Trajectory Generation using Rimless Wheel Model for Walking of Biped Robots

Abstract

The purpose of this paper is generate a natural walking trajectory of a human for biped robots using a rimless wheel model (RWM). To apply the RWM, the central axle of the wheel is matched to the pelvic joint and the spokes to the legs. The leg trajectory is created to mimic the motion of the spokes as they make contacts with the ground while imaginary rimless wheel rotates. A motion of a leg is divided into five-phases (heel contact, toe contact, mid stance, heel off and toe off) similar to human walking. The interval of each phase is adjusted in order to Zero Moment Point (ZMP) is remains in the stable area. Computer simulations shows that the proposed method walks well even with a modest change in the speed.