具有共同分量的相关数据的加权平均

原遵东,王池,林延东,李燕华

计量学报 ›› 2012, Vol. 33 ›› Issue (6) : 577-582.

PDF(390 KB)
PDF(390 KB)
计量学报 ›› 2012, Vol. 33 ›› Issue (6) : 577-582. DOI: 10.3969/j.issn.1000-1158.2012.06.22

具有共同分量的相关数据的加权平均

  • 原遵东,王池,林延东,李燕华
作者信息 +

Weighted Mean for Correlation Data Influenced by Common Components

  • YUAN Zun-dong,WANG Chi,LIN Yan-dong,LI Yan-hua
Author information +
文章历史 +

摘要

对比不相关数据的加权平均法,讨论了对具有共同分量的一组相关数据的加权平均法。分析了比对数据的相关性,将相关数据的不确定度分解为共同分量与不相关分量的组合。提出了具有较好可操作性的估计比对数据溯源链共同分量不确定度的简单方法。结论为加权平均应以不相关分量的不确定度为权重因素;共同分量的不确定度是“固有”的,不随参与平均计算的数据数量的增加而减小;忽略数据相关性,不仅影响加权平均的权重,进而影响平均值,也会导致平均值不确定度计算结果偏小。以压力比对为例分析了共同分量的不确定度的评定以及不同取值结果的差异。

Abstract

By contrasted with the weighted mean method (WMM) for uncorrelated data, the WMM for a set of correlated data with common components is studied. A simple method with manipuility is presented for estimating the uncertainty of common component of traceability chain. The weights for calculating mean, which are derived, are the uncertainties of only uncorrelated components, not the ones of entire data. The uncertainty of common component for the mean is inherent, which is not decreased with increasing the number of averaged data. It not only influences the weights and thus the mean, but also leads an underestimation for the uncertainty of mean to neglect the influence of the correlation on the weights. Take a pressure comparison for instance, the estimations for the uncertainty of common component are illustrated and their results calculated with different estimated values are given.

关键词

计量学 / 不确定度 / 加权平均 / 相关性 / 共同分量 / 比对 / 参考值

Key words

Metrology / Uncertainty / Weighted mean / Correlation / Common component / Comparison / Reference value

引用本文

导出引用
原遵东,王池,林延东,李燕华. 具有共同分量的相关数据的加权平均[J]. 计量学报. 2012, 33(6): 577-582 https://doi.org/10.3969/j.issn.1000-1158.2012.06.22
YUAN Zun-dong,WANG Chi,LIN Yan-dong,LI Yan-hua. Weighted Mean for Correlation Data Influenced by Common Components[J]. Acta Metrologica Sinica. 2012, 33(6): 577-582 https://doi.org/10.3969/j.issn.1000-1158.2012.06.22
中图分类号: TB9   

参考文献

[1]CIPM.Guidelines for CIPM key comparisons [R]. http://www.bipm.org/utils/en/pdf/guidelines.pdf,2009-06-05.
[2]Cox M G. A discussion of approaches for determining a reference value in the analysis of key-comparison data.In:Ciarlini P.   Advanced Mathematical and Computing Tools in Metrology,Vol IV[M].Singapore: World Scientific,2000.
[3]Cox M G. The evaluation of key comparison data [J].Metrologia,2002,39(6):589~595.
[4]Elster C,Link A. Analysis of key comparison data: assessment of current methods for determining a reference value[J].Meas Sci Technol,2001,12(9):1431~1438.
[5]Steele A G, Wood B M,Douglas R J.  Exclusive statistics: simple treatment of the unavoidable correlations from key comparison reference values[J].Metrologia,2001,38(6):483~488.
[6]Cox M G,Ei C,Mana G,Pennecchi F,et al.  The generalized weighted mean of correlated quantities [J].Metrologia,2006,43(4): S268–275.
[7]Collett M A,Cox M G,Esward T J,et al.Aggregating measurement data influenced by common effects[J].Metrologia, 2007,44(5):308-318.
[8]原遵东,王池,林延东.确定量值比对参考值的准则[J].仪器仪表学报, 2010, 31(10):2381-2385.
[9]原遵东,王池,林延东.比对参考值与等效度计算方法探讨[J].计量学报, 2011, 32(1):93-96.

PDF(390 KB)

Accesses

Citation

Detail

段落导航
相关文章

/