Automatic generation of high-dimensional surrogate models using moment quadrature design points

Xuefei Guan

doi:10.1088/3050-2454/ae5966

您当前的位置：

首页 >

文章列表页 >

Automatic generation of high-dimensional surrogate models using moment quadrature design points

Papers | 更新时间：2026-05-08

- Automatic generation of high-dimensional surrogate models using moment quadrature design points
- 利用矩求积设计点自动生成高维替代模型
- Journal of Reliability Science and Engineering Pages: 1-25(2026)
- 作者机构：
  
  Graduate School of China Academy of Engineering Physics, Beijing 100193, People's Republic of China
- 作者简介：
  
  [ "Xuefei Guan is currently a Professor at Graduate School of China Academy of Engineering Physics. His research interests include ultrasonic nondestructive evaluation, fatigue and fracture, and uncertainty quantification methods applied to reliability engineering." ]
- 基金信息：
- DOI：10.1088/3050-2454/ae5966
  CLC：
- Received：20 November 2025，
  
  Revised：2026-02-26，
  
  Accepted：31 March 2026，
  
  Online First：10 April 2026，
  
  Published：2026-06
- 稿件说明：
移动端阅览
Xuefei Guan. 利用矩求积设计点自动生成高维替代模型[J]. 可靠性科学与工程学报（英文）, 2026, 2: 025002.

Xuefei Guan. Automatic generation of high-dimensional surrogate models using moment quadrature design points[J]. Journal of Reliability Science and Engineering, 2026, 2: 025002.
Xuefei Guan. 利用矩求积设计点自动生成高维替代模型[J]. 可靠性科学与工程学报（英文）, 2026, 2: 025002. DOI： 10.1088/3050-2454/ae5966.

Xuefei Guan. Automatic generation of high-dimensional surrogate models using moment quadrature design points[J]. Journal of Reliability Science and Engineering, 2026, 2: 025002. DOI： 10.1088/3050-2454/ae5966.

摘要

本研究提出了一种基于矩求积设计点的确定且明确的代理模型构建方法。与自适应或随机采样方法不同，该框架直接从输入概率密度函数的统计矩中推导出设计点，确保了可重复性、最优多项式精度，以及对涉及任意分布（包括非均匀分布，原则上也包括经验分布）的低维和高维问题的广泛适用性。该方法具备用于基生成、系数回归以及可选项剪枝的自动化流程，避免了依情况而定的参数调整。此外，它允许直接基于同一组模型评估结果来获取不确定性度量和全局灵敏度指数。大量的数值算例，包括标准基准测试函数以及涉及多达100个随机变量的工程应用，表明与传统代理建模方法相比，该方法具有高精度和高计算效率。

Abstract

This study proposes a deterministic and definitive methodology for surrogate model construction based on moment quadrature design points. Different from adaptive or randomly sampling approaches

the proposed framework derives design points directly from the statistical moments of the input probability density functions

ensuring reproducibility

optimal polynomial exactness

and broad applicability to low- and high-dimensional problems involving arbitrary distributions

including nonuniform distributions and

in principle

empirical ones. It features an automatic pipeline for basis generation

coefficient regression

and optional term pruning

avoiding case-dependent tuning. Moreover

it allows direct evaluation of uncertainty metrics and global sensitivity indices to be obtained using the same set of model evaluations. Extensive numerical examples

including standard benchmark functions and engineering applications involving up to 100 random variables

demonstrate that the proposed method achieves high accuracy and computational efficiency compared with conventional surrogate modeling approaches.

关键词

Keywords

references

Cui L-X , Du Y-M and Sun C-P 2025 On quantum reliability characterizing systematic errors in quantum sensing J. Reliab. Sci. Eng. 1 015004

Liu Z , Lesselier D , Sudret B and Wiart J 2020 Surrogate modeling based on resampled polynomial chaos expansions Reliab. Eng. Syst. Saf. 202 107008

Weng Y-Y , Liu T , Zhang X-Y and Zhao Y-G 2025 Probability density estimation of polynomial chaos and its application in structural reliability analysis Reliab. Eng. Syst. Saf. 253 110537

Shang X , Wang L , Fang H , Lu L and Zhang Z 2024 Active learning of ensemble polynomial chaos expansion method for global sensitivity analysis Reliab. Eng. Syst. Saf. 249 110226

Kontolati K , Loukrezis D , Dos Santos K R , Giovanis D G and Shields M D 2022 Manifold learning-based polynomial chaos expansions for high-dimensional surrogate models Int. J. Uncertainty Quantif. 12 39 – 64

Liu J and Jiang C 2023 Surrogate modeling for high dimensional uncertainty propagation via deep kernel polynomial chaos expansion Appl. Math. Modell. 122 167 – 86

Sharma H , Novák L and Shields M 2024 Physics-constrained polynomial chaos expansion for scientific machine learning and uncertainty quantification Comput. Methods Appl. Mech. Eng. 431 117314

Mueller J N , Sargsyan K , Daniels C J and Najm H N 2025 Polynomial chaos surrogate construction for random fields with parametric uncertainty SIAM/ASA J. Uncertain. Quantif. 13 1 – 29

Zong C , Shi M , Li Q , Xue T , Song X , Li X an d Chen D 2023 Sealing design optimization of nuclear pressure relief valves based on the polynomial chaos expansion surrogate model Nucl. Eng. Technol. 55 1382 – 99

Zhu X , Broccardo M and Sudret B 2023 Seismic fragility analysis using stochastic polynomial chaos expansions Probab. Eng. Mech. 72 103413

Zuhal L R , Zakaria K , Palar P S , Shimoyama K and Liem R P 2021 Polynomial-chaos–kriging with gradient information for surrogate modeling in aerodynamic design AIAA J. 59 2950 – 67

Marinescu M 2024 Explaining and connecting Kriging with Gaussian process regression (arXiv:2408.02331)

Xu Y , Renteria A and Wang P 2022 Adaptive surrogate models with partially observed information Reliab. Eng. Syst. Saf. 225 108566

Marrel A and Iooss B 2024 Probabilistic surrogate modeling by Gaussian process: a new estimation algorithm for more robust prediction Reliab. Eng. Syst. Saf. 247 110120

Abaei M M , Leira B J , Sævik S and BahooToroody A 2024 Integrating physics-based simulations with gaussian processes for enhanced safety assessment of offshore installations Reliab. Eng. Syst. Saf. 249 110235

Wu J and Wan L 2024 Reliability sensitivity analysis for RBSMC: a high-efficiency multiple response Gaussian process model Reliab. Eng. Syst. Saf. 243 109812

Tran V N and Kim J 2022 Robust and efficient uncertainty quantification for extreme events that deviate significantly from the training dataset using polynomial chaos-kriging J. Hydrol. 609 127716

Lüthen N , Marelli S and Sudret B 2023 A spectral surrogate model for stochastic simulators computed from trajectory samples Comput. Methods Appl. Mech. Eng. 406 115875

Teng D , Feng Y-W , Lu C , Liu J-Q and Chen J-Y 2024 Vectorial generative adversarial surrogate modeling reliability evaluation framework for engineering structural systems Reliab. Eng. Syst. Saf. 247 110076

Wang R-Z , Gu H-H , Liu Y , Miura H , Zhang X-C and Tu S-T 2023 Surrogate-modeling-assisted creep-fatigue reliability assessment in a low-pressure turbine disc considering multi-source uncertainty Reliab. Eng. Syst. Saf. 240 109550

Lu C , Teng D , Chen J-Y , Fei C-W and Keshtegar B 2023 Adaptive vectorial surrogate modeling framework for multi-objective reliability estimation Reliab. Eng. Syst. Saf. 234 109148

Xu Y , Kohtz S , Boakye J , Gardoni P and Wang P 2023 Physics-informed machine learning for reliability and systems safety applications: state of the art and challenges Reliab. Eng. Syst. Saf. 230 108900

Fuhg J N , Fau A and Nackenhorst U 2021 State-of-the-art and comparative review of adaptive sampling methods for Kriging Arch. Comput. Methods Eng. 28 2689 – 747

Lataniotis C , Marelli S and Sudret B 2018 Extending Classical Surrogate Modelling to Ultrahigh Dimensional Problems through Supervised Dimensionality Reduction: A Data-driven Approach ( ETH Zurich, Chair of Risk, Safety and Uncertainty Quantification )

Zhao L , Wang P , Song B , Wang X and Dong H 2020 An efficient Kriging modeling method for high-dimensional design problems based on maximal information coefficient Struct. Multidiscipl. Optim. 61 39 – 57

Cheng G H , Younis A , Haji Hajikolaei K and Gary Wang G 2015 Trust region based mode pursuing sampling method for global optimization of high dimensional design problems J. Mech. Des. 137 021407

Bai F , Zou D and Wei Y 2024 A surrogate-assisted evolutionary algorithm with clustering-based sampling for high-dimensional expensive blackbox optimization J. Glob. Optim. 89 93 – 115

Zhao Y and Liu J 2025 An efficient solution method for inverse problems with high-dimensional model uncertainty parameters Struct. Multidiscipl. Optim. 68 18

Shan S and Wang G G 2010 Survey of modeling and optimization strategies to solve high-dimensional design problems with computationally-expensive black-box functions Struct. Multidiscipl. Optim. 41 219 – 41

Bouhlel M A , Bartoli N , Otsmane A and Morlier J 2016 Improving Kriging surrogates of high-dimensional design models by partial least squares dimension reduction Struct. Multidiscipl. Optim. 53 935 – 52

Chu H , Li W , Sun D , Liu Z , Zheng W , Xu J and Cheng Q 2025 AK-MCS structural reliability analysis method based on improved active learning function J. Reliab. Sci. Eng. 1 025003

Cao L and Zhao Y 2024 Uncertainty quantification for structural response field with ultra-high dimensions Int. J. Mech. Sci. 271 109110

Wu S , Hu J , Cao L , Huang K , Ma Y and Tang J 2026 Non-intrusive smoothed stochastic finite element method with multi-source uncertainties Int. J. Mech. Sci. 311 111163

Wang T , Li J , Lu D , Dong Y , Tan Y and Li Z 2024 A point mapping strategy-based sparse grid integration method for statistical moments estimation and structural reliability analysis Comput. Methods Appl. Mech. Eng. 430 117238

Shen S , Cheng J , Liu Z , Tan J and Zhang D 2025 Bayesian inference-assisted reliability analysis framework for robotic motion systems in future factories Reliab. Eng. Syst. Saf. 258 110894

Chen Y , Lu Z , Guo Y , Li X and Wu X 2025 Bayesian updating based method with dimension reduction integral technique for efficiently estimating multi-mode system failure probability function Comput. Struct. 316 107859

Guan X 2023 Moment quadrature method for uncertainty quantification of fatigue damage prognosis Int. J. Fatigue 172 107654

Guan X 2023 A moment quadrature method for uncertainty quantification of three-dimensional crack propagation via extremely few model runs Comput. Methods Appl. Mech. Eng. 411 116061

Guan X 2024 Sparse moment quadrature for uncertainty modeling and quantification Reliab. Eng. Syst. Saf. 241 109665

Gong T , He J and Guan X 2026 Continuously nested moment quadrature for uncertainty quantification of black-box models Probab. Eng. Mech. 82 103892

Mysovskikh I P 1968 On the construction of cubature formulas with the smallest number of nodes Doklady Akademii Nauk ( Russian Academy of Sciences )

Golub G H and Welsch J H 1969 Calculation of Gauss quadrature rules Math. Comput. 23 221 – 30

Sanchez-Dehesa J 1978 The spectrum of Jacobi matrices in terms of its associated weight function J. Comput. Appl. Math. 4 275 – 83

De Branges L 1959 The Stone-Weierstrass theorem Proc. Am. Math. Soc. 10 822 – 4

He J , Gao S and Gong J 2014 A sparse grid stochastic collocation method for structural reliability analysis Struct. Saf. 51 29 – 34

Li P-P , Lu Z-H and Zhao Y-G 2022 An effective and efficient method for structural reliability considering the distributional parametric uncertainty Appl. Math. Modell. 106 507 – 23

Zech J and Schwab C 2020 Convergence rates of high dimensional Smolyak quadrature ESAIM Math. Model. Numer. Anal. 54 1259 – 307

Guan X 2025 The moment quadrature method for global sensitivity analysis Reliab. Eng. Syst. Saf. 265 111460

Saves P , Lafage R , Bartoli N , Diouane Y , Bussemaker J , Lefebvre T , Hwang J T , Morlier J and Martins J R 2024 SMT 2.0: a surrogate modeling toolbox with a focus on hierarchical and mixed variables Gaussian processes Adv. Eng. Softw. 188 103571

Barzilai J and Borwein J M 1988 Two-point step size gradient methods IMA J. Numer. Anal. 8 141 – 8

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Accelerating subset simulation with surrogate model ensuring unbiased failure probability estimation

Related Author

Yuanzhuo Ma

Chuang Li

Binbin Li

Related Institution

School of Renewable Energy, Hohai University

School of Electrical and Power Engineering, Hohai University

State Key Laboratory of Biobased Transportation Fuel Technology, ZJU-UIUC Institute, Zhejiang University

⁰