BAYESIAN SAMPLING PLANS WITH PROGRESSIVE CENSORING AND WARRANTY POLICY

Abstract

The paper investigates the design of Bayesian sampling plan for the exponential lifetime model under progressive type-II censoring, in which items are manufactured in batches and sold to consumers with a general rebate warranty policy. Assume that the mean lifetime of items is random and varies from lot to lot. A cost model consists of the cost per item on test, the cost per item of test time and the costs of rejecting and accepting an item is established, and an algorithm is provided to determine the optimal Bayesian sampling plan which minimizes the expected average cost per lot. The use of the proposed method is illustrated by numerical results and an example. A sensitivity study is conducted to evaluate the influences of the removal scheme and using incorrect estimates for the hyper-parameters on the proposed sampling plans. The proposed method can be extended to the Weibull lifetime model as its shape parameter is known.