Bootstrapping Computation of Availability for a Repairable System with Standby Subject to Imperfect Switching

摘要

This article deals with the availability behavior of a repairable system in which standby switched to primary is subject to breakdowns. The time-to-failure of the four primary and two standby units are assumed to be exponentially and generally distributed. In addtion, the repair time of service station follow four common distributions: exponential (EXP), Gamma (G), Uniform (U), and Mixture (M). We use a recursive method, and the supplementary variable technique to develop the steady-state availability, A_v. The estimator A_v is strongly consistent and asymptotically normal. The interval estimations of A_v are constructed by five bootstrap approaches: standard bootstrap confidence interval (SB), the percentile bootstrap confidence interval (PB), the bias-corrected percentile bootstrap confidence interval (BCPB), the bias-corrected and accelerated confidence interval (BCa), and bootstrap pivot confidence interval (BP). Finally, some simulation computations are conducted in order to describe the performances of A_v on various interval estimation by calculating the coverage percentage and the average length of intervals.