An Optimized Algorithm Based on Energy Efficiency for Gait Planning of Humanoid Robots

摘要

An energy-efficient gait planning method is proposed by imitating natural human walking in this paper. This method uses a 3D Linear Inverted Pendulum Model (LIPM) to improve the walking efficiency for the biped robot, which utilizes the allowable Zero Moment Point (ZMP) regions and slight adjustment of the Center of Mass (COM). The novel gait parameters optimization (NGPO) algorithm is proposed to obtain effective gait parameters that can synthesize optimal gait for minimizing energy consumed by the actuators of humanoid robot joints. Moreover, physical constraints, geometrical constraints, friction force limit and yaw moment limit need to be taken into consideration, ensuring feasible walking trajectory. The validity of the algorithm is verified by simulation, and the gait planning is implemented on XT humanoid robot.