A Smooth and Efficient Gait Planning for Humanoids Based on Human ZMP

Based on the ZMP (zero moment point) trajectory and the walking data of human, a new method is proposed to improve the robot walking smoothness as well as to save energy. Firstly, a measurement system is designed to measure the data of humans including the ZMP trajectory and the waist trajectory. Secondly, a new gait planning method which includes presetting the allowable ZMP region is proposed through analyzing human data. Thirdly, the new planning method is applied to the multi-link model based gait planning method. Finally, the feasibility of the proposed method is verified by simulation and experiments.