Adaptive Strategies for Quadruped Robot to Climb High-slope Terrain without Priori Information

摘要

Dynamically traversing high-slope terrain is a challenging task for quadrupedal locomotion, especially in situations without priori information about the terrain it transverses. Different from locomotion on flat ground, a robot on high-slope terrain has to perceive the inclination of the terrain with proprioceptive sensors, and take appropriate strategies to adapt itself to the slope. In this paper, an adaptive method for the quadruped robot to negotiate high-slope terrain is proposed. A least square optimization-based terrain estimation algorithm is developed to estimate the slope angle of the terrain. The torso pitch, foothold locations, and the leg swing trajectory are adjusted according to the estimated slope angle, and a model-based whole-body controller is designed to control for the robot with dynamic trot gait. Experiments demonstrated the effectiveness of the proposed method on a quadrupedal robot to climb up a 41° slope and down a 32° slope.