Μέτρα σπουδαιότητας στη θεωρία αξιοπιστίας

Master Thesis
Author
Δαμηλάτης, Νικόλαος Κ.
Date
2007-11-26View/ Open
Abstract
Several measures of component importance have been introduced in the literature. Component importance measures are important tools to evaluate and rank the impact in system’s quality (reliability) of individual components or even of the impact of the topology of each position within a system. They also have been widely used as tools for identifying system weaknesses, and prioritize improvement activities, even to assign probabilities in an optimal fashion (optimal arrangement) during the design of the system. This dissertation presents all well-known measures of importance and some less-widely used. It gives formulae for their numerical evaluation and the corresponding formulae for the class of Markov Chain Imbeddable Systems. Finally, it calculates the majority of importance measures for each system, and proceed to the ordering of the components (in terms of each specific importance measure), in the cases where this is feasible.