Dynamic Balance Control Algorithm of a Six-Legged Walking Robot, Little Crabster

This paper describes a center of pressure (CoP) control algorithm for maintaining dynamic balance of a six-legged walking robot, Little Crabster. Little Crabster is a ground walking test robot that has been manufactured as part of the process to develop an underwater six-legged walking robot named Crabster for underwater exploration and precise operations. Little Crabster is utilized to develop a walking algorithm with high walking stability in challenging environments. Although Little Crabster fundamentally has good walking stability with a large support polygon formed by its six legs, it is necessary to solve the body shaking problem, which lowers the visibility during walking, and enhance its waking stability under strong tidal currents and uneven terrains in underwater surroundings. Therefore, in this paper, sensory feedback controllers are designed to control the CoP, a criterion of dynamic balance, by modeling Little Crabster as an inverted pendulum model with a spring and a damper. In particular, a controller switching algorithm that leads to a smooth transition among the CoP controllers is proposed to cope with model parameter changes according to the variable number of supporting legs during walking. Finally, the performances of the CoP controllers and the controller switching algorithm are verified through several experiments using Little Crabster.