Sequential Testing of Hypotheses Concerning the Reliability of a System Modeled by a Two-Parameter Weibull Distribution

摘要

Monte Carlo Analysis techniques are used for the development of test plans concerning hypothesized system reliabilities. Systems under consideration are those in which component failure rates are best described by the two-parameter Weibull probability density function. The statistical test employed is Wald's sequential probability radio test using independent, asymptotically computed variances (Cramer-Rao Lower Bound Technique). The null hypothesis, HO: HO: RO =.90, is tested against alternative reliabilities of .854, .810, and .729. Three pairs of alpha and beta risk levels are considered for each test ((.1,.1), (.2,2), (.1,.2). A truncation decision for the sequential test is made at 1.5 times the fixed sample test size for the same conditions. One thousand Monte Carlo repetitions are used for these test procedures. (Author)