Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

A framework for data-driven solution and parameter estimation of PDEs using conditional generative adversarial networks

Journal Article · · Nature Computational Science
 [1];  [2];  [1];  [3];  [4];  [2];  [1]
  1. Cornell Univ., Ithaca, NY (United States)
  2. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
  3. Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
  4. Univ. of Hawaii at Manoa, Honolulu, HI (United States)

We employ and adapt the image-to-image translation concept based on conditional generative adversarial networks (cGAN) for learning a forward and an inverse solution operator of partial differential equations (PDEs). We focus on steady-state solutions of coupled hydromechanical processes in heterogeneous porous media and present the parameterization of the spatially heterogeneous coefficients, which is exceedingly difficult using standard reduced-order modeling techniques. We show that our framework provides a speed-up of at least 2,000 times compared to a finite-element solver and achieves a relative root-mean-square error (r.m.s.e.) of less than 2% for forward modeling. For inverse modeling, the framework estimates the heterogeneous coefficients, given an input of pressure and/or displacement fields, with a relative r.m.s.e. of less than 7%, even for cases where the input data are incomplete and contaminated by noise. The framework also provides a speed-up of 120,000 times compared to a Gaussian prior-based inverse modeling approach while also delivering more accurate results.

Research Organization:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Basic Energy Sciences (BES); USDOE Laboratory Directed Research and Development (LDRD) Program; USDOE National Nuclear Security Administration (NNSA)
Grant/Contract Number:
AC52-07NA27344
OSTI ID:
1843114
Alternate ID(s):
OSTI ID: 1975643
Report Number(s):
LLNL-JRNL--823007; 1035531
Journal Information:
Nature Computational Science, Journal Name: Nature Computational Science Journal Issue: 12 Vol. 1; ISSN 2662-8457
Publisher:
Springer NatureCopyright Statement
Country of Publication:
United States
Language:
English

References (41)

Large-scale hydraulic tomography and joint inversion of head and tracer data using the Principal Component Geostatistical Approach (PCGA) journal July 2014
Scalable subsurface inverse modeling of huge data sets with an application to tracer concentration breakthrough data from magnetic resonance imaging: SCALABLE INVERSE MODELING WITH A HUGE MRI DATA SET journal July 2016
Improved characterization of heterogeneous permeability in saline aquifers from transient pressure data during freshwater injection journal May 2017
An adaptive and efficient greedy procedure for the optimal training of parametric reduced-order models: ADAPTIVE GREEDY PROCEDURE FOR THE OPTIMAL TRAINING OF PARAMETRIC ROMS journal September 2014
A POD-selective inverse distance weighting method for fast parametrized shape morphing: A POD-selective IDW method for fast parametrized shape morphing
  • Ballarin, Francesco; D'Amario, Alessandro; Perotto, Simona
  • International Journal for Numerical Methods in Engineering, Vol. 117, Issue 8 https://doi.org/10.1002/nme.5982
journal November 2018
Multimodal Unsupervised Image-to-Image Translation book January 2018
Certified Reduced Basis Methods for Parametrized Partial Differential Equations book January 2016
U-Net: Convolutional Networks for Biomedical Image Segmentation
  • Ronneberger, Olaf; Fischer, Philipp; Brox, Thomas
  • Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III https://doi.org/10.1007/978-3-319-24574-4_28
book November 2015
Finite Element Solvers for Biot’s Poroelasticity Equations in Porous Media journal October 2020
Fast Large-Scale Joint Inversion for Deep Aquifer Characterization Using Pressure and Heat Tracer Measurements journal September 2017
Data-driven reduced order modeling of poroelasticity of heterogeneous media based on a discontinuous Galerkin approximation journal June 2021
Benchmarks for single-phase flow in fractured porous media journal January 2018
Component-wise reduced order model lattice-type structure design journal August 2021
Well productivity evaluation in deformable single-fracture media journal September 2020
Non-intrusive reduced order modeling of nonlinear problems using neural networks journal June 2018
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations journal February 2019
Gradient-based constrained optimization using a database of linear reduced-order models journal December 2020
Space–time reduced order model for large-scale linear dynamical systems with application to Boltzmann transport problems journal January 2021
Enriched Galerkin discretization for modeling poroelasticity and permeability alteration in heterogeneous porous media journal February 2021
The mixed Deep Energy Method for resolving concentration features in finite strain hyperelasticity journal February 2022
Recent developments in fast and scalable inverse modeling and data assimilation methods in hydrology journal December 2020
PBBFMM3D: A parallel black-box algorithm for kernel matrix-vector multiplication journal August 2021
An investigation of hydromechanical effect on well productivity in fractured porous media using full factorial experimental design journal October 2019
When good statistical models of aquifer heterogeneity go bad: A comparison of flow, dispersion, and mass transfer in connected and multivariate Gaussian hydraulic conductivity fields: FLOW, DISPERSION, AND MASS TRANSFER journal March 2003
Efficient Monte Carlo With Graph‐Based Subsurface Flow and Transport Models journal May 2018
Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators journal March 2021
Fast inverse design of microstructures via generative invariance networks journal March 2021
Adaptive sampling strategies for non‐intrusive POD‐based surrogates journal May 2013
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network conference July 2017
Image-to-Image Translation with Conditional Adversarial Networks conference July 2017
AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks conference June 2018
High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs conference June 2018
Semantic Image Synthesis With Spatially-Adaptive Normalization conference June 2019
Interpreting the Latent Space of GANs for Semantic Face Editing conference June 2020
Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks conference October 2017
Optimal Model Management for Multifidelity Monte Carlo Estimation journal January 2016
Generative adversarial networks journal October 2020
Upscaling two-phase flow in naturally fractured reservoirs journal November 2009
Physics-informed neural networks for solving nonlinear diffusivity and Biot’s equations journal May 2020
Deconvolution and Checkerboard Artifacts journal October 2016
Extended Physics-Informed Neural Networks (XPINNs): A Generalized Space-Time Domain Decomposition Based Deep Learning Framework for Nonlinear Partial Differential Equations journal June 2020

Cited By (1)