Acceptance reliability sampling plan for items from heterogeneous populations

Existing reliability sampling plans reported in the literature assume that the items belong to homogeneous populations. However, at many practical instances, a population of the manufactured items is a mixed one. In this paper, we consider the case when a population consists of items that belong to two subpopulations, that is, the subpopulation with better reliability characteristics and that with worse reliability characteristics. Thus, a variables acceptance reliability sampling plan is proposed and its properties are investigated under this assumption. A sequential numerical algorithm for finding the parameters of the proposed sampling plan is developed and applied. Furthermore, the lifetime of the population before the acceptance test and that of the population which have passed the sampling test are stochastically compared.