一种基于可传递置信模型的分布式决策方法  

The Method to Distributed Decision Based on the Transferable Belief Model

作　　者：范波[1] 潘泉[1] 张洪才[1] 

机构地区：[1]西北工业大学自动控制系,西安710072

出　　处：《系统仿真学报》2004年第11期2622-2624,2627,共4页Journal of System Simulation

摘　　要：证据推理理论是研究在不确定情况下智能决策的有效方法之一。将证据推理引入到分布式决策系统,提出了一种基于可传递置信模型的分布式智能决策方法,从信息融合技术的角度构建了agent模型以及分布式决策系统的框架,并且分析了分布式决策系统的信息提取和合成算法以及决策制定算法。通过借鉴信息融合技术中证据推理的可传递置信模型和算法,使分布式决策的研究方法有进一步的发展。在仿真中采用机器人足球仿真比赛作为平台进行研究。Evidential reasoning theory becomes one of the effective methods to study the intelligent decision in situation with uncertainty. The evidence reasoning is introduced to distributed decision system. The method is presented to distributed intelligent decision based on the transferable belief model. The agent model and distributed decision system architecture is built from the point of view of information fusion. The algorithms of informations extracting and combination and of decision-making are analyzed. This method is applied to the Robot Soccer simulation match.

关 键 词：证据推理 可传递置信模型 分布式决策 多智能体 

分 类 号：TP14[自动化与计算机技术—控制理论与控制工程] TP181[自动化与计算机技术—控制科学与工程]