考虑决策者情绪的暴雨动态贝叶斯网络情景推演

Consider the Heavy Rain Dynamic Bayesian Network Scenario Deduction of Decision Makers

为处理“7·20”河南省特大暴雨灾害的复杂演化过程,验证应急管理措施的科学性和有效性,利用动态贝叶斯网络,运用证据理论和情绪更新机制,提出了一种主客观因素相结合的情景推演方法,较同类方法更加注重对决策者情绪等主观因素的研究。结果表明:动态贝叶斯网络能有效仿真情景推演的动态变化过程,改进DS证据理论能减少方法的主观性并处理演化过程不确定性,情绪更新机制能有效地量化决策主体的情绪变化情况。

In order to deal with the complex evolutionary process of the“July 20”heavy rainfall disaster in Henan province and verify the scientific and effective emergency management measures,by using dynamic Bayesian network and applying evidence theory and emotional re-newal mechanism,this paper proposed a scenario inference model which combined subjective and objective factors,and paid more attention to the research of decision-maker's emotion and other subjective factors.The results show that the dynamic Bayesian network can effectively simulate the dynamic process of scenario deduction,and the improved DS evidence theory can reduce the subjectivity of the model and deal with the uncertainty of the evolution process,the emotion renewal mechanism can effectively quantify the emotion change of decision-makers.