Research Of Decision-Making System For Soccer Robots In RoboCup Middle-Size Matches

The research on soccer robot has obtained rapid development in recent years, which causes the item of many academics cross to receive more and more attention. As a new research realm, it has many contents to be worth studying. This article' s research object is the middle-size soccer robot. It emphatically analyze and study the core system of robot — decision-making system.Soccer Robot system is made of four child systems , which includes vision system, decision system, wireless communication system, movement control system . In this thesis, the design of decision-making system is emphasized. The decision-making system enables robots to kick a ball with "the brains" sufficiently. It may say, the decision-making system is a soccer team's soul. In the soccer robot system, the decision-making system' s mission is to analysis the data of field which spreads from vision system, to make a judgment about the circumstance of the field game, to form game strategy, and to plan robots' path through roles' decision-making model. We can see, the quality of the decision-making influences game result directly.At present, as a key step of game result, the decision-making system' s some technical difficulties have not still obtained complete solution, moreover the game result and watching also set a higher request to the decision-making technology. This article has been clear about decision-making system' s mission. Based on this foundation, it mainly introduces the decision system' s overall structure, the roles assignment algorithm and the design of decision-making model. This article innovation spot has two: in the robot roles assignment , it is proposed to achieve matching between robots and roles with dimidiate graph match algorithm; and this article structures various roles' state automaton decision-making model, the language which is produced with the automaton expresses robots' path.