基于ML-II方法的k/n-系统Bayes可靠性评估  

Bayes reliability assessment for k-out-of-n system based on ML-II method

作　　者：程皖民[1] 冯静[1] 周经伦[1] 孙权[1] 

机构地区：[1]国防科技大学信息系统与管理学院系统工程系,长沙410073

出　　处：《电光与控制》2007年第1期22-24,共3页Electronics Optics & Control

基　　金：863高科技项目(No.2003AA845023)

摘　　要：n中取k系统(简称k/n-系统)是工程实践中应用最广泛的系统类型之一。为了在系统现场试验样本量很小的情况下进行可靠性评估,首先利用次序统计量推导了k/n-系统寿命分布的密度函数,并给出了模型参数的第二类极大似然估计(ML-Ⅱ估计);然后给出了k/n-系统Bayes可靠性评估的一般步骤;仿真实例表明了方法的可行性。： k - out - of- n system （k/n - system） is widely used in engineering. Reliability assessment of traditional statistical method can only utilize on - the - spot test information, which may result in that long test duration and lots of samples are needed to reach the given confidence level for high reliability system. To implement reliability evaluation with few samples on the spot of test, a new method based on Bayesian theory is proposed, each sort of information, including the information of each component, should be made full use of for reliability analysis. Firstly, the distribution function of system life is deduced based on order statistics theory and the second maximum likelihood estimation for the parameters in the model are proposed, Steps to assess the system reliability are given. Examples illustrate the feasibility of the method.

关 键 词：可靠性评估 k/n-系统 BAYES 次序统计量 第二类极大似然估计 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计]