基于感度分布均值的火工品发火可靠性小样本评估方法

利用参数化方法估计火工品的发火可靠性时，标准差估计偏差常使评估结果冒进；通过非参数化方法估计火工品的发火可靠性时，试验样本量通常比较大。为解决以上问题，以应力-强度干涉模型为理论基础，将参数化评估方法和非参数化评估方法相结合，提出一种基于感度分布均值的火工品发火可靠性小样本评估方法。该方法只需利用感度分布均值就能在置信水平不变条件下，将工作刺激量处的验证试验转化为低刺激量处的验证试验，从而大大降低试验样本量。应用大样本的步进法和所提新方法对某型针刺雷管和某型电点火头进行对比验证试验。结果表明：两种方法的可靠性评估结果一致，所提小样本评估方法是合理的，可应用于机械火工品、电火工品的发火可靠性评估。

A Small Sample Assessment Method for Firing Reliability of Explosive Initiator Based on the Mean of Sensitivity Distribution

When the parametric methods are used to estimate the firing reliability of explosive initiator， the bias in estimation of the standard deviation often makes the evaluated results excessive. When the nonparametric methods are used to estimate the firing reliability of explosive initiator，the test sample size is usually large. A small sample reliability assessment method for firing reliability of explosive initiator based on the mean of sensitivity distribution is proposed by using the stress-strain interference model and combining the parametric methods with the nonparametric methods. The proposed method only uses the mean of sensitivity distribution to convert the verification test at the working stimulus to the verification test at the low stimulus at a constant confidence level，thereby greatly reducing the test sample size. The large sample run-down method and the proposed small sample method are used to conduct the comparative tests on a certain type of stab detonator and a certain type of fuse head. The result shows that the reliability evaluation results of the two methods are consistent. The proposed assessment method is reasonable and can be applied to the reliability assessment of mechanical and electric explosive initiators.