摘要
为建立一套新型的机器人系统 ,以满足机器人足球比赛的要求 ,设计并实现了清华机器人足球队。在介绍国际机器人足球比赛 (Robo Cup)环境的基础上 ,按照智能程度由低到高的顺序 ,论述了从动作层到决策层的实现方法。包括机器人的整体结构 ,运用多层前馈神经元网络学习个人技术 ,运用优先级方法设计机器人的顶层决策算法来实现多个机器人的合作与对抗。机器人系统完全满足 Robo Cup比赛的要求 ,并在中国 Robo Cup2 0 0 0锦标赛中取得了好成绩。
The Tsinghua Robot Soccer Team was designed to meet the requirements of the robot soccer games for the Robot World Cup Tournament (RoboCup). A method is presented for constructing individual skill layers and decision making layers according to intelligent levels. The research work includes the structural design of the robots, the implementation of individual skills through multi layer BP neural networks, and the application of priority algorithms to achieve cooperation between robots as well as competition between each other. The Tsinghua Robot Soccer Team succeeded in the Chinese RoboCup2000 Tournament. This paper provides information for other researchers to develop robot soccer teams and other processes multi agent domains.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第7期94-97,共4页
Journal of Tsinghua University(Science and Technology)
