摘 要:在逐次截尾樣本下,研究電子元件混聯(lián)系統(tǒng)可靠性指標的估計問題。將Bayes方法和極大似然法相結(jié)合,在平方損失下,獲得部件失效率、系統(tǒng)可靠度和平均壽命的經(jīng)驗Bayes 估計。最后給出隨機模擬例子,說明該方法的正確性。結(jié)果表明可靠性指標的經(jīng)驗Bayes估計值精度較高。
關(guān)鍵詞:混聯(lián)系統(tǒng);可靠性指標;逐次截尾樣本;經(jīng)驗Bayes估計
EB Estimation of the Reliability Performance for Compound System with Electronic
[J]Components Based on Progressively Censored Sample
SHI Xiaolin
(Xi′an Institute of Post and Telecommunications,Xi′an,710121,China
Abstract:Based on progressively censored sample,the problems of estimating the reliability performance for compound system with electronic components are studied by using Bayes and MLE approaches.Under the squared loss,the formulae to calculate Empirical Bayes (EB estimation of the failure rate,the reliability function and mean lifetime for the system are given.Finally,an illustrative numerical example is examined by means of the Monte-Carlo simulation.It is shown that the method proposed in this paper is correct.The results prove that empirical Bayes estimation of reliability performance has high precision.
eywords:compound system;reliability performance;progressively censored sample;empirical Bayes estimation
1 引 言
混聯(lián)系統(tǒng)的可靠性分析是系統(tǒng)可靠性研究的重要內(nèi)容之一?;炻?lián)系統(tǒng)是指由串聯(lián)系統(tǒng)和并聯(lián)系統(tǒng)混合而成的系統(tǒng),最常見的混聯(lián)系統(tǒng)有串-并聯(lián)系統(tǒng)和并-串聯(lián)系統(tǒng)2種。關(guān)于混聯(lián)系統(tǒng)的可靠性分析文獻上已有討論。文獻[1]針對含增長單元的串并聯(lián)系統(tǒng)的特點,提出系統(tǒng)可靠性參數(shù)先驗分布的方法,并給出系統(tǒng)可靠度先驗分布的計算公式;文獻[2]利用信息論方法給出串并聯(lián),并串聯(lián)系統(tǒng)可靠性綜合評定方法;文獻[3]利用最優(yōu)化方法給出串并聯(lián)系統(tǒng)可靠度函數(shù)的最優(yōu)分配方案;文獻[4]利用信息論方法提出小子樣下系統(tǒng)可靠性置信區(qū)間估計的新方法;文獻[5]假定元件的失效率為常數(shù)給出混聯(lián)系統(tǒng)可靠度與平均壽命的計算方法。然而在工程實際中,元件失效率往往是未知的,特別是新研制產(chǎn)品。
基于這種事實,本文將失效率當作隨機變量,利用逐次截尾數(shù)據(jù),討論混聯(lián)系統(tǒng)可靠度及平均壽命的經(jīng)驗Bayes估計,并給出隨機模擬例子。