A new hybrid Bayesian network approach for modeling reliability - Ifsttar Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

A new hybrid Bayesian network approach for modeling reliability

Résumé

In this paper, a hybrid discrete-continuous Graphical Duration Models is proposed. Since the interest of the Weibull density was demonstrated for reliability analysis, this paper focuses on the use of Weibull densities for modeling sojourn times in each state of the system. This extension of the standard GDM requires a specific structure that we call Weibull-Hybrid Graphical Duration Models (W-HGDM). The main contribution of this study lays in the proposal of a specific inference algorithm for such hybrid networks. Finally, comparisons of reliability estimation will be proposed for both standard GDM and W-HGDM.
Fichier principal
Vignette du fichier
doc00029174.pdf (468.28 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01816353 , version 1 (15-06-2018)

Identifiants

Citer

Florence Petiet, Olivier Francois, Laurent Bouillaut. A new hybrid Bayesian network approach for modeling reliability. ESREL2018, European Safety and Reliability Conference, Jun 2018, Trondheim, Norway. pp.765-771, ⟨10.1201/9781351174664-95⟩. ⟨hal-01816353⟩
108 Consultations
275 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More