在简单恒加和简单步加试验中关于等间距的若干研究

Researches on the Equal Intervals of the Simple Constant-stress Accelerated Life Test and the Simple Step-stress Accelerated Life Test

在一般的统计问题中,通常假设随机样本的观测值为一个确定的数值,然而在一些实际问题中由于条件所限,不能长期连续进行观测,得到的样本往往是观测值落入某些区间内的个数,即分组数据或区间型数据.针对分组数据在简单恒加试验和简单步加试验的场合,给出了等间距的估计,并证明简单恒加试验中的等间距精度优于简单步加试验.

In general statistics, we always assume that the observation value of the random sample is a set numerical value. However, in some practical problems, there exists some limitation so that we can not observe the sample for a long time. Then, the number of observation value we get falls in some intervals. We call such data grouped data, or inverval data. Based on the grouped data, this essay gets the estimation of the equal intervals of the simple constant - stress accelerated life test and the simple step - stress accelerated life test. The result proves that the equal interval in the simple constant - stress accelerated life test is more efficacious than that in the simple step -stress accelerated life test.