Modeling of Rolling Locomotion on a Reconfigurable Quadruped Robot with Servos for State Estimation

Abstract

Motivated by the need for accurately monitoring a condition of a reconfigurable robot in autonomy, in this paper, we present a model of rolling locomotion on a reconfigurable quadruped robot, Scorpio, for state estimation. The quadruped robot can crawl on uneven grounds ensuring high degree of stability and mobility and roll on the flat ground and downhills offering greater speed and efficiency. For the state estimation of the rolling quadruped robot, its model combines servo characteristics of leg joints and considers collision between the robot and the ground and following discontinuous velocity variation. We verify the adequacy of the robot model through numerical simulations. The simulations results show that the rolling quadruped robot can roll continually on the flat ground receiving target angles of leg joints as input and demonstrate the adequacy of the proposed robot model.