摘要
探讨了以美国地震灾害风险评估模型为例的建筑物易损性模型的建立方法及地震随机事件损失组合的方法。由于应用对象的不同,如用于工程项目风险的评估或是对保险资产风险的评估,其易损性模型建立的方式具有较大的差异。在工程应用中,注重对建筑结构的分析,往往使用非线性解析方法如能力频谱法等以得到结构的易损性或脆弱曲线;而保险行业常用历史数据并运用统计方法等获取所需的易损性曲线。由于工程应用的易损性曲线有更好的精准性,越来越多保险用模型采用此方法。此外,工程应用常对单一地震事件进行评估,而保险行业的模型往往涉及对大量随机事件的评估。对不同的建筑物易损性建模方法、规范标准及适用范围进行了探讨,并对随机事件损失组合方法中如何计算损失的单次超越概率(OEP)和累计超越概率(AEP)进行了介绍。
The paper introduced the methodology for establishing a vulnerability model of building structures for risk assessment of US earthquake and the combination of random earthquake events. The approach for a building＇s vulnerability model varies based on＇ the application of the risk assessment, such as for engineering projects or for insurance properties. In the engineering application, it focuses more on the analysis of building＇s structure, and it obtains the vulnerability or fragility curves using non-linear methods such as capacity-spectrum method. In the insurance application, however, it builds up building＇s vulnerability curves using statistics tools and based on quantitative historical claimed data. The vulnerability curves for engineering application often provide better precisions; therefore, more and more insurance models use this kind of curves. Further more, the model for engineering assessment often deals with a single earthquake scenario, while the model for insurance usually processes multiple events at a time. The paper discusses the methodology of establishing different kinds of vulnerability models, the related standard and guidance, and their application field. The paper also discussed the statistical methods of combining losses for multiple earthquake events, the methods includes average annual losses （ AAL）, occurrence exceeding probability （OEP） and accumulative exceeding probability （AEP）.
出处
《灾害学》
CSCD
北大核心
2012年第3期116-121,共6页
Journal of Catastrophology
基金
科技部国际科技合作项目(2010DFB20880)
北京师范大学地表过程与资源生态国家重点实验室重点项目
上海市科委上海浦江人才计划项目(11PJ1405500)
关键词
地震
风险评估
建筑结构易损性
脆弱曲线
随机事件
超越概率
美国
earthquake
risk assessment
building＇s vulnerability
fragility curves
random events
exceeding probability
US
