Multiscale graph neural network autoencoders for interpretable scientific machine learning
From MaRDI portal
Publication:6087967
Abstract: The goal of this work is to address two limitations in autoencoder-based models: latent space interpretability and compatibility with unstructured meshes. This is accomplished here with the development of a novel graph neural network (GNN) autoencoding architecture with demonstrations on complex fluid flow applications. To address the first goal of interpretability, the GNN autoencoder achieves reduction in the number nodes in the encoding stage through an adaptive graph reduction procedure. This reduction procedure essentially amounts to flowfield-conditioned node sampling and sensor identification, and produces interpretable latent graph representations tailored to the flowfield reconstruction task in the form of so-called masked fields. These masked fields allow the user to (a) visualize where in physical space a given latent graph is active, and (b) interpret the time-evolution of the latent graph connectivity in accordance with the time-evolution of unsteady flow features (e.g. recirculation zones, shear layers) in the domain. To address the goal of unstructured mesh compatibility, the autoencoding architecture utilizes a series of multi-scale message passing (MMP) layers, each of which models information exchange among node neighborhoods at various lengthscales. The MMP layer, which augments standard single-scale message passing with learnable coarsening operations, allows the decoder to more efficiently reconstruct the flowfield from the identified regions in the masked fields. Analysis of latent graphs produced by the autoencoder for various model settings are conducted using using unstructured snapshot data sourced from large-eddy simulations in a backward-facing step (BFS) flow configuration with an OpenFOAM-based flow solver at high Reynolds numbers.
Cites work
- A comparison of neural network architectures for data-driven reduced-order modeling
- Adversarial sampling of unknown and high-dimensional conditional distributions
- Cluster-based reduced-order modelling of a mixing layer
- Dynamic mode decomposition of numerical and experimental data
- Exascale applications: skin in the game
- Experimental data-based reduced-order model for analysis and prediction of flame transition in gas turbine combustors
- Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders
- Multi-level convolutional autoencoder networks for parametric prediction of spatio-temporal dynamics
- Neural-network learning of SPOD latent dynamics
- Nonlinear mode decomposition with convolutional neural networks for fluid dynamics
- Self-sustained oscillations and vortex shedding in backward-facing step flows: Simulation and linear instability analysis
- Solution of the implicitly discretized fluid flow equations by operator- splitting
- The finite volume method in computational fluid dynamics. An advanced introduction with OpenFOAM and Matlab
- Turbulence modeling in the age of data
- Turbulent Flows
- Two calculation procedures for steady, three-dimensional flows with recirculation
Cited in
(6)- GINNs: graph-informed neural networks for multiscale physics
- A graph convolutional autoencoder approach to model order reduction for parametrized PDEs
- Multiscale graph neural networks with adaptive mesh refinement for accelerating mesh-based simulations
- An implicit GNN solver for Poisson-like problems
- Interpretable a-posteriori error indication for graph neural network surrogate models
- Convolutional Symmetric AutoEncoders: enhancing latent stability via differential geometry
This page was built for publication: Multiscale graph neural network autoencoders for interpretable scientific machine learning
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6087967)
