近红外光谱技术快速检测烟草花中主要活性成分含量的研究

打开文本图片集

中图分类号：TS411 文献标识码：A文章编号：1007-5119（2026）01-0116-08

Abstract：Toachievetherapidandnondestructivedetectionof themainactiveingredients （reducingsugars，totalflavonoids， chlorogenicacid，rutinndembatriendiolintbaccoflowers，analytcalmodelfortesefivomponentsastablsdvia near-infrared（NR）spectroscopyAtotalof196tobaccoflowersampleswithdiferentoriginsandvarietiesacross Chinawereused asresearch materials.Model performance was further improved byscreening spectral preprocessing methods，optimizing spectral ranges，andadjustingthenumberofprincipalcomponents.Teoptimalmodelforeachconstituentwasobtaiedusing intealcrossvalidation.Resultssowedatfortecalibaionmodelsoftefveaciveingedients，eoelationcoefetsofthelibation set （Rc） ranged from 0.957 to 0.986， and the root mean square errors of calibration （RMSEC） ranged from 0.025% to 0.995% ； the correlation coefficients of cross-validation （Rcv） ranged from O.901 to 0.979，and the root mean square errors of cross-validation （RMSECV） ranged from 0.030% to 1.350% .External validation results demonstrated that the relative standard deviations （RSDs） of all five models were less than 10% . Additionally， t⋅ -testsrevealed no significant difference between the values predicted by NIR spectroscopyandthoseobtainedvia raditionalchemicalanalysis.Thisconfimsthat theproposedNRspectroscopymethodehibits highacuracyandreliabilityandenablesrapid andnondestructivedetectionof the five mainactive ingredients intobaccoflowers， thus providing technical support for the efcient development and utilization of tobacco flower resources.

Keywords： tobaccoflower;near-infraredspectroscopy （NIRS）；reducing sugar;total flavonoids；chlorogenicacid；rutin; cembratriene-diol （CBT）

烟草NicotianatabacumL.作为全球重要的经济作物，其叶片主要用于卷烟生产。（剩余13467字）