Physics-Informed Neural Operator for Learning Partial Differential Equations (2021-11-06T00:00:00.000000Z)

TL;DR

PINO is the first hybrid approach incorporating data and PDE constraints at different resolutions to learn the operator of a given family of parametric Partial Differential Equations (PDE), and achieves a high-fidelity reconstruction of the ground-truth operator.

Abstract

In this paper, we propose physics-informed neural operators (PINO) that combine training data and physics constraints to learn the solution operator of a given family of parametric Partial Differential Equations (PDE). PINO is the first hybrid approach incorporating data and PDE constraints at different resolutions to learn the operator. Specifically, in PINO, we combine coarse-resolution training data with PDE constraints imposed at a higher resolution. The resulting PINO model can accurately approximate the ground-truth solution operator for many popular PDE families and shows no degradation in accuracy even under zero-shot super-resolution, i.e., being able to predict beyond the resolution of training data. PINO uses the Fourier neural operator (FNO) framework that is guaranteed to be a universal approximator for any continuous operator and discretization convergent in the limit of mesh refinement. By adding PDE constraints to FNO at a higher resolution, we obtain a high-fidelity reconstruction of the ground-truth operator. Moreover, PINO succeeds in settings where no training data is available and only PDE constraints are imposed, while previous approaches, such as the Physics-Informed Neural Network (PINN), fail due to optimization challenges, e.g., in multi-scale dynamic systems such as Kolmogorov flows.

Authors

K. Azizzadenesheli

6 papers

Anima Anandkumar

11 papers

Zong-Yi Li

4 papers

TL;DR

Abstract

Authors

References65 items

Spherical Fourier Neural Operators: Learning Stable Dynamics on the Sphere

Real-time High-resolution CO2 Geological Storage Prediction using Nested Fourier Neural Operators

Nonlinear Reconstruction for Operator Learning of PDEs with Discontinuities

Fourier Neural Operator with Learned Deformations for PDEs on General Geometries

Spectral Neural Operators

The Cost-Accuracy Trade-Off In Operator Learning With Neural Networks

Generic Lithography Modeling with Dual-band Optics-Inspired Neural Networks

On the influence of over-parameterization in manifold based surrogates and deep neural operators

FourCastNet: A Global Data-driven High-resolution Weather Model using Adaptive Fourier Neural Operators

Meta-Auto-Decoder for Solving Parametric Partial Differential Equations

A comprehensive and fair comparison of two neural operators (with practical extensions) based on FAIR data

FC-based shock-dynamics solver with neural-network localized artificial-viscosity assignment

Neural Operator: Learning Maps Between Function Spaces

Data-informed deep optimization

MOD-Net: A Machine Learning Approach via Model-Operator-Data Network for Solving PDEs

Physics-informed neural networks (PINNs) for fluid mechanics: a review

Learning the solution operator of parametric partial differential equations with physics-informed DeepONets

Physics-informed machine learning: case studies for weather and climate modelling

A learning-based multiscale method and its application to inelastic impact problems

Machine learning–accelerated computational fluid dynamics

HypoSVI: Hypocenter inversion with Stein variational inference and Physics Informed Neural Networks

Understanding and Mitigating Gradient Flow Pathologies in Physics-Informed Neural Networks

Fourier Neural Operator for Parametric Partial Differential Equations

A physics-informed operator regression framework for extracting data-driven continuum models

Self-Adaptive Physics-Informed Neural Networks using a Soft Attention Mechanism

When and why PINNs fail to train: A neural tangent kernel perspective

Locally adaptive activation functions with slope recovery for deep and physics-informed neural networks

Solver-in-the-Loop: Learning from Differentiable Physics to Interact with Iterative PDE-Solvers

Multipole Graph Neural Operator for Parametric Partial Differential Equations

The ERA5 global reanalysis

The Random Feature Model for Input-Output Maps between Banach Spaces

86 PFLOPS Deep Potential Molecular Dynamics simulation of 100 million atoms with ab initio accuracy

PhyGeoNet: Physics-informed geometry-adaptive convolutional neural networks for solving parameterized steady-state PDEs on irregular domain

NSFnets (Navier-Stokes flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations

Neural Operator: Graph Kernel Network for Partial Differential Equations

Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations

Towards Physics-informed Deep Learning for Turbulent Flow Prediction

DeepXDE: A Deep Learning Library for Solving Differential Equations

Physics Informed Extreme Learning Machine (PIELM) - A rapid method for the numerical solution of partial differential equations

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

Machine Learning for Fluid Mechanics

Variational training of neural network approximations of solution maps for physical models

Learning to Optimize Multigrid PDE Solvers

Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations

Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data

A Multiscale Neural Network Based on Hierarchical Matrices

PDE-Net: Learning PDEs from Data

The Deep Ritz Method: A Deep Learning-Based Numerical Algorithm for Solving Variational Problems

DGM: A deep learning algorithm for solving partial differential equations

Solving high-dimensional partial differential equations using deep learning

Adam: A Method for Stochastic Optimization

Accurate, high-order representation of complex three-dimensional surfaces via Fourier continuation analysis

Stability and Convergence of the Crank-Nicolson/Adams-Bashforth scheme for the Time-Dependent Navier-Stokes Equations

The 2D lid-driven cavity problem revisited

Are Neural Operators Really Neural Operators? Frame Theory Meets Operator Learning

Accelerating Carbon Capture and Storage Modeling using Fourier Neural Operators

Markov Neural Operators for Learning Chaotic Systems

Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries

NVIDIA SimNet™: An AI-Accelerated Multi-Physics Simulation Framework

Ml-pde: A framework for a machine learning enhanced pde solver

LIMITATIONS OF PHYSICS INFORMED MACHINE LEARNING FOR NONLINEAR TWO-PHASE TRANSPORT IN POROUS MEDIA

Model reduction and neural networks for parametric pde(s)

MCMC Methods for Functions : Modifying Old Algorithms to Make Them Faster

Deeponet: Learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators

Lid Cavity flow

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names