Decision making of an omni-directional mobile robot by using artificial potential method with integration-type sonar ring

This study presents a novel decision making method for autonomous mobile robots to reach a goal and avoid obstacles simultaneously based on the artificial potential method. The proposed method is built so as to resolve the problems of high computational and economical costs, and the restriction of work field in conventional methods. There are two specialties to achieve the aims. First, the proposed method adopts the local polar coordinate system whose origin is the center of the robot, and generates each potential field by using only local information from sensors on the mobile robot. Second, we use an integration-type sonar ring that has been proposed by our research group to recognize obstacles and generating the repulsive potential field, and RGB camera to locate the goal and generate the attractive potential field. After generating both potential fields, they are summed up to generate the composed potential field, and then the target angle is decided based on the potential field. By using the proposed method, the computational and economical cost can be reduced and the robot can move without any advance preparations. The validity of the proposal method is investigated in various environments by applying the method to an autonomous mobile robot and results show the effectiveness of this approach.