摘要
针对灾害链风险评估的问题,本文提出一种基于复杂网络结构的灾害链风险评估模型的建模方法.首先,应用复杂网络结构对灾害链的演化特征进行了表征,并对灾害链的作用机理进行了数学描述.然后,在对供应链风险评估方法和模式分析的基础上,以复杂网络为载体,运用贝叶斯公式得到各节点灾害损失等级的联合概率分布,构建了灾害链的风险评估模型,提出用子节点风险损失的期望值来表征子节点在父节点作用下的风险度量模式.最后以珠海市台风灾害链为实例进行了评估,结果表明所建立的模型与实际情况具有较高的吻合性.说明以灾害链为中心构建的灾害链风险评估模型考虑了空间和时间上的叠加,并对灾种间的作用关系进行了厘清,真实、定量、准确地刻画灾害链式演变过程所带来的风险,能为区域性灾害风险管理提供可靠的理论依据.
Considering the expansion of disaster losses caused by disasters chain effects, to establish the risk assessment model of disaster chain, this paper proposed a modeling method of disaster chain risk assessment based on complex network structure. At first, the evolution characteristic of disaster chains was characterized adopting complex network structure. The disaster chain mechanism was mathematical description based on the definition of disaster chain. Then, learning from the risk assessment method and model of supply chains, taking complex network as the carrier, we obtained the joint probability distribution of disaster loss grades of each node using Bayesian formula. And the risk assessment model of disaster chain was built up. The risk measurement mode was proposed that the risk loss expectation values of child nodes should be used to characterize the risk of child nodes acted by parent nodes. At last, using Typhoon disaster chain in Zhuhai as an application instance, we assessed the risk of Typhoon disaster chain. The results indicate that the model established has higher consistency with actual situation. That also shows the risk superposition of space and time has been considered in the assessment model taking disaster chain as the center, and the relationship between the disasters has been distinguished. So it can truly, quantitatively, accurately portray the risk caused by the disaster chain evolution process, and provide a reliable theoretical basis for the regional disaster risk management.
出处
《系统工程理论与实践》
EI
CSSCI
CSCD
北大核心
2015年第2期466-472,共7页
Systems Engineering-Theory & Practice
基金
教育部第二类特色专业项目(TS2318)
关键词
风险评估
灾害链
复杂网络
联合概率分布
贝叶斯公式
risk assessment
disaster chain
complex network
joint probability distribution
Bayesian formula
