Time optimal positioning control of a quadruped robot for trot gait

Abstract

A design method of time-optimal trajectory of quadruped robots from an initial position to a target position is developed for trot gait. The designed trajectory consists of three phases, acceleration phase, steady walk phase and deceleration phase. The time-optimal trajectory is designed in each phase and they are integrated so that the robot reaches the target position. Since the fastest walk trajectory in steady walk has already been obtained in our previous research, the time-optimal trajectories in the acceleration and deceleration phases are designed in a similar way to the previous study. First, all the possible trajectories in the search space with which the velocity of the robot reaches the previously obtained maximum velocity are designed in the state space of the robot. Next, they are checked whether the physical constraints and the constraint imposed by swinging legs are satisfied or not. Then the trajectory with the shortest time to reach the maximum velocity is chosen as the time-optimal trajectory among the trajectories satisfying all the constraints. The obtained time-optimal trajectories are tested by a quadruped walking robot SONY ERS-7 and the effectiveness of the developed method of designing time-optimal trajectory is verified.