论文部分内容阅读
本文从作者对误差分类的设想出发,来理解国际计量局不确定度工作组的INC-1(1980)建议中对不确定度的分类,并探讨其中B类不确定度的一种估算方法。文中提出未定的确定函数误差与未引起测量数据作随机变动的随机误差宜作B类不确定度处理。并设想将确定函数误差视为遍历性平稳随机过程的一个现实,以求得其准分布密度,且给出其解析求法与作图求法。从而可用其准标准差u来表征B类不确定度,必要时可求得相应的准置信因子k。
In this paper, I will introduce the assumptions of error classification to understand the classification of uncertainty in INC-1 (1980) proposal of Uncertainty Working Group of International Bureau of Weights and explore an estimation method of the type B uncertainty. In this paper, we propose that the undetermined deterministic function error and the random error that does not cause the random variation of the measured data should be treated as Class B uncertainty. It is assumed that the determination of function error is a reality of ergodic stochastic process, so as to obtain its quasi-distribution density and give its analytical method and mapping method. Which can be used to quasi-standard deviation u to characterize the type B uncertainty, if necessary, obtain the corresponding quasi-confidence factor k.