免疫算法及其在多組分體系分析中的應用研究.pdf

1、南開大學碩士學位論文免疫算法及其在多組分體系分析中的應用研究姓名：繆玲妮申請學位級別：碩士專業(yè)：化學；分析化學指導教師：邵學廣201105AbstractWiththedevelopmentofanalyticalinstrumentsandcomplexityofanalyticalsystems，itisagreatchallengetoextracttheinformationofspecificcomponentsfromamu

2、iticomponentoverlappingsignaleffectivelyTheapplicationofchemometricmethodsCallbeofgreathelp，allowingtheextractionofhiddeninformationfromthemulticomponentoverlappingsignalforqualitativeandquantitativeanalysisInthisdissert

3、ationthemethodsbasedonimmunealgorithm(n)forresolutionofoverlappingsignalswereproposed，andappliedinresolutionofrealsignals，includingnuclearmagneticresonance(NMR)spectraandgaschromatographymassspectrometry(GCMS)Thecontents

4、ofthisdissertationinclude：一(1)Nonnegativeindependentcomponentanalysis(nonnegativeICA)withknowledgebasedcriteria(KNNICA)andIAwerecombinedforresolutionofmulticomponentoverlappingNMRsignalNonnegativeICAWasemployedtoextractt

5、heindependentcomponents(ICs)frommultieomponentoverlappingsignalDuetothedriftsofthechemicalshiftandnoisesinNMRknowledgebasedcriteriawerealsoemployedformodifyingtheICsobtainedbynonnegativeICAByvirtueofIAtherelativecontents

6、ofthecomponentsinthemixedsamplesofaminoacidswerecalculatedwiththemodifiedICsResultsindicatethattheproposedmethodmaybeapracticalalternativeforidentificationandquantitativedeterminationofthecomponentsinmixturesamplesbyNMRs

7、pectroscopy(2)AnapproachbasedonIAandTFAwagproposedforresolutionofmulticomponentoverlappingGC—MSsignalInthemethodamassspectrallibraryof300pesticideswasorganizedatfirst，thentargetfactoranalysis(TFA)wasemployedfortestingthe

8、existenceofaspecificpesticideinthemulticomponentoverlappingGC—MSsignal，andfinallythechromatographicinformationofthepesticidewasextractedbyanonnegativeimmunealgorithm(NmA)AGCMSsignalofafortycomponentpesticidemixtureeluted