Non-intrusive model reduction of large-scale, nonlinear dynamical systems using deep learning
From MaRDI portal
Publication:2127404
DOI10.1016/J.PHYSD.2020.132614zbMATH Open1489.93013arXiv1911.03808OpenAlexW3035079212MaRDI QIDQ2127404FDOQ2127404
Authors: Han Gao, Jian-Xun Wang, Matthew J. Zahr 
Publication date: 20 April 2022
Published in: Physica D (Search for Journal in Brave)
Abstract: Projection-based model reduction has become a popular approach to reduce the cost associated with integrating large-scale dynamical systems so they can be used in many-query settings such as optimization and uncertainty quantification. For nonlinear systems, significant cost reduction is only possible with an additional layer of approximation to reduce the computational bottleneck of evaluating the projected nonlinear terms. Prevailing methods to approximate the nonlinear terms are code intrusive, potentially requiring years of development time to integrate into an existing codebase, and have been known to lack parametric robustness. This work develops a non-intrusive method to efficiently and accurately approximate the expensive nonlinear terms that arise in reduced nonlinear dynamical system using deep neural networks. The neural network is trained using only the simulation data used to construct the reduced basis and evaluations of the nonlinear terms at these snapshots. Once trained, the neural network-based reduced-order model only requires forward and backward propagation through the network to evaluate the nonlinear term and its derivative, which are used to integrate the reduced dynamical system at a new parameter configuration. We provide two numerical experiments---the dynamical systems result from the semi-discretization of parametrized, nonlinear, hyperbolic partial differential equations---that show, in addition to non-intrusivity, the proposed approach provides more stable and accurate approximations to each dynamical system across a large number of training and testing points than the popular empirical interpolation method.
Full work available at URL: https://arxiv.org/abs/1911.03808
Recommendations
- Operator inference for non-intrusive model reduction of systems with non-polynomial nonlinear terms
- A comprehensive deep learning-based approach to reduced order modeling of nonlinear time-dependent parametrized PDEs
- Neural network closures for nonlinear model order reduction
- Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
- Non-intrusive reduced order modeling of nonlinear problems using neural networks
Artificial neural networks and deep learning (68T07) Large-scale systems (93A15) System structure simplification (93B11) Nonlinear systems in control theory (93C10)
Cites Work
- Title not available (Why is that?)
- On early stopping in gradient descent learning
- An `empirical interpolation' method: Application to efficient reduced-basis discretization of partial differential equations
- Nonlinear model reduction via discrete empirical interpolation
- Evaluating Derivatives
- The complexity of partial derivatives
- Turbulence and the dynamics of coherent structures. I. Coherent structures
- Missing Point Estimation in Models Described by Proper Orthogonal Decomposition
- The GNAT method for nonlinear model reduction: effective implementation and application to computational fluid dynamics and turbulent flows
- A priori hyperreduction method: an adaptive approach
- Reduced basis approximation and a posteriori error estimation for affinely parametrized elliptic coercive partial differential equations. Application to transport and continuum mechanics.
- Diagonally Implicit Runge–Kutta Methods for Stiff O.D.E.’s
- Non-intrusive reduced order modelling of the Navier-Stokes equations
- Convergence rates of the POD-greedy method
- Nonlinear model order reduction based on local reduced-order bases
- Linearized reduced-order models for subsurface flow simulation
- Enhanced linearized reduced-order models for subsurface flow simulation
- Structure-preserving, stability, and accuracy properties of the energy-conserving sampling and weighting method for the hyper reduction of nonlinear finite element dynamic models
- Data-driven operator inference for nonintrusive projection-based model reduction
- Non-intrusive reduced order modeling of nonlinear problems using neural networks
- An adjoint method for a high-order discretization of deforming domain conservation laws for optimization of flow problems
- Data-driven discovery of closure models
- Data-driven science and engineering. Machine learning, dynamical systems, and control
- Non-intrusive reduced order modeling of unsteady flows using artificial neural networks with application to a combustion problem
- Neural network closures for nonlinear model order reduction
- Machine learning for fast and reliable solution of time-dependent differential equations
- Machine learning closures for model order reduction of thermal fluids
- A fully discrete adjoint method for optimization of flow problems on deforming domains with time-periodicity constraints
- Nonintrusive reduced-order modeling of parametrized time-dependent partial differential equations
- An artificial neural network framework for reduced order modeling of transient flows
- Construction of reduced-order models for fluid flows using deep feedforward neural networks
- Energetically optimal flapping wing motions via adjoint-based optimization and high-order discretizations
Cited In (22)
- Deep subspace encoders for nonlinear system identification
- Deep-HyROMnet: a deep learning-based operator approximation for hyper-reduction of nonlinear parametrized PDEs
- The model reduction of the Vlasov–Poisson–Fokker–Planck system to the Poisson–Nernst–Planck system via the Deep Neural Network Approach
- Bifidelity data-assisted neural networks in nonintrusive reduced-order modeling
- Operator inference for non-intrusive model reduction with quadratic manifolds
- Operator learning using random features: a tool for scientific computing
- Neural network closures for nonlinear model order reduction
- Level Set Learning with Pseudoreversible Neural Networks for Nonlinear Dimension Reduction in Function Approximation
- Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
- Data-driven nonintrusive reduced order modeling for dynamical systems with moving boundaries using Gaussian process regression
- Operator inference for non-intrusive model reduction of systems with non-polynomial nonlinear terms
- Transformers for modeling physical systems
- Reduced-order deep learning for flow dynamics. The interplay between deep learning and model reduction
- Symplectic model reduction of Hamiltonian systems using data-driven quadratic manifolds
- A pruning algorithm preserving modeling capabilities for polycrystalline data
- Data-driven reduced order modeling of poroelasticity of heterogeneous media based on a discontinuous Galerkin approximation
- CD-ROM: complemented deep -- reduced order model
- Fast reduced model of non-linear dynamic Euler-Bernoulli beam behaviour
- Non-intrusive parametric hyper-reduction for nonlinear structural finite element formulations
- Network flow problem heuristic reduction using machine learning
- Applied unsupervised learning in model reduction of linear dynamic systems
- Leveraging reduced-order models for state estimation using deep learning
Uses Software
This page was built for publication: Non-intrusive model reduction of large-scale, nonlinear dynamical systems using deep learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2127404)
