Second-order Nonlinear Function Navigation Method for Fast Mobile Robots

There are two different approaches to the mobile robot navigation problem including that of robot soccer: deliberative and reactive. In the deliberative approach, a detailed geometric model of the world is used to create a plan that accomplishes the task. This approach suffers from high computational requirements and lacks of robustness in the face of the world's uncertainty. The reactive approach overcomes the disadvantages of the deliberative approach, but the design that could obtain a desired behavior is difficult. The work reported in this paper overcomes this difficulty by adapting a new navigation method, the limit-cycle navigation method. The defects of the potential field method are also presented. In the proposed method, with the information on the relative position of obstacles, target and robot, the obstacle that affects the robot's path is identified. Then, direction of turn, counter-clockwise or clockwise is decided using the radius of the virtual obstacle. Real experiments ascertain the merits of the proposed method.