Balance Control Strategy of Biped Walking Robot SUBO-1 Based on Force-Position Hybrid Control

Abstract

This paper deals with a balance control strategy based on a force-position hybrid control of biped walking robot, SUBO-1 for stable walking on irregular terrain environments. Walking robots basically require the position control of the leg to accurately land the swing foot on the correct position. Also, they need a force-based balance control to prevent them from falling over on irregular terrain. This study derives the location of the Zero Moment Point (ZMP) to control the Capture Point (CP) for keeping balance of the robot, and additionally obtains the total vertical force required for both feet to maintain the desired height of the Center of Mass (CoM). The vertical force, roll and pitch moments of each foot that optimally satisfies the derived ZMP and the total vertical force of the both feet, are calculated through the Quadratic Programming (QP). To realize the vertical force, roll and pitch moments of each foot, a feedforward control using Jacobian and a PI control using a force/torque sensor at the foot are used simultaneously, and the horizontal forces and yaw moment of each foot are used in Task-Space PD control for positioning of foot. Finally, the proposed method is verified through balance control experiments in double support and single support phases using the biped walking robot, SUBO-1 in the presence of external force and varying ground slope