Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media

In this study, we address the poroelasticity problem in heterogeneous media, which involves a coupled system of equations for fluid pressures and displacements. This problem is crucial in geomechanics for modeling the interaction between fluid flow and deformation in porous media, with applications...

תיאור מלא

שמור ב:

מידע ביבליוגרפי
Главные авторы:	Tyrylgin, A. A., Тырылгин, А. А.
פורמט:	Статья
שפה:	English
יצא לאור:	Pleiades Publishing 2025
נושאים:	Basis functions Online GMsFEM Darcy equation Displacement Heterogeneous media Machine learning Multiscale methods Poroelasticity
גישה מקוונת:	https://dspace.ncfu.ru/handle/123456789/30359
תגים:	הוספת תג אין תגיות, היה/י הראשונ/ה לתייג את הרשומה!

id	ir-123456789-30359
record_format	dspace
spelling	ir-123456789-303592025-04-02T11:17:59Z Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media Tyrylgin, A. A. Тырылгин, А. А. Basis functions Online GMsFEM Darcy equation Displacement Heterogeneous media Machine learning Multiscale methods Poroelasticity In this study, we address the poroelasticity problem in heterogeneous media, which involves a coupled system of equations for fluid pressures and displacements. This problem is crucial in geomechanics for modeling the interaction between fluid flow and deformation in porous media, with applications spanning oil and gas reservoirs, groundwater systems, and geothermal energy production. We introduce an innovative approach by integrating machine learning techniques to train online multiscale basis functions, enhancing the Generalized Multiscale Finite Element Method (GMsFEM). This methodology allows for an adaptive and efficient representation of both macroscopic and local heterogeneities in the system, significantly reducing computational costs. The offline multiscale basis functions are precomputed using local spectral problems, while the online basis functions are dynamically updated using machine learning models trained on local residual data. This approach ensures rapid error reduction and robust convergence, leveraging the computational efficiency of machine learning. We demonstrate the effectiveness of this method through numerical experiments, showcasing its potential in advancing the simulation and modeling of poroelasticity problems in heterogeneous media. 2025-04-02T11:16:54Z 2025-04-02T11:16:54Z 2024 Статья Tyrylgin A., Bai H., Yang Y. Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media // Lobachevskii Journal of Mathematics. - 2024. - 45 (11). - pp. 5437 - 5451. - DOI: 10.1134/S1995080224606696 https://dspace.ncfu.ru/handle/123456789/30359 en Lobachevskii Journal of Mathematics application/pdf application/pdf Pleiades Publishing
institution	СКФУ
collection	Репозиторий
language	English
topic	Basis functions Online GMsFEM Darcy equation Displacement Heterogeneous media Machine learning Multiscale methods Poroelasticity
spellingShingle	Basis functions Online GMsFEM Darcy equation Displacement Heterogeneous media Machine learning Multiscale methods Poroelasticity Tyrylgin, A. A. Тырылгин, А. А. Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
description	In this study, we address the poroelasticity problem in heterogeneous media, which involves a coupled system of equations for fluid pressures and displacements. This problem is crucial in geomechanics for modeling the interaction between fluid flow and deformation in porous media, with applications spanning oil and gas reservoirs, groundwater systems, and geothermal energy production. We introduce an innovative approach by integrating machine learning techniques to train online multiscale basis functions, enhancing the Generalized Multiscale Finite Element Method (GMsFEM). This methodology allows for an adaptive and efficient representation of both macroscopic and local heterogeneities in the system, significantly reducing computational costs. The offline multiscale basis functions are precomputed using local spectral problems, while the online basis functions are dynamically updated using machine learning models trained on local residual data. This approach ensures rapid error reduction and robust convergence, leveraging the computational efficiency of machine learning. We demonstrate the effectiveness of this method through numerical experiments, showcasing its potential in advancing the simulation and modeling of poroelasticity problems in heterogeneous media.
format	Статья
author	Tyrylgin, A. A. Тырылгин, А. А.
author_facet	Tyrylgin, A. A. Тырылгин, А. А.
author_sort	Tyrylgin, A. A.
title	Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
title_short	Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
title_full	Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
title_fullStr	Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
title_full_unstemmed	Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media
title_sort	machine learning for online multiscale model reduction for poroelasticity problem in heterogeneous media
publisher	Pleiades Publishing
publishDate	2025
url	https://dspace.ncfu.ru/handle/123456789/30359
work_keys_str_mv	AT tyrylginaa machinelearningforonlinemultiscalemodelreductionforporoelasticityprobleminheterogeneousmedia AT tyrylginaa machinelearningforonlinemultiscalemodelreductionforporoelasticityprobleminheterogeneousmedia
_version_	1842245824622362624

Machine Learning for Online Multiscale Model Reduction for Poroelasticity Problem in Heterogeneous Media

פריטים דומים