用影响网分析复杂系统关键事件

为分析复杂系统关键事件，将系统事件序列图转换为影响网图，通过着色Petri网进行建模和仿真，采用影响网分析系统事件的敏感性并搜索关键事件集合. 统计着色Petri网模型的仿真输出结果，得出事件发生的条件概率和基准概率，将影响强度与条件概率值的关系线性化，从而推算出影响强度. 该方法避免了影响强度、基准概率等参数由专家指定存在的主观性. 以弹道导弹防御系统为例进行分析，结果表明所述方法直观实用，对复杂系统的决策有一定的参考价值.

Analysis of Key Events in Complex Systems Using Influence Nets

To analyze key events of complex systems, the event sequence diagram is translated to an influence network. By colored petri net (CPN) modeling and simulation, the influence network is used to analyze sensitivity of the system events and search key event sets. According to the output of the CPN model, conditional probability and baseline probability are obtained statistically. The relationship between causal strength and conditional probability is linearized to obtain the causal strength value objectively. We apply the proposed method in a ballistic missile defense system. The results show that the method is straightforward and practical. It is useful in decision-making of complex simulation systems.