【摘要】 在RoboCup仿真比赛中,球员的个体技术是比赛成功的重要因素.在充分分析RoboCup比赛个人技术中的两类技术,即判别函数、动作函数的确定方法的基础上,将模式识别的理论应用于RoboCup仿真比赛的个体技术中.分别使用判别域代数界面方程法确定判别函数,神经网络法确定动作函数.以射门和截球为例,详细叙述了设计过程.最后,使用传统方法和使用模式识别方法的球队在RoboCup仿真比赛中进行比赛,试验证明使用模式识别方法的球队,在比赛中个体技术显著提高.更多还原

【Abstract】 In RoboCup simulation, the technology of individual ballplayer is the basis to win the game. The theory of pattern recognition is applied to the design of individual technology in the game of RoboCup after two kinds of individual technology is analyzed which are the method of differentiate function and neural network function, respectively. The method of algebraic interface equation in discriminated fields is used in making differentiate function and neural network is used in active function. Shooting and intercepting of balls are taken as examples, and their design processes are described in detail. Finally, the game of RoboCup simulation with the team using traditional method vs. the team using pattern recognition method proves the method to be effective.更多还原