Stochastic models of failure in random environments

Authors

  • Anthony Desmond

    1. Department of Mathematics and Staristics University of Guelph Guelph, Ontario NIG 2WI
    Search for more papers by this author
    • Co-winner of the 1983 Pierre Robillard award for the best Ph. D. thesis in Statistics defended in a Canadian university.


Abstract

Stochastic models of failure modes of frequent occurrence in the engineering sciences are considered. The failure-producing stress environment is modelled as a stationary stochastic process. Using theoretical properties of the sample paths of these processes, failure-time distributions which belong to the Birnbaum-Saunders family are obtained. Several examples of particular engineering relevance are treated.

Abstract

Grǎce aux processus stochastiques, il est possible de représenter de manière assez juste de nombreux types de defaillances ďorigine répandue en ingenierie. Un processus stochastique en régime permanent peut notamment servir de modèle pour certaines causes exogènes de défaillance. En faisant appel aux propriétés des chemins échantillonnaux de tels processus, il est possible de calculer les distributions des temps de bris. On constate que celles-ci appartiennent à la famille de lois dite de Birnbaum-Saunders. On présente aussi de nombreux exemples ďintérět pratique.

Ancillary