A supervised neural network for drag prediction of arbitrary 2D shapes in laminar flows at low Reynolds number. (English) Zbl 1521.76097
Summary: Despite the significant breakthrough of neural networks in the last few years, their spreading in the field of computational fluid dynamics is very recent, and many applications remain to explore. In this paper, we explore the drag prediction capabilities of convolutional neural networks for laminar, low-Reynolds number flows past arbitrary 2D shapes. A set of random shapes exhibiting a rich variety of geometrical features is built using Bézier curves. The efficient labelling of the shapes is provided using an immersed method to solve a unified Eulerian formulation of the Navier-Stokes equation. The network is then trained and optimized on the obtained dataset, and its predictive efficiency assessed on several real-life shapes, including NACA airfoils.
MSC:
| 76D05 | Navier-Stokes equations for incompressible viscous fluids |
| 65M99 | Numerical methods for partial differential equations, initial value and time-dependent initial-boundary value problems |
| 68T07 | Artificial neural networks and deep learning |
Keywords:
machine learning; neural networks; convolutional networks; computational fluid dynamics; immersed methodReferences:
| [1] | Tompson J., Schlachter K., Sprechmann P., Perlin K.. Accelerating eulerian fluid simulation with convolutional networks. arXiv preprint: 1607.035972016; |
| [2] | Ling, J.; Kurzawski, A.; Templeton, J., Reynolds averaged turbulence modelling using deep neural networks with embedded invariance, J Fluid Mech, 807, 155-166 (2016) · Zbl 1383.76175 |
| [3] | Tracey, B. D.; Duraisamy, K.; Alonso, J. J., A machine learning strategy to assist turbulence model development, 53rd AIAA Aerospace Sciences Meeting, 1-23 (2015) |
| [4] | Beck A.D., Flad D.G., Munz C.-D.. Deep neural networks for data-driven turbulence models. 2018; arXiv preprint: 1806.04482 |
| [5] | Raissi M., Perdikaris P., Karniadakis G.E.. Physics informed deep learning (Part I): data-driven solutions of nonlinear partial differential equations. arXiv preprint: 1711.105662017;:1-22 |
| [6] | Raissi M., Yazdani A., Karniadakis G.E.. Hidden fluid mechanics: a navier-stokes informed deep learning framework for assimilating flow visualization data. arXiv preprint: 1808.043272018; |
| [7] | Zhang Y., Sung W.-J., Mavris D.. Application of convolutional neural network to predict airfoil lift coefficient. arXiv preprint: 1712.100822017;:1-9 |
| [8] | http://dl.acm.org/citation.cfm?doid=2939672.2939738 |
| [9] | Miyanawala T.P., Jaiman R.K.. An efficient deep learning technique for the navier-stokes equations: application to unsteady wake flow dynamics. arXiv preprint: 1710.090992017;:1-46 |
| [10] | Sofka, M.; Milletari, F.; Jia, J.; Rothberg, A., Fully convolutional regression network for accurate detection of measurement points, Lect Notes Comput Sci, 10553 LNCS, 258-266 (2017) |
| [11] | Ian GoodfellowYoshua Bengio, A. C., The deep learning book (2017), MIT Press |
| [12] | Chollet, F., Deep learning with python (2018), Manning |
| [13] | Caruana, R.; Lawrence, S.; Giles, L., Overfitting in neural nets: backpropagation, conjugate gradient, and early stopping, Adv Neural Inf Process Syst (2001) |
| [14] | Bruchon, J.; Digonnet, H.; Coupez, T., Using a signed distance function for the simulation of metal forming processes: formulation of the contact condition and mesh adaptation, Int J Numer Methods Eng, 78, 8, 980-1008 (2009) · Zbl 1183.74263 |
| [15] | Codina, R.; Soto, O., A numerical model to track two-fluid interfaces based on a stabilized finite element method and the level set technique, Int J Numer Methods Fluids, 40, 293-301 (2002) · Zbl 1010.76053 |
| [16] | van der Pijl, S.; Segal, A.; Vuik, C.; Wesseling, P., A mass-conserving level-set method for modelling of multi-phase flows, Int J Numer Methods Fluids, 47, 339-361 (2005) · Zbl 1065.76160 |
| [17] | Hachem, E.; Feghali, S.; Codina, R.; Coupez, T., Immersed stress method for fluid structure interaction using anisotropic mesh adaptation, Int J Numer Methods Eng, 94, 805-825 (2013) · Zbl 1352.74108 |
| [18] | Jannoun, G.; Hachem, E.; Veysset, J.; Coupez, T., Anisotropic meshing with time-stepping control for unsteady convection-dominated problems, Appl Math Model, 39, 1899-1916 (2001) · Zbl 1443.65205 |
| [19] | Simonyan K., Zisserman A.. Very deep convolutional networks for large-scale image recognition2014;:1-14arXiv preprint: 1409.1556 |
| [20] | He, K.; Zhang, X.; Ren, S.; Sun, J., Deep residual learning for image recognition, Proc IEEE Comput Soc Conf Comput Vis Pattern Recognit, 2016-December, 770-778 (2016) |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
