A Reduced Order Schwarz Method for Nonlinear Multiscale Elliptic Equations Based on Two-Layer Neural Networks

DOI10.4208/JCM.2204-M2021-0311MaRDI QIDQ6197994zbMATH OpenOpenAlexFDO

Authors Shi Chen, Zhiyan Ding, Qin Li, Stephen J. Wright

Publication date 20 February 2024

Published in Journal of Computational Mathematics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2111.02280

zbMATH Keywords

domain decompositionmultiscale elliptic problemneural networksnonlinear homogenization

Mathematics Subject Classification ID

Nonlinear boundary value problems for nonlinear elliptic equations (35J66) Approximation by arbitrary nonlinear expressions; widths and entropy (41A46) Multigrid methods; domain decomposition for boundary value problems involving PDEs (65N55) Artificial neural networks and deep learning (68T07)

Abstract: Neural networks are powerful tools for approximating high dimensional data that have been used in many contexts, including solution of partial differential equations (PDEs). We describe a solver for multiscale fully nonlinear elliptic equations that makes use of domain decomposition, an accelerated Schwarz framework, and two-layer neural networks to approximate the boundary-to-boundary map for the subdomains, which is the key step in the Schwarz procedure. Conventionally, the boundary-to-boundary map requires solution of boundary-value elliptic problems on each subdomain. By leveraging the compressibility of multiscale problems, our approach trains the neural network offline to serve as a surrogate for the usual implementation of the boundary-to-boundary map. Our method is applied to a multiscale semilinear elliptic equation and a multiscale $p$-Laplace equation. In both cases we demonstrate significant improvement in efficiency as well as good accuracy and generalization performance.

Recommendations

• A multi-scale DNN algorithm for nonlinear elliptic equations with multiple scales
• Solving multiscale elliptic problems by sparse radial basis function neural networks
• Solving a class of multi-scale elliptic PDEs by means of Fourier-based mixed physics informed neural networks
• Deep neural network approximation for high-dimensional elliptic PDEs with boundary conditions

Cites work

• A localized orthogonal decomposition method for semi-linear elliptic problems
• A low-rank Schwarz method for radiative transfer equation with heterogeneous scattering coefficient
• A multi-scale DNN algorithm for nonlinear elliptic equations with multiple scales
• A multiscale finite element method for elliptic problems in composite materials and porous media
• A multiscale model reduction method for nonlinear monotone elliptic equations in heterogeneous media
• A Multiscale Neural Network Based on Hierarchical Matrices
• Analysis of the finite element heterogeneous multiscale method for quasilinear elliptic homogenization problems
• Analysis of the heterogeneous multiscale method for elliptic homogenization problems
• Analysis of the Multiscale Finite Element Method for Nonlinear and Random Homogenization Problems
• Analytic regularity and polynomial approximation of parametric and stochastic elliptic PDE's
• BCR-net: A neural network based on the nonstandard wavelet form
• Controlling oscillations in high-order discontinuous Galerkin schemes using artificial viscosity tuned by neural networks
• Convergence of a multiscale finite element method for elliptic problems with rapidly oscillating coefficients
• Convergence of a Nonconforming Multiscale Finite Element Method
• Data-driven deep learning of partial differential equations in modal space
• Deep learning
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• DGM: a deep learning algorithm for solving partial differential equations
• Exploring the locally low dimensional structure in solving random elliptic PDEs
• Finite elements for elliptic problems with stochastic coefficients
• Galerkin Finite Element Approximations of Stochastic Elliptic Partial Differential Equations
• Generalized multiscale finite element methods. Nonlinear elliptic equations
• Heterogeneous Multiscale FEM for Diffusion Problems on Rough Surfaces
• Hierarchical matrices: algorithms and analysis
• Homogenization and Two-Scale Convergence
• scientific article; zbMATH DE number 4076515 (Why is no real title available?)
• scientific article; zbMATH DE number 4115880 (Why is no real title available?)
• scientific article; zbMATH DE number 818071 (Why is no real title available?)
• Iterated numerical homogenization for multiscale elliptic equations with monotone nonlinearity
• Learning data-driven discretizations for partial differential equations
• Localization of elliptic multiscale problems
• Meta-learning pseudo-differential operators with deep neural networks
• Metric-based upscaling
• Multi-level Monte Carlo finite element method for elliptic PDEs with stochastic coefficients
• Multigrid with Rough Coefficients and Multiresolution Operator Decomposition from Hierarchical Information Games
• Multiscale Finite Element Methods
• Multiscale finite element methods for nonlinear problems and their applications
• Numerical methods for stochastic computations. A spectral method approach.
• Optimal Local Approximation Spaces for Generalized Finite Element Methods with Application to Multiscale Problems
• Periodic homogenisation of certain fully nonlinear partial differential equations
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Polyharmonic homogenization, rough polyharmonic splines and sparse super-localization
• Preconditioned descent algorithms for \(p\)-Laplacian
• Random sampling and efficient algorithms for multiscale PDEs
• Randomized Sampling for Basis Function Construction in Generalized Finite Element Methods
• Reduced basis finite element heterogeneous multiscale method for high-order discretizations of elliptic homogenization problems
• Reduced basis finite element heterogeneous multiscale method for quasilinear elliptic homogenization problems
• Reduced-order deep learning for flow dynamics. The interplay between deep learning and model reduction
• Reduced-order multiscale modeling of nonlinear p-Laplacian flows in high-contrast media
• Schwarz methods over the course of time
• Solving for high-dimensional committor functions using artificial neural networks
• Solving parametric PDE problems with artificial neural networks
• Sparse high order FEM for elliptic sPDEs
• SwitchNet: a neural Network Model for forward and inverse scattering problems
• The Barron space and the flow-induced function spaces for neural network models
• The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems
• The heterogeneous multiscale methods
• THE PARTITION OF UNITY METHOD
• The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
• Universal approximation bounds for superpositions of a sigmoidal function
• Variational training of neural network approximations of solution maps for physical models
• Viscosity solutions of fully nonlinear second-order elliptic partial differential equations
• Weak adversarial networks for high-dimensional partial differential equations
• Why Finite Element Discretizations Can Be Factored by Triangular Hierarchical Matrices

Cited in

(4)

• Solving multiscale elliptic problems by sparse radial basis function neural networks
• Solving a class of multi-scale elliptic PDEs by means of Fourier-based mixed physics informed neural networks
• A stochastic approach for elliptic problems in perforated domains
• A Neural-network-based multiscale Hybridizable Discontinuous Galerkin method for solving PDEs in porous media

This page was built for publication: A Reduced Order Schwarz Method for Nonlinear Multiscale Elliptic Equations Based on Two-Layer Neural Networks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6197994)