Non‐intrusive reduced‐order modeling using convolutional autoencoders

DOI10.1002/NME.7072MaRDI QIDQ6092270zbMATH OpenOpenAlexFDO

Authors Krzysztof J. Fidkowski, Kevin J. Maki

Publication date 23 November 2023

Published in International Journal for Numerical Methods in Engineering (Search for Journal in Brave)

Copyright license Creative Commons Attribution 4.0 International

Full work available at URL https://arxiv.org/abs/2204.08280

zbMATH Keywords

deep learningmachine learningmodel reductionproper orthogonal decompositionautoencodersnon-intrusive

Mathematics Subject Classification ID

Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65Mxx) Incompressible viscous fluids (76Dxx) Numerical linear algebra (65Fxx)

Abstract: The use of reduced-order models (ROMs) in physics-based modeling and simulation almost always involves the use of linear reduced basis (RB) methods such as the proper orthogonal decomposition (POD). For some nonlinear problems, linear RB methods perform poorly, failing to provide an efficient subspace for the solution space. The use of nonlinear manifolds for ROMs has gained traction in recent years, showing increased performance for certain nonlinear problems over linear methods. Deep learning has been popular to this end through the use of autoencoders for providing a nonlinear trial manifold for the solution space. In this work, we present a non-intrusive ROM framework for steady-state parameterized partial differential equations (PDEs) that uses convolutional autoencoders (CAEs) to provide a nonlinear solution manifold and is augmented by Gaussian process regression (GPR) to approximate the expansion coefficients of the reduced model. When applied to a numerical example involving the steady incompressible Navier-Stokes equations solving a lid-driven cavity problem, it is shown that the proposed ROM offers greater performance in prediction of full-order states when compared to a popular method employing POD and GPR over a number of ROM dimensions.

Cites work

• A calculation procedure for heat, mass and momentum transfer in three- dimensional parabolic flows
• A comprehensive deep learning-based approach to reduced order modeling of nonlinear time-dependent parametrized PDEs
• Adaptive \(h\)-refinement for reduced-order models
• Advanced Lectures on Machine Learning
• Advances in natural language processing
• An efficient algorithm for constructing optimal design of computer experiments
• Deep learning
• Efficient non linear model reduction via a least-squares Petrov-Galerkin projection and compressive tensor approximations
• Galerkin v. least-squares Petrov-Galerkin projection in nonlinear model reduction
• Learning representations by back-propagating errors
• Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
• Multi-level convolutional autoencoder networks for parametric prediction of spatio-temporal dynamics
• Non-intrusive reduced order modeling of nonlinear problems using neural networks
• Non-intrusive reduced-order modeling using uncertainty-aware deep neural networks and proper orthogonal decomposition: application to flood modeling
• Nonlinear model order reduction based on local reduced-order bases
• Nonlinear model reduction via discrete empirical interpolation
• Numerical study of the turbulent flow past an airfoil with trailing edge separation
• Reduced order modeling for nonlinear structural analysis using Gaussian process regression
• Reducing the Dimensionality of Data with Neural Networks
• Scikit-learn: machine learning in Python

Cited in

(2)

• Theory-guided auto-encoder for surrogate construction and inverse modeling
• A long short-term memory embedding for hybrid uplifted reduced order models

This page was built for publication: Non‐intrusive reduced‐order modeling using convolutional autoencoders

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6092270)