Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data

DOI10.1016/j.cma.2019.112732zbMath1442.76096arXiv1906.02382OpenAlexW2948230027MaRDI QIDQ2176917

Han Gao, Shaowu Pan, Jian-Xun Wang, Luning Sun

Publication date: 6 May 2020

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

Full work available at URL: https://arxiv.org/abs/1906.02382

zbMATH Keywords

neural networksNavier-Stokesuncertainty quantificationphysics-informed machine learninglabel-freecardiovascular flows

Mathematics Subject Classification ID

Applications to the sciences (65Z05) Physiological flows (76Z05) Basic methods in fluid mechanics (76M99)

Related Items (94)

Multi-Fidelity Machine Learning Applied to Steady Fluid Flows ⋮ GINNs: graph-informed neural networks for multiscale physics ⋮ Active training of physics-informed neural networks to aggregate and interpolate parametric solutions to the Navier-Stokes equations ⋮ Deep learning of free boundary and Stefan problems ⋮ PhyGeoNet: physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain ⋮ Physics-Driven Learning of the Steady Navier-Stokes Equations using Deep Convolutional Neural Networks ⋮ DeepM\&Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks ⋮ Using neural networks to accelerate the solution of the Boltzmann equation ⋮ Mutual information for explainable deep learning of multiscale systems ⋮ Physics-informed machine learning for reduced-order modeling of nonlinear problems ⋮ Simple computational strategies for more effective physics-informed neural networks modeling of turbulent natural convection ⋮ Learning time-dependent PDEs with a linear and nonlinear separate convolutional neural network ⋮ Thermodynamically consistent physics-informed neural networks for hyperbolic systems ⋮ When and why PINNs fail to train: a neural tangent kernel perspective ⋮ Simulation of the 3D hyperelastic behavior of ventricular myocardium using a finite-element based neural-network approach ⋮ A general neural particle method for hydrodynamics modeling ⋮ A sample-efficient deep learning method for multivariate uncertainty qualification of acoustic-vibration interaction problems ⋮ Physics informed neural networks for continuum micromechanics ⋮ CAN-PINN: a fast physics-informed neural network based on coupled-automatic-numerical differentiation method ⋮ Learning finite element convergence with the multi-fidelity graph neural network ⋮ Residual-based adaptivity for two-phase flow simulation in porous media using physics-informed neural networks ⋮ Surrogate and inverse modeling for two-phase flow in porous media via theory-guided convolutional neural network ⋮ Scientific machine learning through physics-informed neural networks: where we are and what's next ⋮ ReF-nets: physics-informed neural network for Reynolds equation of gas bearing ⋮ A bi-fidelity surrogate modeling approach for uncertainty propagation in three-dimensional hemodynamic simulations ⋮ Improved deep neural networks with domain decomposition in solving partial differential equations ⋮ Physics-informed machine learning for surrogate modeling of wind pressure and optimization of pressure sensor placement ⋮ A physics-informed neural network technique based on a modified loss function for computational 2D and 3D solid mechanics ⋮ Multifidelity deep operator networks for data-driven and physics-informed problems ⋮ VI-DGP: a variational inference method with deep generative prior for solving high-dimensional inverse problems ⋮ Deep learning-accelerated computational framework based on physics informed neural network for the solution of linear elasticity ⋮ One-dimensional ice shelf hardness inversion: clustering behavior and collocation resampling in physics-informed neural networks ⋮ Dynamic analysis on optical pulses via modified PINNs: soliton solutions, rogue waves and parameter discovery of the CQ-NLSE ⋮ Physics-data combined machine learning for parametric reduced-order modelling of nonlinear dynamical systems in small-data regimes ⋮ Data-driven spatiotemporal modeling for structural dynamics on irregular domains by stochastic dependency neural estimation ⋮ Deep learning phase‐field model for brittle fractures ⋮ Isogeometric analysis-based physics-informed graph neural network for studying traffic jam in neurons ⋮ Deep capsule encoder–decoder network for surrogate modeling and uncertainty quantification ⋮ Physics-informed data-driven model for fluid flow in porous media ⋮ Adaptive weighting of Bayesian physics informed neural networks for multitask and multiscale forward and inverse problems ⋮ A metalearning approach for physics-informed neural networks (PINNs): application to parameterized PDEs ⋮ A physics-informed convolutional neural network for the simulation and prediction of two-phase Darcy flows in heterogeneous porous media ⋮ Physics-informed neural networks with adaptive localized artificial viscosity ⋮ Predicting continuum breakdown with deep neural networks ⋮ Learning Markovian Homogenized Models in Viscoelasticity ⋮ Physics informed WNO ⋮ Solving seepage equation using physics-informed residual network without labeled data ⋮ Isogeometric neural networks: a new deep learning approach for solving parameterized partial differential equations ⋮ Transfer learning based physics-informed neural networks for solving inverse problems in engineering structures under different loading scenarios ⋮ Physics-based self-learning spiking neural network enhanced time-integration scheme for computing viscoplastic structural finite element response ⋮ SeismicNET: physics-informed neural networks for seismic wave modeling in semi-infinite domain ⋮ Variable separated physics-informed neural networks based on adaptive weighted loss functions for blood flow model ⋮ Investigation of Low and High-Speed Fluid Dynamics Problems Using Physics-Informed Neural Network ⋮ Label-free learning of elliptic partial differential equation solvers with generalizability across boundary value problems ⋮ NeuralUQ: A Comprehensive Library for Uncertainty Quantification in Neural Differential Equations and Operators ⋮ On the improvement of the extrapolation capability of an iterative machine-learning based RANS framework ⋮ Physics-integrated neural differentiable (PiNDiff) model for composites manufacturing ⋮ Embedding physical knowledge in deep neural networks for predicting the phonon dispersion curves of cellular metamaterials ⋮ An equivariant neural operator for developing nonlocal tensorial constitutive models ⋮ Solving free-surface problems for non-shallow water using boundary and initial conditions-free physics-informed neural network (bif-PINN) ⋮ Automatic boundary fitting framework of boundary dependent physics-informed neural network solving partial differential equation with complex boundary conditions ⋮ Finite basis physics-informed neural networks (FBPINNs): a scalable domain decomposition approach for solving differential equations ⋮ Adaptive transfer learning for PINN ⋮ Hard enforcement of physics-informed neural network solutions of acoustic wave propagation ⋮ An Adaptive Physics-Informed Neural Network with Two-Stage Learning Strategy to Solve Partial Differential Equations ⋮ Spectral operator learning for parametric PDEs without data reliance ⋮ A complete physics-informed neural network-based framework for structural topology optimization ⋮ AONN: An Adjoint-Oriented Neural Network Method for All-At-Once Solutions of Parametric Optimal Control Problems ⋮ Differentiable hybrid neural modeling for fluid-structure interaction ⋮ Physical informed neural networks with soft and hard boundary constraints for solving advection-diffusion equations using Fourier expansions ⋮ Microstructure-informed probability-driven point-particle model for hydrodynamic forces and torques in particle-laden flows ⋮ Forecasting of nonlinear dynamics based on symbolic invariance ⋮ An immersed boundary neural network for solving elliptic equations with singular forces on arbitrary domains ⋮ Physics-informed multi-LSTM networks for metamodeling of nonlinear structures ⋮ Learning nonlocal constitutive models with neural networks ⋮ On the eigenvector bias of Fourier feature networks: from regression to solving multi-scale PDEs with physics-informed neural networks ⋮ Physics-informed neural network for modelling the thermochemical curing process of composite-tool systems during manufacture ⋮ Theory-guided auto-encoder for surrogate construction and inverse modeling ⋮ Efficient uncertainty quantification for dynamic subsurface flow with surrogate by theory-guided neural network ⋮ Discretizationnet: a machine-learning based solver for Navier-Stokes equations using finite volume discretization ⋮ Deep-learning-based surrogate flow modeling and geological parameterization for data assimilation in 3D subsurface flow ⋮ Machine learning for metal additive manufacturing: predicting temperature and melt pool fluid dynamics using physics-informed neural networks ⋮ Understanding and Mitigating Gradient Flow Pathologies in Physics-Informed Neural Networks ⋮ Flow over an espresso cup: inferring 3-D velocity and pressure fields from tomographic background oriented Schlieren via physics-informed neural networks ⋮ Frame-independent vector-cloud neural network for nonlocal constitutive modeling on arbitrary grids ⋮ Variational Inference Formulation for a Model-Free Simulation of a Dynamical System with Unknown Parameters by a Recurrent Neural Network ⋮ PhyCRNet: physics-informed convolutional-recurrent network for solving spatiotemporal PDEs ⋮ Mosaic flows: a transferable deep learning framework for solving PDEs on unseen domains ⋮ Physics-informed graph neural Galerkin networks: a unified framework for solving PDE-governed forward and inverse problems ⋮ A representative volume element network (RVE-net) for accelerating RVE analysis, microscale material identification, and defect characterization ⋮ Physics-based self-learning recurrent neural network enhanced time integration scheme for computing viscoplastic structural finite element response ⋮ Optimal control of PDEs using physics-informed neural networks ⋮ A physics-informed learning approach to Bernoulli-type free boundary problems ⋮ Data-driven soliton solutions and model parameters of nonlinear wave models via the conservation-law constrained neural network method

Uses Software

• OpenFOAM
• Theano
• MLMSRBF
• TensorFlow
• DiffSharp
• PyTorch
• Adam
• DGM
• Deep Speech

Cites Work

• Unnamed Item
• Unnamed Item
• Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations
• Machine learning strategies for systems with invariance properties
• Occam's razor
• 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
• Deep UQ: learning deep neural network surrogate models for high dimensional uncertainty quantification
• DGM: a deep learning algorithm for solving partial differential equations
• Solving the Kolmogorov PDE by means of deep learning
• Structured Bayesian Gaussian process latent variable model: applications to data-driven dimensionality reduction and high-dimensional inversion
• A proof that rectified deep neural networks overcome the curse of dimensionality in the numerical approximation of semilinear heat equations
• Physics-constrained deep learning for high-dimensional surrogate modeling and uncertainty quantification without labeled data
• Adversarial uncertainty quantification in physics-informed neural networks
• Quantifying total uncertainty in physics-informed neural networks for solving forward and inverse stochastic problems
• A composite neural network that learns from multi-fidelity data: application to function approximation and inverse PDE problems
• Simulator-free solution of high-dimensional stochastic elliptic partial differential equations using deep neural networks
• Adaptive activation functions accelerate convergence in deep and physics-informed neural networks
• Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
• Neural algorithm for solving differential equations
• A Survey of Projection-Based Model Reduction Methods for Parametric Dynamical Systems
• A Stochastic Radial Basis Function Method for the Global Optimization of Expensive Functions
• Nonlinear Model Reduction via Discrete Empirical Interpolation
• Uncertainty Quantification and Polynomial Chaos Techniques in Computational Fluid Dynamics
• Spectral Methods for Uncertainty Quantification
• Flows in Stenotic Vessels
• Model Reduction and Approximation
• Data-Driven Science and Engineering
• Solving the quantum many-body problem with artificial neural networks
• Deep learning of vortex-induced vibrations
• The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
• Predicting the output from a complex computer code when fast approximations are available
• Solving high-dimensional partial differential equations using deep learning
• Model Order Reduction in Fluid Dynamics: Challenges and Perspectives
• A GENERAL FRAMEWORK FOR ENHANCING SPARSITY OF GENERALIZED POLYNOMIAL CHAOS EXPANSIONS
• Model Reduction for Transport-Dominated Problems via Online Adaptive Bases and Adaptive Sampling
• Stability of Discrete Empirical Interpolation and Gappy Proper Orthogonal Decomposition with Randomized and Deterministic Sampling Points
• Overcoming the curse of dimensionality in the numerical approximation of semilinear parabolic partial differential equations
• A Proof that Artificial Neural Networks Overcome the Curse of Dimensionality in the Numerical Approximation of Black–Scholes Partial Differential Equations

This page was built for publication: Surrogate modeling for fluid flows based on physics-constrained deep learning without simulation data