Abstract
This paper presents a step-stress accelerated life test for Weibull distribution with non-constant shape parameter. Proposed model is developed for two stress levels. While changing the stress, one expects that mean time to failure (MTTF) would decrease. Therefore, we choose MTTF as a nonnegative and monotonically decreasing function of stress. Assuming the scale parameter fixed, the shape parameter is function of stress through the link function of MTTF. On the basis of the proposed model, we describe the maximum likelihood (ML) estimation procedure to estimate the parameters involved in the model. Expression of the Fisher's information matrix has been derived. A simulation is carried out. The reliability and percentile estimates with their confidence limits at used stress level are obtained. For given values of the parameters, the optimum stress change time has been obtained by minimizing the variance of a stated percentile.
Get full access to this article
View all access options for this article.