Skip to Main content Skip to Navigation
Theses

Assistance au calage de modèles numériques en hydraulique fluviale -- Apports de l'intelligence artificielle

Abstract : Model calibration aims at simulating reference events thanks to an adjustment of physically-based parameters. This thesis proposes a knowledge-based systems approach to this task. After a definition of key concepts, a state of the art and an analysis of the different objects, procedures and reasonings applied to achieve this task are presented. These elements are formalized and integrated within a computer-aided calibration support system. The knowledge is structured into three levels : generic knowledge, knowledge specific to 1D river hydraulics, and knowledge about the advanced use of the simulation code employed. Two real-life case studies are then handled on the basis of available data and intended model application. This thesis allowed the capitalization of know-how which led to build an operational hydraulic model calibration support prototype and to develop artificial intelligence tools both dedicated to calibration and independent from the selected scientific domain.
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00010185
Contributor : Jean-Philippe Vidal <>
Submitted on : Friday, September 16, 2005 - 9:14:50 PM
Last modification on : Monday, February 8, 2021 - 2:32:06 PM
Long-term archiving on: : Friday, April 2, 2010 - 10:18:21 PM

Identifiers

  • HAL Id : tel-00010185, version 1

Collections

Citation

Jean-Philippe Vidal. Assistance au calage de modèles numériques en hydraulique fluviale -- Apports de l'intelligence artificielle. Interface homme-machine [cs.HC]. Institut National Polytechnique de Toulouse - INPT, 2005. Français. ⟨tel-00010185⟩

Share

Metrics

Record views

719

Files downloads

5828