Issue |
ESAIM: ProcS
Volume 73, 2023
CEMRACS 2021 - Data Assimilation and Reduced Modeling for High Dimensional Problems
|
|
---|---|---|
Page(s) | 89 - 106 | |
DOI | https://doi.org/10.1051/proc/202373089 | |
Published online | 30 August 2023 |
A component-based data assimilation strategy with applications to vascular flows*
1
Laboratoire Analyse, Géométrie et Applications, CNRS UMR 7539, USPN
2
Laboratoire de Mathématiques de Reims, CNRS UMR 9008, URCA
3
École Polytechnique Fédérale de Lausanne (EPFL), CSQI-MATH
4
IMB, UMR 5251, Univ. Bordeaux, 33400, Talence, France; INRIA, Inria Bordeaux Sud-Ouest, Team MEMPHIS, 33400, Talence, France
We present a parameterized-background data-weak (PBDW) approach [Y Maday, AT Patera, JD Penn, M Yano, Int J Numer Meth Eng, 102(5), 933–965] to the steady-state variational data assimilation (DA) problem for systems modeled by partial differential equations (PDEs) and characterized by multiple interconnected components, with emphasis on vascular flows. We focus on the problem of reconstructing the state of the system in one specific component, based on local measurements. The PBDW approach does not require the solution of any PDE model at prediction stage (projection-by-data) and, as such, enables local state estimates on single components, as long as good background and update spaces for the estimation can be constructed. We discuss the application of PBDW to a two-dimensional steady Navier-Stokes problem for a family of parameterized geometries, and investigate instead the effects of enforcing no-slip boundary conditions and incompressibility constraints on the background and update spaces to enhance the state estimation. Furthermore, we show an actionable strategy to train local reduced-order bases (ROBs) for the background space that can later be used for DA tasks.
email: bui@math.univ-paris13.fr
email: pierre.mollo@univ-reims.fr
email: fabio.nobile@epfl.ch
email: tommaso.taddei@inria.fr
© EDP Sciences, SMAI 2023
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.