CONTROL STRATEGY ABOUT HUMANOID ROBOT STABLE LOCOMOTION ON UNEVEN GROUND

SHUAI Mei,FU Chenglong,YANG Xiangdong,CHEN Ken
DOI: https://doi.org/10.3901/jme.2006.08.001
2006-01-01
Journal of Mechanical Engineering
Abstract:Since a humanoid robot tends to tip over easily while walking on uneven terrain, an online regulating control algorithm is proposed based on Kane collision theory, to prevent humanoid robot from tipping over while walking on uneven ground. This regulating algorithm can compute the dynamic model of the robot rapidly, and can estimate the values of the impacting force acting on leg by ground. These data of impacting force can be used to adjust the gait of legs and torso for a humanoid robot to pass over the scraggy obstacle successfully. By using this algorithm, walking simulations of online control on a 5-link humanoid model are performed in the MATLAB/Simulink environment and the effectiveness of the proposed algorithm is illustrated.
What problem does this paper attempt to address?