Coordinator : H.
Garnier
Members : T.
Bastogne, D.
Brie, O.
Cervellin, E-H.
Djermoune, M.
Gilson-Bagrel, S.
Huraux, A.
Richard, P.
Sibille
PhD students : S.
Dobre
System identification, or experimental data-based modelling, is one of the major concerns to the majority of scientific disciplines. It involves at the same time a scientific methodology and a set of techniques aiming at determining mathematical models able to reproduce as accurately as possible the dynamic behavior of a physical, chemical, biological,... or environmental system.
This methodology, initially developed within the automatic control community, is also used in other fields. It is based on system theory and uses various tools coming from applied mathematics, statistical signal processing, information theory as well as physics. It is concretized by signal processing algorithms of experimental data.
The activities of this project relate to the theoretical and applied aspects of system identification. Special attention are currently given to the following themes:
 |
 |
Identification of continuous-time linear models; |
|
|
Identification in closed-loop; |
|
|
Identification of errors-in-variables models. |
The know-how of the team members are capitalised in a toolbox called CONTSID.
The current application areas are:
|
|
Experimental modelling of environmental systems: rainfall-flow modelling, pollutant transport modelling in rivers; |
|
|
Experimental modelling of biological systems: we are currently investigating the contributions of system identification for the experimental modelling of the mechanisms related to the photodynamic therapy of cancerous cells; |
|
|
Experimental modelling and monitoring of MIG arc welding process. |
To know more about the current activities carried out by the members of the
System
Identification project:
|
|
to know more |
|
|
or contact Prof. Hugues Garnier
Centre de Recherche en Automatique de Nancy (CRAN) Tél. : 33 3 83 68 44 90 |