Kragten法评估测量不确定度探讨

Study of measurement uncertainty through Kragten method

利用Kragten数值方法对医学检验中血清淀粉酶(amylase,AMY)活性浓度测量模型的合成标准不确定度(u_(c))进行计算,得到的结果与基于不确定度传播率的方法(GUM)基本一致,方法间计算的偏差均来自与该测量模型输出量呈非线性的输入量的贡献.推导出了非线性输入量引入的不确定度分量的解析式,表明Kragten法和含一阶项的GUM法计算结果的相对偏差仅与各输入量的相对不确定度呈函数关系.考查了逐渐增大的样品体积相对不确定度值(u(V_(S))/V_(S))对Kragten法计算的u_(c)影响趋势,并与GUM法进行了比较.当相对不确定度的值增大而使测量结果非线性地取决于1个或多个输入量时,Kragten法仅给出了估计的标准不确定度,此时GUM法中的高阶项部分不应该被忽略.本研究对将Kragten法用于此类非线性模型的不确定度的评估,具有一定的参考和指导意义.

In the process of measurement uncertainty evaluation,the calculation of combined standard uncertainty is of vital importance.Kragten method is a numerical calculation method to obtain the combined standard uncertainty from the input standard uncertainty according to the known measurement model.When the simple form of GUM method can not be directly used or it is not easy to decompose the original model and then use the simple form for calculation,Kragten method provides a simple and practical mean for the calculation of the combined standard uncertainty.In this work,with serum amylase(AMY)activity measurement in medical laboratory as model,the combined standard uncertainty(u_(c))was evaluated by using Kragten method.The obtained results were close to those by GUM method.The deviations between the two methods were from the contribution of inputs which were nonlinear to the output of the measurement model.The results show that the relative deviation of Kragten method and GUM method with first order term is only a function of the relative uncertainty of each input.The influence of increasing relative uncertainty of sample volume(u(V_(S))/V_(S))on u_(c)calculated by Kragten method was investigated and compared with that by GUM method.When the relative uncertainty increases and the measurement result depends on one or more inputs nonlinearly,the Kragten method only provides the estimated standard uncertainty,and the high-order part of GUM method should not be ignored.Using Kragten method to calculate the combined uncertainty,the approximate numerical method of differential method is used to avoid the process of deriving the partial derivative formula and calculating the sensitivity coefficient by GUM method.The Kragten method can be applied to even complex measurement expressions.Monte Carlo method(MCM)can provide the estimated output distribution,so it can better reflect the real“confidence interval”.When GUM method,Kragten method and MCM are used together,it is helpful to formulate appropriate solutions,because they can focus on different aspects of the problem.This study has a reference and guiding significance for using Kragten method to evaluate the uncertainty of this kind of nonlinear model.