成分分析中测量不确定度的两种评定方法比较及Top-down法有效性验证  

Comparison of two evaluation methods for measurement uncertainty in component analysis and validation of Top-down method

作　　者：杏艳 杜祯宇 范爽 李玉武 杨震[1] 张霖琳 XING Yan;DU Zhenyu;FAN Shuang;LI Yuwu;YANG Zhen;ZHANG Linlin(Shaanxi Province Environmental Monitoring Station,Xi'an 710061,China;Environmental Development Centre of the Ministry of Ecology and Environment,Beijing 100029,China;National Research Center for Environmental Analysis and Measurements,Beijing 100029,China;Guohuan Oasis(Gu'an)Environmental Technology Co.,Ltd.,Langfang 065500,China;China National Environmental Monitoring Centre,Beijing 100012,China)

机构地区：[1]陕西省环境监测中心站,陕西西安710061 [2]生态环境部环境发展中心,北京100029 [3]国家环境分析测试中心,北京100029 [4]国环绿洲(固安)环境科技有限公司,河北廊坊065500 [5]中国环境监测总站,北京100012

出　　处：《冶金分析》2023年第4期10-17,共8页Metallurgical Analysis

基　　金：国家环境保护重金属污染监测重点实验室基金(SKLMHM 202202);生态环境部标准制修订项目(2021-21);陕西省自然科学基础研究计划(2022JM-091);湖湘高层次人才聚集工程-创新团队项目(2021RC5030)。

摘　　要：RB/T 141—2018《化学检测领域测量不确定度评定利用质量控制和方法确认数据评定不确定度》是近年来中国合格评定国家认可委员会大力推行的测量不确定度评定Top-down法标准。其方法原理是在期间精密度测量条件下,将影响成分分析测试结果的各种因素归纳为实验室质控数据的“偏倚值”和“期间精密度(标准偏差)”,合并这两个分量后,作为测试结果不确定度的估计量。方法简便,可操作性强。以全国分析检测人员能力培训委员会系列培训教材《成分分析中的数理统计及不确定度评定概要》(基于Bottom-up法原理的GUM法)和文献中的38个评定示例为基础数据,比较了Top-down和GUM两种方法结果。为充分利用文献数据,方便方法比较,Top-down法计算公式中的“偏倚分量”和“期间精密度”分别采用校准曲线拟合误差和重复测量精密度计算。结果表明,两种方法结果有可比性。用Top-down法结果除GUM法结果,两者比值稳健统计平均值为0.91;Top-down法结果略大于GUM法结果。用Top-down法评估了14个标准样品实验室测定结果的扩展不确定度,并与标准样品标准值扩展不确定度进行了比较。结果表明,以标准样品标准值扩展不确定度为基准,两者比值平均值为1.05,验证了Top-down法的适用性。最后,讨论了两种方法的特点和适用范围,期望能帮助基层检测人员熟练使用Top-down评定方法,并将其应用于实验室质控数据,与相同检测项目的GUM法结果进行比较及验证。RB/T 141-2018 Evaluation of measurement uncertainty in the chemical testing field-Appling quality control and method validation data to evaluate measurement uncertainty is the Top-down method standard for measurement uncertainty evaluation,which is vigorously promoted by China National Accreditation Service for Conformity Assessment in recent years.The principle of this method is that under the condition of intermediate precision measurement,various factors affecting the test results of component analysis are summed up as‘bias value’and‘intermediate precision(standard deviation)’of the internal quality control data of the laboratory,and these two components are combined as the estimation of the uncertainty of the test results.The method is simple and easily operated.The results of two methods(i.e.,Top-down method and GUM method)were compared based on the data of 38 evaluation examples in The Summary of Mathematical Statistics and Uncertainty Evaluation in Component Analysis(which was the training material of National Training Committee for Analyzing&Testing Personnel,NTC)and in some literatures.In order to make full use of the literature data and facilitate the comparison of methods,the‘bias value’and‘intermediate precision components’in the calculation formula of Top-down method were calculated by fitting error of calibration curve and repeated measurement precision.The results showed that the results of two methods were comparable.The results of GUM method were divided by the results of Top-down method,and the average ratio in robust statistics was 0.91.The results of Top-down method were slightly larger than those of GUM method.The extended uncertainty of 14 standard samples measured in laboratory was evaluated by the Top-down method,and then compared with the extended uncertainty of the standard value.It was found that the average value of ratio was 1.05 when the extended uncertainty of the standard value was used as basis,which verified the applicability of the Top-down method.Finally,the cha

关 键 词：Top-down法 Bottom-up法 GUM法 测量不确定度评定 成分分析 

分 类 号：O657.31[理学—分析化学]