Natural gait generation of biped robot based on captured human motion image

This paper proposes a new method of generating a natural and stable gait pattern for biped robots, imitating a human gait. For this purpose, human gait patterns are analyzed in terms of torque in the sagittal as well as ZMP in the frontal plane. These two gait parameters are used as the object function of the modified genetic algorithm which determines the joint angles of the 5-link biped robot. The modified genetic algorithm selects the individual as the solution which can generate a robotpsilas gait having the gait parameters most similar with the human ones. Consequently, the natural and stable walking pattern for the biped robot is obtained, as proved by the experiments.