Artificial neural network subgrid models of 2-D compressible magnetohydrodynamic turbulence

Abstract

We explore the suitability of deep learning to capture the physics of subgrid-scale ideal magnetohydrodynamics turbulence of 2-D simulations of the magnetized Kelvin-Helmholtz instability. We produce simulations at different resolutions to systematically quantify the performance of neural network models to reproduce the physics of these complex simulations. We compare the performance of our neural networks with gradient models, which are extensively used in the extensively in the magnetohydrodynamic literature. Our findings indicate that neural networks significantly outperform gradient models at reproducing the effects of magnetohydrodynamics turbulence. To the best of our knowledge, this is the first exploratory study on the use of deep learning to learn and reproduce the physics of magnetohydrodynamics turbulence.

arXiv comment24 pages, 17 figures

References in corpus (19)

• Deep Learning in Neural Networks: An Overview
• Deep learning for time series classification: a review
• Machine Learning for Fluid Mechanics
• Binary neutron-star mergers: a review of Einstein's richest laboratory
• Efficient magnetic-field amplification due to the Kelvin-Helmholtz instability in binary neutron star mergers
• A neural network approach for the blind deconvolution of turbulent flows
• Physical consistency of subgrid-scale models for large-eddy simulation of incompressible turbulent flows
• Enabling real-time multi-messenger astrophysics discoveries with deep learning
• Energy transfer in compressible magnetohydrodynamic turbulence
• Large-Eddy Simulations of Magnetohydrodynamic Turbulence in Heliophysics and Astrophysics
• MHD Turbulence
• A priori analysis on deep learning of subgrid-scale parameterizations for Kraichnan turbulence
• Gradient sub-grid-scale model for relativistic MHD Large Eddy Simulations
• Simflowny 2: An upgraded platform for scientific modeling and simulation
• A nonlinear structural subgrid-scale closure for compressible MHD Part II: a priori comparison on turbulence simulation data
• Comparative statistics of selected subgrid-scale models in large eddy simulations of decaying, supersonic MHD turbulence
• A nonlinear structural subgrid-scale closure for compressible MHD Part I: derivation and energy dissipation properties
• Nonlinear closures for scale separation in supersonic magnetohydrodynamic turbulence
• Deep learning for turbulent channel flow

Cited by in corpus (11)

• Physics-guided deep learning framework for predictive modeling of the Reynolds stress anisotropy
• Deconvolutional artificial neural network models for large eddy simulation of turbulence
• Convergence of Artificial Intelligence and High Performance Computing on NSF-supported Cyberinfrastructure
• Applications of physics informed neural operators
• Interpreted machine learning in fluid dynamics: Explaining relaminarisation events in wall-bounded shear flows
• Magnetohydrodynamics with Physics Informed Neural Operators
• Interpretable AI forecasting for numerical relativity waveforms of quasi-circular, spinning, non-precessing binary black hole mergers
• Advances in Machine and Deep Learning for Modeling and Real-time Detection of Multi-Messenger Sources
• Physics-Informed Machine Learning for Modeling Turbulence in Supernovae
• Identification of high order closure terms from fully kinetic simulations using machine learning
• AI and extreme scale computing to learn and infer the physics of higher order gravitational wave modes of quasi-circular, spinning, non-precessing binary black hole mergers