CAJ | 학술논문

目的建立蒙特卡罗法(MCM)评定参考测量程序不确定度的方法学.方法方法学建立.根据JCGM 101:2008建立MCM评定不确定度的方法学,并以γ-谷氨酰基转移酶(GGT)参考测量程序为例,通过MATLAB软件计算GGT活性浓度、不确定度与包含区间；GUM法评定其不确定度作为对照.结果 GUM法评定时,标本A测量结果为(95.8 ±2.4)U/L(k =2),包含区间为[93.4U/L,98.2 U/L],标本B为(180.0±3.9)U/L(k=2),包含区间为[176.1 U/L,183.9 U/L]；MCM评定时,标本A测量结果为(95.8±2.4) U/L,包含区间为[93.4 U/L,98.2 U/L],标本B为(180.0±3.9) U/L,包含区间为[176.2 U/L,183.8 U/L]；MCM模拟输出量为正态分布.结论当输出量是正态分布时,MCM与GUM法评定结果相近；当输出量分布未知时,可将MCM作为GUM方法的验证或确认贵法.
목적 건립몽특잡라법(MCM)평정삼고측량정서불학정도적방법학.방법 방법학건립.근거JCGM 101:2008건립MCM평정불학정도적방법학,병이γ-곡안선기전이매(GGT)삼고측량정서위례,통과MATLAB연건계산GGT활성농도、불학정도여포함구간；GUM법평정기불학정도작위대조.결과 GUM법평정시,표본A측량결과위(95.8 ±2.4)U/L(k =2),포함구간위[93.4U/L,98.2 U/L],표본B위(180.0±3.9)U/L(k=2),포함구간위[176.1 U/L,183.9 U/L]；MCM평정시,표본A측량결과위(95.8±2.4) U/L,포함구간위[93.4 U/L,98.2 U/L],표본B위(180.0±3.9) U/L,포함구간위[176.2 U/L,183.8 U/L]；MCM모의수출량위정태분포.결론 당수출량시정태분포시,MCM여GUM법평정결과상근；당수출량분포미지시,가장MCM작위GUM방법적험증혹학인귀법.
Objective To establish the methodology of uncertainty evaluation with reference measurement procedure by Monte Carlo method (MCM).Methods According to JCGM 101:2008,we established the methodology of uncertainty evaluation by MCM in the example of GGT reference measurement procedure.We could calculate an estimate of GGT concentration,the associated standard uncertainty and a coverage interval with a specified probability by MATLAB software,setting the uncertainty evaluation by GUM method as a control.Results When the uncertainty was evaluated by GUM method,the results of sample A and Sample B were (95.8 ±2.4) U/L (k =2) with coverage interval [93.4 U/L,98.2 U/L] and (180.0 ± 3.9) U/L (k =2) with coverage interval [176.1 U/L,183.9 U/L] respectively,while using MCM method,the uncertainty evaluation result of sample A and Sample B were (95.8 ± 2.4) U/L (k =2) with coverage interval [93.4 U/L,98.2 U/L] and (180.0 ± 3.9) U/L with symmetrical 95％ coverage interval [176.2 U/L,183.8 U/L].The output quantity simulated by MCM was normal distributed.Conclusions When the distribution of the output quantity is normal,the measurement uncertainty evaluated by both MCM and GUM method is nearly the same.When the distribution of the output quantity is unknown,MCM can be used as a verification of GUM method.