Reliability analysis and functional design using Bayesian networks generated automatically by an “Idea Algebra” framework

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Currently, the application of intelligent tools and decision methods in reliability, dependability, and maintenance analysis shows an increasing trend due to the complexity of the systems. The Bayesian Network (BN)-based methods are very efficient for this kind of analysis, but the process of constructing the BN is very routine and time-consuming and requires a lot of human effort. One good solution is to construct a proper BN model of a system, with the guidance of a semantic method. Thus, we introduce a novel methodology that automates the BN generation process for reliability analysis directly from the system's description. The method uses an engineering design representation technique to create a BN and allows to evaluate it automatically. The created GeNIe files can be edited and reused for further analysis which increases reusability of engineering design data. For the validation of the developed method, it was applied to an automotive powertrain system. Finally, the Bayesian Networks of 25 different automobiles were evaluated and tested with sensitivity analysis.

Original languageEnglish
Pages (from-to)211-225
Number of pages15
JournalReliability Engineering and System Safety
Volume180
DOIs
Publication statusPublished - Dec 1 2018

Fingerprint

Bayesian networks
Reliability analysis
Algebra
Powertrains
Reusability
Automobiles
Sensitivity analysis
Semantics

Keywords

  • Bayesian networks
  • Engineering design
  • Idea Algebra
  • Reliability analysis

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Industrial and Manufacturing Engineering

Cite this

@article{35fb8dd1a8ae4b18b5c7aa4d2f259df2,
title = "Reliability analysis and functional design using Bayesian networks generated automatically by an “Idea Algebra” framework",
abstract = "Currently, the application of intelligent tools and decision methods in reliability, dependability, and maintenance analysis shows an increasing trend due to the complexity of the systems. The Bayesian Network (BN)-based methods are very efficient for this kind of analysis, but the process of constructing the BN is very routine and time-consuming and requires a lot of human effort. One good solution is to construct a proper BN model of a system, with the guidance of a semantic method. Thus, we introduce a novel methodology that automates the BN generation process for reliability analysis directly from the system's description. The method uses an engineering design representation technique to create a BN and allows to evaluate it automatically. The created GeNIe files can be edited and reused for further analysis which increases reusability of engineering design data. For the validation of the developed method, it was applied to an automotive powertrain system. Finally, the Bayesian Networks of 25 different automobiles were evaluated and tested with sensitivity analysis.",
keywords = "Bayesian networks, Engineering design, Idea Algebra, Reliability analysis",
author = "Andas Amrin and Vasileios Zarikas and Christos Spitas",
year = "2018",
month = "12",
day = "1",
doi = "10.1016/j.ress.2018.07.020",
language = "English",
volume = "180",
pages = "211--225",
journal = "Reliability Engineering and System Safety",
issn = "0951-8320",
publisher = "Elsevier",

}

TY - JOUR

T1 - Reliability analysis and functional design using Bayesian networks generated automatically by an “Idea Algebra” framework

AU - Amrin, Andas

AU - Zarikas, Vasileios

AU - Spitas, Christos

PY - 2018/12/1

Y1 - 2018/12/1

N2 - Currently, the application of intelligent tools and decision methods in reliability, dependability, and maintenance analysis shows an increasing trend due to the complexity of the systems. The Bayesian Network (BN)-based methods are very efficient for this kind of analysis, but the process of constructing the BN is very routine and time-consuming and requires a lot of human effort. One good solution is to construct a proper BN model of a system, with the guidance of a semantic method. Thus, we introduce a novel methodology that automates the BN generation process for reliability analysis directly from the system's description. The method uses an engineering design representation technique to create a BN and allows to evaluate it automatically. The created GeNIe files can be edited and reused for further analysis which increases reusability of engineering design data. For the validation of the developed method, it was applied to an automotive powertrain system. Finally, the Bayesian Networks of 25 different automobiles were evaluated and tested with sensitivity analysis.

AB - Currently, the application of intelligent tools and decision methods in reliability, dependability, and maintenance analysis shows an increasing trend due to the complexity of the systems. The Bayesian Network (BN)-based methods are very efficient for this kind of analysis, but the process of constructing the BN is very routine and time-consuming and requires a lot of human effort. One good solution is to construct a proper BN model of a system, with the guidance of a semantic method. Thus, we introduce a novel methodology that automates the BN generation process for reliability analysis directly from the system's description. The method uses an engineering design representation technique to create a BN and allows to evaluate it automatically. The created GeNIe files can be edited and reused for further analysis which increases reusability of engineering design data. For the validation of the developed method, it was applied to an automotive powertrain system. Finally, the Bayesian Networks of 25 different automobiles were evaluated and tested with sensitivity analysis.

KW - Bayesian networks

KW - Engineering design

KW - Idea Algebra

KW - Reliability analysis

UR - http://www.scopus.com/inward/record.url?scp=85050722408&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85050722408&partnerID=8YFLogxK

U2 - 10.1016/j.ress.2018.07.020

DO - 10.1016/j.ress.2018.07.020

M3 - Article

VL - 180

SP - 211

EP - 225

JO - Reliability Engineering and System Safety

JF - Reliability Engineering and System Safety

SN - 0951-8320

ER -