PhyCRNet: physics-informed convolutional-recurrent network for solving spatiotemporal PDEs

DOI10.1016/j.cma.2021.114399OpenAlexW3173549089WikidataQ114196863 ScholiaQ114196863MaRDI QIDQ2072500

Yanyan Li

Publication date: 26 January 2022

Published in: Computer Methods in Applied Mechanics and Engineering (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2106.14103

zbMATH Keywords

partial differential equationsphysics-informed deep learningconvolutional-recurrent learningencoder-decoderhard-encoding of I/BCsresidual connection

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems (65M99)

Related Items

CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method, Physics-informed PointNet: a deep learning solver for steady-state incompressible flows and thermal fields on multiple sets of irregular geometries, PhySR: physics-informed deep super-resolution for spatiotemporal data, Discovering stochastic partial differential equations from limited data using variational Bayes inference, Variable linear transformation improved physics-informed neural networks to solve thin-layer flow problems, A new method for solving nonlinear partial differential equations based on liquid time-constant networks, Deep convolutional Ritz method: parametric PDE surrogates without labeled data, SeismicNET: physics-informed neural networks for seismic wave modeling in semi-infinite domain, Physics-integrated neural differentiable (PiNDiff) model for composites manufacturing, Physics-informed ConvNet: learning physical field from a shallow neural network, An unsupervised latent/output physics-informed convolutional-LSTM network for solving partial differential equations using peridynamic differential operator

Uses Software

• L-BFGS
• PyTorch
• Adam
• DeepXDE
• torchdiffeq
• PDE-Net
• ConvPDE-UQ
• PhyCNN
• PhyGeoNet
• DeepONet
• NSFnets
• DiscretizationNet

Cites Work

• Isogeometric analysis: CAD, finite elements, NURBS, exact geometry and mesh refinement
• On the limited memory BFGS method for large scale optimization
• DeepONet
• Multilayer feedforward networks are universal approximators
• Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification
• The Deep Ritz Method: a deep learning-based numerical algorithm for solving variational problems
• Machine learning in cardiovascular flows modeling: predicting arterial blood pressure from non-invasive 4D flow MRI data using physics-informed neural networks
• A physics-informed operator regression framework for extracting data-driven continuum models
• The neural particle method - an updated Lagrangian physics informed neural network for computational fluid dynamics
• Discretizationnet: a machine-learning based solver for Navier-Stokes equations using finite volume discretization
• A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics
• Non-invasive inference of thrombus material properties with physics-informed neural networks
• NSFnets (Navier-Stokes flow nets): physics-informed neural networks for the incompressible Navier-Stokes equations
• Physics-informed neural networks for high-speed flows
• Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data
• Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
• ConvPDE-UQ: convolutional neural networks with quantified uncertainty for heterogeneous elliptic partial differential equations on varied domains
• Data driven governing equations approximation using deep neural networks
• Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
• Modeling the dynamics of PDE systems with physics-constrained deep auto-regressive networks
• Physics-informed multi-LSTM networks for metamodeling of nonlinear structures
• A physics-constrained data-driven approach based on locally convex reconstruction for noisy database
• An energy approach to the solution of partial differential equations in computational mechanics via machine learning: concepts, implementation and applications
• The effect of prior probabilities on quantification and propagation of imprecise probabilities resulting from small datasets
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Prediction of aerodynamic flow fields using convolutional neural networks
• Machine Learning for Fluid Mechanics
• Deep Hidden Physics Models: Deep Learning of Nonlinear Partial Differential Equations
• Deep learning of vortex-induced vibrations
• Solving high-dimensional partial differential equations using deep learning
• Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations
• MFNets: MULTI-FIDELITY DATA-DRIVEN NETWORKS FOR BAYESIAN LEARNING AND PREDICTION
• DeepXDE: A Deep Learning Library for Solving Differential Equations
• Sparse dynamics for partial differential equations
• Learning data-driven discretizations for partial differential equations
• Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks