Details
| Original language | English |
|---|---|
| Article number | 118752 |
| Number of pages | 18 |
| Journal | Engineering structures |
| Volume | 318 |
| Early online date | 13 Aug 2024 |
| Publication status | Published - 1 Nov 2024 |
Abstract
This paper proposes an innovative model updating technique that thoroughly considers the interrelations among multivariate output features. The approach involves developing a novel uncertainty quantification metric, termed Sub-Convex Similarity. A specialised data preprocessing operator is proposed to reveal the inherent distributional attributes of multivariate datasets through a sequencing pre-processing. To manage the inherent randomness associated with sample location dispersion, we propose a binning algorithm based on the equal-bin-datapoints principle. This method allows for the quantification of multidimensional stochastic data without the need to calculate the joint probability distribution function. Utilising convex hull theory, sub-regional boundaries are established within each bin to reveal multivariate dataset characteristics. Sub-Convex Similarity serves as a metric for quantifying both interval-based and stochastic uncertainties, measuring discrepancies between simulated and experimental datasets in the context of both interval and stochastic model updating. The proposed model updating framework employs the sparrow search algorithm, a swarm intelligence-based optimization mechanism. The effectiveness and broad applicability of this approach are demonstrated through case studies involving a three-degree-of-freedom mass-spring system and a finite element model of a satellite, addressing multivariate uncertainties.
Keywords
- Convex hull, Model updating, Sub-convex similarity, Sub-interval similarity, Uncertainty quantification
ASJC Scopus subject areas
- Engineering(all)
- Civil and Structural Engineering
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Engineering structures, Vol. 318, 118752, 01.11.2024.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - A sub-convex similarity-based model updating method considering multivariate uncertainties
AU - Zhao, Yanlin
AU - Sun, Bing
AU - Bi, Sifeng
AU - Beer, Michael
AU - Moens, David
N1 - Publisher Copyright: © 2024 Elsevier Ltd
PY - 2024/11/1
Y1 - 2024/11/1
N2 - This paper proposes an innovative model updating technique that thoroughly considers the interrelations among multivariate output features. The approach involves developing a novel uncertainty quantification metric, termed Sub-Convex Similarity. A specialised data preprocessing operator is proposed to reveal the inherent distributional attributes of multivariate datasets through a sequencing pre-processing. To manage the inherent randomness associated with sample location dispersion, we propose a binning algorithm based on the equal-bin-datapoints principle. This method allows for the quantification of multidimensional stochastic data without the need to calculate the joint probability distribution function. Utilising convex hull theory, sub-regional boundaries are established within each bin to reveal multivariate dataset characteristics. Sub-Convex Similarity serves as a metric for quantifying both interval-based and stochastic uncertainties, measuring discrepancies between simulated and experimental datasets in the context of both interval and stochastic model updating. The proposed model updating framework employs the sparrow search algorithm, a swarm intelligence-based optimization mechanism. The effectiveness and broad applicability of this approach are demonstrated through case studies involving a three-degree-of-freedom mass-spring system and a finite element model of a satellite, addressing multivariate uncertainties.
AB - This paper proposes an innovative model updating technique that thoroughly considers the interrelations among multivariate output features. The approach involves developing a novel uncertainty quantification metric, termed Sub-Convex Similarity. A specialised data preprocessing operator is proposed to reveal the inherent distributional attributes of multivariate datasets through a sequencing pre-processing. To manage the inherent randomness associated with sample location dispersion, we propose a binning algorithm based on the equal-bin-datapoints principle. This method allows for the quantification of multidimensional stochastic data without the need to calculate the joint probability distribution function. Utilising convex hull theory, sub-regional boundaries are established within each bin to reveal multivariate dataset characteristics. Sub-Convex Similarity serves as a metric for quantifying both interval-based and stochastic uncertainties, measuring discrepancies between simulated and experimental datasets in the context of both interval and stochastic model updating. The proposed model updating framework employs the sparrow search algorithm, a swarm intelligence-based optimization mechanism. The effectiveness and broad applicability of this approach are demonstrated through case studies involving a three-degree-of-freedom mass-spring system and a finite element model of a satellite, addressing multivariate uncertainties.
KW - Convex hull
KW - Model updating
KW - Sub-convex similarity
KW - Sub-interval similarity
KW - Uncertainty quantification
UR - http://www.scopus.com/inward/record.url?scp=85200996357&partnerID=8YFLogxK
U2 - 10.1016/j.engstruct.2024.118752
DO - 10.1016/j.engstruct.2024.118752
M3 - Article
AN - SCOPUS:85200996357
VL - 318
JO - Engineering structures
JF - Engineering structures
SN - 0141-0296
M1 - 118752
ER -