Reconstruction of three-dimensional porous media using generative adversarial neural networks
Реконструкция трёхмерных пористых сред с использованием генеративных состязательных нейронных сетей
2017-10-23
SCID: 54.1/aexy3qa3
Discuss with AI
adversarial learninggenerative adversarial networkspore morphologysingle-phase permeabilitythree-dimensional porous media
Figures from the paper
Abstract (AI)
To evaluate the variability of multiphase flow properties of porous media at the pore scale, it is necessary to acquire a number of representative samples of the void-solid structure. While modern x-ray computer tomography has made it possible to extract three-dimensional images of the pore space, assessment of the variability in the inherent material properties is often experimentally not feasible. We present a method to reconstruct the solid-void structure of porous media by applying a generative neural network that allows an implicit description of the probability distribution represented by three-dimensional image data sets. We show, by using an adversarial learning approach for neural networks, that this method of unsupervised learning is able to generate representative samples of porous media that honor their statistics. We successfully compare measures of pore morphology, such as the Euler characteristic, two-point statistics, and directional single-phase permeability of synthetic realizations with the calculated properties of a bead pack, Berea sandstone, and Ketton limestone. Results show that generative adversarial networks can be used to reconstruct high-resolution three-dimensional images of porous media at different scales that are representative of the morphology of the images used to train the neural network. The fully convolutional nature of the trained neural network allows the generation of large samples while maintaining computational efficiency. Compared to classical stochastic methods of image reconstruction, the implicit representation of the learned data distribution can be stored and reused to generate multiple realizations of the pore structure very rapidly.
Key Findings
1
A generative adversarial neural network reconstructs three-dimensional porous-media solid–void structures from image datasets through unsupervised adversarial learning.
2
Synthetic realizations successfully represent bead packs, Berea sandstone, and Ketton limestone across different scales and high spatial resolutions.
3
The fully convolutional architecture generates large porous-media samples computationally efficiently while preserving training-image morphology.
4
The generated porous-media samples reproduce key morphological and transport statistics, including Euler characteristic, two-point statistics, and directional single-phase permeability.
5
The learned implicit data distribution can be stored and reused to produce multiple pore-structure realizations much faster than classical stochastic reconstruction methods.
Research Object
three-dimensional porous media solid–void structures, including bead pack, Berea sandstone, and Ketton limestone
Research Subject
reconstruction and statistical representativeness of pore morphology and flow properties in generated porous-media realizations
Publication Details
Publication Date
2017-10-23
Journal
Publisher
ISSN
Cited by
587
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest