PSI - Issue 43

Available online at www.sciencedirect.com Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000 Available online at www.sciencedirect.com ScienceDirect Structural Integrity Procedia 00 (2022) 000 – 000

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

ScienceDirect

Procedia Structural Integrity 43 (2023) 29–34

© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of MSMF10 organizers. © 20 23 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review under the responsibility of MSMF10 organizers. Abstract Seismic exploration is a standard method of prospecting for hydrocarbon deposits. The major goal is to correctly loc lize the spatial posit on of the fractured inclusion and stimate its para eters. In this work we investigate the applicability of machine-learning algorithms fo th s problem. T gen rate synthetic s ismogra s the model of a deformable solid medium containing slip planes with nonlinear slip conditions on t m is used. The explicit-implicit scheme is applied f r btaining the n merical solution f a constitutive system of equations. The prob em of the seismic w ve propagation in an inhomogeneous fract ed g ol gical mod l based on the well-known Marmous 2 model in a two-dimensio al case is considered. Deep onvolutional neur l networks are used to provide a fast solution for the invers problem of restoring he parameters of the fractu ed inclusion b sed on the sur ac meas rements. The neural network archite ture follows a segme tation approach and t rgets primar ly a spatial localization of the fractured inclusion. However, the mechanical parameters of the inclusion are also estimated using a single run of the same network. © 20 23 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Peer-review under the responsibility of MSMF10 organizers. 10th International Conference on Materials Structure and Micromechanics of Fracture Fractured inclusion localization and characterization based on deep convolutional neural networks Golubev V.I. a,b, , Nikitin I.S. a , Vasyukov A.V. b , Nikitin A.D. a * a Institute of Computer Aided Design of RAS, 2-nd Brestskaya str., 19/18, Moscow, 123056, Russia b Moscow Institute of Physics and Technology (National Research University), Institutskiy per., 9, Dolgoprudny, Moscow, 141701, Russia Abstract Seismic exploration is a standard method of prospecting for hydrocarbon deposits. The major goal is to correctly localize the spatial position of the fractured inclusion and estimate its parameters. In this work we investigate the applicability of machine-learning algorithms for this problem. To generate synthetic seismograms the model of a deformable solid medium containing slip planes with nonlinear slip conditions on them is used. The explicit-implicit scheme is applied for obtaining the numerical solution of a constitutive system of equations. The problem of the seismic wave propagation in an inhomogeneous fractured geological model based on the well-known Marmousi2 model in a two-dimensional case is considered. Deep convolutional neural networks are used to provide a fast solution for the inverse problem of restoring the parameters of the fractured inclusion based on the surface measurements. The neural network architecture follows a segmentation approach and targets primarily a spatial localization of the fractured inclusion. However, the mechanical parameters of the inclusion are also estimated using a single run of the same network. 10th International Conference on Materials Structure and Micromechanics of Fracture Fractured inclusion localization and characterization based on deep convolutional neural networks Golubev V.I. a,b, , Nikitin I.S. a , Vasyukov A.V. b , Nikitin A.D. a * a Institute of Computer Aided Design of RAS, 2-nd Bre skaya str., 19/18, Moscow, 123056, Russia b Moscow Institute of Physics and Technology (National Research University), Institutskiy per., 9, Dolgoprudny, Moscow, 141701, Russia

Keywords: Seismic waves; fractured media; continuum models; machine learning. Keywords: Seismic waves; fractured media; continuum models; machine learning.

1. Introduction This paper uses deep convolutional neural networks for the inverse problem of the fractured inclusion location. This approach was successfully applied in the recent works (Yang et al., 2019; Das et al., 2019; Dujardin et al., 2020; 1. Introduction This paper uses deep convolutional neural networks for the inverse problem of the fractured inclusion location. This approach was successfully applied in the recent works (Yang et al., 2019; Das et al., 2019; Dujardin et al., 2020;

* Corresponding author. Tel.: +7-926-457-2707. E-mail address: nikitin_alex@bk.ru * Correspon ing author. Tel.: +7-926-457-2707. E-mail address: nikitin_alex@bk.ru

2452-3216 © 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of MSMF10 organizers. 2452-3216 © 2023 The Authors. Published by Elsevier B.V. This is an ope access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of MSMF10 organizers.

2452-3216 © 2023 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of MSMF10 organizers. 10.1016/j.prostr.2022.12.230