Abstract:Nearly all model-reduction techniques project the governing equations onto a linear subspace of the original state space. Such subspaces are typically computed using methods such as balanced truncation, rational interpolation, the reduced-basis method, and (balanced) proper orthogonal decomposition (POD). Unfortunately, restricting the state to evolve in a linear subspace imposes a fundamental limitation to the accuracy of the resulting reduced-order model (ROM). In particular, linear-subspace ROMs can be expected to produce low-dimensional models with high accuracy only if the problem admits a fast decaying Kolmogorov n-width (e.g., diffusion-dominated problems). Unfortunately, many problems of interest exhibit a slowly decaying Kolmogorov n-width (e.g., advection-dominated problems). To address this, we propose a novel framework for projecting dynamical systems onto nonlinear manifolds using minimum-residual formulations at the time-continuous and time-discrete levels; the former leads to manifold Galerkin projection, while the latter leads to manifold least-squares Petrov–Galerkin (LSPG) projection. We perform analyses that provide insight into the relationship between these proposed approaches and classical linear-subspace reduced-order models; we also derive a posteriori discrete-time error bounds for the proposed approaches. In addition, we propose a computationally practical approach for computing the nonlinear manifold, which is based on convolutional autoencoders from deep learning. Finally, we demonstrate the ability of the method to significantly outperform even the optimal linear-subspace ROM on benchmark advection-dominated problems, thereby demonstrating the method's ability to overcome the intrinsic n-width limitations of linear subspaces.

An autoencoder‐based reduced‐order model for eigenvalue problems with application to neutron diffusion

Neural Network Emulation of Spatio-temporal Data Using Linear and Nonlinear Dimensionality Reduction

Subspace method based on neural networks for eigenvalue problems

Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders

Non-intrusive reduced order modeling of natural convection in porous media using convolutional autoencoders: comparison with linear subspace techniques

Eigenvalue initialisation and regularisation for Koopman autoencoders

A Deep Learning approach to Reduced Order Modelling of Parameter Dependent Partial Differential Equations

Balanced Neural ODEs: nonlinear model order reduction and Koopman operator approximations

Reduced-order modeling of advection-dominated systems with recurrent neural networks and convolutional autoencoders

Reduced basis method for non-symmetric eigenvalue problems: application to the multigroup neutron diffusion equations

Data-driven reduced order modeling of environmental hydrodynamics using deep autoencoders and neural ODEs

Learning a Reduced Basis of Dynamical Systems using an Autoencoder

Operator SVD with Neural Networks via Nested Low-Rank Approximation

A practical existence theorem for reduced order models based on convolutional autoencoders

Non-linear Manifold Reduced-Order Models with Convolutional Autoencoders and Reduced Over-Collocation Method

Neural networks based approach for computing eigenvectors and eigenvalues of symmetric matrix

Model Order Reduction for the 1D Boltzmann-BGK Equation: Identifying Intrinsic Variables Using Neural Networks

Learning Nonlinear Projections for Reduced-Order Modeling of Dynamical Systems using Constrained Autoencoders

Solving high-dimensional eigenvalue problems using deep neural networks: A diffusion Monte Carlo like approach

Hyper-Reduced Autoencoders for Efficient and Accurate Nonlinear Model Reductions

A graph convolutional autoencoder approach to model order reduction for parametrized PDEs