基于性能的地震工程通用框架:结构可靠性与韧性的集成评估

准大一 · 土木工程 × 地震工程 × 概率与统计 —— 结构可靠性与韧性评估精读材料
📄 原文:arXiv:2606.12448 🗓️ 2026年5月30日发布 🏫 arXiv 预印本(physics.geo-ph) ✅ 基于性能的地震工程 × 马尔可夫链 × 韧性 🔊 附英文摘要朗读音频

一、论文档案

英文标题A generalized framework for performance-based earthquake engineering: integrated assessment of structural reliability and resilience
中文标题基于性能的地震工程通用框架:结构可靠性与韧性的集成评估
作者C. NArdin, S. Marelli, B. Sudret, M. Broccardo(按素材原文转录,机构未在素材中标注)
发布时间2026年5月30日(v1)|分类:physics.geo-ph(地球物理)、stat.CO(计算统计)、stat.ME(统计方法论)
原文链接arxiv.org/abs/2606.12448 · 全文 HTML 版
一句话概括把「损伤累积」与「灾后恢复」直接写进结构状态的连续时间马尔可夫链,用一张生成矩阵同时算可靠性与韧性——并证明经典 PEER-PBEE 框架只是它的一个特例。
💡 为什么选这篇给你:① 地震工程是土木工程最硬核的方向,「性能化设计」是行业主流范式;② 这篇把「可靠性」和「韧性」第一次放进同一个数学框架,思想密度高但主线清晰;③ 用支撑框架和基础隔震两个实际结构算例验证,还给了谱方法的直观解释。

二、核心术语表(先扫一遍再读正文)

英文术语中文大白话解释
PBEE (performance-based earthquake engineering)基于性能的地震工程按明确的性能目标(如损伤限值、功能恢复时间)来设计和评估结构的工程范式。
seismic hazard地震危险性场地在未来遭遇不同强度地震动的可能性。
structural reliability结构可靠性结构在服役期内保持不失效(不倒塌)的能力,通常用失效概率或可靠度指标 β 衡量。
structural resilience结构韧性结构震后恢复到原有功能水平的速度与程度。
Poissonian exceedance assumption泊松超越假设假设地震按泊松过程独立发生、性能用「超越阈值」描述,隐含「每次震后自动恢复如新」。
renewability / memorylessness可更新性 / 无记忆性「每次地震后不确定性归零重来」以及「未来只与当前状态有关」的隐含假设。
CTMC (continuous-time Markov chain)连续时间马尔可夫链状态随时间连续演化的随机过程,状态间按速率矩阵(生成矩阵)转移。
generator matrix生成矩阵CTMC 的核心算子,一个矩阵同时编码地震危险性、状态相关易损性和恢复速率。
fragility function易损性函数给定地震动强度下结构达到某一损伤状态的条件概率函数。
hazard curve危险性曲线地震动强度指标与年超越概率之间的关系曲线。
failure probability / reliability index β失效概率 / 可靠度指标 β可靠性的两种量化指标,由瞬态系统动力学导出。
spectral analysis谱分析研究生成矩阵特征值/特征向量的数学工具,用于高效、透明地计算可靠性指标与韧性指标。
quasi-stationary distribution拟平稳分布系统在吸收(倒塌)之前条件状态占用的渐近分布。
braced frame支撑框架通过斜向支撑抵抗水平地震力的钢框架结构。
base-isolated system基础隔震结构用隔震支座把上部结构与地面运动解耦、显著降低地震响应的结构体系。
ETAS modelETAS(流行病型余震序列)模型描述主震后余震丛集特性的地震活动性模型,可用于非齐次地震输入。

三、摘要中英对照(精读核心)

🎧 音频在文末,可先听一遍原文再读;每个英文句都配了逐句翻译。

摘要 Abstract

EN · 原文
Assessing structural performance under seismic hazard requires accounting for both damage accumulation and post-event recovery.
CN · 翻译
评估地震危险性下的结构性能,需要同时考虑损伤累积震后恢复
EN · 原文
In current performance-based earthquake engineering (PBEE), recovery is generally treated as a post-processing attribute, while structural performance is modeled using Poissonian exceedance assumptions that imply renewability and memorylessness.
CN · 翻译
在当前基于性能的地震工程(PBEE)中,恢复通常被当作后处理属性,而结构性能用泊松超越假设建模——这隐含了「可更新性」和「无记忆性」。
EN · 原文
These assumptions hinder a unified treatment of reliability and resilience under repeated seismic loading.
CN · 翻译
这些假设阻碍了在重复地震荷载下对可靠性与韧性的统一处理
EN · 原文
This study proposes a generalized PBEE framework in which damage and recovery are embedded directly into the system dynamics through a continuous-time Markov chain.
CN · 翻译
本研究提出一个通用化 PBEE 框架:通过连续时间马尔可夫链把损伤与恢复直接嵌入系统动力学
EN · 原文
A single generator matrix governs state-dependent transitions, providing a unified description of structural reliability and resilience while remaining compatible with standard PBEE metrics.
CN · 翻译
一张生成矩阵控制状态相关转移,在兼容标准 PBEE 指标的同时,为结构可靠性与韧性提供统一描述。
EN · 原文
Time-dependent failure probabilities and reliability indices are derived from the transient system dynamics, whereas resilience is quantified through the expected fraction of operational time before collapse.
CN · 翻译
时变失效概率可靠度指标由瞬态系统动力学导出;韧性则用倒塌前预期运行时间占比来量化。
EN · 原文
The framework exploits the spectral properties of the generator matrix to compute both metrics efficiently and transparently.
CN · 翻译
框架利用生成矩阵的谱性质,高效且透明地计算两类指标。
EN · 原文
The methodology is illustrated on a three-state example and applied to two structural archetypes: a braced frame and a base-isolated system.
CN · 翻译
方法先用三状态算例说明,再应用于两种结构原型:支撑框架基础隔震系统
EN · 原文
Results show that recovery dynamics can strongly affect long-term resilience even when conventional reliability measures exhibit limited sensitivity, emphasizing the need to explicitly account for recovery in life-cycle seismic performance assessment.
CN · 翻译
结果表明:即使常规可靠性指标敏感性有限,恢复动力学也可能强烈影响长期韧性——这强调了在全寿命地震性能评估中显式考虑恢复的必要性。

关键词 Keywords:Performance-Based Earthquake Engineering 基于性能的地震工程 | Structural Reliability 结构可靠性 | Resilience 韧性 | Continuous-Time Markov Chain 连续时间马尔可夫链 | Generator Matrix 生成矩阵 | Spectral Analysis 谱分析

四、引言精选(为什么这个问题重要)

① 范式转变:从「只看损伤」到「损伤 + 停工 + 恢复」

EN · 原文
In performance-based earthquake engineering (PBEE), the paradigm is evolving from an exclusive focus on structural damage to a comprehensive view that includes downtime, functionality loss, and recovery [1, 2, 3]. This change reflects a fundamental paradigm shift: post-earthquake impacts are determined not only by the extent of physical damage but also by the time required to restore acceptable performance, or to get back to normal. For communities, businesses, and critical infrastructure operators, this recovery timeline can be as consequential as the damage itself, often dictating the long-term social, economic, and operational repercussions of an event.
CN · 翻译
在基于性能的地震工程(PBEE)中,范式正从只关注结构损伤转向包含停工时间、功能损失与恢复的综合视角。这是一次根本性范式转变:震后影响不仅取决于物理损伤程度,还取决于恢复到可接受性能所需的时间。对社区、企业和关键基础设施运营者来说,恢复时间线与损伤本身同样重要,往往决定了事件的长期社会、经济与运营后果。

② 经典框架的隐含假设:每次地震后「满血复活」

EN · 原文
Der Kiureghian [4] showed that the Poissonian representation of seismic occurrence, together with the ergodicity assumptions embedded in the PEER-PBEE framework [5, 6, 7, 8], effectively implies immediate recovery and renewal of uncertainties after each event. As a result, the framework is not suited, in its classical form, to represent cumulative damage and finite, non-instantaneous recovery, and therefore does not readily admit the integration of structural resilience within a single coherent framework. Here, the qualifier structural is used deliberately to denote the capacity of a structure to return to its original undamaged configuration while preserving the same level of performance.
CN · 翻译
Der Kiureghian [4] 指出:地震发生的泊松表示加上 PEER-PBEE 框架中的遍历性假设,实际上隐含了「每次事件后立即恢复、不确定性归零重来」。因此,经典形式的框架无法表示累积损伤和有限、非瞬时的恢复,也就难以把结构韧性纳入一个统一连贯的框架。这里「结构」一词是刻意使用的限定语,指结构恢复到原始无损状态并保持同等性能水平的能力。

③ 本文做法:把 PEER-PBEE 装进连续时间马尔可夫链

EN · 原文
Motivated by these observations, the present work generalizes the PEER-PBEE framework by embedding its modular architecture within the broader class of continuous-time Markov chains (CTMCs). In doing so, it extends the framework from seismic risk assessment, here understood as structural reliability under seismic loading, to an integrated treatment of risk and resilience, while preserving its original modus operandi.
CN · 翻译
受这些观察驱动,本文把 PEER-PBEE 的模块化架构嵌入更广泛的连续时间马尔可夫链(CTMC)类中,从而将其从「地震风险评估」(即地震荷载下的结构可靠性)扩展为风险与韧性的集成处理,同时保留其原有运作方式。

④ 为什么韧性必须显式建模:恢复可能持续数周、数月、数年

EN · 原文
Within this generalized setting, structural resilience acquires a direct representation within PBEE. Whereas the classical formulation expresses performance mainly in terms of mean rates of exceeding prescribed damage or loss thresholds, structural resilience depends on the temporal evolution of repair and restoration after an event [9]. Because recovery may unfold over weeks, months, or years, its explicit representation is essential to capture the persistence of degraded structural performance and the associated broader consequences [10, 2]. This motivates a shift from event-based exceedance measures to a formulation in which performance is described through the stochastic evolution of system states in time, including both damage accumulation and recovery.
CN · 翻译
在通用化框架下,结构韧性在 PBEE 中获得了直接表示。经典表述主要用「超越规定损伤/损失阈值的平均发生率」刻画性能,而韧性取决于震后修复与恢复的时间演化。由于恢复可能持续数周、数月甚至数年,显式表示恢复对于捕捉「性能持续劣化」及其广泛后果至关重要。这促使范式从「基于事件的超越度量」转向「用系统状态随时间的随机演化(同时包含损伤累积与恢复)描述性能」。

⑤ 工程实践:FEMA HAZUS 的恢复函数

EN · 原文
Over the past decade, substantial progress has been made in the incorporation of recovery into seismic loss assessment, both through the development of conceptual and probabilistic frameworks and through the implementation of widely used engineering software. In particular, many community- or regional-scale resilience studies rely on recovery functions to estimate repair and recovery times conditional on occupancy class and damage state. A prominent example is the FEMA HAZUS methodology [11], which provides a national database of generalized damage, loss, and recovery relationships for large-scale planning and risk assessment, especially within U.S. practice.
CN · 翻译
过去十年,把恢复纳入地震损失评估取得了实质进展——既有概念性与概率框架的发展,也有广泛使用的工程软件实现。许多社区或区域尺度的韧性研究依赖恢复函数,按占用类别和损伤状态估计修复与恢复时间。典型代表是 FEMA HAZUS 方法学 [11],它提供了面向大规模规划与风险评估的全国性「损伤-损失-恢复」关系数据库,尤其服务于美国实践。

⑥ 软件工具链:ATC-138、SP3、PELICUN、PACT

EN · 原文
At the same time, analytical studies have formalized recovery modeling, providing the mathematical foundations that underpin these tools. For example, Cassottana et al. [17] introduced a hybrid recovery function, that is a power-law multiplied by an exponential term, providing the first explicit mathematical definition and property set for such functions. Iervolino et al.
CN · 翻译
在这些进展之上,美国应用技术委员会主导的 ATC-138 项目 [14] 为现代建筑提供基于恢复的设计工具与指南。其方法已在多个平台实现:商业 SP3 平台 [15]、开源 PELICUN Python 包 [16],以及把 FEMA P-58 建筑清单与恢复建模结合的集成软件 PACT [13]——把概率化、基于序列的恢复框架转化为实际应用。
💡 这是全文最有味道的一句“In fact, in this generalized setting, the classical PEER-PBEE formulation is shown to be a special case.”——把统治行业几十年的框架「降格」为自己的特例,是数学化框架最有说服力的时刻:不是推翻旧框架,而是把它装进更大的盒子里

五、论文贡献(3 个要点)

EN · 原文
This study presents a continuous-time Markov chain (CTMC) formulation that generalizes the classical PEER-PBEE framework by embedding damage accumulation and recovery directly into the stochastic evolution of the structural state, thereby relaxing the Poissonian assumption of perfect renewability after seismic events.
CN · 翻译
1. CTMC 通用化框架。把损伤累积与恢复直接嵌入结构状态的随机演化,从而放松「震后完美更新」的泊松假设,将经典 PEER-PBEE 框架推广为特例。
EN · 原文
A key finding of the industrial-scale applications is the fundamental decoupling between reliability and resilience: long-term failure probabilities and reliability indices β\beta are governed almost exclusively by the seismic damage-to-collapse rates λ02\lambda_{02} and λ12\lambda_{12}, and are largely insensitive to the recovery rate μ10\mu_{10} across the full range of plausible repair scenarios. Resilience, by contrast, is strongly controlled by the competition between damage accumulation and recovery, as reflected in the conditional occupation fraction 𝒬0\mathcal{Q}_{0} and the standardized resilience index ρ\rho.
CN · 翻译
2. 关键发现:可靠性与韧性「解耦」。长期失效概率与可靠度指标 β 几乎只由「损伤→倒塌」速率 λ02、λ12 决定,对恢复速率 μ10 在全部合理修复场景下都基本不敏感。
EN · 原文
The proposed framework preserves the modular and non-intrusive nature of PBEE: existing fragility functions and hazard curves are embedded without modification, and recovery is introduced as a complementary rate-based component within the same architecture.
CN · 翻译
3. 模块化、非侵入式。现有易损性函数与危险性曲线无需修改即可嵌入,恢复以「补充的速率型组件」加入同一架构——工程上已有的成果全部保留。

六、结论中英对照

EN · 原文
The infinitesimal generator matrix 𝑸\bm{Q} unifies seismic hazard, state-dependent fragility, and recovery rates within a single operator, from which both reliability and resilience metrics are derived in a computationally efficient and physically interpretable manner.
CN · 翻译
无穷小生成矩阵 𝑸 把地震危险性、状态相关易损性与恢复速率统一进单一算子,可靠性与韧性指标都从中以高效计算、物理解释清晰的方式导出。
EN · 原文
The spectral analysis of the transient generator 𝑸T\bm{Q}_{T} provides an additional layer of insight: the dominant eigenvalue σ1\sigma_{1} sets the characteristic timescale of absorption, the quasi-stationary distribution 𝒒\bm{q} describes the conditional damage state occupancy prior to collapse, and the asymptotic resilience ℛ0∞=1−w0q0\mathcal{R}_{0}^{\infty}=1-w_{0}q_{0} links the macroscopic resilience metric directly to these spectral quantities. The large spectral ratios |σ2/σ1||\sigma_{2}/\sigma_{1}| observed in both case studies (10210^{2}–10510^{5}) confirm that the quasi-stationary approximation is accurate well within standard engineering time horizons, making the spectral approach a practical and transparent tool for resilience assessment.
CN · 翻译
对瞬态生成矩阵 𝑸T 的谱分析提供了额外洞见:主特征值 σ1 决定吸收的特征时间尺度;拟平稳分布 𝒒 描述倒塌前条件损伤状态的占用;渐近韧性 ℛ₀^∞ = 1 − w₀q₀ 把宏观韧性指标直接联系到这些谱量。两个算例中观测到的巨大谱比 |σ2/σ1|(10²–10⁵)证实:在标准工程时间尺度内拟平稳近似足够精确,谱方法成为实用而透明的韧性评估工具。
EN · 原文
Several extensions are natural within this setting. Time-varying generators driven by clustered seismicity models, such as the Epidemic-Type Aftershock Sequence (ETAS) model, would extend the formulation to non-homogeneous seismic input, particularly relevant for systems subject to rapid post-mainshock damage accumulation. The state-space representation also extends naturally to networked and portfolio-level assessments, enabling joint reliability and resilience quantification at the infrastructure scale. Future work will explore these directions, further consolidating the role of generalized PBEE formulations in supporting resilient design and management of civil infrastructure under realistic seismic hazard scenarios.
CN · 翻译
在该框架下自然延伸出若干方向:由丛集地震活动模型(如 ETAS)驱动的时变生成矩阵,可把公式推广到非齐次地震输入,尤其适合主震后快速损伤累积的系统;状态空间表示还可自然扩展到网络与组合层面的评估,实现基础设施尺度的可靠性与韧性联合量化。未来工作将探索这些方向,进一步巩固通用化 PBEE 公式在真实地震危险场景下支持韧性设计与管理的角色。

七、编者解读:这篇论文到底讲了什么(大白话版)

  1. 第一遍(10 分钟):只读摘要和术语表,回答三个问题——问题是什么?方法是什么?结果是什么?
  2. 第二遍(20 分钟):读引言 + 结论,重点体会「泊松假设错在哪」「可靠性与韧性为什么解耦」。
  3. 第三遍(30 分钟):读引言中「恢复建模的工程实践」(HAZUS、FEMA P-58、REDi、ATC-138),感受这篇论文在工程脉络里的位置;跳过所有公式,遇到矩阵、特征值等术语回查术语表。

八、给准大一的阅读路线图 & 延伸方向

📖 怎么读这篇论文(三遍法)

  1. 第一遍(10 分钟):只读摘要和术语表,回答三个问题——问题是什么?方法是什么?结果是什么?
  2. 第二遍(20 分钟):读引言 + 结论,重点体会「泊松假设错在哪」「可靠性与韧性为什么解耦」。
  3. 第三遍(30 分钟):读引言中「恢复建模的工程实践」(HAZUS、FEMA P-58、REDi、ATC-138),感受这篇论文在工程脉络里的位置;跳过所有公式,遇到矩阵、特征值等术语回查术语表。

🚀 这个方向你能延伸做什么

九、英文摘要朗读(练听力用)

先盲听一遍→再看对照稿→再听一遍。目标是听出核心概念(Poissonian、renewability、memorylessness、continuous-time Markov chain、generator matrix、spectral)和两个结构名词(braced frame、base-isolated system)。