火工品可靠性试验数据的综合分析方法  

THE SYNTHETICALLY ANALYTICAL METHOD FOR DATA SETS ON PYROTECHNICS RELIABILITY TEST

作　　者：田玉斌[1] 李国英[2] 房永飞[1] 

机构地区：[1]北京理工大学数学系,北京100081 [2]中国科学院数学与系统科学研究院,北京100080

出　　处：《系统科学与数学》2006年第2期147-158,共12页Journal of Systems Science and Mathematical Sciences

基　　金：国家自然科学基金(10371126;10271013)资助课题

摘　　要：升降法试验数据和固定刺激量下的成败型试验数据,是两种最常见的火工品可靠性试验数据．我们应用Markov链,研究了升降法试验数据下,感度分布参数的极大似然估计的特性．在此基础上,应用Bootstrap方法和Bayes方法,给出了综合分析两种试验数据的方法．最后,将该方法应用于520底火的可靠性鉴定,得出了有益的结论．Two types of data sets from an up-and-down test and a success-and-failure test at a specified stimulus level are the most commonly used data sets in pyrotechnics reliability tests. Based on a data set from an up-and-down test, we use a Markov chain to study the characteristics of the maximum likelihood estimators for the parameters in a response distribution. This leads to the development of a method which can analyze two types of data sets synthetically by using the Bayes method and the Bootstrap method. Finally, we apply the proposed method to appraise the reliability of a type of primer numbered as 520 and obtain the useful results.

关 键 词：升降法试验 MARKOV链 BOOTSTRAP方法 极大似然估计 可靠度下限. 

分 类 号：TJ45[兵器科学与技术—火炮、自动武器与弹药工程]