Optimal reliability acceptance sampling plan based on accelerated life test data

Abstract

In constant stress accelerated life testing, a unit is tested at a fixed stress level until failure or the termination time point of test. Assuming a log linear life-stress relationship, two different acceptance sampling plans namely: 1) acceptance sampling based on asymptotic distribution of the estimator process capability index (ĈL); 2) acceptance sampling based on exact distribution of ĈL, are developed for exponential distribution based on data obtained from the progressive type-II censoring. Asymptotic and exact distributions of the estimators of the parameters of exponential distribution are obtained. The parameters of the sampling scheme are obtained by minimising expected total testing cost. Some numerical results and comparisons of obtained sampling schemes are presented in tables to illustrate our plans.