兵器试验中的异常值检验

Outliers Testing in Weapon Test

兵器试验中由于试验条件以及测量设备的原因,测量数据可能会存有异常值。阐述了2种适用于兵器试验的异常值检验方法—格拉布斯检验法和狄克逊检验法,分析了2种方法在不同条件下的异常值检验模型,并结合兵器试验的数据处理实例进行了分析判断。通过实例应用表明:这2种检验方法合理有效,是兵器试验中异常值检验的有效方法。二者具有不同的适用范围,格拉布斯法适用于剔除一个异常值,狄克逊法适用于剔除多个异常值。

For the condition and reason of the measuring device in weapon test, there may be many outli- ers of the measurement data. Two applicable methods for outliers testing in weapon test are discussed- Grubbs test and Dixon test. The two methods of outliers testing model under different conditions are built, and the application examples of the two methods are analyzed. It is predicted that the two methods are rea- sonable and effective for outliers testing in weapon test. Both have different application ranges. Gruhbs method is applicable to test an outlier, whereas Dixon method is applicable to test multiple outliers.