摘要： Agent理论和技术是当前AI研究中的热点。其中，基于目标驱动的Agent计算是一个尚待解决的关键性问题。文章提出了目标驱动中的依赖和协商机理，以Robocup为测试床建立了面向足球抢断的动态目标驱动模型，并基于该模型分析了在特定足球阵形下借助依赖与协商形成抢断策略的过程。

关键词: Agent, 目标驱动, RoboCup

Abstract: Theories and technologies about Agent play an important role in AI (atificial ielligence) research, among these, the goal-driven computing is still an unsolved key problem. This paper gives a dynamic goal-driven model, which built on the test-bed of RoboCup to simulate soccer scramble, and uses it to analyze the process of forming a scramble tactic with the help of reliant and negotiatory relationships in a certain information .

中图分类号: 

• TP393.02