基于扩展事件图的因果分析方法及在仿真中的应用

通过因果分析对仿真结果作出因果解释是仿真非常重要的一个环节.为了能够支持因果分析,首先提出扩展事件图,以重构仿真中发生的事件以及事件之间的因果关系;然后根据扩展事件图提出若干分析算法,通过对图模型的概率参数的可能性和主要性分析,获取因果知识;最后以一个攻防对抗仿真为例,对以上方法进行了必要的验证.

Extended Event Graph Based Causal Analysis and Its Application in Simulation

Causal explanation of simulation results through causal analysis is one of important parts of simulation.To support causal analysis,extended event graph(EEG) is put forward to rebuild the events and their causal relations happened in simulation.Then,some analysis algorithms based on the EEGs are proposed,from which user can gain causal knowledge through possibility analysis and importance analysis of probability parameter in graph model.At last,these methods are validated by an engagement simulation.