Condition-based risk assessment and management for infrastructure retrofit in the face of uncertainty

社会生活和经济增长严重依赖关键的基础设施系统，加拿大的资产价值约为1.1万亿美元。然而，根据2016年加拿大基础设施报告卡，35%的基础设施资产处于一般、较差或非常差的状态，原因是严重的环境攻击（如腐蚀）和反复发生的低强度或中等强度的灾害事件（如地震）。这种退化，以结构/材料的渐进退化和基于冲击的物理损伤的形式，降低了工程结构的性能。由于认识到在极端危险事件中它们的故障或功能丧失可能造成毁灭性的生命安全和经济后果，近年来对现有基础设施系统进行基于条件的风险评估和管理（例如，通过最佳改造）已成为一项迫切需要。拟议研究计划的目标是为面对不确定性的“第二代”改造战略制定基于条件的风险评估和管理（CRAM）框架。该框架集成了受危险或年龄影响的基础设施系统的状态识别（即，估计当前状态的概率）、考虑相关不确定性来源的风险评估，以及风险管理的风险知情优化改造设计。条件识别和风险评估将为改造提供信息，并且将开发一种统一的风险知情改造或修复设计方法，使用优化来帮助提高结构性能。这项拟议的研究计划将重点放在现有的基础设施，如桥梁和建筑物遭受地震危险。四个目标或主题是开发(A)具有退化性能的工程结构的详细非线性有限元（FE）模型；(B)用于状态识别的非线性有限元模型更新技术；(C)考虑相关不确定性（包括模型参数不确定性）的风险评估的不确定性传播工具；(D)面对不确定性时用于降低风险的统一的最佳改造方法。该项目的成功成果将成为将CRAM框架扩展和应用于更广泛的工程风险问题的长期研究计划的里程碑，为社会提供风险管理解决方案。

The societal living and economic growth rely heavily on critical infrastructural systems, with the assets value of about $1.1 trillion in Canada. However, according to the 2016 Canadian Infrastructure Report Card, 35% of those infrastructure assets are in fair, poor or very poor condition, due to aggressive environmental attacks (e.g., corrosion) and recurring low- or moderate-intensity hazard events (e.g., earthquakes). The degradation, in the form of progressive structural/material deteriorations and shock-based physical damage, reduces the performance of engineering structures. In recognition of devastating life safety and economic consequences potentially caused by their failure or loss of functionality in extreme hazard events, the condition-based risk assessment and management (e.g., through optimum retrofit) of existing infrastructure systems has become an emergent need in recent years.The goal of the proposed research program is to develop a condition-based risk assessment and management (CRAM) framework for the “second-generation” retrofit strategy in the face of uncertainty. This framework integrates condition identification (i.e., estimating the current state probabilistically) of the hazard- or age-affected infrastructure systems, risk assessment with pertinent sources of uncertainty considered, and risk-informed optimum retrofit design for risk management. The condition identification and risk assessment will provide information on retrofit, and a unified risk-informed retrofit or repair design approach using optimization will be developed to aid in enhancing the structural performance.This proposed research program will focus on existing infrastructure such as bridges and buildings subjected to seismic hazard. The four aims or themes are developing (A) detailed nonlinear finite element (FE) models of engineering structures with degrading properties; (B) nonlinear FE model updating techniques used for condition identification; (C) an uncertainty propagation tool for risk assessment accounting for pertinent uncertainties including model parameter uncertainty; and (D) a unified optimum retrofit approach used for risk-mitigation in the face of uncertainty. The successful outcome of this project will reach a milestone in the long-term research program on extending and applying the CRAM framework to a wide range of engineering risk problems, providing society with risk management solutions.