Abstract
Turbulence is the paradigm of complex systems, notably for its non-equilibrium multi-scale character. Since it is rarely possible to have access to direct simulations of turbulence, the development of efficient models is crucial. The use of machine learning tools has recently received ample attention, yet learning such a complex system is a challenging task. In the present work, we explore the limitations of machine learning in describing high-order statistics, which are essential for understanding turbulence structure. We will focus on shell models of turbulence. They have a manageable dimension and display the same characteristics as Navier–Stokes turbulence, such as Kolmogorov scaling and intermittency. We aim to develop a robust model using machine learning strategies, focusing on analysing recurrent and multi-layer networks. We have found that the standard protocol for training a recurrent neural network is not satisfactory for the description of rare events. The standard training approach creates a discrepancy between what the model learns and what the model has to be able to do. We propose a new training method called Forecast Training. Our extensive statistical analysis shows that this method significantly improves performance. We can train a model that gives excellent agreement with the actual truth, even with limited training data.
Similar content being viewed by others
Data Availability Statement
All data are available upon request. The source code can be found at https://github.com/raudh1/ShellModels.
Notes
Bias terms are omitted.
References
B. Stephen, Pope, Turbulent flows (Cambridge University Press, Cambridge, 2000)
U. Frisch, Turbulence: The Legacy of A (N. Kolmogorov. Cambridge University Press, Cambridge, 1995). https://doi.org/10.1017/CBO9781139170666
A.S. Monin, A.M. Yaglom, Statistical fluid mechanics: mechanics of turbulence. Courier Corporation, (2013)
A.N. Kolmogorov, The local structure of turbulence in incompressible viscous fluid for very large reynolds numbers. Proc. R. Soc. London Ser. A Math. Phys. Sci. 434(1890), 9–13 (1991). https://doi.org/10.1098/rspa.1991.0075
R. Benzi, G. Paladin, G. Parisi, A. Vulpiani, On the multifractal nature of fully developed turbulence and chaotic systems. J. Phys. A Math. Gen. 17(18), 3521 (1984). https://doi.org/10.1088/0305-4470/17/18/021
L.D. Landau, E.M. Lifshitz, Fluid Mechanics: Landau and Lifshitz: Course of Theoretical Physics, Volume 6, volume 6. Elsevier, (2013)
T. Bohr, M.H. Jensen, G. Paladin, and A. Vulpiani. Dynamical Systems Approach to Turbulence. Cambridge Nonlinear Science Series. Cambridge University Press, (1998). https://doi.org/10.1017/CBO9780511599972
S. Bengio, O. Vinyals, N. Jaitly, N. Shazeer. Scheduled sampling for sequence prediction with recurrent neural networks. Advances in neural information processing systems, 28, (2015)
S.L. Brunton, B.R. Noack, P. Koumoutsakos, Machine learning for fluid mechanics. Ann. Rev. Fluid Mech. 52, 477–508 (2020). https://doi.org/10.1146/annurev-fluid-010719-060214
S. Cheng, R. Morel, E. Allys, B. Ménard, S. Mallat, Scattering spectra models for physics. arXiv preprint arXiv:2306.17210, (2023)
F. Borra, M. Baldovin, Using machine-learning modeling to understand macroscopic dynamics in a system of coupled maps. Chaos Interd. J. Nonlinear Sci. (2021). https://doi.org/10.1063/5.0036809
J. Pathak, B. Hunt, M. Girvan, L. Zhixin, E. Ott, Model-free prediction of large spatiotemporally chaotic systems from data: A reservoir computing approach. Phys. Rev. Lett. 120(2), 024102 (2018). https://doi.org/10.1103/PhysRevLett.120.024102
M. Buzzicotti, F. Bonaccorso, P. Clark Di Leoni, L. Biferale, Reconstruction of turbulent data with deep generative models for semantic inpainting from turb-rot database. Phys. Rev. Fluids 6(5), 050503 (2021). https://doi.org/10.1103/PhysRevFluids.6.050503
P.C.D. Leoni, A. Mazzino, L. Biferale, Synchronization to big data: Nudging the navier-stokes equations for data assimilation of turbulent flows. Phys. Rev. X 10(1), 011023 (2020). https://doi.org/10.1103/PhysRevX.10.011023
L. Biferale, Shell models of energy cascade in turbulence. Ann. Rev. Fluid Mech. 35, 441–468 (2003). https://doi.org/10.1146/annurev.fluid.35.101101.161122
P.D. Ditlevsen, Turbulence and Shell Models (Cambridge University Press, Cambridge, 2010)
M.A. Bucci, O. Semeraro, A. Allauzen, S. Chibbaro, L. Mathelin, Curriculum learning for data-driven modeling of dynamical systems. Eur. Phys. J. E 46(3), 12 (2023). https://doi.org/10.1140/epje/s10189-023-00269-8
V. Lvov, E. Podivilov, A. Pomyalov, I. Procaccia, D. Vandembroucq, Improved shell model of turbulence. Phys. Rev. E 58, 03 (1998). https://doi.org/10.1103/PhysRevE.58.1811
M. Yamada, K. Ohkitani, Lyapunov spectrum of a chaotic model of three-dimensional turbulence. J. Phys. Soc. Jpn. 56(12), 4210–4213 (1987). https://doi.org/10.1143/JPSJ.56.4210
F. Cecconi, A. Vulpiani, Approximation of chaotic systems in terms of markovian processes. Phys. Lett. A 201(4), 326–332 (1995). https://doi.org/10.1016/0375-9601(95)00286-C
M.H. Jensen, G. Paladin, A. Vulpiani, Intermittency in a cascade model for three-dimensional turbulence. Phys. Rev. A 43, 798–805 (1991). https://doi.org/10.1103/PhysRevA.43.798
D. Pisarenko, L. Biferale, D. Courvoisier, U. Frisch, M. Vergassola, Further results on multifractality in shell models. Phys. Fluids A Fluid Dyn. 5(10), 2533–2538 (1993). https://doi.org/10.1063/1.858766
J.L. McClelland, D.E. Rumelhart, PDP Research. Group, et al, Parallel Distributed Processing, Volume 2: Explorations in the Microstructure of Cognition: Psychological and Biological Models, volume 2. MIT press, (1987)
S. Hochreiter, J. Schmidhuber, Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997). https://doi.org/10.1162/neco.1997.9.8.1735
A.J. Robinson, F. Fallside, The utility driven dynamic error propagation network. Technical Report CUED/F-INFENG/TR.1, Engineering Department, Cambridge University, Cambridge, UK, (1987)
R.J. Williams, D. Zipser, A learning algorithm for continually running fully recurrent neural networks. Neural Comput. 1(2), 270–280 (1989). https://doi.org/10.1162/neco.1989.1.2.270
Y. Bengio, P. Simard, P. Frasconi, Learning long-term dependencies with gradient descent is difficult. IEEE Trans. Neural Netw. 5(2), 157–166 (1994). https://doi.org/10.1109/72.279181
R. Pascanu, T. Mikolov, Y. Bengio, On the difficulty of training recurrent neural networks. In Proceedings of the 30th International Conference on Machine Learning, ICML 2013, Atlanta, GA, USA, 16-21 June 2013, volume 28 of JMLR Proceedings, pages 1310–1318. JMLR.org, (2013). URL: http://jmlr.org/proceedings/papers/v28/pascanu13.html
S. Wiseman, A.M. Rush, Sequence-to-sequence learning as beam-search optimization. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1296–1306, Austin, Texas, November 2016. Association for Computational Linguistics. URL: https://aclanthology.org/D16-1137, https://doi.org/10.18653/v1/D16-1137
H. Daumé, J. Langford, D. Marcu, Search-based structured prediction. Mach. Learn. 75(3), 297–325 (2009). https://doi.org/10.1007/s10994-009-5106-x
R. Leblond, J.B. Alayrac, A. Osokin, S. Lacoste-Julien, SEARNN: Training RNNs with global-local losses. In International Conference on Learning Representations, (2018). URL: https://openreview.net/forum?id=HkUR_y-RZ
M. Sangiorgio, F. Dercole, Robustness of lstm neural networks for multi-step forecasting of chaotic time series. Chaos Solitons Fract. 139, 110045 (2020). https://doi.org/10.1016/j.chaos.2020.110045
C. Tallec, Y. Ollivier, Can recurrent neural networks warp time? arXiv:1804.11188 (2018)
P.N. Brown, G.D. Byrne, A.C. Hindmarsh, Vode: a variable-coefficient ode solver. SIAM J. Sci. Stat. Comput. 10, 1038–1051 (1989). https://doi.org/10.1137/0910062
Z.S. She, E. Leveque, Universal scaling laws in fully developed turbulence. Phys. Rev. Lett. 72, 336–339 (1994). https://doi.org/10.1103/PhysRevLett.72.336
P. Castiglione, M. Falcioni, A. Lesne, A. Vulpiani, Chaos and Coarse Graining in Statistical Mechanics (Cambridge University Press, Cambridge, 2008)
B. Mehlig, Machine Learning with Neural Networks: An Introduction for Scientists and Engineers (Cambridge University Press, Cambridge, 2021)
T. Pfaff, M. Fortunato, A. Sanchez-Gonzalez, P.W. Battaglia, Learning mesh-based simulation with graph networks. arXiv:2010.03409, (2020)
A. Sanchez-Gonzalez, N. Heess, J.T. Springenberg, J. Merel, M. Riedmiller, R. Hadsell, P. Battaglia, Graph networks as learnable physics engines for inference and control. In International Conference on Machine Learning, pages 4470–4479. PMLR, (2018)
J. Su, J. Kempe, D. Fielding, N. Tsilivis, M. Cranmer, S. Ho, Adversarial noise injection for learned turbulence simulations. In Machine Learning for Physical Sciences Workshop, Advances in Neural Information Processing Systems, (2022)
Funding
This study was funded by ANR (Agence Nationale de la Recherche) within the framework of the SPEED (Simulating Physical PDEs Efficiently with Deep Learning) project ANR-20-CE23-0025-01.
Author information
Authors and Affiliations
Corresponding author
Additional information
Guest editors: S. Chibbaro, M.-C. Firpo, G. Napoli and S. Randoux.
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
Noto, D., Allauzen, A. & Chibbaro, S. An efficient training method to learn a model of turbulence. Eur. Phys. J. Plus 139, 298 (2024). https://doi.org/10.1140/epjp/s13360-024-05056-8
Received:
Accepted:
Published:
DOI: https://doi.org/10.1140/epjp/s13360-024-05056-8