Neural-network learning of SPOD latent dynamics

DOI10.1016/J.JCP.2022.111475MaRDI QIDQ2168295zbMATH OpenOpenAlexWikidataFDO

Authors Andrea Lario, Romit Maulik, Oliver T. Schmidt, Gianluigi Rozza, Gianmarco Mengaldo

Publication date 31 August 2022

Published in Journal of Computational Physics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2110.09218

zbMATH Keywords

deep learningdynamical systemsneural networksreduced order modeling

Mathematics Subject Classification ID

Numerical methods for ordinary differential equations (65Lxx) Turbulence (76Fxx) Time-dependent statistical mechanics (dynamic and nonequilibrium) (82Cxx)

Recommendations

• Learning proper orthogonal decomposition of complex dynamics using heavy-ball neural ODEs
• Frequency-time analysis, low-rank reconstruction and denoising of turbulent flows using SPOD
• From snapshots to modal expansions -- bridging low residuals and pure frequencies
• Construction of reduced-order models for fluid flows using deep feedforward neural networks
• Sparse identification of nonlinear dynamics with low-dimensionalized flow representations

Cites work

• scientific article; zbMATH DE number 3430284 (Why is no real title available?)
• Broadband reconstruction of inhomogeneous turbulence using spectral proper orthogonal decomposition and Gabor modes
• Data-driven POD-Galerkin reduced order model for turbulent flows
• Frequency-time analysis, low-rank reconstruction and denoising of turbulent flows using SPOD
• Fundamentals of recurrent neural network (RNN) and long short-term memory (LSTM) network
• Modelling of jet noise: a perspective from large-eddy simulations
• Non-intrusive reduced order modeling of nonlinear problems using neural networks
• Parametric non-intrusive model order reduction for flow-fields using unsupervised machine learning
• Physics-informed machine learning for reduced-order modeling of nonlinear problems
• Recurrent neural network closure of parametric POD-Galerkin reduced-order models based on the Mori-Zwanzig formalism
• Spectral proper orthogonal decomposition and its relationship to dynamic mode decomposition and resolvent analysis
• Statistical Mechanics of Nonequilibrium Liquids
• Time-series learning of latent-space dynamics for reduced-order model closure
• Transport, Collective Motion, and Brownian Motion
• Turbulence and the dynamics of coherent structures. I. Coherent structures

Cited in

(5)

• Unlocking massively parallel spectral proper orthogonal decompositions in the PySPOD package
• Operator inference with roll outs for learning reduced models from scarce and low-quality data
• Neural-network learning of SPOD latent dynamics
• A digital twin framework for civil engineering structures
• Multiscale graph neural network autoencoders for interpretable scientific machine learning

Describes a project that uses

Uses Software

• PyParSVD
• PySPOD
• torchdiffeq
• Tensor2Tensor

This page was built for publication: Neural-network learning of SPOD latent dynamics

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2168295)