Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2025.103734
Luis Sánchez , Andrew J. Simpkin , Norma Bargary
A methodology is proposed to solve stochastic differential equations (SDEs) using B-spline chaos and Kalman Filters. The Fokker–Planck equation is employed to model physical processes involving nonlinear structures, non-Gaussian distributions, and multimodal distributions. B-spline chaos is used to approximate the solution of the SDE, while state estimation is achieved using Ensemble and Unscented Kalman Filter algorithms. To validate the approach, a time series of temperature and humidity data collected from manufacturing sensors is used, demonstrating the accuracy of the methodology in reconstructing the true system states. The proposed method is specifically designed for solving SDEs related to known physical processes. Additionally, numerical comparisons with existing approaches are presented to highlight the advantages in terms of performance and accuracy.
{"title":"B-splines chaos and Kalman Filters for solving a stochastic differential equation","authors":"Luis Sánchez , Andrew J. Simpkin , Norma Bargary","doi":"10.1016/j.probengmech.2025.103734","DOIUrl":"10.1016/j.probengmech.2025.103734","url":null,"abstract":"<div><div>A methodology is proposed to solve stochastic differential equations (SDEs) using B-spline chaos and Kalman Filters. The Fokker–Planck equation is employed to model physical processes involving nonlinear structures, non-Gaussian distributions, and multimodal distributions. B-spline chaos is used to approximate the solution of the SDE, while state estimation is achieved using Ensemble and Unscented Kalman Filter algorithms. To validate the approach, a time series of temperature and humidity data collected from manufacturing sensors is used, demonstrating the accuracy of the methodology in reconstructing the true system states. The proposed method is specifically designed for solving SDEs related to known physical processes. Additionally, numerical comparisons with existing approaches are presented to highlight the advantages in terms of performance and accuracy.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103734"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143146060","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2025.103733
Weiyan Liu, Xunru Yin, Zhongjin Guo, Shan Jiang
A stochastic stabilization control strategy is proposed for multi-degree-of-freedom (MDOF) quasi integrable and non-resonant Hamiltonian systems under combined Gaussian and Poisson white noise excitation. At first, the averaged Itô stochastic differential equations (SDEs) for controlled first integrals are derived from the system motion equations by using the stochastic averaging method. Second, the dynamical programming equation of the averaged system with undetermined cost function is established based on the dynamical programming principle. The optimal control law is given through solving the dynamical programming equation. Third, the asymptotic Lyapunov stability with probability one of the controlled system is analyzed approximately by evaluating the largest Lyapunov exponent of the averaged system. Finally, the cost function and optimal control forces are determined based on the requirement of stabilizing the system. An example is delivered to illustrate the application and effectiveness of the proposed stochastic stabilization control strategy.
{"title":"Stochastic stabilization of quasi integrable and non-resonant Hamiltonian systems under combined Gaussian and Poisson white noise excitation","authors":"Weiyan Liu, Xunru Yin, Zhongjin Guo, Shan Jiang","doi":"10.1016/j.probengmech.2025.103733","DOIUrl":"10.1016/j.probengmech.2025.103733","url":null,"abstract":"<div><div>A stochastic stabilization control strategy is proposed for multi-degree-of-freedom (MDOF) quasi integrable and non-resonant Hamiltonian systems under combined Gaussian and Poisson white noise excitation. At first, the averaged Itô stochastic differential equations (SDEs) for controlled first integrals are derived from the system motion equations by using the stochastic averaging method. Second, the dynamical programming equation of the averaged system with undetermined cost function is established based on the dynamical programming principle. The optimal control law is given through solving the dynamical programming equation. Third, the asymptotic Lyapunov stability with probability one of the controlled system is analyzed approximately by evaluating the largest Lyapunov exponent of the averaged system. Finally, the cost function and optimal control forces are determined based on the requirement of stabilizing the system. An example is delivered to illustrate the application and effectiveness of the proposed stochastic stabilization control strategy.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103733"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147056","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2025.103731
Ziyu Tao , Duo Zhang , Desi Tu , Lingshan He , Chao Zou
Train-induced vibrations are of increasing concern as railway tracks and buildings are located closer to each other. In this study, a field measurement campaign was carried out where train-induced accelerations were monitored at the ground surface. Despite the uniformity of the train and track, field measurements unveil discrepancies in train-induced ground-borne vibrations at the given observation point. It underscores the need for a probabilistic evaluation of train-induced environmental vibrations. This study quantifies the uncertainties in train-induced ground-borne vibration transmissions and achieves a probabilistic assessment of the amplification phenomenon during ground-borne vibration transmissions. A transfer function-based model accounting for the train-induced ground-borne vibration transmission is developed and applied to the case study to justify its validity. Based on the validated prediction model, Monte Carlo method is subsequently adopted for quantifying uncertainties in train-induced ground-borne vibration transmissions caused by the spatially varied soil properties and by the variation in train speed, separately. Variations in soil properties are simulated using log-normally distributed random fields, while the train speed is simulated using a uniformly distributed random variable. The amplitude ratio between a pair of observation points is used to characterize ground-borne vibration transmissions. Burr distribution fitting is applied to the calculated samples of amplitude ratios, and during this process, the null hypothesis test is not rejected at the 5% significance level. The proposed methodology enables the determination of the amplification probability during train-induced ground-borne vibration transmissions. In addition, it aids reliability analysis and site screening in assessing building serviceability.
{"title":"Prediction of train-induced ground-borne vibration transmission considering parametric uncertainties","authors":"Ziyu Tao , Duo Zhang , Desi Tu , Lingshan He , Chao Zou","doi":"10.1016/j.probengmech.2025.103731","DOIUrl":"10.1016/j.probengmech.2025.103731","url":null,"abstract":"<div><div>Train-induced vibrations are of increasing concern as railway tracks and buildings are located closer to each other. In this study, a field measurement campaign was carried out where train-induced accelerations were monitored at the ground surface. Despite the uniformity of the train and track, field measurements unveil discrepancies in train-induced ground-borne vibrations at the given observation point. It underscores the need for a probabilistic evaluation of train-induced environmental vibrations. This study quantifies the uncertainties in train-induced ground-borne vibration transmissions and achieves a probabilistic assessment of the amplification phenomenon during ground-borne vibration transmissions. A transfer function-based model accounting for the train-induced ground-borne vibration transmission is developed and applied to the case study to justify its validity. Based on the validated prediction model, Monte Carlo method is subsequently adopted for quantifying uncertainties in train-induced ground-borne vibration transmissions caused by the spatially varied soil properties and by the variation in train speed, separately. Variations in soil properties are simulated using log-normally distributed random fields, while the train speed is simulated using a uniformly distributed random variable. The amplitude ratio between a pair of observation points is used to characterize ground-borne vibration transmissions. Burr distribution fitting is applied to the calculated samples of amplitude ratios, and during this process, the null hypothesis test is not rejected at the 5% significance level. The proposed methodology enables the determination of the amplification probability during train-induced ground-borne vibration transmissions. In addition, it aids reliability analysis and site screening in assessing building serviceability.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103731"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Porous hydrostatic gas bearing (PHGB) utilizes porous materials as restrictors and is widely recognized in mechanical equipment and scientific instruments due to their exceptional stability and load capacity. At present, the design of PHGB relies on deterministic models to calculate bearing capacity and stiffness, and the adjustment of parameters such as air supply pressure and bearing clearance mainly depends on experience. However, uncertainties related to compressor performance, material properties, and manufacturing errors are inevitably introduced in the practical applications, which can significantly affect the design performance of PHGBs. To address these challenges, this paper presents a global sensitivity analysis to identify the sensitive factors causing variations in the mechanical properties of PHGBs. First, a PHGB model is developed based on the Darcy and continuity equations, and its predictive accuracy for bearing characteristics is validated. Subsequently, a global sensitivity analysis method employing sparse polynomial chaos expansion is introduced to quantitatively assess the impact of uncertainties such as supply pressure, bearing length, diameter, clearance, and eccentricity on load capacity and mass flow rate. This analysis identifies the most critical uncertain parameters influencing the mechanical performance of PHGBs. The insights gained from this study will enable designers to comprehensively understand the mechanical performance of bearings under uncertainty while reducing computational costs, thus providing a valuable theoretical foundation for PHGB analysis and design.
{"title":"Global sensitivity analysis of design variables for porous hydrostatic gas bearings considering uncertainty","authors":"Yihua Wu , Lixiong Cao , Jiachang Tang , Mingqi Tian","doi":"10.1016/j.probengmech.2024.103722","DOIUrl":"10.1016/j.probengmech.2024.103722","url":null,"abstract":"<div><div>Porous hydrostatic gas bearing (PHGB) utilizes porous materials as restrictors and is widely recognized in mechanical equipment and scientific instruments due to their exceptional stability and load capacity. At present, the design of PHGB relies on deterministic models to calculate bearing capacity and stiffness, and the adjustment of parameters such as air supply pressure and bearing clearance mainly depends on experience. However, uncertainties related to compressor performance, material properties, and manufacturing errors are inevitably introduced in the practical applications, which can significantly affect the design performance of PHGBs. To address these challenges, this paper presents a global sensitivity analysis to identify the sensitive factors causing variations in the mechanical properties of PHGBs. First, a PHGB model is developed based on the Darcy and continuity equations, and its predictive accuracy for bearing characteristics is validated. Subsequently, a global sensitivity analysis method employing sparse polynomial chaos expansion is introduced to quantitatively assess the impact of uncertainties such as supply pressure, bearing length, diameter, clearance, and eccentricity on load capacity and mass flow rate. This analysis identifies the most critical uncertain parameters influencing the mechanical performance of PHGBs. The insights gained from this study will enable designers to comprehensively understand the mechanical performance of bearings under uncertainty while reducing computational costs, thus providing a valuable theoretical foundation for PHGB analysis and design.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103722"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147077","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2024.103720
Zixin Liu , Zhangjun Liu , Xinxin Ruan , Bohang Xu
The novel dimension-reduction methods based on spectral representation for simulating both univariate and multivariate non-stationary stochastic processes are addressed. Initially, the original spectral (decomposition) representation of univariate (multivariate) non-stationary stochastic processes is derived. A unified expression for the original spectral decomposition integrating Cholesky decomposition and eigen decomposition is presented. Further, two typical random orthogonal functions associated with the standard orthogonal random variables in the original spectral (decomposition) representation are defined, resulting in the conventional Monte Carlo methods, namely the random amplitude method and the random phase method, are readily obtained. Meanwhile, two updated random orthogonal functions are introduced, enabling the dimension-reduction methods based on the random amplitude and the random phase respectively. The analysis above establishes that the original spectral (decomposition) representation forms the unified theoretical foundation for both the conventional Monte Carlo and the developed dimension-reduction methods. Essentially, both approaches are specific cases of the original spectral form. However, they share the same necessary conditions for simulation, though their sufficient conditions differ. Consequently, the dimension-reduction methods just require merely one to three elementary random variables for simulating univariate and multivariate non-stationary stochastic processes, significantly reducing the randomness of simulation from the level of thousands to an extremely low degree. Finally, numerical examples including the comparisons of simulation accuracy and efficiency between the conventional Monte Carlo methods and the dimension-reduction methods thoroughly substantiate the effectiveness and superiority of the latter.
{"title":"Advances in dimension-reduction methods for simulating univariate and multivariate non-stationary stochastic processes via spectral representation","authors":"Zixin Liu , Zhangjun Liu , Xinxin Ruan , Bohang Xu","doi":"10.1016/j.probengmech.2024.103720","DOIUrl":"10.1016/j.probengmech.2024.103720","url":null,"abstract":"<div><div>The novel dimension-reduction methods based on spectral representation for simulating both univariate and multivariate non-stationary stochastic processes are addressed. Initially, the original spectral (decomposition) representation of univariate (multivariate) non-stationary stochastic processes is derived. A unified expression for the original spectral decomposition integrating Cholesky decomposition and eigen decomposition is presented. Further, two typical random orthogonal functions associated with the standard orthogonal random variables in the original spectral (decomposition) representation are defined, resulting in the conventional Monte Carlo methods, namely the random amplitude method and the random phase method, are readily obtained. Meanwhile, two updated random orthogonal functions are introduced, enabling the dimension-reduction methods based on the random amplitude and the random phase respectively. The analysis above establishes that the original spectral (decomposition) representation forms the unified theoretical foundation for both the conventional Monte Carlo and the developed dimension-reduction methods. Essentially, both approaches are specific cases of the original spectral form. However, they share the same necessary conditions for simulation, though their sufficient conditions differ. Consequently, the dimension-reduction methods just require merely one to three elementary random variables for simulating univariate and multivariate non-stationary stochastic processes, significantly reducing the randomness of simulation from the level of thousands to an extremely low degree. Finally, numerical examples including the comparisons of simulation accuracy and efficiency between the conventional Monte Carlo methods and the dimension-reduction methods thoroughly substantiate the effectiveness and superiority of the latter.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103720"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147078","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2024.103705
Minhyeok Ko, Konstantinos G. Papakonstantinou
This paper presents in detail the originally developed Quadratic Point Estimate Method (QPEM), aimed at efficiently and accurately computing the first four output moments of probabilistic distributions, using sample (or sigma) points, with , the number of input random variables. The proposed QPEM particularly offers an effective, superior, and practical alternative to existing sampling and quadrature methods for low- and moderately-high-dimensional problems. Detailed theoretical derivations are provided proving that the proposed method can achieve a fifth or higher-order accuracy for symmetric input distributions. Various numerical examples, from simple polynomial functions to nonlinear finite element analyses with random field representations, support the theoretical findings and further showcase the validity, efficiency, and applicability of the QPEM, from low- to high-dimensional problems.
{"title":"Quadratic point estimate method for probabilistic moments computation","authors":"Minhyeok Ko, Konstantinos G. Papakonstantinou","doi":"10.1016/j.probengmech.2024.103705","DOIUrl":"10.1016/j.probengmech.2024.103705","url":null,"abstract":"<div><div>This paper presents in detail the originally developed Quadratic Point Estimate Method (QPEM), aimed at efficiently and accurately computing the first four output moments of probabilistic distributions, using <span><math><mrow><mn>2</mn><msup><mrow><mi>n</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>+</mo><mn>1</mn></mrow></math></span> sample (or sigma) points, with <span><math><mi>n</mi></math></span>, the number of input random variables. The proposed QPEM particularly offers an effective, superior, and practical alternative to existing sampling and quadrature methods for low- and moderately-high-dimensional problems. Detailed theoretical derivations are provided proving that the proposed method can achieve a fifth or higher-order accuracy for symmetric input distributions. Various numerical examples, from simple polynomial functions to nonlinear finite element analyses with random field representations, support the theoretical findings and further showcase the validity, efficiency, and applicability of the QPEM, from low- to high-dimensional problems.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103705"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147080","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2025.103729
Hugh J. Kinnear, F.A. DiazDelaO
Subset Simulation is a Markov chain Monte Carlo method used to compute small failure probabilities in structural reliability problems. This is done by iteratively sampling from nested subsets in the input space of a performance function, i.e. a function describing the behaviour of a physical system. When the performance function has features such as multimodality or rapidly changing output, it is not uncommon for Subset Simulation to suffer from ergodicity problems. To address these problems, this paper proposes a new framework that enhances Subset Simulation with niching, a concept from the field of evolutionary multimodal optimisation. Niching subset simulation dynamically partitions the input space using support vector machines, and recursively begins anew in each set of the partition. A new niching technique, which uses community detection methods and is specifically designed for high-dimensional problems, is also introduced. It is shown that Niching Subset Simulation is robust against ergodicty problems and can also offer additional insight into the topology of challenging reliability problems.
{"title":"Niching subset simulation","authors":"Hugh J. Kinnear, F.A. DiazDelaO","doi":"10.1016/j.probengmech.2025.103729","DOIUrl":"10.1016/j.probengmech.2025.103729","url":null,"abstract":"<div><div>Subset Simulation is a Markov chain Monte Carlo method used to compute small failure probabilities in structural reliability problems. This is done by iteratively sampling from nested subsets in the input space of a performance function, i.e. a function describing the behaviour of a physical system. When the performance function has features such as multimodality or rapidly changing output, it is not uncommon for Subset Simulation to suffer from ergodicity problems. To address these problems, this paper proposes a new framework that enhances Subset Simulation with niching, a concept from the field of evolutionary multimodal optimisation. Niching subset simulation dynamically partitions the input space using support vector machines, and recursively begins anew in each set of the partition. A new niching technique, which uses community detection methods and is specifically designed for high-dimensional problems, is also introduced. It is shown that Niching Subset Simulation is robust against ergodicty problems and can also offer additional insight into the topology of challenging reliability problems.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103729"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143146062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In engineering practice, the parametric uncertainty and correlation may coexist in the powertrain mounting system (PMS). An effective robust-based design optimization approach is proposed for uncertain PMS based on full vehicle model, where both the parametric uncertainty and correlation are considered. The uncertain parameters of PMS are firstly treated as probabilistic variables, and the Unscented Transformation Inspired (UTI) transformation is introduced to quantify the correlation of uncertain parameters. Then, to perform the uncertainty and correlation analysis, the UTI-Monte Carlo (UMC) method is developed based on UTI transformation and Monte Carlo sampling to estimate the means, standard deviations, variation ranges and correlation coefficients of PMS responses. Meanwhile, an efficient method named UTI-Arbitrary Polynomial Chaos Expansion (UAPCE) method is derived for the uncertainty and correlation analysis of PMS responses by combining UTI transformation and arbitrary polynomial chaos expansion. Next, an optimization model considering parametric uncertainty and correlation is formulated to perform the robust-based design of PMS, in which the weight coefficients of optimization components are calculated by principal component analysis. Finally, the numerical example is investigated to verify the effectiveness of the proposed methods.
{"title":"Robust-based design optimization of powertrain mounting system based on full vehicle model involving parametric uncertainty and correlation","authors":"Hui Lü , Jiaming Zhang , Xiaoting Huang , Wen-Bin Shangguan","doi":"10.1016/j.probengmech.2024.103726","DOIUrl":"10.1016/j.probengmech.2024.103726","url":null,"abstract":"<div><div>In engineering practice, the parametric uncertainty and correlation may coexist in the powertrain mounting system (PMS). An effective robust-based design optimization approach is proposed for uncertain PMS based on full vehicle model, where both the parametric uncertainty and correlation are considered. The uncertain parameters of PMS are firstly treated as probabilistic variables, and the Unscented Transformation Inspired (UTI) transformation is introduced to quantify the correlation of uncertain parameters. Then, to perform the uncertainty and correlation analysis, the UTI-Monte Carlo (UMC) method is developed based on UTI transformation and Monte Carlo sampling to estimate the means, standard deviations, variation ranges and correlation coefficients of PMS responses. Meanwhile, an efficient method named UTI-Arbitrary Polynomial Chaos Expansion (UAPCE) method is derived for the uncertainty and correlation analysis of PMS responses by combining UTI transformation and arbitrary polynomial chaos expansion. Next, an optimization model considering parametric uncertainty and correlation is formulated to perform the robust-based design of PMS, in which the weight coefficients of optimization components are calculated by principal component analysis. Finally, the numerical example is investigated to verify the effectiveness of the proposed methods.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103726"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2025.103737
Cuong Pham Van Le , Sawekchai Tangaramvong , Van Thu Huynh , Bach Do , Quang-Viet Vu , Wei Gao
The traditional approach to parameter identification of constitutive material models involves an iterative trial-and-error process that requires numerous costly finite element (FE) analyses as solvers for the corresponding discretized forward problems. This computational burden is particularly severe for composite structures like concrete-filled steel tube (CFST) columns with ultra-high-strength concrete. This paper presents an effective machine learning-based method to automate the identification of constitutive parameters for circular CFST (CCFST) short columns. A global sensitivity analysis leveraging a Gaussian process regression (GPR) model investigates the influence of each parameter on the column response, thereby providing the most influential parameters. Then, Bayesian optimization (BO) combined with a comprehensive learning particle swarm optimization (CLPSO) finds an optimal set of these parameters as the solution to a deterministic inverse problem formulated for the columns. Specifically, the CLPSO maximizes a highly non-convex expected improvement acquisition function formulated in each iteration of BO. By calibrating the numerical simulations of 14 CCFST columns against their experimental tests, BO delivers accurate parameters while avoiding the need for extensive FE analyses. The identified parameters reliably predict the responses of CCFST columns, demonstrating the accuracy and efficiency of the proposed identification method.
{"title":"Combined surrogate-assisted Bayesian optimization and comprehensive learning PSO method for parameter identification of ultra-high strength circular CFST short columns","authors":"Cuong Pham Van Le , Sawekchai Tangaramvong , Van Thu Huynh , Bach Do , Quang-Viet Vu , Wei Gao","doi":"10.1016/j.probengmech.2025.103737","DOIUrl":"10.1016/j.probengmech.2025.103737","url":null,"abstract":"<div><div>The traditional approach to parameter identification of constitutive material models involves an iterative trial-and-error process that requires numerous costly finite element (FE) analyses as solvers for the corresponding discretized forward problems. This computational burden is particularly severe for composite structures like concrete-filled steel tube (CFST) columns with ultra-high-strength concrete. This paper presents an effective machine learning-based method to automate the identification of constitutive parameters for circular CFST (CCFST) short columns. A global sensitivity analysis leveraging a Gaussian process regression (GPR) model investigates the influence of each parameter on the column response, thereby providing the most influential parameters. Then, Bayesian optimization (BO) combined with a comprehensive learning particle swarm optimization (CLPSO) finds an optimal set of these parameters as the solution to a deterministic inverse problem formulated for the columns. Specifically, the CLPSO maximizes a highly non-convex expected improvement acquisition function formulated in each iteration of BO. By calibrating the numerical simulations of 14 CCFST columns against their experimental tests, BO delivers accurate parameters while avoiding the need for extensive FE analyses. The identified parameters reliably predict the responses of CCFST columns, demonstrating the accuracy and efficiency of the proposed identification method.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103737"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143194890","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-01-01DOI: 10.1016/j.probengmech.2024.103725
Xiang Yu , Zhuxin Li , Yuke Wang , Rui Pang , Xiaolong Lv , Meng Fu
Dams are inevitably built on a deep overburden constrained by site conditions. Moreover, the spatial variability of the geomechanical parameter in overburden tends to significantly affect the mechanical state of the dam foundation anti-seepage structure. In this study, random field theory was combined with finite element analysis to consider the spatial variability of geotechnical parameters in the overburden. Gaussian autocorrelation function and spectral representation were used in random field simulations followed by stochastic finite element calculation. Subsequently, the deformation modulus in Duncan-Chang E-B model was selected as a random parameter in combination with an engineering example. The tensile stress at the top of the anti-seepage structure, horizontal displacement at the top of the cutoff wall, and compressive stress of the cutoff wall were analyzed using statistical laws and the probability distribution tests of the mean value, maximum value, exceedance probability, and 95% confidence interval limit. The results show that when the spatial variability of geomechanical parameter in overburden is not considered, the stress and deformation of the anti-seepage structure are underestimated. Probability distribution statistics of the anti-seepage structure were different from those of geomechanical parameters. The horizontal displacement at the top of the cutoff wall demonstrated a stronger sensitivity to the coefficient of variation than to correlation distance. Therefore, numerical simulations considering the spatial variability of geomechanical parameter in overburden can reasonably reflect the stress and deformation of anti-seepage structure.
{"title":"Stress–deformation analysis of concrete anti-seepage structure in earth-rock dam on overburden considering spatial variability of geomechanical parameter","authors":"Xiang Yu , Zhuxin Li , Yuke Wang , Rui Pang , Xiaolong Lv , Meng Fu","doi":"10.1016/j.probengmech.2024.103725","DOIUrl":"10.1016/j.probengmech.2024.103725","url":null,"abstract":"<div><div>Dams are inevitably built on a deep overburden constrained by site conditions. Moreover, the spatial variability of the geomechanical parameter in overburden tends to significantly affect the mechanical state of the dam foundation anti-seepage structure. In this study, random field theory was combined with finite element analysis to consider the spatial variability of geotechnical parameters in the overburden. Gaussian autocorrelation function and spectral representation were used in random field simulations followed by stochastic finite element calculation. Subsequently, the deformation modulus in Duncan-Chang E-B model was selected as a random parameter in combination with an engineering example. The tensile stress at the top of the anti-seepage structure, horizontal displacement at the top of the cutoff wall, and compressive stress of the cutoff wall were analyzed using statistical laws and the probability distribution tests of the mean value, maximum value, exceedance probability, and 95% confidence interval limit. The results show that when the spatial variability of geomechanical parameter in overburden is not considered, the stress and deformation of the anti-seepage structure are underestimated. Probability distribution statistics of the anti-seepage structure were different from those of geomechanical parameters. The horizontal displacement at the top of the cutoff wall demonstrated a stronger sensitivity to the coefficient of variation than to correlation distance. Therefore, numerical simulations considering the spatial variability of geomechanical parameter in overburden can reasonably reflect the stress and deformation of anti-seepage structure.</div></div>","PeriodicalId":54583,"journal":{"name":"Probabilistic Engineering Mechanics","volume":"79 ","pages":"Article 103725"},"PeriodicalIF":3.0,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143147079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号