ASTM D6122-2010 多变量联机和实验室基于红外分光光度计的分析仪系统的性能确认的标准实施规程

作者:标准资料网 时间:2024-05-21 09:15:52  浏览:8792   来源:标准资料网

【英文标准名称】:StandardPracticeforValidationofthePerformanceofMultivariateOnline,At-Line,andLaboratoryInfraredSpectrophotometerBasedAnalyzerSystems
【原文标准名称】:多变量联机和实验室基于红外分光光度计的分析仪系统的性能确认的标准实施规程
【标准号】:ASTMD6122-2010
【标准状态】:现行
【国别】:美国
【发布日期】:2010
【实施或试行日期】:
【发布单位】:美国材料与试验协会(US-ASTM)
【起草单位】:D02.25
【标准类型】:(Practice)
【标准水平】:()
【中文主题词】:
【英文主题词】:controlchart;infraredanalyzer;infraredspectrophotometers;IRspectroscopy;multivariateprocess;NIRspectroscopy;statisticalqualityassurance;validation;Controlcharttechnique;Infraredspectrophotometry;IRspectroscopy;Multivariateanalyses;
【摘要】:Theprimarypurposeofthispracticeistopermittheusertovalidatenumericalvaluesproducedbyamultivariate,infraredornear-infraredlaboratoryorprocess(onlineorat-line)analyzercalibratedtomeasureaspecificchemicalconcentration,chemicalproperty,orphysicalproperty.Thevalidatedanalyzerresultsareexpectedtobeequivalent,overdiversesampleswhosespectraareneitheroutliersornearestneighborinliers,tothoseproducedbytheprimarytestmethodtowithincontrollimitsestablishedbycontrolchartsfortheprespecifiedstatisticalconfidencelevel.Proceduresaredescribedforverifyingthattheinstrument,themodel,andtheanalyzersystemarestableandproperlyoperating.Amultivariateanalyzersysteminherentlyutilizesamultivariatecalibrationmodel.Inpracticethemodelbothimplicitlyandexplicitlyspanssomesubsetofthepopulationofallpossiblesamplesthatcouldbeinthecompletemultivariatesamplespace.Themodelisapplicableonlytosamplesthatfallwithinthesubsetpopulationusedinthemodelconstruction.Asamplemeasurementcannotbevalidatedunlessapplicabilityisestablished.Applicabilitycannotbeassumed.Outlierdetectionmethodsareusedtodemonstrateapplicabilityofthecalibrationmodelfortheanalysisoftheprocesssamplespectrum.Theoutlierdetectionlimitsarebasedonhistoricalaswellastheoreticalcriteria.Theoutlierdetectionmethodsareusedtoestablishwhethertheresultsobtainedbyananalyzerarepotentiallyvalid.Thevalidationproceduresarebasedonmathematicaltestcriteriathatindicatewhethertheprocesssamplespectrumiswithintherangespannedbytheanalyzersystemcalibrationmodel.Ifthesamplespectrumisanoutlier,theanalyzerresultisinvalid.Ifthesamplespectrumisnotanoutlier,thentheanalyzerresultisvalidprovidingthatallotherrequirementsforvalidityaremet.Additional,optionaltestsmaybeperformedtodetermineiftheprocesssamplespectrumfallsinasparselypopulatedregionofthemultivariatespacecoveredbythecalibrationset,toofarfromneighboringcalibrationspectratoensuregoodinterpolation.Forexample,suchnearestneighbortestsarerecommendedifthecalibrationsamplespectraarehighlyclustered.Thispracticedoesnotdefinemathematicalcriteriatodeterminefromaspectroscopicmeasurementofasamplewhetherthesample,themodel,ortheinstrumentisthecauseofanoutliermeasurement.Thustheoperatorwhoismeasuringsamplesonaroutinebasiswillfindcriteriaintheoutlierdetectionmethodtodeterminewhetherasamplemeasurementlieswithintheexpectedcalibrationspace,butwillnothavespecificinformationastothecauseoftheoutlierwithoutadditionaltesting.1.1Thispracticecoversrequirementsforthevalidationofmeasurementsmadebylaboratoryorprocess(onlineorat-line)near-ormid-infraredanalyzers,orboth,usedinthecalculationofphysical,chemical,orqualityparameters(thatis,properties)ofliquidpetroleumproducts.Thepropertiesarecalculatedfromspectroscopicdatausingmultivariatemodelingmethods.Therequirementsincludeverificationofadequateinstrumentperformance,verificationoftheapplicabilityofthecalibrationmodeltothespectrumofthesampleundertest,andverificationofequivalencebetweentheresultcalculatedfromtheinfraredmeasurementsandtheresultproducedbytheprimarytestmethodusedforthedevelopmentofthecalibrationmodel.Whenthereisadequatevariationinpropertylevel,thestatisticalmethodologyofPracticeD6708isusedtoprovidegeneralvalidationofthisequivalenceoverthecompleteoperatingrangeoftheanalyzer.Forcaseswherethereisinadequatepropertyvariation,methodologyfor......
【中国标准分类号】:N33
【国际标准分类号】:17_180_30
【页数】:30P.;A4
【正文语种】:英语