Generally, diagnosis methods use one kind of knowledge only (i.e., a model, a numerical data base, a base of previously solved cases, and so on). Even if these methods perform well on simple enough systems, they are restrictively efficient for complex systems like cars. In practice, it is impossible or too expensive to acquire all the knowledge necessary to their implementation (complete modelling of the system, cases representing all the possible breakdowns, and so on). The approach followed in this paper is to use different and heterogeneous kinds of knowledge into an integrated diagnosis tool. This is done by using different diagnosis methods in particular model-based reasoning, pattern recognition and semantic information retrieval. The MODE framework then enables all these methods to collaborate defining the required inputs and the possible outputs of each method and proposing a general scheduler sequencing the application of each method depending on the available data and the user decisions. This framework is illustrated on real data issued from the automotive domain.
Diagnostic a base de connaissances heterogenes: application au domaine de l'automobile
Heterogene wissensbasierte Diagnostik: eine Anwendung im Fahrzeugbau
Heterogeneous knowledge based diagnosis: an application to the automotive domain
2009
6 Seiten, 4 Bilder, 8 Quellen
Aufsatz (Zeitschrift)
Französisch
wissensbasiertes System , Programmdiagnose , Modelluntersuchung , numerisches Verfahren , Automobilindustrie , Anwendung im Fahrzeugbau , Ausfallanalyse , Zuverlässigkeit , Diagnoseverfahren , Mustererkennung , Datenmodell , Echtzeitdatenverarbeitung , Kraftfahrzeugelektronik , Fahrzeugwartung , Instandhaltungsplanung , Diagnose (Technik) , Datenbanknutzung , Wartungseinrichtung
Les applications les plus recentes du thixomoulage dans le domaine de l'automobile
Tema Archiv | 2001
|Diagnostic et Analyse Electroniques de l'Automobile
Tema Archiv | 1972
|ACTUALITÉS - L'automobile << croquée >>
Online Contents | 2001
|Europäisches Patentamt | 2020
|SLUB | 2011
|