地下建筑火灾风险的贝叶斯网络评估系统设计

摘要

Aiming at the problem that the uncertainty factors existing in underground building fire can affect the accuracy of fire risk quantitative assessment,a risk assessment method based on Bayesian network is proposed.Firstly,a Bayesian network model is established in accordance with the ideal assumption of underground building fire scene.Then,this model is modified considering the effect of uncertainty factors and used to calculate the probabilities of the fire in different development stages for fire risk assessment.Finally,a fire risk assessment system based on Bayesian network is developed using Visual Basic + Matlab mixed programming method.This system can be used by evaluation people with no rich professional knowledge,so it is convenient for a wide range of applications.%针对地下建筑火灾过程中存在的不确定性因素会影响到火灾风险定量评估准确性的问题,提出了基于贝叶斯网络的风险评估方法:首先,根据地下建筑火灾场景的理想假设建立贝叶斯网络模型;然后,考虑不确定因素的影响对模型进行修正,利用经修正的模型计算火灾发展到不同阶段的概率,进而评估火灾风险;最后,提出了利用Visual Basic+ Matlab混合编程的方法开发基于贝叶斯网络的火灾风险评估系统,降低对评估人员专业知识的要求,便于方法的推广应用.