Deep Learning of Dynamic Subsurface Flow via Theory-guided Generative Adversarial Network

  title={Deep Learning of Dynamic Subsurface Flow via Theory-guided Generative Adversarial Network},
  author={Tianhao He and Dongxiao Zhang},
Generative adversarial network (GAN) has been shown to be useful in various applications, such as image recognition, text processing and scientific computing, due its strong ability to learn complex data distributions. In this study, a theory-guided generative adversarial network (TgGAN) is proposed to solve dynamic partial differential equations (PDEs). Different from standard GANs, the training term is no longer the true data and the generated data, but rather their residuals. In addition… Expand
Theory-guided full convolutional neural network: An efficient surrogate model for inverse problems in subsurface contaminant transport
  • Tianhao He, Nanzhe Wang, Dongxiao Zhang
  • Computer Science
  • Advances in Water Resources
  • 2021
A theory-guided full convolutional neural network (TgFCNN) model is proposed to solve inverse problems in subsurface contaminant transport and demonstrates strong generalization and extrapolation abilities, and satisfactory accuracy when estimating unknown contaminant source parameters, as well as the permeability field. Expand
Theory-guided hard constraint projection (HCP): a knowledge-based data-driven scientific machine learning method
Machine learning models have been successfully used in many scientific and engineering fields. However, it remains difficult for a model to simultaneously utilize domain knowledge and experimental… Expand


Physics-Based Generative Adversarial Models for Image Restoration and Beyond
An algorithm to directly solve numerous image restoration problems (e.g., image deblurring, image dehazing, and image deraining) by generative models with adversarial learning within the GAN framework. Expand
Physics-Informed Generative Adversarial Networks for Stochastic Differential Equations
The accuracy and effectiveness of PI-GANs in solving SDEs for up to 30 dimensions is demonstrated, but in principle, PI-gans could tackle very high dimensional problems given more sensor data with low-polynomial growth in computational cost. Expand
Enforcing constraints for interpolation and extrapolation in Generative Adversarial Networks
This work suggests ways to enforce given constraints in the output of a GAN generator both for interpolation and extrapolation (prediction), and provides examples of linear and nonlinear systems of differential equations to illustrate the various constructions. Expand
Enforcing statistical constraints in generative adversarial networks for modeling chaotic dynamical systems
This work presents a statistical constrained generative adversarial network that has great potential for being an alternative to the explicit modeling of closures or parameterizations for unresolved physics, which are known to be a major source of uncertainty in simulating multi-scale physical systems. Expand
Generative Adversarial Nets
We propose a new framework for estimating generative models via an adversarial process, in which we simultaneously train two models: a generative model G that captures the data distribution, and a… Expand
SeqGAN: Sequence Generative Adversarial Nets with Policy Gradient
Modeling the data generator as a stochastic policy in reinforcement learning (RL), SeqGAN bypasses the generator differentiation problem by directly performing gradient policy update. Expand
Deep Learning of Subsurface Flow via Theory-guided Neural Network
Numerical results demonstrate that the Theory-guided Neural Network model achieves much better predictability, reliability, and generalizability than ANN models due to the physical/engineering constraints in the former. Expand
InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
Experiments show that InfoGAN learns interpretable representations that are competitive with representations learned by existing fully supervised methods. Expand
Physics-Constrained Deep Learning for High-dimensional Surrogate Modeling and Uncertainty Quantification without Labeled Data
This paper provides a methodology that incorporates the governing equations of the physical model in the loss/likelihood functions of the model predictive density and the reference conditional density as a minimization problem of the reverse Kullback-Leibler (KL) divergence. Expand
Improved Training of Wasserstein GANs
This work proposes an alternative to clipping weights: penalize the norm of gradient of the critic with respect to its input, which performs better than standard WGAN and enables stable training of a wide variety of GAN architectures with almost no hyperparameter tuning. Expand