白酒红外光谱特征提取与分类应用分析-analysis on extraction and classification of infrared spectral characteristics of liquor.docxVIP

AbstractInfraredSpectroscopyisasubstancecomponentanalysismethodinthefieldofanalyticalchemistry.Becauseofconvenientsamplecollection,itiswidelyusedinChemical,foodscienceandmanyotherfieldsnow.Liquorisacomplexliquidmixture.Analysingliquorbyquantityaswellasquality,whichbasesoninfraredspectroscopy,isasignificantmethodwhenidentifying,anti-counterfeiting,trasingthequalityofLiquorproducts.Liquorinfraredspectroscopycanrapidlyidentifyliquoralcoholcontent,flavorandotherinformation,achievingthepurposeofreal-timedetectionofliquorquality.Thisthesisstudiestheinfraredspectroscopyofthedifferentbrandsandseriesliquor.Themaincontentsareasfollows:Usingmulti-scalefeatureselectionalgorithmtoselectthefeaturewavebandfromtheoriginalliquorinfraredspectroscopy.Buildthefeaturedatabaseoftheliquor.Usingpartialleastsquaresregressionmethodtotrainthealcoholregressionmodel.Thenpredictthealcoholquantitativelybytheregressionmodel.Usingsupportvectormachinemethodtotraintheclassificationmodelofthebrands,andclassifyliquorbrands.Formthestudieswecanconcludethatmulti-scalefeatureselectionalgorithmcanextractthevalidinformationaccurately.It’susefulforfurtherresearchonthewhitespirit.BasedonthemachinelearningmethodsandPLSregression,theclassificationaccuracyratereach99%,andtheregressionerrorislessthan1°.Keywords：InfraredspectroscopyMulti-scalefeaturePLSSVM目录摘要......................................................................................................IAbstract......................................................................................................II1绪论1.1研究背景与意义............................................................................................................(1)1.2国内外研究现状............................................................................................................(2)1.3论文主要研究内容.......................................................................................................(5)2红外光谱分析原理2.1红外光谱...........................................................................