HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Journal articles

A twin-decoder structure for incompressible laminar flow reconstruction with uncertainty estimation around 2D obstacles

Abstract : Over the past few years, deep learning methods have proved to be of great interest for the computational fluid dynamics community, especially when used as surrogate models, either for flow reconstruction, turbulence modeling, or for the prediction of aerodynamic coefficients. Overall exceptional levels of accuracy have been obtained but the robustness and reliability of the proposed approaches remain to be explored, particularly outside the confidence region defined by the training dataset. In this contribution, we present an autoencoder architecture with twin decoder for incompressible laminar flow reconstruction with uncertainty estimation around 2D obstacles. The proposed architecture is trained over a dataset composed of numerically-computed laminar flows around 12,000 random shapes, and naturally enforces a quasi-linear relation between a geometric reconstruction branch and the flow prediction decoder. Based on this feature, two uncertainty estimation processes are proposed, allowing either a binary decision (accept or reject prediction), or proposing a confidence interval along with the flow quantities prediction (u, v, p). Results over dataset samples as well as unseen shapes show a strong positive correlation of this reconstruction score to the mean-squared error of the flow prediction. Such approaches offer the possibility to warn the user of trained models when provided input shows too large deviation from the training data, making the produced surrogate model conservative for fast and reliable flow prediction.
Complete list of metadata

https://hal.mines-ales.fr/hal-03538448
Contributor : Administrateur Imt - Mines Alès Connect in order to contact the contributor
Submitted on : Monday, January 24, 2022 - 9:15:03 AM
Last modification on : Thursday, February 3, 2022 - 7:00:38 PM
Long-term archiving on: : Monday, April 25, 2022 - 6:20:41 PM

File

 Restricted access
To satisfy the distribution rights of the publisher, the document is embargoed until : 2022-07-15

Please log in to resquest access to the document

Identifiers

Relations

Citation

J. Chen, J. Viquerat, Frederic Heymes, Elie Hachem. A twin-decoder structure for incompressible laminar flow reconstruction with uncertainty estimation around 2D obstacles. Neural Computing and Applications, Springer Verlag, In press, ⟨10.1007/s00521-021-06784-z⟩. ⟨hal-03538448⟩

Share

Metrics

Record views

35