首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于影响图的多Agent决策问题研究
引用本文:王浩.基于影响图的多Agent决策问题研究[J].合肥工业大学学报(自然科学版),2005,28(9):1112-1116.
作者姓名:王浩
作者单位:合肥工业大学,计算机与信息学院,安徽,合肥,230009
基金项目:安徽省自然科学基金资助项目(03042305)
摘    要:影响图是决策问题的图形表示,它是在贝叶斯网络基础上增加了决策结点和效用结点。文章讨论了影响图理论在多Agent建模与决策方面的应用;提出了联合分层影响图和动态贝叶斯网络来实现多Agent的实时决策的方法,该方法已在Robcup球员建模和决策方面初步得到应用。

关 键 词:影响图  贝叶斯网络  多Agent系统  决策
文章编号:1003-5060(2005)09-1112-05
修稿时间:2005年7月11日

Study of the decision problem of multi-agent systems based on IDs
WANG Hao.Study of the decision problem of multi-agent systems based on IDs[J].Journal of Hefei University of Technology(Natural Science),2005,28(9):1112-1116.
Authors:WANG Hao
Abstract:Influence Diagrams(IDs) are graphical knowledge representation of decision problems and can be viewed as an extension of Bayesian Networks(BNs) with additional node types for decision and utility.In this paper,the application of Influence Diagrams in modeling and decision-making of multi-agent systems is discussed.The method that combines hierarchical Influence Diagrams and Dynamic Bayesian Networks to process real-time decision problems of agents is presented,and it is applied to decision-making in Robocup Simulation Contest.
Keywords:influence diagrams  bayesian networks  multi-agent system  decision-making
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号