用联合样本法确定直升机零件的疲劳分散性

在直升机型号的疲劳合格鉴定中,由于经费和时间的限制,同一种零部件的试件数通常不超过6个。由这些试验结果构成的样本在统计学上叫小样本。一般认为,由这种小样本确定的平均值还比较准确,但确定的标准差可能偏离真值甚远。本文介绍一种把小样本试验数据联合起来构成大样本,比较精确地确定标准差的方法。

ESTIMATION OF THE FATIGUE STRENGTH SCATTER FOR HELICOPTER COMPONENTS BY A POOLING SAMPLE METHOD

Due to economic and time constraints the number of the same components tested does no usually exceed six in fatigue qualification of helicopter componemts. The sample consisted of such test data is called the "small sample" in statistics. It is generally recognised that the mean estimated on the basis of such a small sample is more exact, but the standard deviation estimated on that basis may be quite far from the true value. The paper presents an approach wich consistutes large samples by pooling small samples, and estimates log standard deviations (or coefficients of variation) of fatigue strength for helicopter components more exactly.