Data-Driven, Physics-Based Feature Extraction from Fluid Flow Fields using Convolutional Neural Networks

DOI10.4208/CICP.OA-2018-0035zbMath1475.76075arXiv1802.00775WikidataQ128463350 ScholiaQ128463350MaRDI QIDQ5161377

Eric G. Paterson, Heng Xiao, Jin-Long Wu, Carlos A. Michelén Ströfer

Publication date: 29 October 2021

Published in: Communications in Computational Physics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1802.00775

zbMATH Keywords

machine learninghorseshoe vortexconvolutional neural networkfeature identificationwing-body junctionairfoil flow field

Mathematics Subject Classification ID

Learning and adaptive systems in artificial intelligence (68T05) General aerodynamics and subsonic flows (76G25) Basic methods in fluid mechanics (76M99)

Related Items (8)

MOD-Net: A Machine Learning Approach via Model-Operator-Data Network for Solving PDEs ⋮ Using machine learning to detect the turbulent region in flow past a circular cylinder ⋮ Better Approximations of High Dimensional Smooth Functions by Deep Neural Networks with Rectified Power Units ⋮ Frequency Principle: Fourier Analysis Sheds Light on Deep Neural Networks ⋮ A Multi-Scale DNN Algorithm for Nonlinear Elliptic Equations with Multiple Scales ⋮ Multi-Scale Deep Neural Network (MscaleDNN) for Solving Poisson-Boltzmann Equation in Complex Domains ⋮ Construction of reduced-order models for fluid flows using deep feedforward neural networks ⋮ A Tailored Convolutional Neural Network for Nonlinear Manifold Learning of Computational Physics Data Using Unstructured Spatial Discretizations

Uses Software

• Theano
• Faster R-CNN

Cites Work

• Flow over periodic hills -- numerical and experimental study in a wide range of Reynolds numbers
• Feature-driven Cartesian adaptive mesh refinement for vortex-dominated flows
• Highly resolved large-eddy simulation of separated flow in a channel with streamwise periodic constrictions
• Deep Potential: A General Representation of a Many-Body Potential Energy Surface
• Learning representations by back-propagating errors
• Shock Wave Drag Reduction

This page was built for publication: Data-Driven, Physics-Based Feature Extraction from Fluid Flow Fields using Convolutional Neural Networks