A real-time role assignment mechanism for five-on-five robot soccer competition

Abstract

In the robot soccer competition, we concern not only the accuracy of the image processing, but also the role that each robot should play. In this paper, we firstly apply the reflex behavior method and univector field method for the soccer robot to accomplish the obstacle-avoidance and goal-reaching missions. Then the role assignment mechanism (RAM) is proposed to determine all the worthy location for every soccer robot. The worthy locations include the best shooting, obstruction, intercepting, assisted defending and defending positions. Both computer simulations and the real-time five-on-five robot soccer game illustrate the feasibility and effectiveness of the proposed schemes.