【作者】 王翔；

【导师】 荣莉莉；

【作者基本信息】 大连理工大学 ， 系统工程， 2011， 硕士

【摘要】 当前,各种灾害事件频发,且往往呈现链式结构不断演化的态势,所造成的危害和影响远比单一灾害事件大而深远。越来越多灾害链的实例使得人们认识到,从灾害链的角度进行灾害风险研究,可以更加有效的进行灾前准备和灾中处理,以减少由灾害连锁带来的损失变得尤为重要。所以,对灾害链的风险评估问题,成为摆在人们面前亟待解决的问题。针对目前关于灾害链的风险评估方面的研究还比较少,本文从供应链和事故链总结出了链式风险评估的两种模式,借鉴这些模式,同时对区域灾害链风险进行了系统分析,提出了适合于灾害链风险评估的模式。并对灾害链发生的区域环境进行分析,构建了-个区域灾害链的影响因素指标体系,在此基础上提出区域灾害链风险评估模型,来定量计算灾害链的风险值。构成该模型的因素有：一个灾害事件引起另一个事件的概率；灾害链上灾害事件的损失；在灾害链网上一个灾害事件引发另一事件对其它事件的影响强度,即边的脆弱度；以及它们共同度量灾害链风险的方式,即评估模式。与其它单一灾种的风险评估模型最大的不同在于,将灾害链网中边的脆弱度引进来,可以较好的体现灾害事件之间引发关系的相互影响程度。并给出了一个实例,当某一灾害在某一区域发生时,引发了多条基于该区域特性的灾害链,运用上述模型,对每条灾害链进行风险度量,得到一个灾害链风险排序,为灾害预警和断链减灾工作做出决策支持。针对上述模型当中灾害链的脆弱性,进行了详细的分析。在灾害链构成的网络中,边表示的是灾害事件之间的引起关系,基于网络拓扑结构度量边的脆弱性,就是运用网络抗毁性的研究成果,衡量整个网络的脆弱环节。故边的脆弱性是某一网络的固有属性,其表达该边在整个网络中的薄弱程度,在灾害链网络中反映的是灾害事件之间引发关系的相互影响程度,是度量灾害链风险的一个宏观参数。边的脆弱性度量基于网络抗毁性的研究,研究网络抗毁性就是找到该网络中的脆弱环节,而在灾害链网络中,基于节点的不可移除性,边的脆弱性即为网络的脆弱环节。同时,参考连通系数的定义,并综合考虑了介数因素,本文给出了一个计算边的脆弱度的算法,并以2008年雪灾为例,验证了这种算法的有效性。更多还原

【Abstract】 These years, all disasters happens frequently, and evolving as chain structure, often lead to further harm impact than a single disaster event. More and more instances of disaster chain makes people realize that starting from the point of view of disaster chain, disaster risk can be more effective in pre-disaster preparedness and disaster handling. So, how to assess risk of disaster chain is a very urgent problem to be solved.Aiming at the risk assessment about disasters chain research also less, From the supply chain and accident chain, this paper sum up 2 models on chained risk assessment. Draw on these patterns, and propose a more appropriate chain risk assessment model for disaster chain assessment. And analyze regional environment on which disaster chain happen, construct a indexes system that describe influence factors of region disaster chain, put forward a risk assessment model of regional disaster chain, to calculate risk value. Composition factors of the model are:the probability of a disaster cause another; the loss of an event on a disaster chain; the influence intensity of a disaster cause another to other events, that is, edge’s vulnerability; the mode of they measure risk of disaster chain together, that is, assess model. The biggest difference with other risk assessment models of single disaster is, introducing edge’s vulnerability of network of disaster chain, can better reflect interaction degree of causing relationship between disaster events. An example is given, when a disaster occurs in a region, trigger a number of disaster chain based on regional characteristics, we use the model, measure each disaster chain risk, can get an order of disaster chain risk, then do some decision-making support for disaster warning, chain bread and disaster reduction.Aiming at disaster chain’s vulnerability in above model, carry out a detailed analysis. In a network constituted by disaster chain, edge express a caused relationship between disaster events, it is to use the research findings of network invulnerability, that measure vulnerability of edge based on network topology. Therefore, vulnerability of edge is network’s inherent properties, it shows the edge’s weak degree in the whole network, reflects interaction of caused relationship between disaster events, is a macroparameter that can measure disaster chain’s risk. The measure of edge’s vulnerability is based on research of network’s invulnerability, which is to find the weak link in the network. In disaster network, base on the particularity that nodes can’t be removed in disaster network, edge’s vulnerability is the weak link of the network. In the meanwhile, an algorithm is given to calculate the vulnerability of edge, taking snowstorm in 2008 for example, verify the effectiveness of this algorithm.更多还原