Solving inverse-PDE problems with physics-aware neural networks

Abstract

We propose a novel composite framework to find unknown fields in the context of inverse problems for partial differential equations (PDEs). We blend the high expressibility of deep neural networks as universal function estimators with the accuracy and reliability of existing numerical algorithms for partial differential equations as custom layers in semantic autoencoders. Our design brings together techniques of computational mathematics, machine learning and pattern recognition under one umbrella to incorporate domain-specific knowledge and physical constraints to discover the underlying hidden fields. The network is explicitly aware of the governing physics through a hard-coded PDE solver layer in contrast to most existing methods that incorporate the governing equations in the loss function or rely on trainable convolutional layers to discover proper discretizations from data. This subsequently focuses the computational load to only the discovery of the hidden fields and therefore is more data efficient. We call this architecture Blended inverse-PDE networks (hereby dubbed BiPDE networks) and demonstrate its applicability for recovering the variable diffusion coefficient in Poisson problems in one and two spatial dimensions, as well as the diffusion coefficient in the time-dependent and nonlinear Burgers' equation in one dimension. We also show that this approach is robust to noise.

arXiv comment39 pages, 17 figures

References in corpus (30)

• Adam: A Method for Stochastic Optimization
• TensorFlow: Large-Scale Machine Learning on Heterogeneous Distributed Systems
• Improving neural networks by preventing co-adaptation of feature detectors
• Deep Convolutional Neural Network for Inverse Problems in Imaging
• DeepXDE: A deep learning library for solving differential equations
• Artificial Neural Networks for Solving Ordinary and Partial Differential Equations
• DGM: A deep learning algorithm for solving partial differential equations
• Solving high-dimensional partial differential equations using deep learning
• Hidden Physics Models: Machine Learning of Nonlinear Partial Differential Equations
• Physics Informed Deep Learning (Part I): Data-driven Solutions of Nonlinear Partial Differential Equations
• Machine Learning of Linear Differential Equations using Gaussian Processes
• PDE-Net 2.0: Learning PDEs from Data with A Numeric-Symbolic Hybrid Deep Network
• Physics Informed Deep Learning (Part II): Data-driven Discovery of Nonlinear Partial Differential Equations
• Learning to Predict the Cosmological Structure Formation
• PDE-Net: Learning PDEs from Data
• Data-driven discovery of PDEs in complex datasets
• Data driven approximation of parametrized PDEs by Reduced Basis and Neural Networks
• Active learning of constitutive relation from mesoscopic dynamics for macroscopic modeling of non-Newtonian flows
• From Dark Matter to Galaxies with Convolutional Networks
• A physics-aware, probabilistic machine learning framework for coarse-graining high-dimensional systems in the Small Data regime
• Coercing Machine Learning to Output Physically Accurate Results
• Deep Neural Network Computes Electron Densities and Energies of a Large Set of Organic Molecules Faster than Density Functional Theory (DFT)
• HIGAN: Cosmic Neutral Hydrogen with Generative Adversarial Networks
• The Neural Network Approach to Inverse Problems in Differential Equations
• Enforcing constraints for interpolation and extrapolation in Generative Adversarial Networks
• Neural network augmented inverse problems for PDEs
• Unsupervised Deep Learning Algorithm for PDE-based Forward and Inverse Problems
• Zernike Basis to Cartesian Transformations
• A parallel Voronoi-based approach for mesoscale simulations of cell aggregate electropermeabilization
• Self-supervised Learning with Physics-aware Neural Networks I: Galaxy Model Fitting

Cited by in corpus (16)

• Locally adaptive activation functions with slope recovery term for deep and physics-informed neural networks
• Physics Informed Neural Fields for Smoke Reconstruction with Sparse Data
• Algorithmically-Consistent Deep Learning Frameworks for Structural Topology Optimization
• Optimizing a DIscrete Loss (ODIL) to solve forward and inverse problems for partial differential equations using machine learning tools
• Short-term traffic prediction using physics-aware neural networks
• An unsupervised learning approach to solving heat equations on chip based on Auto Encoder and Image Gradient
• Imaging Conductivity from Current Density Magnitude using Neural Networks
• Solving the Pulsar Equation using Physics-Informed Neural Networks
• GoRINNs: Godunov-Riemann Informed Neural Networks for Learning Hyperbolic Conservation Laws
• Physics-Integrated Variational Autoencoders for Robust and Interpretable Generative Modeling
• Error-correcting neural networks for semi-Lagrangian advection in the level-set method
• Predicting Change, Not States: An Alternate Framework for Neural PDE Surrogates
• Physical machine learning outperforms "human learning" in Quantum Chemistry
• Stabilizing and Solving Unique Continuation Problems by Parameterizing Data and Learning Finite Element Solution Operators
• TAEN: A Model-Constrained Tikhonov Autoencoder Network for Forward and Inverse Problems
• Solver-in-the-Loop joint operator learning: fractional Laplace-Beltrami features for interface reconstruction