Bondgraph Robot Soccer Simulation for Minimum Time Attacker Maneuvers

Dynamic model of a differential drive soccer player using bondgraph method is presented. Minimum time maneuvers for the attacker behavior based on the model derived using bondgraph methodology are studied. The paper presents a new cooperative attacking behavior for two agents in a real time robot soccer game. This behavior consists of four hierarchy levels, the intelligent level, path planning level, path following and velocity controller level. This paper reports the Simulation and experimental results of applying those hierarchy levels on the AUS robot soccer team.