基于电子鼻的白酒品质检测-李静

FOODANDFERMENTATIONINDUSTRIES1602015Vol.41No.4Total328DOI10．13995/j．cnki．11－1802/ts．201504030*121212123121412222141223312000zNoseTM。“”。6100%。、。[e-mail]。*21206051BY20130155-222014－07－182015－01－12、、、、1。、、2。、、1%～2%3。。ZhangQiny4、5、6GC。GCGC。7。Buratti8－9、Garca10。Lozanoc1197%。12。ZHANG4、1314－17TGS、18GC-Flash。6zNoseTM“”。zNoseTM。11.11.1.1、、、、、117102。2～61。17。1.1.220154143281611Table1CharacteristicsoftheFenjiusamples//%10.25+“”353250142310242414645166260.55507//zNoseTM4200ESTMicroSense4.011DB-51minzNoseTM5～20C5～C207。1.21.2.1zNoseTM420060℃10℃40℃180℃150℃。0.03m/s。21200Counts。40mL98mm、28mm10mL。1h。3。1.2.2954。2021。1。2。2Table2qualitystandardsoftheFenjiusamples、4、3、2、11543214321343213321344213543213432123212432121.3MATLABMATLAB7.0TheMathworksInc.MAUSAPCA。SAS/STATSAS9.1SASInstituteIncCaryNCUSA。22.1zNoseTM4200。1617zNoseTM42003MicroSense4.01。C5～C14zNoseTM。retentiontimeRTpeakareaPA。19。1～4。5C10。11“”2～65。1“”2FOODANDFERMENTATIONINDUSTRIES1622015Vol.41No.4Total3281～91～91Fig．1StandardfigerprintoftheFenjiusamples5“”C10。2416～9C11～C14。2.2sumofPAzNoseTM4200。2。“”2295.8“”1416137Counts、。170.3102、314、5、6“”692043Counts4、5、6。2、。2“”2、3、1、4、6、5。“”“”。zNoseTM4200“”1min。2Fig．2SumofthepeakareasandscoreofthesensoryevaluationP＜0.052.3principalcomponentanalysisPCA22。。1710299。85.3%7.7%93%PC1、PC2。3PC1、PC261717。3PCAFig．3PCAscoresplotoftheFenjiusamples36。2、3。4、5、6PC1PC2。12015414328163。4PC1、PC22、4、5。12、4、52、46～9。2、4。。4Fig．4LoadingsplotoftheFenjiusamples2.4canonicaldiscriminantanalysisCDAFisher。5。102178414183。CDA5143。56。100%。53CDAPCA。CDAPCA。5Fig．5CDAanalysisoftheFenjiusamples3PCA、CDA6。zNoseTM42005。“”。PCA、93%2、4。CDA100%。、、。。1.、J．2007343106－110．2.J．200021614－17．3.FOODANDFERMENTATIONINDUSTRIES1642015Vol.41No.4Total328J．20014106－110．4ZHANGQXIECZANGSetal．Identificationandpat-ternrecognitionanalysisofChineseliquorsdopednanoZnOgassensorarrayJ．SensorsandActuators.BChemical20051102370－376．5.J．20123310184－189．6.GS-MSJ．2002295225－26．7.J．2012388114－118．8BurattiSBenedettiSScampicchioM.CharacterizationandclassificationofItalianBarberwinesbyusinganelectronicnoseandanamperometricelectronictongueJ．AnalyticaChimicaActa20045251133－139．9BurattiSBallabioDBenedettiSetal．PredictionofItalianredwinesensorialdescriptorsfromelectronicnoseelec-tronictongueandspectraphotometricmeasurementsbymeansofGeneticAlgorithmregressionmodelsJ．FoodChemistry2007100211－218．10GarcaMAleixandreMGutirrezJetal．ElectronicnoseforwinediscriminationJ．SensorsandActuatorsBChemical20061132911－916．11LozanocJArroyobTSantosaJPetal．ElectronicnoseforwineageingdetectionJ．SensorsandActuatorsBChemical20081331180－186．12.J．2011373163－167．13.J．201425－9．14.PCAWilksJ．2007285849－85215.J．2005266565－65616.J．2003194151－154．17.J．200425229－32．18.J．2008538－44．19.J．2013393100－104．20GB/T10781.2－2006．S．21.J．20091194－96．22.SASM．2013QualityinspectionofChineseliquorbasedonelectronicnoseLIJingSONGFei-huPUHong-jieLIZhen-feng1SchoolofMechanicalEngineeringJiangnanUniversityWuxi214122China2JiangsuKeyLaboratoryofAdvancedFoodManufacturingEquipmentandTechnologyJiangnanUniversityWuxi214122China3ShaoxingQuelandMechatronicsTechnologyCo.Ltd.Shaoxing312000ChinaABSTRACTInordertoimprovequalityinspectionbasedonconventionalsensoryevaluationtheliquorsampleswithdifferentqualitylevelsinthesamecompanywereanalyzedwithzNoseelectronicnoseandthesensoryevalua-tion．Thestructuresofthecharacteristicpeaksintheelectronicnosefingerprintswereanalyzed．‘SumofPA’param-etersdescribedthequalitylevelaccuratelyandwerearrangedinorderaccordingtothesensoryevaluation．Theresultsshowedthat‘SumofPA’parametercouldaccuratelydescribetheliquorqualitylevel．PrincipalComponentAnalysisPCAandCanonicalDiscriminantAnalysisCDAwereadoptedfordataanalyses．CDAhadagreaterdistinctioncomparedtoPDAanditsidentificationratereached100%．Themodelofqualityinspectionofliquorbasedonelec-tronicnosecanquicklyandaccuratelyexpressthequalitylevelandcharacteristicstyleofliquor．Itisfeasibleforquality-onlineinspectiontoimproveliquorproduction．KeywordsliquorelectronicnoseCDAqualityinspection