Skip to Main content Skip to Navigation

Fiabilité des systèmes embarqués

Abstract : The work presented in this doctoral thesis deals with systems reliability estimation. Reliability assesment is becoming an integral part of the design process of complex systems in order to highlight potential risk areas so they can be dealt with at the design stage of the project. Indeed, the early control of system specifications allows diminishing operating ( either financial or security ) risks. Since systems must be more and more reliable and offer longer guarantees, it is necessary to check the compliance of their performances as early as possible. Testing prototypes is an appropriate approach for evaluating the reliability of a system before it is mass-produced. This process requires long testing times and huge numbers of prototypes since the latter are more and more reliable, therefore extremely diminishing the probability of failure. One solution for diminishing testing times and therefore their costs is based on accelerated life testing. ALT is achieved by subjecting the test units to stresses wich are more severe than the normal operating conditions, in order to accelerate damage and failure occurence ( the failure mechanisms must be the same as under normal operating conditions ) and thus reducing the time needed for obtaining behavioral characteristics of parts for normal operating conditions. We proposed using parametric and semi-parametric standard accelerated life models in test data analysis. In order to obtain accurate estimations, we used Bayesian approach, allowing the use of all existent knowledge on the reliability of the studied part. We proposed thus some estimation methods, a technique to construct the conjugate "a priori" distribution together with an analysis method in order to check if "a priori" knowledge and tests results are compatible. Finally, we developed some reliability estimation methods taking into account both use and environment variables.
Document type :
Complete list of metadatas
Contributor : Anne-Marie Plé <>
Submitted on : Tuesday, March 30, 2010 - 1:02:36 PM
Last modification on : Friday, October 23, 2020 - 4:36:14 PM
Long-term archiving on: : Wednesday, September 8, 2010 - 8:42:43 PM


  • HAL Id : tel-00468219, version 1


Sorin Voiculescu. Fiabilité des systèmes embarqués. Sciences de l'ingénieur [physics]. Université d'Angers, 2009. Français. ⟨tel-00468219⟩



Record views


Files downloads