Efficient reliability analysis of slopes in spatially variable soils with active learning-assisted bootstrap polynomial chaos expansion

Research output: Contribution to journalArticleResearchpeer review

Authors

  • Kang Liao
  • Xiaoyan Zhao
  • Yiping Wu
  • Fasheng Miao
  • Yutao Pan
  • Michael Beer

Research Organisations

External Research Organisations

  • Southwest Jiaotong University
  • China University of Geosciences
  • Norwegian University of Science and Technology (NTNU)
View graph of relations

Details

Original languageEnglish
Article number107022
Number of pages11
JournalComputers and geotechnics
Volume179
Early online date28 Dec 2024
Publication statusPublished - Mar 2025

Abstract

Evaluating the reliability of slopes with spatial variability is a challenging issue, especially when the failure probably of the target event is at a low level, because of unaffordable computational cost required in such cases. In this context, an adaptive surrogate model-based approach, namely active learning-assisted bootstrap polynomial chaos expansion, is proposed to alleviate the above computational burden. The proposed approach extends the traditional polynomial chaos expansion by introducing the bootstrap resampling method so that it can deal with reliability issues smoothly and provide a feasible configuration environment to support the active learning algorithm. The computational efficiency can thus be greatly improved by adaptively searching for the most informative samples to train the surrogate model through iterative program. Two spatially varying soil slopes are studied to illustrate the validity of the active learning-assisted bootstrap polynomial chaos expansion. The results show that the proposed approach has superior advantages in terms of efficiency and accuracy, and it is also suitable for handling problems with complex parameter configurations, including high dimensionality and cross-correlation. Besides, the proposed approach has potential in addressing geotechnical engineering problems with low probability levels.

Keywords

    Active learning algorithm, Bootstrap polynomial chaos expansion, Spatial variability, Surrogate model

ASJC Scopus subject areas

Cite this

Efficient reliability analysis of slopes in spatially variable soils with active learning-assisted bootstrap polynomial chaos expansion. / Liao, Kang; Zhao, Xiaoyan; Wu, Yiping et al.
In: Computers and geotechnics, Vol. 179, 107022, 03.2025.

Research output: Contribution to journalArticleResearchpeer review

Liao K, Zhao X, Wu Y, Miao F, Pan Y, Beer M. Efficient reliability analysis of slopes in spatially variable soils with active learning-assisted bootstrap polynomial chaos expansion. Computers and geotechnics. 2025 Mar;179:107022. Epub 2024 Dec 28. doi: 10.1016/j.compgeo.2024.107022
Download
@article{ee49f082e4a74de3a83a5aef54dfefdc,
title = "Efficient reliability analysis of slopes in spatially variable soils with active learning-assisted bootstrap polynomial chaos expansion",
abstract = "Evaluating the reliability of slopes with spatial variability is a challenging issue, especially when the failure probably of the target event is at a low level, because of unaffordable computational cost required in such cases. In this context, an adaptive surrogate model-based approach, namely active learning-assisted bootstrap polynomial chaos expansion, is proposed to alleviate the above computational burden. The proposed approach extends the traditional polynomial chaos expansion by introducing the bootstrap resampling method so that it can deal with reliability issues smoothly and provide a feasible configuration environment to support the active learning algorithm. The computational efficiency can thus be greatly improved by adaptively searching for the most informative samples to train the surrogate model through iterative program. Two spatially varying soil slopes are studied to illustrate the validity of the active learning-assisted bootstrap polynomial chaos expansion. The results show that the proposed approach has superior advantages in terms of efficiency and accuracy, and it is also suitable for handling problems with complex parameter configurations, including high dimensionality and cross-correlation. Besides, the proposed approach has potential in addressing geotechnical engineering problems with low probability levels.",
keywords = "Active learning algorithm, Bootstrap polynomial chaos expansion, Spatial variability, Surrogate model",
author = "Kang Liao and Xiaoyan Zhao and Yiping Wu and Fasheng Miao and Yutao Pan and Michael Beer",
note = "Publisher Copyright: {\textcopyright} 2024 Elsevier Ltd",
year = "2025",
month = mar,
doi = "10.1016/j.compgeo.2024.107022",
language = "English",
volume = "179",
journal = "Computers and geotechnics",
issn = "0266-352X",
publisher = "Elsevier BV",

}

Download

TY - JOUR

T1 - Efficient reliability analysis of slopes in spatially variable soils with active learning-assisted bootstrap polynomial chaos expansion

AU - Liao, Kang

AU - Zhao, Xiaoyan

AU - Wu, Yiping

AU - Miao, Fasheng

AU - Pan, Yutao

AU - Beer, Michael

N1 - Publisher Copyright: © 2024 Elsevier Ltd

PY - 2025/3

Y1 - 2025/3

N2 - Evaluating the reliability of slopes with spatial variability is a challenging issue, especially when the failure probably of the target event is at a low level, because of unaffordable computational cost required in such cases. In this context, an adaptive surrogate model-based approach, namely active learning-assisted bootstrap polynomial chaos expansion, is proposed to alleviate the above computational burden. The proposed approach extends the traditional polynomial chaos expansion by introducing the bootstrap resampling method so that it can deal with reliability issues smoothly and provide a feasible configuration environment to support the active learning algorithm. The computational efficiency can thus be greatly improved by adaptively searching for the most informative samples to train the surrogate model through iterative program. Two spatially varying soil slopes are studied to illustrate the validity of the active learning-assisted bootstrap polynomial chaos expansion. The results show that the proposed approach has superior advantages in terms of efficiency and accuracy, and it is also suitable for handling problems with complex parameter configurations, including high dimensionality and cross-correlation. Besides, the proposed approach has potential in addressing geotechnical engineering problems with low probability levels.

AB - Evaluating the reliability of slopes with spatial variability is a challenging issue, especially when the failure probably of the target event is at a low level, because of unaffordable computational cost required in such cases. In this context, an adaptive surrogate model-based approach, namely active learning-assisted bootstrap polynomial chaos expansion, is proposed to alleviate the above computational burden. The proposed approach extends the traditional polynomial chaos expansion by introducing the bootstrap resampling method so that it can deal with reliability issues smoothly and provide a feasible configuration environment to support the active learning algorithm. The computational efficiency can thus be greatly improved by adaptively searching for the most informative samples to train the surrogate model through iterative program. Two spatially varying soil slopes are studied to illustrate the validity of the active learning-assisted bootstrap polynomial chaos expansion. The results show that the proposed approach has superior advantages in terms of efficiency and accuracy, and it is also suitable for handling problems with complex parameter configurations, including high dimensionality and cross-correlation. Besides, the proposed approach has potential in addressing geotechnical engineering problems with low probability levels.

KW - Active learning algorithm

KW - Bootstrap polynomial chaos expansion

KW - Spatial variability

KW - Surrogate model

UR - http://www.scopus.com/inward/record.url?scp=85213277938&partnerID=8YFLogxK

U2 - 10.1016/j.compgeo.2024.107022

DO - 10.1016/j.compgeo.2024.107022

M3 - Article

AN - SCOPUS:85213277938

VL - 179

JO - Computers and geotechnics

JF - Computers and geotechnics

SN - 0266-352X

M1 - 107022

ER -

By the same author(s)