A numerical method, based on the design of two artificial neural networks, is presented in order to approximate the viscosity and density features of fluids from the eigenvalues of the Stokes operator. The finite element method is used to solve the direct problem by training a first artificial neural network. A nonlinear map of eigenvalues of the Stokes operator as a function of the viscosity and density of the fluid under study is then obtained. This relationship is later inverted and refined by training a second artificial neural network, solving the aforementioned inverse problem. Numerical examples are presented in order to show the effectiveness and the limitations of this methodology.
Accepté le :
Publié le :
Sebastián Ossandón 1 ; Mauricio Barrientos 1 ; Camilo Reyes 2
@article{CRMECA_2018__346_1_39_0, author = {Sebasti\'an Ossand\'on and Mauricio Barrientos and Camilo Reyes}, title = {Neural network solution to an inverse problem associated with the eigenvalues of the {Stokes} operator}, journal = {Comptes Rendus. M\'ecanique}, pages = {39--47}, publisher = {Elsevier}, volume = {346}, number = {1}, year = {2018}, doi = {10.1016/j.crme.2017.11.006}, language = {en}, }
TY - JOUR AU - Sebastián Ossandón AU - Mauricio Barrientos AU - Camilo Reyes TI - Neural network solution to an inverse problem associated with the eigenvalues of the Stokes operator JO - Comptes Rendus. Mécanique PY - 2018 SP - 39 EP - 47 VL - 346 IS - 1 PB - Elsevier DO - 10.1016/j.crme.2017.11.006 LA - en ID - CRMECA_2018__346_1_39_0 ER -
%0 Journal Article %A Sebastián Ossandón %A Mauricio Barrientos %A Camilo Reyes %T Neural network solution to an inverse problem associated with the eigenvalues of the Stokes operator %J Comptes Rendus. Mécanique %D 2018 %P 39-47 %V 346 %N 1 %I Elsevier %R 10.1016/j.crme.2017.11.006 %G en %F CRMECA_2018__346_1_39_0
Sebastián Ossandón; Mauricio Barrientos; Camilo Reyes. Neural network solution to an inverse problem associated with the eigenvalues of the Stokes operator. Comptes Rendus. Mécanique, Volume 346 (2018) no. 1, pp. 39-47. doi : 10.1016/j.crme.2017.11.006. https://comptes-rendus.academie-sciences.fr/mecanique/articles/10.1016/j.crme.2017.11.006/
[1] Numerical solution of the Navier–Stokes equations, Math. Comput., Volume 22 (1968), pp. 745-762
[2] Implementation of Finite Element Methods for Navier–Stokes Equations, Springer-Verlag, Berlin, 1981
[3] The Finite Element Method: Linear Static and Dynamic Finite Element Analysis, Dover Publications, Mineola, NY, USA, 2000
[4] Variational multiscale residual-based turbulence modeling for large eddy simulation of incompressible flows, Comput. Methods Appl. Mech. Eng., Volume 197 (2007), pp. 173-201
[5] Eigenvalue approximation by mixed and hybrid methods, Math. Comput., Volume 36 (1981), pp. 427-453
[6] A mesh free method based on radial basis functions for the eigenvalues of transient Stokes equations, Eng. Anal. Bound. Elem., Volume 36 (2012), pp. 1555-1559
[7] A numerical existence proof of nodal lines for the first eigenfunction of the plate equation, Arch. Math., Volume 66 (1996), pp. 420-427
[8] Are there localized eddies in the trihedral corners of the Stokes eigenmodes in cubical cavity?, Comput. Fluids, Volume 43 (2011), pp. 98-101
[9] A feed-forward neural network for solving Stokes problem, Acta Appl. Math., Volume 116 (2011), pp. 55-64
[10] Artificial neural network method for solving the Navier–Stokes equations, Neural Comput. Appl., Volume 26 (2015), pp. 765-773
[11] Regularization theory and neural networks architectures, J. Neural Comput., Volume 7 (1995), pp. 219-269
[12] On the neural network calculation of the Lamé coefficients through eigenvalues of the elasticity operator, C. R. Mecanique, Volume 344 (2016), pp. 113-118
[13] Neural network solution of an inverse problem associated with the Dirichlet eigenvalues of the anisotropic Laplace operator, Comput. Math. Appl., Volume 72 (2016), pp. 1153-1163
[14] Neural network approach for the calculation of potential coefficients in quantum mechanics, Comput. Phys. Commun., Volume 214 (2017), pp. 31-38
[15] Approximation of nonlinear systems with radial basis function neural networks, IEEE Trans. Neural Netw., Volume 12 (2001) no. 1, pp. 1-15
[16] Eigenvalue problems (P.G. Ciarlet; J.-L. Lions, eds.), Handbook of Numerical Analysis, vol. II: Finite Element Methods (Part 1), North-Holland, Amsterdam, 1991
[17] Mixed Finite Element Methods and Applications, Springer Series in Computational Mathematics, vol. 44, Springer, Heidelberg, 2013
[18] Mixed and Hybrid Finite Element Methods, Springer Series in Computational Mathematics, vol. 15, Springer-Verlag, 1991
[19] Finite Element Methods for Navier–Stokes Equations, Springer Verlag, 1986
[20] Approximation and eigenvalue extrapolation of Stokes eigenvalue problem by nonconforming finite element methods, Appl. Math. Czech., Volume 54 (2009), pp. 1-15
[21] Asymptotic expansions and extrapolations of eigenvalues for the Stokes problem by mixed finite element methods, J. Comput. Appl. Math., Volume 215 (2008), pp. 127-141
Cité par Sources :
Commentaires - Politique