Research And Implement On Simulation Robot Soccer Attacking Strategy

In recent years, robot soccer tournament has become one of the focuses of AI and robotics. Because the hardware is very expensive and mainly technologies are included in robot soccer simulation game, robot soccer simulation game has become an important embranchment of robot soccer tournament. The thesis is research on simulation robot soccer tournament.Simulation robot soccer tournament is the same as the decision-making of robot soccer game. And a new multilayer antagonizing strategy model which include stratagem layer, tactics layer, action layer is proposed by the analogy of all kinds of decision-making and the soccer robot basic characteristic.Stratagem layer is a global, unified scheming and directive decision-making layer. In stratagem layer, a new way that according to the game the strategy of tactics is changed is proposed in the article. The tactics is a very important layer in the decision-making and is a connecting link between the preceding and the following. Two mainly part—formation and role assignment, path-planning and anti-collision are included in the tactics layer. In the part of formation and role assignment, many familiar ways of formation and role assignment are analyzed, and at last, the author give a new way—PSO. In the part of path-planning and anti-collision, the thesis analyzes the characteristic of robot soccer path-planning and anti-collision and introduces some accustomed ways. In the end, the way of the shortest tangent path-planning and anti-collision that is used in our robot soccer system is given. In action layer, the basic actions are detailed divided and the implement ways of the basic actions are given. All kinds of ways of kickball, such as perpendicular bisector kickball method, vector field kickball method, are introduced and the implement methods are given, too. All ways of shoot are analyzed and a simple dead angle way of shoot that is used in this system is proposed. With the discovery learning, the penalty kick strategy and free ball strategy of opponent are discovered. And according to the learning knowledge, an effective strategy of penalty kick is presented and a new method of free ball is proposed.