Abstract:This paper presents a new deep learning data-driven model for predicting structure dependent pore-fluid velocity fields in rock. The model is based on a Convolutional Auto-Encoder (CAE) artificial neural network capable of learning from image data generated by direct numerical simulations of fluid flow through pore-structures, such as by Lattice Boltzmann or molecular dynamics methods. The main novelty of the model in comparison to previous CAE-based data-driven approaches consists of three parts. The first is a methodology for decomposing the full-domain of the porous media into sub-regions, or "sub-domains", in order to reduce the overall size of the CAE, batch process the sub-domains in parallel, and enable the CAE to learn local and generalizable nonlinear mappings of pore-fluid velocities. The second consists of embedding the finite difference solutions of the incompressible Navier-Stokes and continuity equations into convolutional layers prior to the CAE in order to provide the CAE with knowledge of fluid dynamics physics (PhyFlow). The third main novelty is that the training of the CAE is regularized with a hierarchical loss function that encourages the learning of fluid flow patterns (in a way similar to ranked modes in principal component analysis), ranking from most to least important. This is shown to increase the stability in learning, reduce over-fitting, and promote interpretability of the CAE neural network layers (HierCAE). The comprehensive new data-driven model, which we call the PhyFlow-HierCAE model, is shown to exhibit improved accuracy and generalizability of flow field predictions over conventional CAE models, attributable to the embedded physical knowledge and the hierarchical regularization, as well as realize orders of magnitude speed-ups in computation times as a surrogate for the direct numerical simulations. Examples of training and forward predictions on unseen pore-structures are provided and evaluated for data from Lattice Boltzmann and molecular dynamics simulations of pore-fluid flow. The model is shown to be a fast and accurate emulator (or "surrogate") for predicting effective permeability of unseen pore-structures based on learning from relatively small direct numerical simulation datasets.

Modeling of Coupled Turbulent Channel Porous Media Flow through a Deep Autoencoder Echo State Network Framework

Non-intrusive, transferable model for coupled turbulent channel-porous media flow based upon neural networks

Modelling spatiotemporal turbulent dynamics with the convolutional autoencoder echo state network

Deep Convolutional Recurrent Autoencoders for Flow Field Prediction

Data-driven modeling of unsteady flow based on deep operator network

Unsteady Flow-Field Forecasting Leveraging a Hybrid Deep-Learning Architecture

A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media

Echo state networks for modeling turbulent convection

Combined space-time reduced-order model with 3D deep convolution for extrapolating fluid dynamics

Deep learning-based predictive modelling of transonic flow over an aerofoil

Reduced-Order Modeling of Steady and Unsteady Flows with Deep Neural Networks

Three-dimensional autoencoder for the flow field reconstruction of an inclined circular disk

Convolutional autoencoder for the spatiotemporal latent representation of turbulence

Reduced-order modeling for stochastic large-scale and time-dependent flow problems using deep spatial and temporal convolutional autoencoders

Applications of autoencoder in reducedGorder modeling of flow field

Deep convolutional encoder-decoder networks for uncertainty quantification of dynamic multiphase flow in heterogeneous media

Predicting Transonic Flowfields in Non-Homogeneous Unstructured Grids Using Autoencoder Graph Convolutional Networks

Combined space–time reduced-order model with three-dimensional deep convolution for extrapolating fluid dynamics

A Deep Learning Approach For Epistemic Uncertainty Quantification Of Turbulent Flow Simulations

DeepCFD: Efficient Steady-State Laminar Flow Approximation with Deep Convolutional Neural Networks