Structural Safety最新文献_第7页

In Memoriam of Ove Dalager Ditlevsen 纪念爱·达拉格·迪特莱夫森

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-21 DOI: 10.1016/j.strusafe.2025.102585

Armen Der Kiureghian

引用次数: 0

Response probability distribution estimation of expensive computer simulators: A Bayesian active learning perspective using Gaussian process regression 昂贵计算机模拟器的响应概率分布估计：使用高斯过程回归的贝叶斯主动学习视角

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-19 DOI: 10.1016/j.strusafe.2025.102579

Chao Dang , Marcos A. Valdebenito , Nataly A. Manque , Jun Xu , Matthias G.R. Faes

Estimation of the response probability distributions of computer simulators subject to input random variables is a crucial task in many fields. However, achieving this task with guaranteed accuracy remains an open computational challenge, especially for expensive-to-evaluate computer simulators. In this work, a Bayesian active learning perspective is presented to address the challenge, which is based on the use of the Gaussian process (GP) regression. First, estimation of the response probability distributions is conceptually interpreted as a Bayesian inference problem, as opposed to frequentist inference. This interpretation provides several important benefits: (1) it quantifies and propagates discretization error probabilistically; (2) it incorporates prior knowledge of the computer simulator, and (3) it enables the effective reduction of numerical uncertainty in the solution to a prescribed level. The conceptual Bayesian idea is then realized by using the GP regression, where we derive the posterior statistics of the response probability distributions in semi-analytical form and also provide a numerical solution scheme. Based on the practical Bayesian approach, a Bayesian active learning (BAL) method is further proposed for estimating the response probability distributions. In this context, the key contribution lies in the development of two crucial components for active learning, i.e., stopping criterion and learning function, by taking advantage of the posterior statistics. It is empirically demonstrated by five numerical examples that the proposed BAL method can efficiently estimate the response probability distributions with desired accuracy.

在输入随机变量的情况下，计算机模拟器响应概率分布的估计是许多领域的一项重要任务。然而，在保证精度的情况下实现这一任务仍然是一个开放的计算挑战，特别是对于昂贵的评估计算机模拟器。在这项工作中，提出了一个基于高斯过程（GP）回归的贝叶斯主动学习视角来解决这一挑战。首先，响应概率分布的估计在概念上被解释为贝叶斯推理问题，而不是频率推理问题。这种解释提供了几个重要的好处：(1)它量化和传播离散误差的概率；(2)它结合了计算机模拟器的先验知识，(3)它能够有效地将解决方案中的数值不确定性降低到规定的水平。然后利用GP回归实现了概念贝叶斯思想，在GP回归中，我们以半解析形式导出了响应概率分布的后验统计量，并提供了数值解方案。在实际贝叶斯方法的基础上，进一步提出了一种估计响应概率分布的贝叶斯主动学习（BAL）方法。在这种情况下，关键的贡献在于主动学习的两个关键组成部分的发展，即停止准则和学习函数，利用后验统计的优势。通过5个数值算例的经验证明，所提出的BAL方法能够有效地估计出响应概率分布，并具有理想的精度。

{"title":"Response probability distribution estimation of expensive computer simulators: A Bayesian active learning perspective using Gaussian process regression","authors":"Chao Dang , Marcos A. Valdebenito , Nataly A. Manque , Jun Xu , Matthias G.R. Faes","doi":"10.1016/j.strusafe.2025.102579","DOIUrl":"10.1016/j.strusafe.2025.102579","url":null,"abstract":"<div><div>Estimation of the response probability distributions of computer simulators subject to input random variables is a crucial task in many fields. However, achieving this task with guaranteed accuracy remains an open computational challenge, especially for expensive-to-evaluate computer simulators. In this work, a Bayesian active learning perspective is presented to address the challenge, which is based on the use of the Gaussian process (GP) regression. First, estimation of the response probability distributions is conceptually interpreted as a Bayesian inference problem, as opposed to frequentist inference. This interpretation provides several important benefits: (1) it quantifies and propagates discretization error probabilistically; (2) it incorporates prior knowledge of the computer simulator, and (3) it enables the effective reduction of numerical uncertainty in the solution to a prescribed level. The conceptual Bayesian idea is then realized by using the GP regression, where we derive the posterior statistics of the response probability distributions in semi-analytical form and also provide a numerical solution scheme. Based on the practical Bayesian approach, a Bayesian active learning (BAL) method is further proposed for estimating the response probability distributions. In this context, the key contribution lies in the development of two crucial components for active learning, i.e., stopping criterion and learning function, by taking advantage of the posterior statistics. It is empirically demonstrated by five numerical examples that the proposed BAL method can efficiently estimate the response probability distributions with desired accuracy.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102579"},"PeriodicalIF":5.7,"publicationDate":"2025-02-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A novel simulation method for the multivariate non-stationary non-Gaussian wind speed based on KL expansion and translation process theory 基于KL展开和平移过程理论的多变量非平稳非高斯风速模拟新方法

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-18 DOI: 10.1016/j.strusafe.2025.102584

Fengbo Wu , Yu Wu , Ning Zhao

Accurate simulation of multivariate non-stationary non-Gaussian wind speed is the premise of evaluating the response of nonlinear structures. The methods based on Karhunen-Loève (KL) expansion and translation process method are extensively applied to predict non-stationary non-Gaussian simulation because it is easy for use and has relatively satisfactory simulation efficiency. However, these methods perform poorly in simulating the non-stationary strongly non-Gaussian process, especially the wind speed processes with highly skewed or bimodal distributions. This study comprehensively utilizes the KL expansion, the maximum entropy methods (MEM), and piecewise Hermite polynomial model (PHPM) to formulate a novel approach for simulating multivariate non-stationary non-Gaussian wind speed. In this method, the KL expansion is firstly used to generate the non-stationary Gaussian process. Then, a new strategy, the MEM is used to approximate the probability density function (PDF) of the target process which is then used to establish PHPM, is proposed to achieve the accurate and efficient simulation of non-stationary non-Gaussian process. The numerical results show that the proposed method has better simulation accuracy than traditional KL-based methods for non-stationary strongly non-Gaussian wind speed processes, especially the processes with highly skewed or bimodal distributions. Note that the proposed method can also be applied to simulate other non-Gaussian non-stationary excitations such as the wind pressure processes influenced by complex effects such as interference effect.

准确模拟多变量非平稳非高斯风速是评价非线性结构响应的前提。基于karhunen - lo (KL)展开法和平移过程法的非平稳非高斯模拟预测方法因其易于使用和具有比较满意的模拟效率而被广泛应用于非平稳非高斯模拟预测。然而，这些方法在模拟非平稳的强非高斯过程中表现不佳，特别是在高度偏态或双峰分布的风速过程中。综合运用KL展开、最大熵方法（MEM）和分段Hermite多项式模型（PHPM），提出了一种模拟多变量非平稳非高斯风速的新方法。该方法首先利用KL展开生成非平稳高斯过程。然后，提出了一种新的策略，即利用MEM近似目标过程的概率密度函数（PDF），然后利用该函数建立PHPM，以实现对非平稳非高斯过程的精确和高效模拟。数值结果表明，对于非平稳强非高斯风速过程，特别是具有高度偏态或双峰分布的风速过程，所提出的方法比传统的基于kl的方法具有更好的模拟精度。注意，所提出的方法也可以应用于模拟其他非高斯非平稳激励，如受干扰效应等复杂效应影响的风压过程。

{"title":"A novel simulation method for the multivariate non-stationary non-Gaussian wind speed based on KL expansion and translation process theory","authors":"Fengbo Wu , Yu Wu , Ning Zhao","doi":"10.1016/j.strusafe.2025.102584","DOIUrl":"10.1016/j.strusafe.2025.102584","url":null,"abstract":"<div><div>Accurate simulation of multivariate non-stationary non-Gaussian wind speed is the premise of evaluating the response of nonlinear structures. The methods based on Karhunen-Loève (KL) expansion and translation process method are extensively applied to predict non-stationary non-Gaussian simulation because it is easy for use and has relatively satisfactory simulation efficiency. However, these methods perform poorly in simulating the non-stationary strongly non-Gaussian process, especially the wind speed processes with highly skewed or bimodal distributions. This study comprehensively utilizes the KL expansion, the maximum entropy methods (MEM), and piecewise Hermite polynomial model (PHPM) to formulate a novel approach for simulating multivariate non-stationary non-Gaussian wind speed. In this method, the KL expansion is firstly used to generate the non-stationary Gaussian process. Then, a new strategy, the MEM is used to approximate the probability density function (PDF) of the target process which is then used to establish PHPM, is proposed to achieve the accurate and efficient simulation of non-stationary non-Gaussian process. The numerical results show that the proposed method has better simulation accuracy than traditional KL-based methods for non-stationary strongly non-Gaussian wind speed processes, especially the processes with highly skewed or bimodal distributions. Note that the proposed method can also be applied to simulate other non-Gaussian non-stationary excitations such as the wind pressure processes influenced by complex effects such as interference effect.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102584"},"PeriodicalIF":5.7,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143464184","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advanced terrain-adaptive tropical cyclone wind field modeling using deep learning for infrastructure resilience planning 基于深度学习的高级地形自适应热带气旋风场建模，用于基础设施弹性规划

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-18 DOI: 10.1016/j.strusafe.2025.102580

Yilin Shi , Naiyu Wang , Bruce R. Ellingwood

Tropical cyclones pose significant threats to the resilience of coastal communities, underscoring the need for reliable wind field models to support robust hazard analyses. Parametric wind models (PWMs), despite their computational efficiency, often fall short in capturing intricate wind-terrain interactions, leading to inaccurate resilience evaluations for spatially-distributed civil infrastructure systems situated in complex terrains. This study introduces an innovative approach that integrates the strengths of numerical wind models to handle intricate terrain features into PWMs through a deep learning-based Convolutional Neural Network for Terrain Modification (CNN-TM). The CNN-TM model, trained over 3 million km² of numerically simulated high-resolution wind fields, enhances terrain representation in PWMs by generating 450 m-resolution terrain-modified wind fields for both wind speed and direction. The accuracy and efficiency of this integration are validated across multiple scales: grid (∼0.2 km²), patch (∼506 km²), and region (∼34,000 km²). Applications during Typhoon Hagupit (2020) in Zhejiang Province, China, demonstrate its practical effectiveness across a 105,000 km² area. By leveraging deep learning to synergize numerical and parametric models, the CNN-TM model addresses limitations of traditional PWMs and provides a robust tool for resilience-oriented decision-making for infrastructure systems in coastal regions characterized by complex terrains.

热带气旋对沿海社区的恢复能力构成重大威胁，强调需要可靠的风场模型来支持强有力的危害分析。参数风模型（PWMs）尽管计算效率很高，但在捕捉复杂的风-地形相互作用方面往往不足，导致对位于复杂地形的空间分布式民用基础设施系统的弹性评估不准确。本研究介绍了一种创新的方法，通过基于深度学习的卷积神经网络地形改造（CNN-TM），将数值风模型处理复杂地形特征的优势整合到pwm中。CNN-TM模型训练了超过300万平方公里的数值模拟高分辨率风场，通过生成450米分辨率的风速和风向地形修正风场，增强了PWMs中的地形表征。这种整合的准确性和效率在多个尺度上得到了验证：网格（~ 0.2 km2）、斑块（~ 506 km2）和区域（~ 34,000 km2）。在中国浙江省台风黑格比（2020）期间的应用，在105,000平方公里的面积上证明了其实际有效性。通过利用深度学习来协同数值模型和参数模型，CNN-TM模型解决了传统PWMs的局限性，并为具有复杂地形特征的沿海地区基础设施系统的弹性导向决策提供了强大的工具。

{"title":"Advanced terrain-adaptive tropical cyclone wind field modeling using deep learning for infrastructure resilience planning","authors":"Yilin Shi , Naiyu Wang , Bruce R. Ellingwood","doi":"10.1016/j.strusafe.2025.102580","DOIUrl":"10.1016/j.strusafe.2025.102580","url":null,"abstract":"<div><div>Tropical cyclones pose significant threats to the resilience of coastal communities, underscoring the need for reliable wind field models to support robust hazard analyses. Parametric wind models (PWMs), despite their computational efficiency, often fall short in capturing intricate wind-terrain interactions, leading to inaccurate resilience evaluations for spatially-distributed civil infrastructure systems situated in complex terrains. This study introduces an innovative approach that integrates the strengths of numerical wind models to handle intricate terrain features into PWMs through a deep learning-based Convolutional Neural Network for Terrain Modification (CNN-TM). The CNN-TM model, trained over 3 million km<sup>2</sup> of numerically simulated high-resolution wind fields, enhances terrain representation in PWMs by generating 450 m-resolution terrain-modified wind fields for both wind speed and direction. The accuracy and efficiency of this integration are validated across multiple scales: grid (∼0.2 km<sup>2</sup>), patch (∼506 km<sup>2</sup>), and region (∼34,000 km<sup>2</sup>). Applications during Typhoon Hagupit (2020) in Zhejiang Province, China, demonstrate its practical effectiveness across a 105,000 km<sup>2</sup> area. By leveraging deep learning to synergize numerical and parametric models, the CNN-TM model addresses limitations of traditional PWMs and provides a robust tool for resilience-oriented decision-making for infrastructure systems in coastal regions characterized by complex terrains.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102580"},"PeriodicalIF":5.7,"publicationDate":"2025-02-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143480618","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Enhanced sequential directional importance sampling for structural reliability analysis 结构可靠性分析的增强顺序定向重要抽样

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-06 DOI: 10.1016/j.strusafe.2025.102574

Kai Cheng, Iason Papaioannou, Daniel Straub

Sequential directional importance sampling (SDIS) Kai Cheng et al. (2023) is an efficient adaptive simulation method for estimating failure probabilities. It expresses the failure probability as the product of a group of integrals that are easy to estimate, wherein the first one is estimated with Monte Carlo simulation (MCS), and all the subsequent ones are estimated with directional importance sampling. In this work, we propose an enhanced SDIS method for structural reliability analysis. We discuss the efficiency of MCS for estimating the first integral in standard SDIS and propose using Subset Simulation as an alternative method. Additionally, we propose a Kriging-based active learning algorithm tailored to identify multiple roots in certain important directions within a specificed search interval. The performance of the enhanced SDIS is demonstrated through various complex benchmark problems. The results show that the enhanced SDIS is a versatile reliability analysis method that can efficiently and robustly solve challenging reliability problems.

程凯等（2023）是一种有效的故障概率估计自适应仿真方法。它将失效概率表示为一组易于估计的积分的乘积，其中第一个积分用蒙特卡罗模拟（MCS）估计，随后的所有积分都用方向重要抽样估计。在这项工作中，我们提出了一种改进的SDIS方法进行结构可靠性分析。我们讨论了MCS在标准SDIS中估计第一个积分的效率，并提出使用子集模拟作为一种替代方法。此外，我们提出了一种基于kriging的主动学习算法，该算法可以在特定的搜索间隔内识别某些重要方向上的多个根。通过各种复杂的基准测试问题验证了增强的SDIS的性能。结果表明，改进的SDIS是一种通用的可靠性分析方法，能够有效、稳健地解决具有挑战性的可靠性问题。

引用次数: 0

Sensitivity of ship hull reliability considering geometric imperfections and residual stresses 考虑几何缺陷和残余应力的船体可靠性敏感性

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-05 DOI: 10.1016/j.strusafe.2025.102575

Aws Idris, Mohamed Soliman

Initial geometric imperfections and welding-induced residual stresses are inevitable consequences of ship fabrication and manufacturing processes. This paper quantifies the effect of these imperfections, as well as other input parameters, on the reliability of ship hull girders. The paper introduces a comprehensive variance-based sensitivity analysis approach, assisted by artificial neural networks, to characterize the key input parameters influencing the failure probability under different operational conditions. A total of 16 input parameters related to load and capacity quantification are considered in the simulation. The ultimate strength of the hull girder is quantified using a high-fidelity nonlinear finite element model that accounts for initial geometric imperfections and residual stresses. The vertical bending moments acting on the ship during its service life are quantified probabilistically. The results indicate that although it is essential to account for initial geometric imperfections to properly establish the ultimate hull capacity, the uncertainty in their magnitude has low effect on the reliability of the investigated hull. Accordingly, their magnitude can be considered deterministically in the probabilistic simulations. It was also found that the influence of various input parameters on the variability of the ship reliability depends on the considered operational condition.

初始几何缺陷和焊接残余应力是船舶制造过程中不可避免的结果。本文量化了这些缺陷以及其他输入参数对船体大梁可靠性的影响。本文介绍了一种基于方差的综合灵敏度分析方法，并辅以人工神经网络对不同工况下影响失效概率的关键输入参数进行表征。仿真中共考虑了16个与负荷和容量量化相关的输入参数。采用考虑初始几何缺陷和残余应力的高保真非线性有限元模型对船体梁的极限强度进行了量化。对舰船在使用寿命期间所受的垂直弯矩进行了概率量化。结果表明，虽然考虑初始几何缺陷是建立船体极限承载力的必要条件，但其大小的不确定性对船体可靠性的影响很小。因此，在概率模拟中，它们的大小可以被认为是确定性的。研究还发现，各种输入参数对舰船可靠性变异性的影响取决于所考虑的运行状态。

{"title":"Sensitivity of ship hull reliability considering geometric imperfections and residual stresses","authors":"Aws Idris, Mohamed Soliman","doi":"10.1016/j.strusafe.2025.102575","DOIUrl":"10.1016/j.strusafe.2025.102575","url":null,"abstract":"<div><div>Initial geometric imperfections and welding-induced residual stresses are inevitable consequences of ship fabrication and manufacturing processes. This paper quantifies the effect of these imperfections, as well as other input parameters, on the reliability of ship hull girders. The paper introduces a comprehensive variance-based sensitivity analysis approach, assisted by artificial neural networks, to characterize the key input parameters influencing the failure probability under different operational conditions. A total of 16 input parameters related to load and capacity quantification are considered in the simulation. The ultimate strength of the hull girder is quantified using a high-fidelity nonlinear finite element model that accounts for initial geometric imperfections and residual stresses. The vertical bending moments acting on the ship during its service life are quantified probabilistically. The results indicate that although it is essential to account for initial geometric imperfections to properly establish the ultimate hull capacity, the uncertainty in their magnitude has low effect on the reliability of the investigated hull. Accordingly, their magnitude can be considered deterministically in the probabilistic simulations. It was also found that the influence of various input parameters on the variability of the ship reliability depends on the considered operational condition.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102575"},"PeriodicalIF":5.7,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143418857","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Time-dependent reliability analysis for non-differentiable limit state functions due to discrete load processes 离散荷载过程下不可微极限状态函数的时变可靠性分析

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-05 DOI: 10.1016/j.strusafe.2025.102576

Hao-Peng Qiao , Zhao-Hui Lu , Chun-Qing Li , Chao-Huang Cai , Cao Wang

In time-dependent structural reliability analysis it is not always the case that the limit state functions are continuous and differentiable. A commonly encountered case is the structures subjected to instantaneous loads with high intensity, such as seismic actions. This paper intends to propose an efficient method for time-dependent reliability analysis of structures with non-differentiable limit state functions. A new formulation is proposed for outcrossing rate based on the minimum value of limit state functions at the time of outcrossing. The developed method also considers the degradation of structural resistance. It is found in the paper that the instantaneous discrete load with high intensity can induce non-differentiable limit state functions and that the probabilistic information of such instantaneous loads significantly affects the time-dependent failure probabilities. The paper concludes that the proposed method can predict the time-dependent failure probability of structures for non-differentiable limit state functions accurately and efficiently. The proposed method contributes to the body of knowledge of time-dependent reliability with wider practical applications.

在时变结构可靠度分析中，极限状态函数并不总是连续可微的。通常遇到的情况是结构承受高强度的瞬时载荷，例如地震作用。提出了一种具有不可微极限状态函数的结构时变可靠度分析的有效方法。根据异交时极限状态函数的最小值，提出了异交率的新公式。该方法还考虑了结构抗力的退化。本文发现高强度的瞬时离散荷载可以诱发不可微的极限状态函数，并且这种瞬时荷载的概率信息对随时间变化的失效概率有显著影响。结果表明，该方法能准确有效地预测结构不可微极限状态函数的时变失效概率。提出的方法有助于建立时变可靠性的知识体系，具有广泛的实际应用价值。

{"title":"Time-dependent reliability analysis for non-differentiable limit state functions due to discrete load processes","authors":"Hao-Peng Qiao , Zhao-Hui Lu , Chun-Qing Li , Chao-Huang Cai , Cao Wang","doi":"10.1016/j.strusafe.2025.102576","DOIUrl":"10.1016/j.strusafe.2025.102576","url":null,"abstract":"<div><div>In time-dependent structural reliability analysis it is not always the case that the limit state functions are continuous and differentiable. A commonly encountered case is the structures subjected to instantaneous loads with high intensity, such as seismic actions. This paper intends to propose an efficient method for time-dependent reliability analysis of structures with non-differentiable limit state functions. A new formulation is proposed for outcrossing rate based on the minimum value of limit state functions at the time of outcrossing. The developed method also considers the degradation of structural resistance. It is found in the paper that the instantaneous discrete load with high intensity can induce non-differentiable limit state functions and that the probabilistic information of such instantaneous loads significantly affects the time-dependent failure probabilities. The paper concludes that the proposed method can predict the time-dependent failure probability of structures for non-differentiable limit state functions accurately and efficiently. The proposed method contributes to the body of knowledge of time-dependent reliability with wider practical applications.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102576"},"PeriodicalIF":5.7,"publicationDate":"2025-02-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143349034","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sequential and adaptive probabilistic integration for structural reliability analysis 结构可靠性分析的顺序自适应概率集成

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-05 DOI: 10.1016/j.strusafe.2025.102577

Masaru Kitahara , Pengfei Wei

We propose the application of sequential and adaptive probabilistic integration (SAPI) to the estimation of the probability of failure in structural reliability. SAPI was originally developed to explore the posterior distribution and estimate its normalising constant in Bayesian model updating. The principle is to perform probabilistic integration on a sequence of distributions, moving from the prior to the posterior, to learn the normalising constant of each distribution. In structural reliability, SAPI can be used to sample an approximation of the optimal importance sampling (IS) density, and we present a particular choice of the intermediate distributions. The derived SAPI estimator is thus an IS estimator of the thought probability. The numerical uncertainty is propagated using random process sampling, and the induced posterior statistics are used to design a Bayesian active learning strategy. Four numerical examples demonstrate that SAPI outperforms other state-of-the-art active learning reliability methods using sequential Monte Carlo samplers.

提出了序列自适应概率积分法（SAPI）在结构可靠性失效概率估计中的应用。SAPI最初是为了在贝叶斯模型更新中探索后验分布并估计其归一化常数而开发的。其原理是对一系列分布进行概率积分，从先验到后验，以学习每个分布的归一化常数。在结构可靠性中，SAPI可用于抽样最优重要抽样（IS）密度的近似值，并给出了中间分布的特定选择。因此，导出的SAPI估计量是思想概率的is估计量。采用随机过程抽样方法传播数值不确定性，并利用诱导后验统计量设计贝叶斯主动学习策略。四个数值示例表明，SAPI优于使用顺序蒙特卡罗采样器的其他最先进的主动学习可靠性方法。

引用次数: 0

Multi-output stochastic emulation with applications to seismic response correlation estimation 多输出随机仿真及其在地震响应相关估计中的应用

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2025-02-04 DOI: 10.1016/j.strusafe.2025.102578

Sang-ri Yi , Alexandros A. Taflanidis

Stochastic emulation techniques represent a specialized surrogate modeling branch that is appropriate for applications for which the relationship between input and output is stochastic in nature. Their objective is to address the stochastic uncertainty sources by directly predicting the output distribution for a given input. An example of such application, and the focus of this contribution, is the estimation of structural response (engineering demand parameter) distribution in seismic risk assessment. In this case, the stochastic uncertainty originates from the aleatoric variability in the seismic hazard description. Note that this is a different uncertainty-source than the potential parametric uncertainty associated with structural characteristics or explanatory variables for the seismic hazard (for example, intensity measures), that are treated as the parametric input in surrogate modeling context. The key challenge in stochastic emulation pertains to addressing heteroscedasticity in the output variability. Relevant approaches to-date for addressing this challenge have focused on scalar outputs. In contrast, this paper focuses on the multi-output stochastic emulation problem and presents a methodology for predicting the output correlation matrix, while fully addressing heteroscedastic characteristics. This is achieved by introducing a Gaussian Process (GP) regression model for approximating the components of the correlation matrix, and coupling this approximation with a correction step to guarantee positive definite properties for the resultant predictions. For obtaining the observation data to inform the GP calibration, different approaches are examined, relying-or-not on the existence of replicated samples for the response output. Such samples require that, for a portion of the training points, simulations are repeated for the same inputs and different descriptions of the stochastic uncertainty. This information can be readily used to obtain observation for the response statistics (correlation or covariance in this instance) to inform the GP development. An alternative approach is to use as observations noisy covariance samples based on the sample deviations from a primitive mean approximation. These different observation variants lead to different GP variants that are compared within a comprehensive case study. A computational framework for integrating the correlation matrix approximation within the stochastic emulation for the marginal distribution approximation of each output component is also discussed, to provide the joint response distribution approximation.

随机仿真技术代表了一种专门的代理建模分支，适用于输入和输出之间的关系在本质上是随机的应用程序。他们的目标是通过直接预测给定输入的输出分布来解决随机不确定性源。这种应用的一个例子是地震风险评估中结构响应（工程需求参数）分布的估计。在这种情况下，随机不确定性来源于地震灾害描述中的任意变率。请注意，这与与地震危险的结构特征或解释变量（例如，强度测量）相关的潜在参数不确定性是不同的不确定性来源，后者在替代建模上下文中被视为参数输入。随机仿真的关键挑战在于如何处理输出变异性中的异方差。迄今为止应对这一挑战的相关方法主要集中在标量输出上。相比之下，本文主要研究多输出随机仿真问题，并提出了一种预测输出相关矩阵的方法，同时充分解决了异方差特性。这是通过引入高斯过程（GP）回归模型来逼近相关矩阵的组成部分，并将此近似与校正步骤相结合，以保证结果预测的正定性质来实现的。为了获得观测数据以通知GP校准，检查了不同的方法，是否依赖于响应输出的复制样本的存在。这样的样本要求，对于一部分训练点，对相同的输入和不同的随机不确定性描述进行重复模拟。该信息可以很容易地用于获得响应统计数据的观察结果（在本例中为相关性或协方差），以通知GP开发。另一种方法是使用基于样本偏离原始均值近似的噪声协方差样本作为观测值。这些不同的观测变量导致不同的GP变量，在一个全面的案例研究中进行比较。本文还讨论了在随机仿真中对各输出分量的边际分布近似积分相关矩阵近似的计算框架，以提供联合响应分布近似。

{"title":"Multi-output stochastic emulation with applications to seismic response correlation estimation","authors":"Sang-ri Yi , Alexandros A. Taflanidis","doi":"10.1016/j.strusafe.2025.102578","DOIUrl":"10.1016/j.strusafe.2025.102578","url":null,"abstract":"<div><div>Stochastic emulation techniques represent a specialized surrogate modeling branch that is appropriate for applications for which the relationship between input and output is stochastic in nature. Their objective is to address the stochastic uncertainty sources by directly predicting the output distribution for a given input. An example of such application, and the focus of this contribution, is the estimation of structural response (engineering demand parameter) distribution in seismic risk assessment. In this case, the stochastic uncertainty originates from the aleatoric variability in the seismic hazard description. Note that this is a different uncertainty-source than the potential parametric uncertainty associated with structural characteristics or explanatory variables for the seismic hazard (for example, intensity measures), that are treated as the parametric input in surrogate modeling context. The key challenge in stochastic emulation pertains to addressing heteroscedasticity in the output variability. Relevant approaches to-date for addressing this challenge have focused on scalar outputs. In contrast, this paper focuses on the multi-output stochastic emulation problem and presents a methodology for predicting the output correlation matrix, while fully addressing heteroscedastic characteristics. This is achieved by introducing a Gaussian Process (GP) regression model for approximating the components of the correlation matrix, and coupling this approximation with a correction step to guarantee positive definite properties for the resultant predictions. For obtaining the observation data to inform the GP calibration, different approaches are examined, relying-or-not on the existence of replicated samples for the response output. Such samples require that, for a portion of the training points, simulations are repeated for the same inputs and different descriptions of the stochastic uncertainty. This information can be readily used to obtain observation for the response statistics (correlation or covariance in this instance) to inform the GP development. An alternative approach is to use as observations noisy covariance samples based on the sample deviations from a primitive mean approximation. These different observation variants lead to different GP variants that are compared within a comprehensive case study. A computational framework for integrating the correlation matrix approximation within the stochastic emulation for the marginal distribution approximation of each output component is also discussed, to provide the joint response distribution approximation.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"115 ","pages":"Article 102578"},"PeriodicalIF":5.7,"publicationDate":"2025-02-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143838094","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Serviceability limit state target reliability for concrete structures 混凝土结构的使用能力极限状态目标可靠性

IF 5.7 1区工程技术 Q1 ENGINEERING, CIVIL

Structural Safety

Pub Date : 2024-12-30 DOI: 10.1016/j.strusafe.2024.102572

Andrew Way , Frederik Bakker , Dirk Proske , Celeste Viljoen

The balance between safety and economy, referred to as target reliability, forms the basis of modern structural design. Target reliability is determined by economic minimisation, either directly through generic cost optimisation or by back calibration to existing practice. However, the currently codified annual target reliability indices for serviceability limit state (SLS), differ by as much as

Δ β = 1.6

between those from generic cost optimisation and back calibration. Various assumptions are made in the generic cost optimisation which may not be appropriate to determine SLS target reliability. Target reliability from back calibration is likely to be closer to actual SLS failure rates, however, no literature exists which details the process or rationale by which the back calibration was performed. It is therefore uncertain if either of these methods produce cost optimal SLS target reliability. This research aims to evaluate currently codified SLS target reliability for cost optimality. SLS failure costs from existing research and engineering practice are used with an amended cost optimisation procedure which overcomes the deficiencies identified in the generic formulation to specifically determine SLS target reliability. The amended cost optimisation also considers parameter variation and decision parameter form for typical SLS cases. Results indicate that overall, the target reliability indices for annual irreversible SLS from back calibration to existing practice (

β = 2.9

) represents the range of considered SLS cases (

2.5 \leq β \leq 3.3

) well, whereas those from generic cost optimisation are notably lower (

1.3 \leq β \leq 2.3

). In some cases, target reliability varied sufficiently from

2.9

to warrant adjustments being made for better cost optimality.

安全与经济之间的平衡，即目标可靠度，是现代结构设计的基础。目标可靠性由经济最小化决定，要么直接通过一般成本优化，要么通过对现有实践的反向校准。然而，目前编制的可用性极限状态（SLS）年度目标可靠性指标在通用成本优化和反向校准之间的差异高达Δβ=1.6。在通用成本优化中做出的各种假设可能不适合确定SLS目标的可靠性。反向校准的目标可靠性可能更接近实际的SLS故障率，然而，没有文献详细说明进行反向校准的过程或基本原理。因此，这两种方法是否产生成本最优的SLS目标可靠性是不确定的。本研究旨在评估目前已编纂的SLS目标的成本最优可靠性。从现有的研究和工程实践中得出的SLS失效成本与修正的成本优化程序一起使用，该程序克服了通用配方中确定的缺陷，具体确定了SLS目标的可靠性。修正后的成本优化还考虑了典型SLS案例的参数变化和决策参数形式。结果表明，总体而言，从反校准到现有实践的年度不可逆SLS目标可靠性指数（β=2.9）很好地代表了考虑的SLS情况范围（2.5≤β≤3.3），而通用成本优化的目标可靠性指数明显较低（1.3≤β≤2.3）。在某些情况下，目标可靠性在2.9之间变化很大，需要进行调整以获得更优的成本。

{"title":"Serviceability limit state target reliability for concrete structures","authors":"Andrew Way , Frederik Bakker , Dirk Proske , Celeste Viljoen","doi":"10.1016/j.strusafe.2024.102572","DOIUrl":"10.1016/j.strusafe.2024.102572","url":null,"abstract":"<div><div>The balance between safety and economy, referred to as target reliability, forms the basis of modern structural design. Target reliability is determined by economic minimisation, either directly through generic cost optimisation or by back calibration to existing practice. However, the currently codified annual target reliability indices for serviceability limit state (SLS), differ by as much as <span><math><mrow><mi>Δ</mi><mi>β</mi><mo>=</mo><mn>1.6</mn></mrow></math></span> between those from generic cost optimisation and back calibration. Various assumptions are made in the generic cost optimisation which may not be appropriate to determine SLS target reliability. Target reliability from back calibration is likely to be closer to actual SLS failure rates, however, no literature exists which details the process or rationale by which the back calibration was performed. It is therefore uncertain if either of these methods produce cost optimal SLS target reliability. This research aims to evaluate currently codified SLS target reliability for cost optimality. SLS failure costs from existing research and engineering practice are used with an amended cost optimisation procedure which overcomes the deficiencies identified in the generic formulation to specifically determine SLS target reliability. The amended cost optimisation also considers parameter variation and decision parameter form for typical SLS cases. Results indicate that overall, the target reliability indices for annual irreversible SLS from back calibration to existing practice (<span><math><mrow><mi>β</mi><mo>=</mo><mn>2.9</mn></mrow></math></span>) represents the range of considered SLS cases (<span><math><mrow><mn>2.5</mn><mo>≤</mo><mi>β</mi><mo>≤</mo><mn>3.3</mn></mrow></math></span>) well, whereas those from generic cost optimisation are notably lower (<span><math><mrow><mn>1.3</mn><mo>≤</mo><mi>β</mi><mo>≤</mo><mn>2.3</mn></mrow></math></span>). In some cases, target reliability varied sufficiently from <span><math><mrow><mn>2.9</mn></mrow></math></span> to warrant adjustments being made for better cost optimality.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"114 ","pages":"Article 102572"},"PeriodicalIF":5.7,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143151076","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0