用于态势估计的贝叶斯网络方法  

Methods of Bayesian Networks for Situation Assessment

作　　者：孙兆林[1] 马志奇[1] 

机构地区：[1]国防科技大学ATR重点实验室,湖南长沙410073

出　　处：《情报指挥控制系统与仿真技术》2005年第4期15-18,共4页Information Command Control System and Simulation Technology

摘　　要：态势估计系统需要对大量的不确定性知识进行处理,不确定性知识表示和推理是态势估计中的研究热点。利用贝叶斯网络技术,可以实现对不确定性知识的处理。为此讨论了贝叶斯网络理论,详细分析了态势估计的功能模型,提出事件检测是态势估计的核心和起点,研究了用于态势估计的贝叶斯网络的构建方法。该方法充分考虑贝叶斯网络的几项基本要素,可以解决态势估计领域中贝叶斯网络的构建问题。贝叶斯网络技术在态势估计领域具有广阔的前景,将极大地推动态势估计系统的发展。We have to process a lot of uncertain knowledge in the situation assessment system. It＇s a hot topic to represent and reason for uncertain knowledge in the situation assessment. We can realize the processing of uncertain knowledge by using the Bayesian networks technology. This paper discusses the theory of Bayesian networks, analyzes the functional model for situation assessment, and presents that event-detection is the start point and kernel for situation assessment. Also, we research on the method to construct Bayesian networks for situation assessment. The method takes the basic elements of the Bayesian networks into account and realizes the construction of the networks in the situation assessment domain. The Bayesian networks technology has wide prospect in the situation assessment domain, and it will greatly promote the development of the situation assessment system.

关 键 词：态势估计 贝叶斯网络 事件检测 

分 类 号：O212.8[理学—概率论与数理统计]