导弹战斗部任务可靠性评估方法分析

为评估技术状态不断变化的导弹战斗部的可靠性，提出了适用于成败型产品的AMSAA模型评估方法和Gompertz模型评估方法。并对某型战斗部研制阶段的试验数据进行了评估，将评估结果与传统评估方法得到的结果进行了比对。结果表明，用二项分布评估方法得到的置信下限小于用超几何分布评估方法得到的置信下限，并且都低于用AMSAA模型评估方法得到的置信下限；用传统评估方法得到的点估计值小于用Gompertz模型评估方法得到的点估计值。着重分析了用不同方法对同一试验数据进行评估时得出不同结果的原因。结果表明，由于考虑了产品的可靠性变化趋势，用AMSAA模型评估方法和Gompertz模型评估方法得出的结果更合理。

Analyzing Several Evaluation Methods for Mission Reliability of Missile Warhead

In order to evaluate the reliability of the missile warhead whose configuration is subject to change, two evaluation methods based on AMSAA model and Gompertz model respectively, which are suitable for the success or failure product, are presented. These evaluation methods are applied to estimate the reliability of a warhead by adopting source data which is collected from its development phase, and the reliability values obtaining from two model evaluation methods are compared with the ones obtaining from the traditional evaluation methods. It is found that the lower limit value of confidence obtaining from the binomial distribution evaluation method is less than the one obtaining from the hyper-geometric distribution evaluation method, and these two lower limits is less than the one obtaining from the AMSAA model evaluation method, and the point evaluation value obtaining from the traditional evaluation method is less than the one obtaining from the Gompertz model evaluation method. The reason for obtaining different values from the different evaluation methods using the same test data is analyzed. The conclusion is that the values of reliability obtaining from the AMSAA model method and Gompertz model method are more credible, because the changing of reliability has been taken into account in both methods.