Data driven approximation of parametrized PDEs by reduced basis and neural networks
From MaRDI portal
Publication:782002
DOI10.1016/J.JCP.2020.109550zbMATH Open1437.65224arXiv1904.01514OpenAlexW2930017973WikidataQ114163503 ScholiaQ114163503MaRDI QIDQ782002FDOQ782002
Luca Pegolotti, Simone Deparis, Niccolò Dal Santo
Publication date: 21 July 2020
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: We are interested in the approximation of partial differential equations with a data-driven approach based on the reduced basis method and machine learning. We suppose that the phenomenon of interest can be modeled by a parametrized partial differential equation, but that the value of the physical parameters is unknown or difficult to be directly measured. Our method allows to estimate fields of interest, for instance temperature of a sample of material or velocity of a fluid, given data at a handful of points in the domain. We propose to accomplish this task with a neural network embedding a reduced basis solver as exotic activation function in the last layer. The reduced basis solver accounts for the underlying physical phenomenonon and it is constructed from snapshots obtained from randomly selected values of the physical parameters during an expensive offline phase. The same full order solutions are then employed for the training of the neural network. As a matter of fact, the chosen architecture resembles an asymmetric autoencoder in which the decoder is the reduced basis solver and as such it does not contain trainable parameters. The resulting latent space of our autoencoder includes parameter-dependent quantities feeding the reduced basis solver, which -- depending on the considered partial differential equation -- are the values of the physical parameters themselves or the affine decomposition coefficients of the differential operators.
Full work available at URL: https://arxiv.org/abs/1904.01514
machine learningparameter identificationNavier-Stokes equationsreduced basis methoddeep neural networksparametrized PDEs
Artificial neural networks and deep learning (68T07) Navier-Stokes equations for incompressible viscous fluids (76D05) Dependence of solutions to PDEs on initial and/or boundary data and/or on parameters of PDEs (35B30) Numerical methods for partial differential equations, boundary value problems (65N99)
Cites Work
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Reduced Basis Methods for Partial Differential Equations
- Learning Deep Architectures for AI
- An `empirical interpolation' method: Application to efficient reduced-basis discretization of partial differential equations
- Nonlinear Model Reduction via Discrete Empirical Interpolation
- On the stability of the reduced basis method for Stokes equations in parametrized domains
- Approximation by superpositions of a sigmoidal function
- A reduced basis element method for the steady Stokes problem
- Certified Reduced Basis Methods for Parametrized Partial Differential Equations
- A Petrov-Galerkin reduced basis approximation of the Stokes equation in parameterized geometries
- An efficient computational framework for reduced basis approximation and a posteriori error estimation of parametrized Navier-Stokes flows
- Approximation properties of a multilayered feedforward artificial neural network
- Supremizer stabilization of POD-Galerkin approximation of parametrized steady incompressible Navier-Stokes equations
- Non-intrusive reduced order modeling of nonlinear problems using neural networks
- The Deep Ritz Method: a deep learning-based numerical algorithm for solving variational problems
- An algebraic least squares reduced basis method for the solution of nonaffinely parametrized Stokes equations
- Non-intrusive reduced order modeling of unsteady flows using artificial neural networks with application to a combustion problem
- Hyper-reduced order models for parametrized unsteady Navier-Stokes equations on domains with variable shape
- Efficient model reduction of parametrized systems by matrix discrete empirical interpolation
- Deep learning in high dimension: Neural network expression rates for generalized polynomial chaos expansions in UQ
- Optimal Approximation with Sparsely Connected Deep Neural Networks
Cited In (30)
- CAS4DL: Christoffel adaptive sampling for function approximation via deep learning
- Solving parametric partial differential equations with deep rectified quadratic unit neural networks
- Numerical solutions of boundary problems in partial differential equations: a deep learning framework with Green's function
- AI‐enhanced iterative solvers for accelerating the solution of large‐scale parametrized systems
- The Random Feature Model for Input-Output Maps between Banach Spaces
- Operator learning using random features: a tool for scientific computing
- Towards a machine learning pipeline in reduced order modelling for inverse problems: neural networks for boundary parametrization, dimensionality reduction and solution manifold approximation
- Deep neural networks and adaptive quadrature for solving variational problems
- Numerical solution of the parametric diffusion equation by deep neural networks
- JAX-DIPS: neural bootstrapping of finite discretization methods and application to elliptic problems with discontinuities
- PDE-Aware Deep Learning for Inverse Problems in Cardiac Electrophysiology
- Conditional variational autoencoder with Gaussian process regression recognition for parametric models
- Linear/ridge expansions: enhancing linear approximations by ridge functions
- Nonlinear Reduced DNN Models for State Estimation
- Retracted: Model order reduction method based on machine learning for parameterized time-dependent partial differential equations
- A deep learning approach to Reduced Order Modelling of parameter dependent partial differential equations
- Error assessment of an adaptive finite elements -- neural networks method for an elliptic parametric PDE
- Surrogate convolutional neural network models for steady computational fluid dynamics simulations
- Parametric model reduction with convolutional neural networks
- A theoretical analysis of deep neural networks and parametric PDEs
- Physics constrained learning for data-driven inverse modeling from sparse observations
- Learning high-dimensional parametric maps via reduced basis adaptive residual networks
- Data assimilation predictive GAN (DA-PredGAN) applied to a spatio-temporal compartmental model in epidemiology
- Be greedy and learn: efficient and certified algorithms for parametrized optimal control problems
- Learning time-dependent PDEs with a linear and nonlinear separate convolutional neural network
- A non-intrusive model order reduction approach for parameterized time-domain Maxwell's equations
- Solving inverse-PDE problems with physics-aware neural networks
- Addressing discontinuous root-finding for subsequent differentiability in machine learning, inverse problems, and control
- DeepBND: a machine learning approach to enhance multiscale solid mechanics
- A comparison of neural network architectures for data-driven reduced-order modeling
Uses Software
This page was built for publication: Data driven approximation of parametrized PDEs by reduced basis and neural networks
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q782002)
