Combining Numerical Simulations, Artificial Intelligence and Intelligent Sampling Algorithms to Build Surrogate Models and Calculate the Probability of Failure of Urban Tunnels

Vinícius Resende Domingues, Luan Carlos de Sena Monteiro Ozelim, André Pacheco de Assis and André Luís Brasil Cavalcante
Additional contact information
Vinícius Resende Domingues: Department of Civil and Environmental Engineering, University of Brasilia, Brasilia 70910-900, Brazil
Luan Carlos de Sena Monteiro Ozelim: Department of Civil and Environmental Engineering, University of Brasilia, Brasilia 70910-900, Brazil
André Pacheco de Assis: Department of Civil and Environmental Engineering, University of Brasilia, Brasilia 70910-900, Brazil
André Luís Brasil Cavalcante: Department of Civil and Environmental Engineering, University of Brasilia, Brasilia 70910-900, Brazil

Sustainability, 2022, vol. 14, issue 11, 1-29

Abstract: When it is necessary to evaluate, with a probabilistic approach, the interaction of urban tunnels with neighboring structures, computational power is an important challenge for numerical models. Thus, intelligent sampling algorithms can be allies in obtaining a better knowledge of the result’s domain, even if in possession of a smaller number of samples. In any case, when sampling is limited, the evaluation of the risks is also restricted. In this context, artificial intelligence (AI) can fill an important gap in risk analysis by interpolating results and generating larger samples quickly. The goal of the AI algorithm is to find an approximation function (also called a surrogate model) that reproduces the original numerical simulation behavior and can be evaluated much faster. This function is constructed by performing multiple simulations at special points obtained by intelligent sampling techniques. This paper used a hypothetical case to validate the methodological proposal. It concerns the sequential excavation of a tunnel, about three diameters deep, interacting with a seven-story building. First, the three-dimensional numerical model (FEM) was solved deterministically, and then its domain and mesh were refined. After that, another 170 solutions were numerically obtained from FEM software, strategically sampling the random variables involved. Sequentially, based on 31 artificial intelligence techniques, it was evaluated which variables were of greatest importance in predicting the magnitude of vertical displacement in the foundation elements of a surrounding building. Then, once the most important variables were selected, the 31 artificial intelligence techniques were again trained and tested to define the one with the least R-squared. Finally, using this best-fit algorithm, it was possible to calculate the probability of failure using massive samples, with sizes on the order of 10 7 . These samples were used to illustrate the convergence of the Simple Monte Carlo Sampling (MC) and Latin Hypercube Sampling (LHS). The main contribution of this paper is methodological; therefore, this new procedure can be aggregated to state-of-the-art risk assessment methodologies in tunnel-related problems.

Keywords: numerical simulations; artificial intelligence; intelligent sampling algorithms; surrogate models; probability of failure (search for similar items in EconPapers)
JEL-codes: O13 Q Q0 Q2 Q3 Q5 Q56 (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (1)

Downloads: (external link)
https://www.mdpi.com/2071-1050/14/11/6385/pdf (application/pdf)
https://www.mdpi.com/2071-1050/14/11/6385/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jsusta:v:14:y:2022:i:11:p:6385-:d:822448

Access Statistics for this article

Sustainability is currently edited by Ms. Alexandra Wu

More articles in Sustainability from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().