a striding controller with contact consistency is designed to track desired orbit energy, which answers the former two questions. To answer the third question, the hip’s horizontal locomotion is selected as the parameter to generate the desired trajectories in task space, and a task space controller without calculating the inversion of inertial matrix is presented. Simulation experiments are implemented on a virtual 5-link point foot biped robot. The results validate the walking pattern generation and control strategy which can realize periodic gait cycles with no slipping or falling and withstand a push of up to 10 Ns.

Important Links:



Go Back