针对机器人足球仿真比赛,提出了一种新的多智能体系统结构,它包括基于行为的双层动态智能体结构和决策算法以及基于角色的动态多智能体协作模型.论文首次提出了自信函数这个概念,并运用此概念来平衡决策算法和多智能体之间的协作.该结构应用于中南大学云麓仿真球队中,在中国机器人足球RoboCup仿真组比赛中取得了较好的成绩、
This paper presents a new multi-agent system architecture for the RoboCup simulation league. It includes a behavior-based two-layer dynamic agent architecture, a behavior-based decision algorithm, and a role-based dynamic multi-agent cooperation model. A self-confidence function is proposed for the first time and is used to balance decisions and multi-agent cooperation. This architecture has been implemented in the CSU_Yunlu team of Central South University, which obtained excellent result in the simulation league of the RoboCup 2002 Chinese Championship.