The Neural Network shifted-Proper Orthogonal Decomposition: a Machine Learning Approach for Non-linear Reduction of Hyperbolic Equations

Models with dominant advection always posed a difficult challenge for projection-based reduced order modelling. Many methodologies that have recently been proposed are based on the pre-processing of the full-order solutions to accelerate the Kolmogorov N−width decay thereby obtaining smaller linear subspaces with improved accuracy. These methods however must rely on the knowledge of the characteristic speeds in phase space of the solution, limiting their range of applicability to problems with explicit functional form for the advection field. In this work we approach the problem of automatically detecting the correct pre-processing transformation in a statistical learning framework by implementing a deep-learning architecture. The purely data-driven method allowed us to generalise the existing approaches of linear subspace manipulation to non-linear hyperbolic problems with unknown advection fields. The proposed algorithm has been validated against simple test cases to benchmark its performances and later successfully applied to a multiphase simulation. Keywords— Deep Neural Networks (DNNs), shifted-POD (sPOD), Non-linear hyperbolic equations, Reduced Order Modelling (ROM), Multiphase simulation

[1]  Donsub Rim,et al.  Transport Reversal for Model Reduction of Hyperbolic Partial Differential Equations , 2017, SIAM/ASA J. Uncertain. Quantification.

[2]  Gianluigi Rozza,et al.  Overcoming slowly decaying Kolmogorov n-width by transport maps: application to model order reduction of fluid dynamics and fluid-structure interaction problems , 2019, ArXiv.

[3]  Kenji Kashima,et al.  Nonlinear model reduction by deep autoencoder of noise response data , 2016, 2016 IEEE 55th Conference on Decision and Control (CDC).

[4]  Tsuyoshi Murata,et al.  {m , 1934, ACML.

[5]  Diana Baader,et al.  Theoretical Numerical Analysis A Functional Analysis Framework , 2016 .

[6]  A. Quarteroni,et al.  Reduced Basis Methods for Partial Differential Equations: An Introduction , 2015 .

[7]  Davide Torlo,et al.  Model Reduction for Advection Dominated Hyperbolic Problems in an ALE Framework: Offline and Online Phases , 2020, ArXiv.

[8]  M. Darwish,et al.  The Finite Volume Method in Computational Fluid Dynamics: An Advanced Introduction with OpenFOAM® and Matlab , 2015 .

[9]  Lalit K. Mestha,et al.  A deep learning framework for model reduction of dynamical systems , 2017, 2017 IEEE Conference on Control Technology and Applications (CCTA).

[10]  Rémi Abgrall,et al.  Model order reduction for parametrized nonlinear hyperbolic problems as an application to uncertainty quantification , 2018, J. Comput. Appl. Math..

[11]  B. P. Leonard,et al.  A stable and accurate convective modelling procedure based on quadratic upstream interpolation , 1990 .

[12]  Gianluigi Rozza,et al.  Reduced Basis Approximation of Parametrized Advection-Diffusion PDEs with High Péclet Number , 2013, ENUMATH.

[13]  Gianluigi Rozza,et al.  Supremizer stabilization of POD–Galerkin approximation of parametrized steady incompressible Navier–Stokes equations , 2015 .

[14]  G. Rozza,et al.  POD-Galerkin reduced order methods for CFD using Finite Volume Discretisation: vortex shedding around a circular cylinder , 2017, 1701.03424.

[15]  G. G. Stokes "J." , 1890, The New Yale Book of Quotations.

[16]  Benjamin Stamm,et al.  Model Order Reduction for Problems with Large Convection Effects , 2018, Computational Methods in Applied Sciences.

[17]  Volker Mehrmann,et al.  The Shifted Proper Orthogonal Decomposition: A Mode Decomposition for Multiple Transport Phenomena , 2015, SIAM J. Sci. Comput..

[18]  G. W. STEWARTt ON THE EARLY HISTORY OF THE SINGULAR VALUE DECOMPOSITION * , 2022 .

[19]  J. Hesthaven,et al.  Certified Reduced Basis Methods for Parametrized Partial Differential Equations , 2015 .

[20]  Kookjin Lee,et al.  Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders , 2018, J. Comput. Phys..

[21]  A. Quarteroni Numerical Models for Differential Problems , 2009 .

[22]  Kyle T. Mandli,et al.  Model reduction of a parametrized scalar hyperbolic conservation law using displacement interpolation , 2018, 1805.05938.

[23]  Benjamin Peherstorfer,et al.  Model reduction for transport-dominated problems via online adaptive bases and adaptive sampling , 2018, SIAM J. Sci. Comput..

[24]  Annalisa Quaini,et al.  A POD-Galerkin reduced order model for a LES filtering approach , 2020, J. Comput. Phys..

[25]  Tommaso Taddei,et al.  A registration method for model order reduction: data compression and geometry reduction , 2019, SIAM J. Sci. Comput..

[26]  Nirmal J. Nair,et al.  Transported snapshot model order reduction approach for parametric, steady‐state fluid flows containing parameter‐dependent shocks , 2017, International Journal for Numerical Methods in Engineering.

[27]  Gianluigi Rozza,et al.  Stabilized Weighted Reduced Basis Methods for Parametrized Advection Dominated Problems with Random Inputs , 2017, SIAM/ASA J. Uncertain. Quantification.

[28]  Gianluigi Rozza,et al.  Data-Driven POD-Galerkin Reduced Order Model for Turbulent Flows , 2019, J. Comput. Phys..

[29]  Gianluigi Rozza,et al.  Reduced Order Methods for Modeling and Computational Reduction , 2013 .

[30]  Angelo Iollo,et al.  Advection modes by optimal mass transfer. , 2014, Physical review. E, Statistical, nonlinear, and soft matter physics.

[31]  J. Nathan Kutz,et al.  Deep learning in fluid dynamics , 2017, Journal of Fluid Mechanics.

[32]  Albert Cohen,et al.  Kolmogorov widths under holomorphic mappings , 2015, ArXiv.

[33]  Dimitris Drikakis,et al.  Machine-Learning Methods for Computational Science and Engineering , 2020, Comput..

[34]  Zhichao Peng,et al.  A learning-based projection method for model order reduction of transport problems , 2021, ArXiv.

[35]  J. Li,et al.  Numerical simulation of moving contact line problems using a volume-of-fluid method , 2001 .

[36]  F. Sarghini,et al.  Neural networks based subgrid scale modeling in large eddy simulations , 2003 .

[37]  Maciej Balajewicz,et al.  Arbitrary Lagrangian Eulerian framework for efficient projection-based reduction of convection dominated nonlinear flows , 2017 .

[38]  S. P. van den Bosch,et al.  Model reduction methods , 1993 .

[39]  Gianluigi Rozza,et al.  POD–Galerkin monolithic reduced order models for parametrized fluid‐structure interaction problems , 2016 .

[40]  T. Yonomoto,et al.  Numerical Study on an Interface Compression Method for the Volume of Fluid Approach , 2021, Fluids.