军事指挥控制中的贝叶斯决策支持

Bayesian Decision Support for Military Command and Control

军事指挥控制是一个复杂的决策过程,该过程中信息是结合定性的专家知识和定量的融合数据而得到的。贝叶斯网为指挥控制系统提供了一个框架来处理这种专家知识和具体数据相结合的问题,并且影响图作为贝叶斯网的延伸可被用于分析指挥控制系统中某个决策的可行性。进一步地,为了减少所需指定的条件概率的数量并引入时变的概念,提出动态影响网作为一种新的决策支持方法。

The military Command and Control （C2） is a complicated decision process. Moreover, the information acquirement of this decision process is based on the qualitative data of experts＇ knowledge structure and quantitative fusion data. Bayesian Networks （BNs） provide a framework for the C2 system in which to deal with the problems combining the experts＇ knowledge structure and specific data,and Influence Diagrams can be used to analysis the feasibility of some decisions in the C2 system as the extension of BNs. Furthermore, in order to decrease the amount of assigned conditional probabilities and introduce the concept of time-varying,Dynamic Influence Networks is proposed to be used as a new method of decision-making support.