论文部分内容阅读
本文证明了平方相加法则对各种概率分布的误差项极限值的综合是普遍成立的,解决了误差理论中长期争论的问题。本文的分析基于下列三个前提:1.承认可能出现超差,但必须对测量结果的质量影响不大。2.引入超差二次矩代替超差概率(或置信率)作为判断超差严重程度的主要数学指标。3.认为严重偏离正态分布的数值较大的误差项的出现概率很小,并在估计综合方法所得结果的可靠程度时考虑这种概率的影响。
This paper proves that the sum of squared addition rules holds the general value of the limits of the error term of various probability distributions and solves the problem of long-term dispute in error theory. The analysis of this article is based on the following three premises: 1. Recognition of possible over-tolerance, but must have little effect on the quality of the measurement results. 2. The introduction of overcompensated second moment instead of over-poor probability (or confidence rate) as the main mathematical indicators to determine the severity of over-tolerance. 3. The probability of occurrence of errors with larger values deviating from the normal distribution is considered to be small, and the effect of such probabilities is taken into account in estimating the reliability of the results obtained from the synthesis method.