摘要
在制导弹药的验收试验中,对可靠性指标有较高的要求。由于弹药的高价值以及试验的破坏性,使得只能进行小样本试验。分析一个现行的可靠性二次抽样方案,指出在该方案中使用方风险过大,进而提出了2种改进方案:利用子系统数据折合为系统试验数据,成为三次抽样;或利用前若干批数据,使用基于Bayes方法的抽样方案。使用改进的验收方案,可以在较小的样本量下,使生产方风险和使用方风险处于合理的水平。
The reliability index in acceptance test of guidance ammunition is strictly required. Because guidance ammunition is expensive and the destructibility is inevitable in the testing, the sample size is small. A current secondary sampling plan of the reliability is analyzed. In the plan, the consumer＇s risk is too big. Two improved plans were given. Converting subsystem test data into system test data, the second sampling plan is replaced with a third sampling plan; Bayes sampling plan with test data of some former lots is investigated. Using the improved plans, the producer＇s risk and the consumer＇s risk can be controlled at the reasonable level with small sample.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2006年第2期360-362,共3页
Acta Armamentarii
关键词
数理统计学
可靠性
二次抽样
小样本
BAYES方法
mathematical statistics
reliability
secondary sampling plan
small sample
Bayes method
