Ball Passing Strategy Based on a Fuzzy Inference System in RoboCup
Ball passing is an elementary and frequently employed human soccer skill. However, in RoboCup tournament, most of the time robots do not pass to each other; they prefer to dribble the ball towards the goal on their own. Because the current AI system only takes into account a limited amount of possible passing lanes, thus making it impossible to pass on many occasions. This paper presents a ball passing strategy which based on the fuzzy inference system. First, we use an algorithm to figure out all the potential passing lanes.Second, evaluate these lanes by fuzzy rules and pick out the optimal one. With this method AmoiensisNQ soccer simulation team of Xiamen University achieved the championship in the preliminaries of RoboCup that was held in Bremen, Germany in 2006.
RoboCup ball passing fuzzy inference system
XIA Feifei MIAO Kehua
Department of automation University of Xiamen Xiamen, Fujian, China 361005
国际会议
武汉
英文
181-183
2007-07-25（万方平台首次上网日期，不代表论文的发表时间）