A Nonlinear Physics-based Reduced Order Model with Convolutional-based Operator Compression

DOI10.48550/arXiv.2609.24300MaRDI QIDQ7353438arXivFDO

Authors Anna Ivagnes, Giovanni Stabile, Gianluigi Rozza

Publication date 21 September 2026

Published in arXiv

Tldr We present a nonlinear physics-based reduced order model (ROM) that incorporates nonlinearity into the representation of governing operators, using convolutional architectures to capture spatial structure and radial basis functions for online prediction.

Full work available at URL https://doi.org/10.48550/arXiv.2609.24300

arXiv classification

math.NA

Cites work

• A comprehensive deep learning-based approach to reduced order modeling of nonlinear time-dependent parametrized PDEs
• A continuous convolutional trainable filter for modelling unstructured data
• A deep learning approach to Reduced Order Modelling of parameter dependent partial differential equations
• A fast and accurate physics-informed neural network reduced order model with shallow masked autoencoder
• A graph convolutional autoencoder approach to model order reduction for parametrized PDEs
• A reduced-basis element method
• Accelerating Neural ODEs Using Model Order Reduction
• Approximation of high-dimensional parametric PDEs
• Data-driven forecasting of high-dimensional chaotic systems with long short-term memory networks
• Data-driven reduced-order modeling of spatiotemporal chaos with neural ordinary differential equations
• Deep learning-based reduced order models in cardiac electrophysiology
• DOLFIN: automated finite element computing
• Latent dynamics graph convolutional networks for model order reduction of parameterized time-dependent PDEs
• Linearly recurrent autoencoder networks for learning dynamics
• Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
• On latent dynamics learning in nonlinear reduced order modeling
• Physics-informed machine learning for reduced-order modeling of nonlinear problems
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Physics-informed non-intrusive reduced-order modeling of parameterized dynamical systems
• Reduced Order Modeling for Parameterized Time-Dependent PDEs using Spatially and Memory Aware Deep Learning
• Reduced order modeling of parametrized systems through autoencoders and SINDy approach: continuation of periodic solutions

This page was built for publication: A Nonlinear Physics-based Reduced Order Model with Convolutional-based Operator Compression

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7353438)