The Synthetically Analytical Method For Data Sets On Pyrotechnics Reliability Test
火工品可靠性试验数据的综合分析方法

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.