Bilateral Given Censored Reliability Of Statistical Analysis And Application

In this paper, based on type-II doubly censored samples, we study Bayesian estimation and prediction of parameters and reliability indexes from a special sort of reliability life model-exponentiated-weibull distribution.Then we discuss the application of Bayesian analytical approach in mixed system reliability estimation.The empirical Bayes estimations of performances in mixed system are derived also.This paper is made of four parts:Firstly, under type-II doubly censored, we obtain the Bayesian and empirical Bayesian estimations of one parameter and reliability indexes for EW distribution, based on quadratic and LINEX loss function. Thus we prove that this method is correctable by means of Monte-Carlo simulation.Secondly, when the two parameters of EW distribution are unknown, approximate Bayesian estimations of the two parameters, reliability function are given by using the methods of Tierney&Kadane based on type-II doubly censored samples. A numerical simulation results are derived.Thirdly, we discuss the Bayesian prediction of EW distribution. In this case, using Bayesian method prediction intervals are obtained for future observes values in one sample and two-sample problem. And the accuracy of these results is illustrated.Finally, for mixed system, such as series-parallel system and parallel-series system, under type- II doubly censored samples, the empirical Bayesian estimators of reliability performances from the exponentiated distribution life model are considered.