Quantitative assessment of PINN inference on experimental data for gravity currents flows
In this paper, we apply physics informed neural networks (PINNs) to infer velocity and pressure field from light attenuation technique (LAT) measurements for gravity current induced by lock-exchange. In a PINN model, physical laws are embedded in the loss function of a neural network, such that the...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IOP Publishing
2025-01-01
|
| Series: | Machine Learning: Science and Technology |
| Subjects: | |
| Online Access: | https://doi.org/10.1088/2632-2153/adaca0 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1823860594012323840 |
|---|---|
| author | Mickaël Delcey Yoann Cheny Jean Schneider Simon Becker Sébastien Kiesgen De Richter |
| author_facet | Mickaël Delcey Yoann Cheny Jean Schneider Simon Becker Sébastien Kiesgen De Richter |
| author_sort | Mickaël Delcey |
| collection | DOAJ |
| description | In this paper, we apply physics informed neural networks (PINNs) to infer velocity and pressure field from light attenuation technique (LAT) measurements for gravity current induced by lock-exchange. In a PINN model, physical laws are embedded in the loss function of a neural network, such that the model fits the training data but is also constrained to reduce the residuals of the governing equations. PINNs are able to solve ill-posed inverse problems training on sparse and noisy data, and therefore can be applied to real engineering applications. The noise robustness of PINNs and the model parameters are investigated in a 2 dimensions toy case on a lock-exchange configuration, employing synthetic data. Then we train a PINN with experimental LAT measurements and quantitatively compare the velocity fields inferred to particle image velocimetry measurements performed simultaneously on the same experiment. The results state that accurate and useful quantities can be derived from a PINN model trained on real experimental data which is encouraging for a better description of gravity currents. |
| format | Article |
| id | doaj-art-a99b2540705f498794373f69f82f6c9b |
| institution | Kabale University |
| issn | 2632-2153 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IOP Publishing |
| record_format | Article |
| series | Machine Learning: Science and Technology |
| spelling | doaj-art-a99b2540705f498794373f69f82f6c9b2025-02-10T11:13:42ZengIOP PublishingMachine Learning: Science and Technology2632-21532025-01-016101503210.1088/2632-2153/adaca0Quantitative assessment of PINN inference on experimental data for gravity currents flowsMickaël Delcey0https://orcid.org/0000-0003-2225-8531Yoann Cheny1Jean Schneider2Simon Becker3Sébastien Kiesgen De Richter4LEMTA, Université de Lorraine , CNRS, 2, Avenue de la Forêt de Haye, B.P. 160, 54500 Vandœuvre-lés-Nancy, FranceLEMTA, Université de Lorraine , CNRS, 2, Avenue de la Forêt de Haye, B.P. 160, 54500 Vandœuvre-lés-Nancy, FranceLEMTA, Université de Lorraine , CNRS, 2, Avenue de la Forêt de Haye, B.P. 160, 54500 Vandœuvre-lés-Nancy, FranceLEMTA, Université de Lorraine , CNRS, 2, Avenue de la Forêt de Haye, B.P. 160, 54500 Vandœuvre-lés-Nancy, FranceLEMTA, Université de Lorraine , CNRS, 2, Avenue de la Forêt de Haye, B.P. 160, 54500 Vandœuvre-lés-Nancy, France; Institut Universitaire de France (IUF) , Paris, FranceIn this paper, we apply physics informed neural networks (PINNs) to infer velocity and pressure field from light attenuation technique (LAT) measurements for gravity current induced by lock-exchange. In a PINN model, physical laws are embedded in the loss function of a neural network, such that the model fits the training data but is also constrained to reduce the residuals of the governing equations. PINNs are able to solve ill-posed inverse problems training on sparse and noisy data, and therefore can be applied to real engineering applications. The noise robustness of PINNs and the model parameters are investigated in a 2 dimensions toy case on a lock-exchange configuration, employing synthetic data. Then we train a PINN with experimental LAT measurements and quantitatively compare the velocity fields inferred to particle image velocimetry measurements performed simultaneously on the same experiment. The results state that accurate and useful quantities can be derived from a PINN model trained on real experimental data which is encouraging for a better description of gravity currents.https://doi.org/10.1088/2632-2153/adaca0physics informed neural networksgravity currentsexperimental measuresfluid mechanics |
| spellingShingle | Mickaël Delcey Yoann Cheny Jean Schneider Simon Becker Sébastien Kiesgen De Richter Quantitative assessment of PINN inference on experimental data for gravity currents flows Machine Learning: Science and Technology physics informed neural networks gravity currents experimental measures fluid mechanics |
| title | Quantitative assessment of PINN inference on experimental data for gravity currents flows |
| title_full | Quantitative assessment of PINN inference on experimental data for gravity currents flows |
| title_fullStr | Quantitative assessment of PINN inference on experimental data for gravity currents flows |
| title_full_unstemmed | Quantitative assessment of PINN inference on experimental data for gravity currents flows |
| title_short | Quantitative assessment of PINN inference on experimental data for gravity currents flows |
| title_sort | quantitative assessment of pinn inference on experimental data for gravity currents flows |
| topic | physics informed neural networks gravity currents experimental measures fluid mechanics |
| url | https://doi.org/10.1088/2632-2153/adaca0 |
| work_keys_str_mv | AT mickaeldelcey quantitativeassessmentofpinninferenceonexperimentaldataforgravitycurrentsflows AT yoanncheny quantitativeassessmentofpinninferenceonexperimentaldataforgravitycurrentsflows AT jeanschneider quantitativeassessmentofpinninferenceonexperimentaldataforgravitycurrentsflows AT simonbecker quantitativeassessmentofpinninferenceonexperimentaldataforgravitycurrentsflows AT sebastienkiesgenderichter quantitativeassessmentofpinninferenceonexperimentaldataforgravitycurrentsflows |