Contribution à un système de retour d'expérience basé sur le raisonnement à partir de cas conversationnel : application à la gestion des pannes de machines industrielles

Abstract : Faced with the fast technological development of products, incremental innovation of new products, and the mobility of their most experienced staff, companies are seeking to formalize and capitalize on the experiences and know-how of their personnel in order to reuse them later. To deal with these problems, the conversational case based reasoning (CCBR) approach is a potential answer to the question of capitalization and reuse of knowledge. Our research focuses on methods to manage experience feedback (EF). We are placed in the field of experience feedback applied to technical problem solving. Our methodology for creating aided failure diagnosis systems is divided into four phases: the events description, the development of all solutions to failures, the arrangement of a diagnostic aid through fault trees and setting up a computer system. We based our work on the fault tree approach in order to extract tacit knowledge and its formalization. Our objective was to create decision protocols in order to assist the expert in solving an industrial problem. Therefore, we have proposed a formulation and development of conversational cases in diagnosis. These cases must be memorised in a database of cases. To validate our proposal methodology, we have carried out the experimental phase in an industrial company in eastern France. This experiment allowed us to validate our work and highlight its advantages and limitations. Finally, we propose a computer model designed for the company. This model enables failure diagnosis by creating the case in a case base for later utilization
Document type :
Theses
Complete list of metadatas

Cited literature [170 references]  Display  Hide  Download

https://hal.univ-lorraine.fr/tel-01748785
Contributor : Thèses Ul <>
Submitted on : Thursday, March 29, 2018 - 11:49:12 AM
Last modification on : Thursday, June 28, 2018 - 12:10:27 PM
Long-term archiving on : Friday, September 14, 2018 - 7:23:50 AM

File

2009_ARMAGHAN_N.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : tel-01748785, version 1

Collections

Citation

Negar Armaghan. Contribution à un système de retour d'expérience basé sur le raisonnement à partir de cas conversationnel : application à la gestion des pannes de machines industrielles. Autre. Institut National Polytechnique de Lorraine, 2009. Français. ⟨NNT : 2009INPL026N⟩. ⟨tel-01748785⟩

Share

Metrics

Record views

33

Files downloads

563