Velocity tracking algorithm in forward walking using estimated zmp by whole-body control framework

This paper presents a forward walking strategy for humanoid robots using a whole-body control framework. Implementation of walking requires the coordination of movements between the feet and the center of mass (COM) of the system. In addition, the condition of preventing falling needs to be considered at the planning stage and later the robot is controlled to meet this condition upon execution. In this paper, assuming that the planned velocity of the COM and the corresponding foot trajectories are given, generating the desired COM position and velocity by estimating the ZMP of the foot is proposed to account for the stability condition. This strategy provides a way to implement walking without the explicit planning of the COM in the forward direction. The proposed approach is implemented and verified in a simulation of a humanoid robot.