A neural network closure for the Euler-Poisson system based on kinetic simulations

Previous Article
On spectral gaps of growth-fragmentation semigroups in higher moment spaces
KRM Home
This Issue
Next Article
The inviscid limit for the 2D Navier-Stokes equations in bounded domains

doi: 10.3934/krm.2021044

Online First

Online First articles are published articles within a journal that have not yet been assigned to a formal issue. This means they do not yet have a volume number, issue number, or page numbers assigned to them, however, they can still be found and cited using their DOI (Digital Object Identifier). Online First publication benefits the research community by making new scientific discoveries known as quickly as possible.

Readers can access Online First articles via the “Online First” tab for the selected journal.

A neural network closure for the Euler-Poisson system based on kinetic simulations

Léo Bois ^1,2, , Emmanuel Franck ^1,2, , Laurent Navoret ^1,2, and Vincent Vigon ^1,

1.	Institut de Recherche Mathématique Avancée, UMR 7501, Université de Strasbourg et CNRS, 7 rue René Descartes, 67000 Strasbourg, France
2.	INRIA Nancy-Grand Est, TONUS Project, Strasbourg, France

Received October 2020 Revised August 2021 Early access December 2021

This work deals with the modeling of plasmas, which are ionized gases. Thanks to machine learning, we construct a closure for the one-dimensional Euler-Poisson system valid for a wide range of collisional regimes. This closure, based on a fully convolutional neural network called V-net, takes as input the whole spatial density, mean velocity and temperature and predicts as output the whole heat flux. It is learned from data coming from kinetic simulations of the Vlasov-Poisson equations. Data generation and preprocessings are designed to ensure an almost uniform accuracy over the chosen range of Knudsen numbers (which parametrize collisional regimes). Finally, several numerical tests are carried out to assess validity and flexibility of the whole pipeline.

Keywords: Neural networks, plasma, Vlasov-Poisson, closure, fluid models, Knudsen number.

Mathematics Subject Classification: Primary: 35Q31, 65M08, 82D10, 68T07.

Citation: Léo Bois, Emmanuel Franck, Laurent Navoret, Vincent Vigon. A neural network closure for the Euler-Poisson system based on kinetic simulations. Kinetic & Related Models, doi: 10.3934/krm.2021044

References:

[1]	A. Beck, D. Flad and C.-D. Munz, Deep neural networks for data-driven les closure models, J. Comput. Phys., 398 (2019), 108910, 23 pp. doi: 10.1016/j.jcp.2019.108910. Google Scholar
[2]	N. Besse, F. Berthelin, Y. Brenier and P. Bertrand, The multi-water-bag equations for collisionless kinetic modeling, Kinet. Relat. Models, 2 (2009), 39-80. doi: 10.3934/krm.2009.2.39. Google Scholar
[3]	N. Besse and P. Bertrand, Gyro-water-bag approach in nonlinear gyrokinetic turbulence, J. Comput. Phys., 228 (2009), 3973-3995. doi: 10.1016/j.jcp.2009.02.025. Google Scholar
[4]	S. I. Braginskii, Transport phenomena in plasma, Rev. Plasma Phys., 1 (1963), 205. Google Scholar
[5]	A. Brizard, Nonlinear gyrofluid description of turbulent magnetized plasmas, Phys. Fluids B, 4 (1992), 1213-1228. doi: 10.1063/1.860129. Google Scholar
[6]	Z. Cai, Y. Fan and R. Li, Globally hyperbolic regularization of grad's moment system, Communications on Pure and Applied Mathematics, 67 (2014), 464-518. doi: 10.1002/cpa.21472. Google Scholar
[7]	Z. Chang and J. D. Callen, Unified fluid/kinetic description of plasma microinstabilities. part i: Basic equations in a sheared slab geometry, Physics of Fluids B: Plasma Physics, 4 (1992), 1167-1181. doi: 10.1063/1.860125. Google Scholar
[8]	G. F. Chew, M. L. Goldberger, F. E. Low and Y. Nambu, Application of dispersion relations to low-energy meson-nucleon scattering, Phys. Rev., 106 (1957), 1337-1344. doi: 10.1103/PhysRev.106.1337. Google Scholar
[9]	A. Crestetto, N. Crouseilles and M. Lemou, Kinetic/fluid micro-macro numerical schemes for Vlasov-Poisson-BGK equation using particles, Kinet. Relat. Models, 5 (2012) 787–816. doi: 10.3934/krm.2012.5.787. Google Scholar
[10]	N. Crouseilles, P. Degond and M. Lemou, A hybrid kinetic/fluid model for solving the gas dynamics boltzmann–bgk equation, J. Comput. Phys., 199 (2004), 776-808. doi: 10.1016/j.jcp.2004.03.007. Google Scholar
[11]	P. Degond, Macroscopic limits of the boltzmann equation: A review, In Modeling and Computational Methods for Kinetic Equations, Birkhäuser, Boston, MA, 2004, 3–57. Google Scholar
[12]	P. Degond, G. Dimarco and L. Mieussens, A multiscale kinetic–fluid solver with dynamic localization of kinetic effects, J. Comput. Phys., 229 (2010), 4907-4933. doi: 10.1016/j.jcp.2010.03.009. Google Scholar
[13]	O. Desjardins, R. O. Fox and P. Villedieu, A quadrature-based moment method for dilute fluid-particle flows, J. Comput. Phys., 227 (2008), 2514-2539. doi: 10.1016/j.jcp.2007.10.026. Google Scholar
[14]	G. Dimarco and L. Pareschi, Numerical methods for kinetic equations, Acta Numer., 23 (2014), 369-520. doi: 10.1017/S0962492914000063. Google Scholar
[15]	B. Dubroca and J.-L. Feugeas, Etude théorique et numérique d'une hiérarchie de modèles aux moments pour le transfert radiatif, C. R. Acad. Sci. Paris Sér. I Math., 329 (1999), 915–920. doi: 10.1016/S0764-4442(00)87499-6. Google Scholar
[16]	B. Dubroca, J.-L. Feugeas and M. Frank, Angular moment model for the Fokker-Planck equation, The European Physical Journal D, 60 (2010), 301-207. doi: 10.1140/epjd/e2010-00190-8. Google Scholar
[17]	D. Dumoulin and F. Visin, A guide to convolution arithmetic for deep learning, 2016. Google Scholar
[18]	K. Duraisamy, G. Iaccarino and H. Xiao, Turbulence modeling in the age of data, Annu. Rev. Fluid Mech., 51 (2019), 357-377. Google Scholar
[19]	C. K. Garrett and C. D. Hauck, A comparison of moment closures for linear kinetic transport equations: The line source benchmark, Transport Theor. Stat. Phys., 42 (2013), 203-235. doi: 10.1080/00411450.2014.910226. Google Scholar
[20]	H. Grad, On the kinetic theory of rarefied gases, Commun. Pure Appl. Math., 2 (1949), 331-407. doi: 10.1002/cpa.3160020403. Google Scholar
[21]	G. W. Hammett, W. Dorland and F. W. Perkins, Fluid models of phase mixing, landau damping, and nonlinear gyrokinetic dynamics, Phys. Fluids B, 4 (1992), 2052-2061. doi: 10.1063/1.860014. Google Scholar
[22]	G. W. Hammett and F. W. Perkins, Fluid moment models for landau damping with application to the ion-temperature-gradient instability, Phys. Rev. Lett., 64 (1990), 3019-3022. doi: 10.1103/PhysRevLett.64.3019. Google Scholar
[23]	J. Han, C. Ma, Z. Ma and W. E, Uniformly accurate machine learning-based hydrodynamic models for kinetic equations, PNAS, 116 (2019), 21983-21991. doi: 10.1073/pnas.1909854116. Google Scholar
[24]	K. He, X. Zhang, S. Ren and J. Sun, Deep residual learning for image recognition, In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016,770–778. doi: 10.1109/CVPR.2016.90. Google Scholar
[25]	P. Helluy, L. Navoret, N. Pham and A. Crestetto, Reduced Vlasov-Maxwell simulations, C. R. Mécanique, 342 (2014), 619–635. doi: 10.1016/j.crme.2014.06.008. Google Scholar
[26]	P. Hunana, G. P. Zank, M. Laurenza, A. Tenerani, G. M. Webb, M. L. Goldstein, M. Velli and L. Adhikari, New closures for more precise modeling of landau damping in the fluid framework., Physical Review Letters, 121 (2018), 135101. doi: 10.1103/PhysRevLett.121.135101. Google Scholar
[27]	M. Junk, Domain of definition of levermore's five-moment system, J. Stat. Phys., 93 (1998), 1143-1167. doi: 10.1023/B:JOSS.0000033155.07331.d9. Google Scholar
[28]	R. J. Leveque, Finite Volume Methods for Hyperbolic Problems, volume 31. Cambridge University Press, 2002. doi: 10.1017/CBO9780511791253. Google Scholar
[29]	C. D. Levermore, Moment closure hierarchies for kinetic theories, J. Stat. Phys., 83 (1996), 1021-1065. doi: 10.1007/BF02179552. Google Scholar
[30]	C. D. Levermore and W. J. Morokoff, The gaussian moment closure for gas dynamics, SIAM J. Appl. Math., 59 (1998), 72-96. doi: 10.1137/S0036139996299236. Google Scholar
[31]	J. L. Lumley, Toward a turbulent constitutive relation, J. Fluid Mech, 41 (1970), 413-434. doi: 10.1017/S0022112070000678. Google Scholar
[32]	C. Ma, B. Zhu, X.-Q. Xu and W. Wang, Machine learning surrogate models for landau fluid closure, Physics of Plasmas, 27 (2020), 042502, arXiv: 1909.11509. doi: 10.1063/1.5129158. Google Scholar
[33]	G. Manfredi, Density functional theory for collisionless plasmas–equivalence of fluid and kinetic approaches, J. Plasma Phys., 86 (2020). doi: 10.1017/S0022377820000240. Google Scholar
[34]	R. Maulik, N. A. Garland, J. W. Burby, X.-Z. Tang and P. Balaprakash, Neural network representability of fully ionized plasma fluid model closures, Phys. Plasmas, 27 (2020), 072106. doi: 10.1063/5.0006457. Google Scholar
[35]	F. Milletari, N. Navab and S. Ahmadi, V-net: Fully convolutional neural networks for volumetric medical image segmentation, In 2016 Fourth International Conference on 3D Vision (3DV), 2016,565–571. doi: 10.1109/3DV.2016.79. Google Scholar
[36]	C. Negulescu and S. Possanner, Closure of the strongly magnetized electron fluid equations in the adiabatic regime, Multiscale Model. Sim., 14 (2016), 839-873. doi: 10.1137/15M1027309. Google Scholar
[37]	T. Passot, P. L. Sulem and P. Hunana, Extending magnetohydrodynamics to the slow dynamics of collisionless plasmas, Physics of Plasmas, 19 (2012), 082113. doi: 10.1063/1.4746092. Google Scholar
[38]	M. Perin, C. Chandre, P. J. Morrison and E. Tassi, Hamiltonian closures for fluid models with four moments by dimensional analysis, J. Phys. A Math. Theor., 48 (2015), 275501. doi: 10.1088/1751-8113/48/27/275501. Google Scholar
[39]	N. Pham, P. Helluy and A. Crestetto, Space-only hyperbolic approximation of the vlasov equation, ESAIM: Proc., 43 (2013), 17-36. doi: 10.1051/proc/201343002. Google Scholar
[40]	O. Ronneberger, P. Fischer and T. Brox, U-net: Convolutional networks for biomedical image segmentation, In International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2015,234–241. doi: 10.1007/978-3-319-24574-4_28. Google Scholar
[41]	J. Schneider, Entropic approximation in kinetic theory, Esaim Math Model Numer Anal., 38 (2004), 541-561. doi: 10.1051/m2an:2004025. Google Scholar
[42]	C. Shorten and T. M. Khoshgoftaar, A survey on image data augmentation for deep learning, Journal of Big Data, 6 (2019), 60. doi: 10.1186/s40537-019-0197-0. Google Scholar
[43]	P. B. Snyder, G. W. Hammett and W. Dorland, Landau fluid models of collisionless magnetohydrodynamics, Phys. Plasmas, 4 (1997), 3974-3985. doi: 10.1063/1.872517. Google Scholar
[44]	E. Sonnendrücker, Numerical Methods for the Vlasov-Maxwell Equations, 2015. Google Scholar
[45]	H. Struchtrup and M. Torrilhon, Regularization of grad's 13 moment equations: Derivation and linear analysis, Physics of Fluids, 15 (2003), 2668-2680. doi: 10.1063/1.1597472. Google Scholar
[46]	E. Tassi, Hamiltonian closures in fluid models for plasmas, Eur. Phys. J. D, 71 (2017), 269. doi: 10.1140/epjd/e2017-80223-6. Google Scholar
[47]	M. Torrilhon, Hyperbolic moment equations in kinetic gas theory based on multi-variate pearson-iv-distributions, Commun. in Comput. Phys., 7 (2010), 639-673. doi: 10.4208/cicp.2009.09.049. Google Scholar
[48]	M. Torrilhon, Modeling nonequilibrium gas flow based on moment equations, Annual Review of Fluid Mechanics, 48 (2016), 429-458. Google Scholar
[49]	J.-X. Wang, J.-L. Wu and H. Xiao, Physics-informed machine learning approach for reconstructing reynolds stress modeling discrepancies based on dns data, Physic. Rev. Fluids, 2 (2017), 034603. doi: 10.1103/PhysRevFluids.2.034603. Google Scholar
[50]	X.-H. Zhou, J. Han and H. Xiao, Learning nonlocal constitutive models with neural networks, Comput. Methods Appl. Mech. Engrg., 384 (2021), Paper No. 113927, 27 pp. arXiv: 2010.10491. doi: 10.1016/j.cma.2021.113927. Google Scholar

show all references

References:

[1]	A. Beck, D. Flad and C.-D. Munz, Deep neural networks for data-driven les closure models, J. Comput. Phys., 398 (2019), 108910, 23 pp. doi: 10.1016/j.jcp.2019.108910. Google Scholar
[2]	N. Besse, F. Berthelin, Y. Brenier and P. Bertrand, The multi-water-bag equations for collisionless kinetic modeling, Kinet. Relat. Models, 2 (2009), 39-80. doi: 10.3934/krm.2009.2.39. Google Scholar
[3]	N. Besse and P. Bertrand, Gyro-water-bag approach in nonlinear gyrokinetic turbulence, J. Comput. Phys., 228 (2009), 3973-3995. doi: 10.1016/j.jcp.2009.02.025. Google Scholar
[4]	S. I. Braginskii, Transport phenomena in plasma, Rev. Plasma Phys., 1 (1963), 205. Google Scholar
[5]	A. Brizard, Nonlinear gyrofluid description of turbulent magnetized plasmas, Phys. Fluids B, 4 (1992), 1213-1228. doi: 10.1063/1.860129. Google Scholar
[6]	Z. Cai, Y. Fan and R. Li, Globally hyperbolic regularization of grad's moment system, Communications on Pure and Applied Mathematics, 67 (2014), 464-518. doi: 10.1002/cpa.21472. Google Scholar
[7]	Z. Chang and J. D. Callen, Unified fluid/kinetic description of plasma microinstabilities. part i: Basic equations in a sheared slab geometry, Physics of Fluids B: Plasma Physics, 4 (1992), 1167-1181. doi: 10.1063/1.860125. Google Scholar
[8]	G. F. Chew, M. L. Goldberger, F. E. Low and Y. Nambu, Application of dispersion relations to low-energy meson-nucleon scattering, Phys. Rev., 106 (1957), 1337-1344. doi: 10.1103/PhysRev.106.1337. Google Scholar
[9]	A. Crestetto, N. Crouseilles and M. Lemou, Kinetic/fluid micro-macro numerical schemes for Vlasov-Poisson-BGK equation using particles, Kinet. Relat. Models, 5 (2012) 787–816. doi: 10.3934/krm.2012.5.787. Google Scholar
[10]	N. Crouseilles, P. Degond and M. Lemou, A hybrid kinetic/fluid model for solving the gas dynamics boltzmann–bgk equation, J. Comput. Phys., 199 (2004), 776-808. doi: 10.1016/j.jcp.2004.03.007. Google Scholar
[11]	P. Degond, Macroscopic limits of the boltzmann equation: A review, In Modeling and Computational Methods for Kinetic Equations, Birkhäuser, Boston, MA, 2004, 3–57. Google Scholar
[12]	P. Degond, G. Dimarco and L. Mieussens, A multiscale kinetic–fluid solver with dynamic localization of kinetic effects, J. Comput. Phys., 229 (2010), 4907-4933. doi: 10.1016/j.jcp.2010.03.009. Google Scholar
[13]	O. Desjardins, R. O. Fox and P. Villedieu, A quadrature-based moment method for dilute fluid-particle flows, J. Comput. Phys., 227 (2008), 2514-2539. doi: 10.1016/j.jcp.2007.10.026. Google Scholar
[14]	G. Dimarco and L. Pareschi, Numerical methods for kinetic equations, Acta Numer., 23 (2014), 369-520. doi: 10.1017/S0962492914000063. Google Scholar
[15]	B. Dubroca and J.-L. Feugeas, Etude théorique et numérique d'une hiérarchie de modèles aux moments pour le transfert radiatif, C. R. Acad. Sci. Paris Sér. I Math., 329 (1999), 915–920. doi: 10.1016/S0764-4442(00)87499-6. Google Scholar
[16]	B. Dubroca, J.-L. Feugeas and M. Frank, Angular moment model for the Fokker-Planck equation, The European Physical Journal D, 60 (2010), 301-207. doi: 10.1140/epjd/e2010-00190-8. Google Scholar
[17]	D. Dumoulin and F. Visin, A guide to convolution arithmetic for deep learning, 2016. Google Scholar
[18]	K. Duraisamy, G. Iaccarino and H. Xiao, Turbulence modeling in the age of data, Annu. Rev. Fluid Mech., 51 (2019), 357-377. Google Scholar
[19]	C. K. Garrett and C. D. Hauck, A comparison of moment closures for linear kinetic transport equations: The line source benchmark, Transport Theor. Stat. Phys., 42 (2013), 203-235. doi: 10.1080/00411450.2014.910226. Google Scholar
[20]	H. Grad, On the kinetic theory of rarefied gases, Commun. Pure Appl. Math., 2 (1949), 331-407. doi: 10.1002/cpa.3160020403. Google Scholar
[21]	G. W. Hammett, W. Dorland and F. W. Perkins, Fluid models of phase mixing, landau damping, and nonlinear gyrokinetic dynamics, Phys. Fluids B, 4 (1992), 2052-2061. doi: 10.1063/1.860014. Google Scholar
[22]	G. W. Hammett and F. W. Perkins, Fluid moment models for landau damping with application to the ion-temperature-gradient instability, Phys. Rev. Lett., 64 (1990), 3019-3022. doi: 10.1103/PhysRevLett.64.3019. Google Scholar
[23]	J. Han, C. Ma, Z. Ma and W. E, Uniformly accurate machine learning-based hydrodynamic models for kinetic equations, PNAS, 116 (2019), 21983-21991. doi: 10.1073/pnas.1909854116. Google Scholar
[24]	K. He, X. Zhang, S. Ren and J. Sun, Deep residual learning for image recognition, In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016,770–778. doi: 10.1109/CVPR.2016.90. Google Scholar
[25]	P. Helluy, L. Navoret, N. Pham and A. Crestetto, Reduced Vlasov-Maxwell simulations, C. R. Mécanique, 342 (2014), 619–635. doi: 10.1016/j.crme.2014.06.008. Google Scholar
[26]	P. Hunana, G. P. Zank, M. Laurenza, A. Tenerani, G. M. Webb, M. L. Goldstein, M. Velli and L. Adhikari, New closures for more precise modeling of landau damping in the fluid framework., Physical Review Letters, 121 (2018), 135101. doi: 10.1103/PhysRevLett.121.135101. Google Scholar
[27]	M. Junk, Domain of definition of levermore's five-moment system, J. Stat. Phys., 93 (1998), 1143-1167. doi: 10.1023/B:JOSS.0000033155.07331.d9. Google Scholar
[28]	R. J. Leveque, Finite Volume Methods for Hyperbolic Problems, volume 31. Cambridge University Press, 2002. doi: 10.1017/CBO9780511791253. Google Scholar
[29]	C. D. Levermore, Moment closure hierarchies for kinetic theories, J. Stat. Phys., 83 (1996), 1021-1065. doi: 10.1007/BF02179552. Google Scholar
[30]	C. D. Levermore and W. J. Morokoff, The gaussian moment closure for gas dynamics, SIAM J. Appl. Math., 59 (1998), 72-96. doi: 10.1137/S0036139996299236. Google Scholar
[31]	J. L. Lumley, Toward a turbulent constitutive relation, J. Fluid Mech, 41 (1970), 413-434. doi: 10.1017/S0022112070000678. Google Scholar
[32]	C. Ma, B. Zhu, X.-Q. Xu and W. Wang, Machine learning surrogate models for landau fluid closure, Physics of Plasmas, 27 (2020), 042502, arXiv: 1909.11509. doi: 10.1063/1.5129158. Google Scholar
[33]	G. Manfredi, Density functional theory for collisionless plasmas–equivalence of fluid and kinetic approaches, J. Plasma Phys., 86 (2020). doi: 10.1017/S0022377820000240. Google Scholar
[34]	R. Maulik, N. A. Garland, J. W. Burby, X.-Z. Tang and P. Balaprakash, Neural network representability of fully ionized plasma fluid model closures, Phys. Plasmas, 27 (2020), 072106. doi: 10.1063/5.0006457. Google Scholar
[35]	F. Milletari, N. Navab and S. Ahmadi, V-net: Fully convolutional neural networks for volumetric medical image segmentation, In 2016 Fourth International Conference on 3D Vision (3DV), 2016,565–571. doi: 10.1109/3DV.2016.79. Google Scholar
[36]	C. Negulescu and S. Possanner, Closure of the strongly magnetized electron fluid equations in the adiabatic regime, Multiscale Model. Sim., 14 (2016), 839-873. doi: 10.1137/15M1027309. Google Scholar
[37]	T. Passot, P. L. Sulem and P. Hunana, Extending magnetohydrodynamics to the slow dynamics of collisionless plasmas, Physics of Plasmas, 19 (2012), 082113. doi: 10.1063/1.4746092. Google Scholar
[38]	M. Perin, C. Chandre, P. J. Morrison and E. Tassi, Hamiltonian closures for fluid models with four moments by dimensional analysis, J. Phys. A Math. Theor., 48 (2015), 275501. doi: 10.1088/1751-8113/48/27/275501. Google Scholar
[39]	N. Pham, P. Helluy and A. Crestetto, Space-only hyperbolic approximation of the vlasov equation, ESAIM: Proc., 43 (2013), 17-36. doi: 10.1051/proc/201343002. Google Scholar
[40]	O. Ronneberger, P. Fischer and T. Brox, U-net: Convolutional networks for biomedical image segmentation, In International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 2015,234–241. doi: 10.1007/978-3-319-24574-4_28. Google Scholar
[41]	J. Schneider, Entropic approximation in kinetic theory, Esaim Math Model Numer Anal., 38 (2004), 541-561. doi: 10.1051/m2an:2004025. Google Scholar
[42]	C. Shorten and T. M. Khoshgoftaar, A survey on image data augmentation for deep learning, Journal of Big Data, 6 (2019), 60. doi: 10.1186/s40537-019-0197-0. Google Scholar
[43]	P. B. Snyder, G. W. Hammett and W. Dorland, Landau fluid models of collisionless magnetohydrodynamics, Phys. Plasmas, 4 (1997), 3974-3985. doi: 10.1063/1.872517. Google Scholar
[44]	E. Sonnendrücker, Numerical Methods for the Vlasov-Maxwell Equations, 2015. Google Scholar
[45]	H. Struchtrup and M. Torrilhon, Regularization of grad's 13 moment equations: Derivation and linear analysis, Physics of Fluids, 15 (2003), 2668-2680. doi: 10.1063/1.1597472. Google Scholar
[46]	E. Tassi, Hamiltonian closures in fluid models for plasmas, Eur. Phys. J. D, 71 (2017), 269. doi: 10.1140/epjd/e2017-80223-6. Google Scholar
[47]	M. Torrilhon, Hyperbolic moment equations in kinetic gas theory based on multi-variate pearson-iv-distributions, Commun. in Comput. Phys., 7 (2010), 639-673. doi: 10.4208/cicp.2009.09.049. Google Scholar
[48]	M. Torrilhon, Modeling nonequilibrium gas flow based on moment equations, Annual Review of Fluid Mechanics, 48 (2016), 429-458. Google Scholar
[49]	J.-X. Wang, J.-L. Wu and H. Xiao, Physics-informed machine learning approach for reconstructing reynolds stress modeling discrepancies based on dns data, Physic. Rev. Fluids, 2 (2017), 034603. doi: 10.1103/PhysRevFluids.2.034603. Google Scholar
[50]	X.-H. Zhou, J. Han and H. Xiao, Learning nonlocal constitutive models with neural networks, Comput. Methods Appl. Mech. Engrg., 384 (2021), Paper No. 113927, 27 pp. arXiv: 2010.10491. doi: 10.1016/j.cma.2021.113927. Google Scholar

Figure 1. Graph of the composition of the closure. The different operations are (Re)-(Re') resampling, (P) pre-processing, (Sl) slicing, ($ \text{NN}_\theta $) neural network, (R) reconstruction, (P') post-processing and (Sm) smoothing

Download full-size image

Download as PowerPoint slide

Figure 2. Visual representation of the three operations involved in the non-locality: the (Sm) smoothing, the (R) reconstruction and (Sl) slicing mechanism, and the ($ \text{NN}_\theta $) neural network. Highlighted in gray is the part of the data used to compute the output value whose position is indicated by the red line. For better readability, only one of the four vectors of the input $ X $, and only the three windows involved, are represented. These examples use the redundancy parameter $ r = 2 $. Two scenarios are illustrated: on the left, each output value of the neural network depends on relatively few input values, while on the right it depends on relatively many. Also, the position of the output value considered is not the same

Download full-size image

Download as PowerPoint slide

Figure 3. Graph of a 1D V-Net with a window of size $ N = 512 $, a depth $ k = 2 $ and $ \ell = 3 $ levels

Download full-size image

Download as PowerPoint slide

Figure 4. Scheme of the data generation process for one simulation with the kinetic model. With one Knudsen number and one initial condition, we produce 20 entries in the dataset

Download full-size image

Download as PowerPoint slide

Figure 5. Possible functions $ \rho $, $ u $ and $ T $ generated with the process described in Section 4.2

Download full-size image

Download as PowerPoint slide

Figure 6. Scheme of the whole data processing for the training and the prediction processes respectivily. The different operations are (S) standardization, (Sl) slicing, (N) normalization, (I) inverse normalization, (R) reconstruction and (Sm) smoothing

Download full-size image

Download as PowerPoint slide

Figure 7. Slicing of a 1D signal into overlapping windows with different redundancy parameters $ r $. The original signal is represented vertically on the left, and the windows are on the right. The hatched parts of each window are the margins that are ignored when reconstructing the signal. The windows that exceed the span of the original signal are completed by periodicity

Download full-size image

Download as PowerPoint slide

Figure 8. Kernel used for the reconstruction of an entire signal from the windows predicted by the neural network, with the redundancy parameter $ r = 3 $

Download full-size image

Download as PowerPoint slide

Figure 9. Three examples of predictions. Note the different scales of the y-axis: the amplitude of the heat flux tends to increase with the Knudsen number

Download full-size image

Download as PowerPoint slide

Figure 10. Distribution of the relative errors of the neural network and the Navier-Stokes estimation over the test dataset ($ 10, 000 $ predictions)

Download full-size image

Download as PowerPoint slide

Figure 11. Error of the Navier-Stokes estimations and of the network on the test dataset, depending on the Knudsen number and the $ L^2 $ norm of the real heat flux. Each dot corresponds to one entry in the test dataset. For better clarity, only one $ 31^{\text{th}} $ of the data is shown here

Download full-size image

Download as PowerPoint slide

Figure 12. Relative error of the Navier-Stokes estimations and of the network on the test dataset, depending on the Knudsen number. The line represents the median of the error over the 100 entries of the test dataset for each epsilon, and the coloured area the interquartile interval

Download full-size image

Download as PowerPoint slide

Figure 13. Norm of the real heat flux and relative errors of the predicted heat flux and the Navier-Stokes estimation throughout simulations. Median and interquartile interval over 50 simulations

Download full-size image

Download as PowerPoint slide

Figure 14. Median relative error and number of learnable parameters for V-Nets with different hyper-parameters : ($ \ell $) number of levels, ($ d $) depth, and ($ p $) size of the kernels. Points sharing the same color represent networks that differ by only one hyper-parameter

Download full-size image

Download as PowerPoint slide

Figure 15. Example of the evolution of the density, mean velocity, temperature and heat flux during the first time unit with a randomly generated initial condition and $ \varepsilon = 0.1 $, obtained with the kinetic model

Download full-size image

Download as PowerPoint slide

Figure 16. Comparison of the four models on the density and heat flux at $ t = 1 $ with a randomly generated initial condition and $ \varepsilon = 0.1 $

Download full-size image

Download as PowerPoint slide

Figure 17. Examples of the evolution of the electric energy with different initial conditions and different Knudsen numbers

Download full-size image

Download as PowerPoint slide

Figure 18. Distributions over 200 simulations of the relative errors of the three fluid models compared to the kinetic model, measured on the logarithm of the electric energy from $ t = 0 $ to $ t = 8 $. For better readability, the x-axis in in logarithmic scale

Download full-size image

Download as PowerPoint slide

Figure 19. $ L^2 $ relative errors of the fluid models over the kinetic model depending on the Knudsen number. The first plot shows the raw data and the second shows the median and interquartile interval for 20 uniform classes of Knudsen numbers between $ 0.01 $ and $ 1 $

Download full-size image

Download as PowerPoint slide

Figure 20. Example of a heat flux predicted by the neural network, with and without smoothing. The smoothing here uses $ \sigma\simeq0.05 $

Download full-size image

Download as PowerPoint slide

Figure 21. Relative errors of the predictions over the test dataset depending on the quantity $ \sigma $ of smoothing

Download full-size image

Download as PowerPoint slide

Figure 22. Proportion of simulations of the fluid model reaching $ t = 3 $ out of 30 simulations, depending on the quantity $ \sigma $ of smoothing

Download full-size image

Download as PowerPoint slide

Figure 23. Relative errors of the fluid+network model on five different quantities depending on the resolution used. Each point is an average of 10 simulations with always the same 10 different initial conditions and $ \varepsilon = 0.1 $. The network used was trained with data generated with $ N_x = 1\, 024 $. To adapt to different resolutions, option 1 relies on the slicing mechanism to adapt to the different input sizes, while option 2 uses the resampling of the input to $ N_x = 1\, 024 $ and the resampling of the output back to the original resolution. For all resolutions, the errors are computed relative to kinetic simulations with a high resolution ($ N_x = 2\, 048 $, $ N_v = 141 $)

Download full-size image

Download as PowerPoint slide

Figure 24. Example of convergence of the fluid+network model with option 2, towards what seems to be a slightly different solution than the kinetic model

Download full-size image

Download as PowerPoint slide

figure 18, which seems to indicate that the fluid+network model is not affected much by the initial discontinuities">Figure 25. Distributions over 100 simulations of the relative errors of the three fluid models compared to the kinetic model, measured on the logarithm of the electric energy from $ t = 0 $ to $ t = 8 $, with discontinuous initial conditions. For better readability, the x-axis in in logarithmic scale. It looks a lot like figure 18, which seems to indicate that the fluid+network model is not affected much by the initial discontinuities

Download full-size image

Download as PowerPoint slide

Figure 26. Example of the evolution of the density and temperature up to $ t = 0.5 $ with a randomly generated discontinuous initial condition and $ \varepsilon = 0.1 $, obtained with the kinetic model. The initial discontinuities are quickly smoothed out by the dynamic of the system

Download full-size image

Download as PowerPoint slide

Table 1. Hyper-parameters of the reference neural network

Hyper-parameter	Value
size of the input window ($ N $)	512
number of levels ($ \ell $)	5
depth ($ d $)	4
size of the kernels ($ p $)	11
activation function	softplus

Download as excel

Hyper-parameter	Value
size of the input window ($ N $)	512
number of levels ($ \ell $)	5
depth ($ d $)	4
size of the kernels ($ p $)	11
activation function	softplus

[1]	Silvia Caprino, Guido Cavallaro, Carlo Marchioro. Time evolution of a Vlasov-Poisson plasma with magnetic confinement. Kinetic & Related Models, 2012, 5 (4) : 729-742. doi: 10.3934/krm.2012.5.729
[2]	Gang Li, Xianwen Zhang. A Vlasov-Poisson plasma of infinite mass with a point charge. Kinetic & Related Models, 2018, 11 (2) : 303-336. doi: 10.3934/krm.2018015
[3]	Yulia O. Belyaeva, Björn Gebhard, Alexander L. Skubachevskii. A general way to confined stationary Vlasov-Poisson plasma configurations. Kinetic & Related Models, 2021, 14 (2) : 257-282. doi: 10.3934/krm.2021004
[4]	Silvia Caprino, Guido Cavallaro, Carlo Marchioro. A Vlasov-Poisson plasma with unbounded mass and velocities confined in a cylinder by a magnetic mirror. Kinetic & Related Models, 2016, 9 (4) : 657-686. doi: 10.3934/krm.2016011
[5]	Katherine Zhiyuan Zhang. Focusing solutions of the Vlasov-Poisson system. Kinetic & Related Models, 2019, 12 (6) : 1313-1327. doi: 10.3934/krm.2019051
[6]	Blanca Ayuso, José A. Carrillo, Chi-Wang Shu. Discontinuous Galerkin methods for the one-dimensional Vlasov-Poisson system. Kinetic & Related Models, 2011, 4 (4) : 955-989. doi: 10.3934/krm.2011.4.955
[7]	Jack Schaeffer. Global existence for the Vlasov-Poisson system with steady spatial asymptotic behavior. Kinetic & Related Models, 2012, 5 (1) : 129-153. doi: 10.3934/krm.2012.5.129
[8]	Gianluca Crippa, Silvia Ligabue, Chiara Saffirio. Lagrangian solutions to the Vlasov-Poisson system with a point charge. Kinetic & Related Models, 2018, 11 (6) : 1277-1299. doi: 10.3934/krm.2018050
[9]	Zili Chen, Xiuting Li, Xianwen Zhang. The two dimensional Vlasov-Poisson system with steady spatial asymptotics. Kinetic & Related Models, 2017, 10 (4) : 977-1009. doi: 10.3934/krm.2017039
[10]	Meixia Xiao, Xianwen Zhang. On global solutions to the Vlasov-Poisson system with radiation damping. Kinetic & Related Models, 2018, 11 (5) : 1183-1209. doi: 10.3934/krm.2018046
[11]	Jean Dolbeault. An introduction to kinetic equations: the Vlasov-Poisson system and the Boltzmann equation. Discrete & Continuous Dynamical Systems, 2002, 8 (2) : 361-380. doi: 10.3934/dcds.2002.8.361
[12]	Jack Schaeffer. On time decay for the spherically symmetric Vlasov-Poisson system. Kinetic & Related Models, , () : -. doi: 10.3934/krm.2021021
[13]	Gerhard Rein, Christopher Straub. On the transport operators arising from linearizing the Vlasov-Poisson or Einstein-Vlasov system about isotropic steady states. Kinetic & Related Models, 2020, 13 (5) : 933-949. doi: 10.3934/krm.2020032
[14]	Xianglong Duan. Sharp decay estimates for the Vlasov-Poisson and Vlasov-Yukawa systems with small data. Kinetic & Related Models, , () : -. doi: 10.3934/krm.2021049
[15]	Shigeru Takata, Hitoshi Funagane, Kazuo Aoki. Fluid modeling for the Knudsen compressor: Case of polyatomic gases. Kinetic & Related Models, 2010, 3 (2) : 353-372. doi: 10.3934/krm.2010.3.353
[16]	Joackim Bernier, Michel Mehrenberger. Long-time behavior of second order linearized Vlasov-Poisson equations near a homogeneous equilibrium. Kinetic & Related Models, 2020, 13 (1) : 129-168. doi: 10.3934/krm.2020005
[17]	Hyung Ju Hwang, Jaewoo Jung, Juan J. L. Velázquez. On global existence of classical solutions for the Vlasov-Poisson system in convex bounded domains. Discrete & Continuous Dynamical Systems, 2013, 33 (2) : 723-737. doi: 10.3934/dcds.2013.33.723
[18]	Francis Filbet, Roland Duclous, Bruno Dubroca. Analysis of a high order finite volume scheme for the 1D Vlasov-Poisson system. Discrete & Continuous Dynamical Systems - S, 2012, 5 (2) : 283-305. doi: 10.3934/dcdss.2012.5.283
[19]	Dongming Wei. 1D Vlasov-Poisson equations with electron sheet initial data. Kinetic & Related Models, 2010, 3 (4) : 729-754. doi: 10.3934/krm.2010.3.729
[20]	Mihaï Bostan. Asymptotic behavior for the Vlasov-Poisson equations with strong uniform magnetic field and general initial conditions. Kinetic & Related Models, 2020, 13 (3) : 531-548. doi: 10.3934/krm.2020018

2020 Impact Factor: 1.432

Tools

Metrics

Other articles
by authors

on AIMS
Léo Bois Emmanuel Franck Laurent Navoret Vincent Vigon
on Google Scholar
Léo Bois Emmanuel Franck Laurent Navoret Vincent Vigon

2020 Impact Factor: 1.432

Tools

Article outline

Figures and Tables

2020 Impact Factor: 1.432

Tools

Figures and Tables

2020 Impact Factor: 1.432

Tools

Metrics

Other articles
by authors

on AIMS
Léo Bois Emmanuel Franck Laurent Navoret Vincent Vigon
on Google Scholar
Léo Bois Emmanuel Franck Laurent Navoret Vincent Vigon

[Back to Top]