Abstract: 在大型基础设施系统中，不可避免地会出现重大的不确定性，因而，无法确保已建系统的安全性和可靠性。有 功能失效甚至破坏的可能，同时也有相应的规避风险的措施。从实用角度出发，不确定性可以分为两大类: 数据型 ( 又称偶然型) 不确定性和知识型( 又称认知型) 不确定性。前者可以通过对数据的观察来评估，而后者则需要主观 判断。系统地分析了这两类不确定性，得到有关风险预测的信息，为大型基础设施设计提供基于安全( 即规避风 险) 的建议; 对在地震荷载下大型基础设施进行了安全设计; 以韩国某斜拉桥为例，对其进行了数值化评估及最低生 命周期成本设计。

Key words: 风险, 概率, 不确定性, 安全性, 可靠性, 基础设施, 桥梁设计

摘要： Significant uncertainties are invariably present in the design of major infrastructure systems such as long-span bridges and other critical systems． In this light，the safety and reliability of the built system cannot be ensured with certainty or complete confidence; there is some probability of non-performance or failure and，therefore，some measure of risk ( the product of the failure probability and its adverse consequence) ． For practical purposes，uncertainties may be classified into two broad types-the data-based type ( known as aleatory) and the knowledge-based type ( known as epistemic) ; the former can be assessed through observed data，whereas the latter often requires judgments for its assessment． The thrust of the paper is to present the systematic analysis of the two distinct types of uncertainty and to generate the pertinent risk-informed information with suggestions for determining conservative ( i． e． ，risk-averse) safety levels for the design of major infrastructure systems． The approach is presented for the safe design of major infrastructure systems under seismic loading，and illustrated numerically with the evaluation of a cable-stayed bridge in Korea including its minimum life-cycle cost design．

关键词: risk, probability, uncertainty, safety, reliability, infrastructure, bridge design