An Obstacle Avoidance TrajectoryPlanning for a Quadruped Walking Robot Using Vision and PSD Sensor

초록

This paper deals with obstacle avoidance of a quadruped robot with vision system and a distance-measuring sensor. In order to work out the designated plan, it is required to recognize the terrain status and the obstacles using sensor systems, and then it is needed to generate the modified gait trajectory. In this paper we propose an obstacle recognition method for a quadruped robot equipped with sensor systems, and also suggest the walking path generation method to the target at corresponding case. Using this walking path, we can generate the adequate walking trajectories for the legs using genetic algorithm and our walking robot can avoid the obstacles that are located irregularly.