弹道一致性评定中的样本顺次正态性检验

Normality Test Method of Sequential Samples in Evaluation of Trajectory Consistency

为了解决弹道一致性评定中小子样情况下的样本正态性检验问题,提高评定结果的可靠性,推荐采用Shapiro-Wilk或Epps-Pulley正态性检验方法。采用蒙特卡洛随机抽样的方法,补充完善了Epps-Pulley检验中样本容量3≤n≤8时不同置信水平α下的零假设接受临界值;当α=0.05时,对7种典型分布计算了不同样本容量下两种检验方法的检验功效,结果表明检验效能相当。基于上述两种正态性检验方法,提出了一种小样本弹道一致性评定的样本顺次正态性检验方法,并给出了单组与多组情况下的检验流程,研究结果具有一定的工程指导价值。

In order to perform the normality test with small samples in evaluation of trajectory consistency and improve the reliability of evaluation results, the Shapiro-Wilk or Epps-Pulley test method can be used. For the Epps-Pulley test method, the null hypothesis acceptance thresholds under different confidence level a when the sample size is 3≤n≤8 were complemen- ted by Monte Carlo random sampling method. Aimed at seven typical distributions when a= 0.05, the test efficacy of two test methods were calculated and the results show that test effica- cy is basically same. On the basis of two methods mentioned above, a normality test method of sequential samples was proposed to evaluate the trajectory consistency in condition of small samples. The detail test procedure for single group and multiple groups were given. The results have a certain value for engineering application.