本文選題：系統(tǒng)可靠度　切入點：置信下限　出處：《強激光與粒子束》2017年12期 　論文類型：期刊論文

【摘要】：針對壽命服從指數(shù)分布的隨機截尾樣本,研究單元及系統(tǒng)可靠度的評估方法,給出了系統(tǒng)可靠度置信下限估計方法,該方法同樣適用于工程中常見的定時截尾樣本與定數(shù)截尾樣本的系統(tǒng)可靠度評估問題。首先通過分位數(shù)填充算法將隨機截尾數(shù)據(jù)補充成虛擬完全樣本,然后基于信仰推斷,給出系統(tǒng)內(nèi)每個單元失效率的信仰分布,最后根據(jù)指數(shù)分布的特點,從失效率角度出發(fā),采用蒙特卡羅法對單元失效率進行隨機抽樣,結(jié)合系統(tǒng)可靠性模型得到系統(tǒng)失效率的分布,在給定置信水平下即可得到系統(tǒng)失效率上限,進而得到工程上特別關(guān)注的系統(tǒng)可靠度置信下限。
[Abstract]:Based on the random censored samples of exponential distribution of service life, the method of estimating the confidence lower limit of system reliability is given, and the evaluation method of unit and system reliability is given. This method can also be used to evaluate the system reliability of timing truncated samples and deterministic truncated samples. Firstly, random truncated data is supplemented into virtual complete samples by quantile filling algorithm, and then based on belief inference. The belief distribution of each unit failure rate in the system is given. Finally, according to the characteristics of exponential distribution, the Monte Carlo method is used to random sampling the unit failure rate from the point of view of failure rate. Combined with the system reliability model, the distribution of the system failure rate can be obtained, and the upper limit of the system failure rate can be obtained under the given confidence level, and then the lower confidence limit of the system reliability which is of special concern in engineering can be obtained.
【作者單位】： 中國原子能科學(xué)研究院同位素研究所;北京機電工程總體設(shè)計部;
【分類號】：O213
，

本文編號：1641388