Step­Stress Accelerated Life Testing for Pareto Distribution with Type I Censoring

Abstract

This paper presents step­stress accelerated life test for the Pareto distribution with type I censoring. It is assumed that Pareto scale parameter depends on the stress variable and the shape parameter remains constant. It is also assumed that failures occur according to a cumulative exposure modcl(CEM) by Nelson (1990). On the basis of this model we describe maximum likelihood estimation proceuure and expression of the Fisher information matrix has been derived in closed form. A simulation study has been done to obtain the maximum likelihood estimates(MLE), Fisher information matrix and hence the variance­covariance matrix of the parameters. For selected values of the rlesign parameters the optimum low stress as well as the stress change time are obtained which minimize the asymptotic variance of the MLE of a stated percentile.