(Created page with " == Abstract == We develop a wavelet-based three-dimensional convolutional neural network (WCNN3d) for superresolution of coarse-grained data of homogeneous isotropic turbule...") |
|||
| (2 intermediate revisions by the same user not shown) | |||
| Line 2: | Line 2: | ||
== Abstract == | == Abstract == | ||
| − | We develop a wavelet-based three-dimensional convolutional neural network (WCNN3d) for superresolution of coarse-grained data of homogeneous isotropic turbulence. The turbulent flow data are computed by high resolution direct numerical simulation (DNS), while the coarse-grained data are obtained by applying a Gaussian filter to the DNS data. The CNNs are trained with the DNS data and the coarse-grained data. We compare vorticityand velocity-based approaches and assess the proposed | + | We develop a wavelet-based three-dimensional convolutional neural network (WCNN3d) for superresolution of coarse-grained data of homogeneous isotropic turbulence. The turbulent flow data are computed by high resolution direct numerical simulation (DNS), while the coarse-grained data are obtained by applying a Gaussian filter to the DNS data. The CNNs are trained with the DNS data and the coarse-grained data. We compare vorticityand velocity-based approaches and assess the proposed WCNN3d method in terms of flow visualization, enstrophy spectra and probability density functions. |
| + | We show that orthogonal wavelets enhance the efficiency of the learning of CNN. | ||
<pdf>Media:Draft_Content_577186525-2785_paper-4258-document.pdf</pdf> | <pdf>Media:Draft_Content_577186525-2785_paper-4258-document.pdf</pdf> | ||
| − | == | + | == Full Paper == |
<pdf>Media:Draft_Content_577186525-2785_paper-1211-document.pdf</pdf> | <pdf>Media:Draft_Content_577186525-2785_paper-1211-document.pdf</pdf> | ||
Latest revision as of 08:45, 20 July 2022
Abstract
We develop a wavelet-based three-dimensional convolutional neural network (WCNN3d) for superresolution of coarse-grained data of homogeneous isotropic turbulence. The turbulent flow data are computed by high resolution direct numerical simulation (DNS), while the coarse-grained data are obtained by applying a Gaussian filter to the DNS data. The CNNs are trained with the DNS data and the coarse-grained data. We compare vorticityand velocity-based approaches and assess the proposed WCNN3d method in terms of flow visualization, enstrophy spectra and probability density functions. We show that orthogonal wavelets enhance the efficiency of the learning of CNN.
Full Paper
Document information
Published on 05/07/22
Submitted on 05/07/22
Volume 1700 Data Science, Machine Learning and Artificial Intelligence, 2022
DOI: 10.23967/wccm-apcom.2022.013
Licence: CC BY-NC-SA license
Share this document
Keywords
claim authorship
Are you one of the authors of this document?