Terrain estimation and strategy for quadruped robot body posture and COG adjustment

Quadruped robot has gained its popularity due to the adaptation ability of rough terrain. To improve the walking stability and balance of quadrupeds, separate static walking methods are discussed based on specific known terrains. For some unknown terrains, a uniform terrain model is established in this paper. A fuzzy logic based terrain estimation method is then proposed to estimate the terrain parameters. Based on the criterion of maximum working space, a body posture and center of gravity (COG) position adjusting strategy is proposed, and a method to adjust the posture and COG position is derived based on the kinematic analysis. Simulations show that the proposed methods can deal with different terrains and body posture and COG position can be tuned according to the terrain.