Omnidirectional Sensing for Escaping Local Minimum on Potential Field Mobile Robot Path Planning in Corridors Environment

Abstract

Mobile robot path planning using artificial potential field approach is popular for its computational simplicity. However, the conventional artificial field potential approach possesses weakness when the robot is deployed into corridors environment. The approach makes the robot easily trapped in local minimum caused by long obstacles presence, thus makes the robot unable to get to the goal point. In this paper, a method for escaping local minimum in corridors scenario is proposed. The proposed method utilizes the omnidirectional sensor, which has the ability to sense 360 degrees field of view, to get information on obstacles which are surrounding the robot. This information is used for the robot to put a feasible temporary goal to guide the robot to detour the trap. Numerical experiment verified that the proposed method successfully generates a safe path and is able to escape the local minimum trap in corridors environment.