OPTIMAL DESIGN OF FAILURE-CENSORED CONSTANT-STRESS LIFE TEST PLANS FOR THE INVERSE WEIBULL DISTRIBUTION

This paper presents the optimal design of constant-stress partially accelerated life tests (CSPALT) using type-II censored data from the inverse Weibull distribution. The maximum likelihood approach is applied to estimate the distribution parameters and the corresponding factor. In addition, the corresponding confidence interval estimates are obtained. Moreover, optimal CSPALT plans are developed using the D-optimality criterion. That is, the proportion of test units that should be allocated to run under accelerated condition is optimally determined. This proportion is obtained such that the generalized asymptotic variance of the maximum likelihood estimators of the model parameters is minimized. To illustrate the theoretical results presented in this paper, simulation studies are conducted.