Prediction of aerodynamic flow fields using convolutional neural networks

An approximation model based on convolutional neural networks (CNNs) is proposed for flow field predictions. The CNN is used to predict the velocity and pressure field in unseen flow conditions and geometries given the pixelated shape of the object. In particular, we consider Reynolds Averaged Navier–Stokes (RANS) flow solutions over airfoil shapes as training data. The CNN can automatically detect essential features with minimal human supervision and is shown to effectively estimate the velocity and pressure field orders of magnitude faster than the RANS solver, making it possible to study the impact of the airfoil shape and operating conditions on the aerodynamic forces and the flow field in near-real time. The use of specific convolution operations, parameter sharing, and gradient sharpening are shown to enhance the predictive capabilities of the CNN. We explore the network architecture and its effectiveness in predicting the flow field for different airfoil shapes, angles of attack, and Reynolds numbers.

[1]  H. Ng,et al.  A Multilayer Convolutional Encoder-Decoder Neural Network for Grammatical Error Correction , 2018, AAAI.

[2]  Donghyun You,et al.  Prediction of laminar vortex shedding over a cylinder using deep learning , 2017 .

[3]  E. Turkel,et al.  PRECONDITIONING TECHNIQUES IN COMPUTATIONAL FLUID DYNAMICS , 1999 .

[4]  B. V. Leer,et al.  Towards the ultimate conservative difference scheme V. A second-order sequel to Godunov's method , 1979 .

[5]  Yann LeCun,et al.  Deep multi-scale video prediction beyond mean square error , 2015, ICLR.

[6]  Karthik Duraisamy,et al.  Assessment of Transition Model and CFD Methodology for Wind Turbine Flows , 2012 .

[7]  Vladimir G. Kim,et al.  Data‐Driven Shape Analysis and Processing , 2015, Comput. Graph. Forum.

[8]  Karthikeyan Duraisamy,et al.  Studies in Tip Vortex Formation, Evolution and Control , 2005 .

[9]  Wang Ping,et al.  Research on data augmentation for image classification based on convolution neural networks , 2017, 2017 Chinese Automation Congress (CAC).

[10]  Barry Koren,et al.  A robust upwind discretization method for advection, diffusion and source terms , 1993 .

[11]  P. Roe.  CHARACTERISTIC-BASED SCHEMES FOR THE EULER EQUATIONS , 1986 .

[12]  Yann LeCun,et al.  Convolutional Learning of Spatio-temporal Features , 2010, ECCV.

[13]  M. Carter.  Computer graphics: Principles and practice , 1997 .

[14]  Haibin Ling,et al.  Shape Classification Using the Inner-Distance , 2007, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[15]  George E. Karniadakis,et al.  Hidden physics models: Machine learning of nonlinear partial differential equations , 2017, J. Comput. Phys..

[16]  George E. Karniadakis,et al.  Hidden Fluid Mechanics: A Navier-Stokes Informed Deep Learning Framework for Assimilating Flow Visualization Data , 2018, ArXiv.

[17]  James D. Baeder,et al.  Application of the Correlation-based Gamma-Re Theta t Transition Model to the Spalart-Allmaras Turbulence Model , 2011 .

[18]  Guojun Lu,et al.  Review of shape representation and description techniques , 2004, Pattern Recognit..

[19]  Hongkai Zhao,et al.  A fast sweeping method for Eikonal equations , 2004, Math. Comput..

[20]  D. M. Somers.  S814 and S815 Airfoils: October 1991--July 1992 , 2004 .

[21]  Sajib Saha,et al.  Object Class Detection and Classification using Multi Scale Gradient and Corner Point based Shape Descriptors , 2015, ArXiv.

[22]  T. P. Miyanawala,et al.  An Efficient Deep Learning Technique for the Navier-Stokes Equations: Application to Unsteady Wake Flow Dynamics , 2017, 1710.09099.

[23]  V. Leitáo,et al.  Computer Graphics: Principles and Practice , 1995 .

[24]  Yoshua. Bengio,et al.  Learning Deep Architectures for AI , 2007, Found. Trends Mach. Learn..

[25]  Hugues Hoppe,et al.  Poisson surface reconstruction and its applications , 2008, SPM '08.

[26]  Hugues Hoppe.  Surface Reconstruction from Unorganized Points (PhD Thesis) , 1994 .

[27]  M. Goesele,et al.  Floating scale surface reconstruction , 2014, ACM Trans. Graph..

[28]  Thomas H. Pulliam,et al.  Implementation of Preconditioned Dual-Time Procedures in OVERFLOW , 2003 .

[29]  Donghyun You,et al.  Data-driven prediction of unsteady flow over a circular cylinder using deep learning , 2018, Journal of Fluid Mechanics.

[30]  Robert M. Haralick,et al.  Feature normalization and likelihood-based similarity measures for image retrieval , 2001, Pattern Recognit. Lett..

[31]  Paris Perdikaris,et al.  Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations , 2019, J. Comput. Phys..

[32]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[33]  Isaac Amidror,et al.  Scattered data interpolation methods for electronic imaging systems: a survey , 2002, J. Electronic Imaging.

[34]  Wei Li,et al.  Convolutional Neural Networks for Steady Flow Approximation , 2016, KDD.

[35]  T. Pulliam,et al.  A diagonal form of an implicit approximate-factorization algorithm , 1981 .

[36]  Yao Zhang,et al.  Application of Convolutional Neural Network to Predict Airfoil Lift Coefficient , 2017, ArXiv.

[37]  D. M. Somers,et al.  Design and experimental results for the S809 airfoil , 1997 .

[38]  P. Spalart.  A One-Equation Turbulence Model for Aerodynamic Flows , 1992 .

[39]  Richard K. Beatson,et al.  Reconstruction and representation of 3D objects with radial basis functions , 2001, SIGGRAPH.

[40]  D. M. Somers.  Design and experimental results for the S805 airfoil , 1997 .

[41]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[42]  Tony DeRose,et al.  Surface reconstruction from unorganized points , 1992, SIGGRAPH.

[43]  Quoc V. Le,et al.  Searching for Activation Functions , 2018, arXiv.

[44]  J A Sethian,et al.  A fast marching level set method for monotonically advancing fronts. , 1996, Proceedings of the National Academy of Sciences of the United States of America.

[45]  James D. Baeder,et al.  Computational Investigation of Micro Hovering Rotor Aerodynamics , 2006 .

[46]  Karthik Duraisamy,et al.  Turbulence Modeling in the Age of Data , 2018, Annual Review of Fluid Mechanics.

[47]  Oliver Hennigh,et al.  Lat-Net: Compressing Lattice Boltzmann Flow Simulations using Deep Neural Networks , 2017, 1705.09036.

[48]  Ken Perlin,et al.  Accelerating Eulerian Fluid Simulation With Convolutional Networks , 2016, ICML.

[49]  Gang Wang,et al.  Convolutional recurrent neural networks: Learning spatial dependencies for image representation , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).