• DocumentCode
    1855692
  • Title

    Reliability and availability estimation for complex systems: a simple concept and tool

  • Author

    Collas, Gérard

  • Author_Institution
    R&D Dept., Bull, Les Clayes-sous-Bois, France
  • fYear
    1994
  • fDate
    24-27Jan 1994
  • Firstpage
    110
  • Lastpage
    113
  • Abstract
    This paper describes a simple modeling concept and an associated computerized tool for estimating complex repairable availability. The method is intended as a first step in computing dependability metrics for making architecture choices during conceptual system design. The modeling approach is based on classifying element failures as permanent, intermittent and transient and treats both hardware and software failures. It comprises a combination of the top-down reliability block diagram to represent system hierarchy, a simplified Markov chain representation of equipment state transitions, and conditional probability definitions to represent dependence between equipment levels. The key concept is the partitioning of the failure rates in correspondence to the failure types. This permits simpler computation and as a result the method is implemented as a computerized analysis tool in the form of a customized Microsoft Excel menu
  • Keywords
    large-scale systems; probability; reliability theory; software reliability; Markov chain representation; availability estimation; complex repairable availability; complex systems; computerized tool; conceptual system design; conditional probability; customized Microsoft Excel menu; dependability metrics; failure rates; hardware failures; intermittent; partitioning; permanent; reliability; software failures; top-down reliability block diagram; transient; Application software; Availability; Computational modeling; Computer architecture; Hardware; Logistics; Maintenance; Personal communication networks; Software tools;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Reliability and Maintainability Symposium, 1994. Proceedings., Annual
  • Conference_Location
    Anaheim, CA
  • Print_ISBN
    0-7803-1786-6
  • Type

    conf

  • DOI
    10.1109/RAMS.1994.291091
  • Filename
    291091