Abstract
The recent development of high-performance computing enables us to generate spatio-temporal high-resolution data of nonlinear dynamical systems and to analyze them for a deeper understanding of their complex nature. This trend can be found in a wide range of science and engineering, which suggests that detailed investigations on efficient data handling in physical science must be required in the future. This study considers the use of convolutional neural networks (CNNs) to achieve efficient data storage and estimation of scientific big data derived from nonlinear dynamical systems. The CNN is used to reconstruct three-dimensional data from a few numbers of two-dimensional sections in a computationally friendly manner. The present model is a combination of two- and three-dimensional CNNs, which allows users to save only some of the two-dimensional sections to reconstruct the volumetric data. As examples, we consider a flow around a square cylinder at the diameter-based Reynolds number \(Re_D = 300\). We demonstrate that volumetric fluid flow data can be reconstructed with the present method from as few as five sections. Furthermore, we propose a combination of the present CNN-based reconstruction with an adaptive sampling-based super-resolution analysis to augment the data compressibility. Our report can serve as a bridge toward practical data handling for not only fluid mechanics but also a broad range of physical sciences.
Similar content being viewed by others
Availability of Data and Materials
Upon reasonable request.
Code Availability
Available at https://github.com/MitsuMatsu/2D-3D-CNN
References
Moore GE. Cramming more components onto integrated circuits. Electronics. 1965;38:1–4.
Kajishima T, Taira K. Computational fluid dynamics. Cham: Springer International Publishing, 2017;(1).
Brunton SL, Noack BR, Koumoutsakos P. Machine learning for fluid mechanics. Ann Rev Fluid Mech. 2020;52:477–508.
Kutz JN. Deep learning in fluid dynamics. J Fluid Mech. 2017;814:1–4.
Brunton SL, Hemati MS, Taira K. Special issue on machine learning and data-driven methods in fluid dynamics. Theor Comput Fluid Dyn. 2020;34(4):333–7.
Hinton GE, Salakhutdinov RR. Reducing the dimensionality of data with neural networks. Science. 2006;313(5786):504–7.
Taira K, Hemati MS, Brunton SL, Sun Y, Duraisamy K, Bagheri S, Dawson S, Yeh C-A. Modal analysis of fluid flows: applications and outlook. AIAA J. 2020;58(3):998–1022.
Nóvoa A, Magri L. Real-time thermoacoustic data assimilation. J Fluid Mech. 2022;948:A35.
Milano M, Koumoutsakos P. Neural network modeling for near wall turbulent flow. J Comput Phys. 2002;182:1–26.
Glaws A, King R, Sprague M. Deep learning for in situ data compression of large turbulent flow simulations. Phys Rev Fluids. 2020;5:114602.
Morimoto M, Fukami K, Zhang K, Nair AG, Fukagata K. Convolutional neural networks for fluid flow analysis: toward effective metamodeling and low-dimensionalization. Theor Comput Fluid Dyn. 2021;35(5):633–58.
Eivazi H, Le Clainche S, Hoyas S, Vinuesa R. Towards extraction of orthogonal and parsimonious non-linear modes from turbulent flows. Expert Syst Appl. 2022;202:117038.
Le Clainche S, Ferrer E, Gibson S, Cross E, Parente A, Vinuesa R. Improving aircraft performance using machine learning: a review. Aerosp Sci Technol. 2023;138:108354.
Yousif MZ, Yu L, Lim H-C. Physics-guided deep learning for generating turbulent inflow conditions. J. Fluid Mech., 2022;936
Doan NAK, Polifke W, Magri L. Auto-encoded reservoir computing for turbulence learning. In International Conference on Computational Science, 2021;344–351. Springer.
Omata N, Shirayama S. A novel method of low-dimensional representation for temporal behavior of flow fields using deep autoencoder. AIP Adv. 2019;9(1):015006.
Hasegawa K, Fukami K, Murata T, Fukagata K. Machine-learning-based reduced-order modeling for unsteady flows around bluff bodies of various shapes. Theor Comput Fluid Dyn. 2020;34(4):367–83.
Hasegawa K, Fukami K, Murata T, Fukagata K. Data-driven reduced order modeling of flows around two-dimensional bluff bodies of various shapes. ASME-JSME-KSME Joint Fluids Engineering Conference, San Francisco, USA, (Paper 5079), 2019.
Maulik R, Lusch B, Balaprakash P. Reduced-order modeling of advection-dominated systems with recurrent neural networks and convolutional autoencoders. Phys Fluids. 2021;33(3):037106.
Hasegawa K, Fukami K, Murata T, Fukagata K. CNN-LSTM based reduced order modeling of two-dimensional unsteady flows around a circular cylinder at different Reynolds numbers. Fluid Dyn Res. 2020;52:065501.
Nakamura T, Fukami K, Hasegawa K, Nabae Y, Fukagata K. Convolutional neural network and long short-term memory based reduced order surrogate for minimal turbulent channel flow. Phys Fluids. 2021;33:025116.
Fukami K, Murata T, Zhang K, Fukagata K. Sparse identification of nonlinear dynamics with low-dimensionalized flow representations. J Fluid Mech. 2021;926:A10.
Lusch B, Kutz JN, Brunton SL. Deep learning for universal linear embeddings of nonlinear dynamics. Nat Commu. 2018;9(1):4950.
Murata T, Fukami K, Fukagata K. Nonlinear mode decomposition with convolutional neural networks for fluid dynamics. J Fluid Mech. 2020;882:A13.
Xu J, Duraisamy K. Multi-level convolutional autoencoder networks for parametric prediction of spatio-temporal dynamics. Comput Meth Appl Mech Eng. 2020;372:113379.
Liu Y, Ponce C, Brunton SL, Kutz JN. Multiresolution convolutional autoencoders. J Comput Phys. 2023;474:111801.
Muehlebach M, Erichson NB, Mahoney M. Physics-informed autoencoders for lyapunov-stable fluid flow prediction. In Machine Learning and the Physical Sciences Workshop, Conference on Neural Information Processing Systems, 2019.
Fukami K, Nakamura T, Fukagata K. Convolutional neural network based hierarchical autoencoder for nonlinear mode decomposition of fluid field data. Phys Fluids. 2020;32:095110.
Brenner MP, Eldredge JD, Freund JB. Perspective on machine learning for advancing fluid mechanics. Phys Rev Fluids. 2019;4:100501.
Erichson NB, Mathelin L, Yao Z, Brunton SL, Mahoney MW, Kutz JN. Shallow neural networks for fluid flow reconstruction with limited sensors. Proc R Soc A. 2020;476(2238):20200097.
Fukami K, Fukagata K, Taira K. Assessment of supervised machine learning for fluid flows. Theor Comput Fluid Dyn. 2020;34(4):497–519.
Everson R, Sirovich L. Karhunen–loeve procedure for gappy data. J Opt Soc Am. 1995;12(8):1657–64.
Nakamura T, Fukami K, Fukagata K. Identifying key differences between linear stochastic estimation and neural networks for fluid flow regressions. Sci Rep. 2022;12:3726.
Nair NJ, Goza A. Leveraging reduced-order models for state estimation using deep learning. J Fluid Mech. 2020;897:R1.
Fukami K, Maulik R, Ramachandra N, Fukagata K, Taira K. Global field reconstruction from sparse sensors with Voronoi tessellation-assisted deep learning. Nat Mach Intell. 2021;3:945–51.
Fukami K, Fukagata K, Taira K. Super-resolution reconstruction of turbulent flows with machine learning. J Fluid Mech. 2019;870:106–20.
Fukami K, Fukagata K, Taira K. Super-resolution analysis with machine learning for low-resolution flow data. In 11th International Symposium on Turbulence and Shear Flow Phenomena (TSFP11), Southampton, UK, number 208, 2019.
Maulik R, San O. Resolution and energy dissipation characteristics of implicit LES and explicit filtering models for compressible turbulence. Fluids. 2017;2(2):14.
Liu B, Tang J, Huang H, Lu X-Y. Deep learning methods for super-resolution reconstruction of turbulent flows. Phys Fluids. 2020;32:025105.
Kim H, Kim J, Won S, Lee C. Unsupervised deep learning for super-resolution reconstruction of turbulence. J Fluid Mech. 2021;910:A29.
Fukami K, Fukagata K, Taira K. Machine-learning-based spatio-temporal super resolution reconstruction of turbulent flows. J Fluid Mech. 2021;909:A9.
Yousif MZ, Yu L, Lim H-C. High-fidelity reconstruction of turbulent flow from spatially limited data using enhanced super-resolution generative adversarial network. Phys Fluids. 2021;33(12):125119.
Yousif MZ, Yu L, Lim H-C. Super-resolution reconstruction of turbulent flow fields at various reynolds numbers based on generative adversarial networks. Phys Fluids. 2022;34(1):015130.
Yousif MZ, Zhang M, Yu L, Vinuesa R, Lim H-C. A transformer-based synthetic-inflow generator for spatially-developing turbulent boundary layers. J Fluid Mech. 2023;957:A6.
Deng Z, He C, Liu Y, Kim KC. Super-resolution reconstruction of turbulent velocity fields using a generative adversarial network-based artificial intelligence framework. Phys Fluids. 2019;31:125111.
Morimoto M, Fukami K, Fukagata K. Experimental velocity data estimation for imperfect particle images using machine learning. Phys Fluids. 2021;33(8):087121.
Cai S, Zhou S, Xu C, Gao Q. Dense motion estimation of particle images via a convolutional neural network. Exp Fluids. 2019;60:60–73.
Yousif MZ, Yu L, Hoyas S, Vinuesa R, Lim H-C. A deep-learning approach for reconstructing 3D turbulent flows from 2D observation data. Sci Rep. 2023;13:2529.
Huang J, Liu H, Cai W. Online in situ prediction of 3-D flame evolution from its history 2-D projections via deep learning. J Fluid Mech. 2019;875:R2.
Kench S, Cooper SJ. Generating three-dimensional structures from a two-dimensional slice with generative adversarial network-based dimensionality expansion. Nat Mach Intell. 2021;3:299–305.
Bai H, Alam MdM. Dependence of square cylinder wake on Reynolds number. Phys Fluids. 2018;30:015102.
Caltagirone JP. Sur l’interaction fluide-milieu poreux: application au calcul des efforts excerses sur un obstacle par un fluide visqueux. C R Acad Sci Paris. 1994;318:571–7.
LeCun Y, Bottou L, Bengio Y, Haffner P. Gradient-based learning applied to document recognition. Proc IEEE. 1998;86(11):2278–324.
LeCun Y, Bengio Y, Hinton G. Deep learning. Nature. 2015;521:436–44.
Fukami K, Nabae Y, Kawai K, Fukagata K. Synthetic turbulent inflow generator using machine learning. Phys Rev Fluids. 2019;4:064603.
Champion K, Lusch B, Kutz JN, Brunton SL. Data-driven discovery of coordinates and governing equations. Proc Natl Acad Sci USA. 2019;45(116):22445–51.
Kim J, Lee C. Prediction of turbulent heat transfer using convolutional neural networks. J Fluid Mech. 2020;882:A18.
Fukami K, An B, Nohmi M, Obuchi M, Taira K. Machine-learning-based reconstruction of turbulent vortices from sparse pressure sensors in a pump sump. J Fluids Eng. 2022;144(12):121501.
He K, Zhang X, Ren S, Sun J. Deep residual learning for image recognition. Proc IEEE. 2016;1:770–8.
Glaws A, King R, Sprague M. Deep learning for in situ data compression of large turbulent flow simulations. Phys Rev Fluids. 2020;5(11):114602.
Momenifar M, Diao E, Tarokh V, Bragg AD. Dimension reduced turbulent flow data from deep vector quantisers. J Turbul. 2022;23(4–5):232–64.
Nair V, Hinton GE. Rectified linear units improve restricted boltzmann machines. Proc. Int. Conf. Mach. Learn., 2010;807–814.
Bergstra J, Yamins D, Cox D. Making a science of model search: Hyperparameter optimization in hundreds of dimensions for vision architectures. In International Conference on Machine Learning, 2013;28:115–123. PMLR.
Maulik R, Mohan A, Lusch B, Madireddy S, Balaprakash P, Livescu D. Time-series learning of latent-space dynamics for reduced-order model closure. Phys D. 2020;405:132368.
Prechelt L. Automatic early stopping using cross validation: quantifying the criteria. Neural Netw. 1998;11(4):761–7.
Brunton SL, Kutz JN. Data-Driven Science and Engineering: Machine Learning, Dynamical Systems, and Control. Cambridge University Press, pages ISBN, 1108422098, 2019.
Jeong J, Hussain F. On the identification of a vortex. J Fluid Mech. 1995;285:69–94.
Pérez JM, Clainche SL, Vega JM. Reconstruction of three-dimensional flow fields from two-dimensional data. J Comput Phys. 2020;407:109239.
Morimoto M, Fukami K, Zhang K, Fukagata K. Generalization techniques of neural networks for fluid flow estimation. Neural Comput Appl. 2022;34:3647–69.
Nakamura T, Fukagata K. Robust training approach of neural networks for fluid flow state estimations. Int J Heat Fluid Flow. 2022;96:108997.
Keys R. Cubic convolution interpolation for digital image processing. IEEE Trans Acoustics Speech Signal Process. 1981;29(6):1153–60.
Robinson AH, Cherry C. Results of a prototype television bandwidth compression scheme. Proc IEEE. 1967;55(3):356–64.
Fukami K, Hasegawa K, Nakamura T, Morimoto M, Fukagata K. Model order reduction with neural networks: application to laminar and turbulent flows. SN Comput Sci. 2021;2:467.
Ogoke F, Meidani K, Hashemi A, Farimani AB. Graph convolutional networks applied to unstructured flow field data. Mach. Learn.: Sci. Technol., 2021;2:045020.
Kashefi A, Rempe D, Guibas LJ. A point-cloud deep learning framework for prediction of fluid flow fields on irregular geometries. Phys Fluids. 2021;33:027104.
Gao H, Sun L, Wang J-X. PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain. J Comput Phys. 2021;428:110079.
Maulik R, Fukami K, Ramachandra N, Fukagata K, Taira K. Probabilistic neural networks for fluid flow surrogate modeling and data recovery. Phys Rev Fluids. 2020;5:104401.
Morimoto M, Fukami K, Maulik R, Vinuesa R, Fukagata K. Assessments of epistemic uncertainty using gaussian stochastic weight averaging for fluid-flow regression. Phys D. 2022;440:133454.
Acknowledgements
This work was supported by the Japan Society for the Promotion of Science (KAKENHI grant number: 18H03758 and 21H05007). We are also grateful to Professor Shinnosuke Obi, Professor Keita Ando, and Professor Linyu Peng, Professor Takuya Kawata for fruitful discussions. The authors thank Mr. Hikaru Murakami for sharing his DNS code.
Funding
Funded by the Japan Society for the Promotion of Science (KAKENHI grant number: 18H03758 and 21H05007).
Author information
Authors and Affiliations
Contributions
MM: software, validation, investigation, writing—original draft, visualization. TN: software, investigation, writing—original draft. MM: software, investigation, writing—original draft. KaF: conceptualization, software, writing—review and editing, supervision. KoF: conceptualization, resources, writing—review and editing, supervision, project administration, funding acquisition.
Corresponding author
Ethics declarations
Conflict of Interest
The authors declare that they have no confict of interest.
Ethics Approval
Not applicable.
Consent to Participate
Not applicable.
Consent for Publication
Not applicable.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Matsuo, M., Fukami, K., Nakamura, T. et al. Reconstructing Three-Dimensional Bluff Body Wake from Sectional Flow Fields with Convolutional Neural Networks. SN COMPUT. SCI. 5, 306 (2024). https://doi.org/10.1007/s42979-024-02602-0
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s42979-024-02602-0