Data-based models for flow control

Abstract : This thesis presents two experimental data-based model design techniques: the system identification and the data-assimilation. The first allows the construction of a model, based on the system’s input-output signals only. Several flow-control examples, using system identification, are described: flow inside a channel, and behind a cylinder. Data-assimilation is another data-based technique which makes use of a physical model. The assimilated model gives access to physics based interpolation and extrapolation of a scattered data-set. Three examples of mean-velocity measurement assimilation are presented: a turbulent boundary layer, the flow around a cylinder, and the flow over an idealized airfoil. The impact of measurement uncertainty is demonstrated. An extension is developed for both methods to cope with corrupted data.
Complete list of metadatas

Cited literature [156 references]  Display  Hide  Download

https://hal-polytechnique.archives-ouvertes.fr/tel-01097717
Contributor : Nicolas Dovetta <>
Submitted on : Friday, March 4, 2016 - 10:03:16 PM
Last modification on : Wednesday, March 27, 2019 - 4:39:25 PM
Long-term archiving on: Sunday, June 5, 2016 - 10:11:52 AM

Identifiers

  • HAL Id : tel-01097717, version 1

Collections

Citation

Nicolas Dovetta. Data-based models for flow control. Fluids mechanics [physics.class-ph]. Ecole Polytechnique X, 2014. English. ⟨tel-01097717⟩

Share

Metrics

Record views

235

Files downloads

397