Abstract
This article presents a method allowing to integrate the prior knowledge on mechanical component in the failure probability estimation. The proposed method uses the Bayesian method and the Monte-Carlo results obtained by 3 kinds of simulation : Direct, Importance sampling and Conditional expectation. Different methods of prior distribution definition, which is needed for Bayesian inference, are presented using all available information concerning the studied system (First Order Reliability Method (FORM), First Order Second Moment (FOSM), Reliability handbook, …). Two examples illustrate the method and show its efficiency.
Key Words:
Additional information
Notes on contributors
F. Guérin
Fabrice Guerin MS and a PhD in mechanical engineering from the University of Nantes, France. Guerin is currently a professor in the quality and reliability department at the University of Angers (France), where he teaches mechanical technology, statistics, metrology, and reliability. His research interest lies in the reliability assessment of mechanical components using accelerated testing and Bayesian statistics. Since 2002 he has worked on European Project EUREKA/PIDEA/ACLIMAT on the reliability assessment of embedded systems by accelerated testing.
M. Barreau
Mihaela Barreau Engineer in computer science since 1989, obtained the PhD in automatic control and computer science in the University of Angers (France) in 1997. She is associate professor in the Quality & Reliability Engineering Department of the University of Angers. Her interests areas are modeling and optimization methods and mechatronic systems dependability.
A. Charki
Aberafi Charki Doctor and an engineer in mechanical engineering, he was a project leader in the automotive industry and an accreditation manager in the Laboratory section of the French Committee of Accreditation (COFRAC). Since 2004, he has been an assistant professor in the quality and reliability department at the University of Angers (France). He is a researcher in the reliability of mechanical and engineering systems. Moreover, he is the president of the African Committee of Metrology, an association contributing to the development and recognition of industrial competences in Africa. He also assesses laboratories for the French Committee of Accreditation.
A. Todoskoff
Alexis Todoskoff PhD in automatic control and man-machine systems from the University of Valenciennes (France) in 1999. Since 1999, he is an associate professor in automatic control and software engineering in the Quality & Reliability Engineering Department of the University of Angers. His main research field is in the reliability of software systems.