You need JavaScript to view this

Diapo, applying advanced AI methods to diagnosis of PWR reactor coolant pump; Diapo, application de methodes de pointes en intelligence artificielle pour le diagnostic des groupes motopompes primaires des REP

Abstract

Electricite de France has decided to increase the capabilities of its monitoring and diagnostic architecture with the development of an AI system for reactor coolant pump diagnostic support. This development is carried out with the cooperation of the equipment constructor Jeumont Schneider Industries. This diagnostic system will eventually be included in an integrated surveillance architecture. We present the architecture of the system and the basics of the knowledge model used. Main data for diagnosis are provided by sensor data issued by the pump monitoring system. Diagnostic reasoning is based on the cooperation of two main activities : a heuristic search among typical symptomatic situations that leads to the formulation of hypotheses and a ``deep`` causal analysis that consists in backtracking from identified situations up to initial faults or causes. This approach is well fitted to field expert reasoning, and provides powerful diagnostic capabilities that help to overcome conventional limitations of expert systems entirely based on heuristic knowledge. (authors). 9 figs., 11 refs.
Publication Date:
Jan 01, 1993
Product Type:
Technical Report
Report Number:
EDF-93-NJ-00055
Reference Number:
SCA: 210200; PA: AIX-25:028921; EDB-94:051536; NTS-94:018344; SN: 94001178246
Resource Relation:
Other Information: PBD: Jan 1993
Subject:
21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS; PUMPS; REACTOR MONITORING SYSTEMS; PWR TYPE REACTORS; ARTIFICIAL INTELLIGENCE; PRIMARY COOLANT CIRCUITS; 210200; POWER REACTORS, NONBREEDING, LIGHT-WATER MODERATED, NONBOILING WATER COOLED
OSTI ID:
10138779
Research Organizations:
Electricite de France (EDF), 92 - Clamart (France)
Country of Origin:
France
Language:
French
Other Identifying Numbers:
Other: ON: DE94621275; TRN: FR9401180028921
Availability:
OSTI; NTIS (US Sales Only); INIS
Submitting Site:
FRN
Size:
10 p.
Announcement Date:
Jul 05, 2005

Citation Formats

Porcheron, M, and Ricard, B. Diapo, applying advanced AI methods to diagnosis of PWR reactor coolant pump; Diapo, application de methodes de pointes en intelligence artificielle pour le diagnostic des groupes motopompes primaires des REP. France: N. p., 1993. Web.
Porcheron, M, & Ricard, B. Diapo, applying advanced AI methods to diagnosis of PWR reactor coolant pump; Diapo, application de methodes de pointes en intelligence artificielle pour le diagnostic des groupes motopompes primaires des REP. France.
Porcheron, M, and Ricard, B. 1993. "Diapo, applying advanced AI methods to diagnosis of PWR reactor coolant pump; Diapo, application de methodes de pointes en intelligence artificielle pour le diagnostic des groupes motopompes primaires des REP." France.
@misc{etde_10138779,
title = {Diapo, applying advanced AI methods to diagnosis of PWR reactor coolant pump; Diapo, application de methodes de pointes en intelligence artificielle pour le diagnostic des groupes motopompes primaires des REP}
author = {Porcheron, M, and Ricard, B}
abstractNote = {Electricite de France has decided to increase the capabilities of its monitoring and diagnostic architecture with the development of an AI system for reactor coolant pump diagnostic support. This development is carried out with the cooperation of the equipment constructor Jeumont Schneider Industries. This diagnostic system will eventually be included in an integrated surveillance architecture. We present the architecture of the system and the basics of the knowledge model used. Main data for diagnosis are provided by sensor data issued by the pump monitoring system. Diagnostic reasoning is based on the cooperation of two main activities : a heuristic search among typical symptomatic situations that leads to the formulation of hypotheses and a ``deep`` causal analysis that consists in backtracking from identified situations up to initial faults or causes. This approach is well fitted to field expert reasoning, and provides powerful diagnostic capabilities that help to overcome conventional limitations of expert systems entirely based on heuristic knowledge. (authors). 9 figs., 11 refs.}
place = {France}
year = {1993}
month = {Jan}
}