Advanced RESTART method for the estimation of the probability of failure of highly reliable hybrid dynamic systems - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Reliability Engineering and System Safety Année : 2016

Advanced RESTART method for the estimation of the probability of failure of highly reliable hybrid dynamic systems

Résumé

The efficient estimation of system reliability characteristics is of paramount importance for many engineering applications. Real world system reliability modeling calls for the capability of treating systems that are: i) dynamic, ii) complex, iii) hybrid and iv) highly reliable. Advanced Monte Carlo (MC) methods offer a way to solve these types of problems, which are feasible according to the potentially high computational costs. In this paper, the REpetitive Simulation Trials After Reaching Thresholds (RESTART) method is employed, extending it to hybrid systems for the first time (to the authors' knowledge). The estimation accuracy and precision of RESTART highly depend on the choice of the Importance Function (IF) indicating how close the system is to failure: in this respect, proper IFs are here originally proposed to improve the performance of RESTART for the analysis of hybrid systems. The resulting overall simulation approach is applied to estimate the probability of failure of the control system of a liquid holdup tank and of a pump-valve subsystem subject to degradation induced by fatigue. The results are compared to those obtained by standard 2 MC simulation and by RESTART with classical IFs available in the literature. The comparison shows the improvement in the performance obtained by our approach.
Fichier principal
Vignette du fichier
2016_06_04_RESTART_RESS_accepted.pdf (527.51 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01330158 , version 1 (10-06-2016)

Identifiants

Citer

Pietro Turati, Nicola Pedroni, Enrico Zio. Advanced RESTART method for the estimation of the probability of failure of highly reliable hybrid dynamic systems. Reliability Engineering and System Safety, 2016, ⟨10.1016/j.ress.2016.04.020⟩. ⟨hal-01330158⟩
116 Consultations
277 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More