基于Bayes方法的火箭弹存储可靠性分析

Analysis of Reliability of Rocket Storage Based on Bayes Method

对成败型的火箭弹存储可靠性进行分析,建立了一种综合部件存储数据和少量整机试验数据的存储可靠性评估方法。先将部件串联而成的模拟系统作为先验分布的参考模型,然后利用先验矩方法确定先验分布的超参数,最后结合整机试验信息,利用Bayes方法对系统存储可靠性进行评估。通过该方法在某型产品上的应用表明,可显著减少存储试验样本,且具有较高的评估精度。

This paper analyzes the reliability of pass-fail rocket storage,and founds a method of evaluating storage reliability of integrated component storage data and a small amount of entire machine test data.Firstly,the simulated system which connects with components in series is used to be the reference model of transcendental distribution.Secondly,transcendental parameters of transcendental distribution are confirmed using the transcendental moment method.Finally,the reliability of the system storage is evaluated using the Bayes method integrated with the entire machine test information.Using this method to an actual product indicates that the storage test sample is reduced observably and the evaluating precision is high.