Likelihood ratio and dispersive orders of parallel and series systems consisting of dependent multiple-outlier components

Abstract

This article discusses stochastic comparisons of parallel and series systems consisting of dependent multiple-outlier scale components with Archimedean copula as the joint distribution in terms of likelihood ratio and dispersive orders. Under some conditions on the baseline distribution and on the Archimedean copula, likelihood ratio and dispersive orders between parallel and series systems have been established. Several examples have been presented to illustrate the results established and to demonstrate their reliability implications.