Skip to Main content Skip to Navigation
Theses

DIAGNOSTIC LOGIQUE DES SYSTEMES COMPLEXES ET DYNAMIQUES DANS UN CONTEXTE MULTI-AGENT

Abstract : This thesis proposes a methodology to design a reliable diagnosis systems permitting to comprehend the dynamic complex and spatially distributed systems. The proposed results lean on the, one hand, techniques of formal analysis diagnosis or to basis of consistency, that permits to guarantee the exactness of the diagnosis analysis, and on the other hand, on multi-agents paradigm. The proposed algorithms permit to deduct all possible faults for an observed behavior while classifying them according to different criteria of plausibility. Our contribution consisted in proposing a method of diagnosis that pulls left of the two DX approaches (Artificial intelligence community) and FDI (Control community) while distinguishing the phase of detection, that can be done by various and sometimes very sophisticated techniques (state observer, parity relation, signal based,...), of the phase of localization or diagnostic analysis, that must permit to guarantee what can be, and to analyze all available information to deduct the just diagnosis and possible suit of it. Besides, we showed that it was possible to fear uncertainties of decision while transposing the logic of the analysis diagnosis in fuzzy logic. Work presented in this memory has been developed in the setting of the European project called MAGIC (Multi-Agents-Based Diagnostic Data Acquisition and Management in Complex systems). Results are under transfer toward different industries thanks to the industrial partners of the project: the SATE enterprises (System Advanced Technologies Engineering) and SMS-DEMAG.
Complete list of metadata

Cited literature [91 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00008768
Contributor : Samir Touaf <>
Submitted on : Monday, March 14, 2005 - 12:16:15 PM
Last modification on : Wednesday, November 4, 2020 - 3:06:57 AM
Long-term archiving on: : Friday, April 2, 2010 - 9:42:30 PM

Identifiers

  • HAL Id : tel-00008768, version 1

Collections

Citation

Samir Touaf. DIAGNOSTIC LOGIQUE DES SYSTEMES COMPLEXES ET DYNAMIQUES DANS UN CONTEXTE MULTI-AGENT. Automatique / Robotique. Université Joseph-Fourier - Grenoble I, 2005. Français. ⟨tel-00008768⟩

Share

Metrics

Record views

731

Files downloads

4331