近红外光谱法对黔茶6种主要成分的快速测定  

Rapid determination of six major components in Guizhou tea by Near Infrared Spectroscopy

作　　者：石艳梅 伍庆[2] 谭高好 王娇[2] 

机构地区：[1]贵阳市第三实验中学,贵州贵阳550003 [2]贵州师范大学分析测试中心,贵州贵阳550001

出　　处：《贵州科学》2016年第4期74-76,96,共4页Guizhou Science

基　　金：喀斯特地区生物与信息技术协同创新中心(黔教合协同创新字[2014]04);国家新药创新重大专项(2015ZX0910104-007)

摘　　要：以贵州绿茶为研究对象,利用近红外光谱法对茶叶中水分、游离总氨基酸、咖啡碱、儿茶素类总量、茶多酚及水浸出物的含量进行快速同时测定。采用偏最小二乘法结合多元线性回归建立茶叶中水分、游离总氨基酸、咖啡碱、儿茶素类总量、茶多酚及水浸出物的定量校正模型,并用相关系数、交互验证均方差和预测均方差对模型进行评价,验证集中水分、游离总氨基酸、咖啡碱、儿茶素类总量、茶多酚及水浸出物6个内含物成分定量预测模型的相关系数分别为0.971 5、0.910 7、0.970 6、0.922 5、0.964 1、0.903 7;预测均方差(RMSEP)分别为0.574、0.418、0.279、0.959、0.735、0.28。结果表明,该方法简单、准确,更适合于茶叶中主要成分的快速分析。Taking green tea in Guizhou for research object. The contents of water,total free amino acids,caffeine,cate-chins, tea polyphenols and water extracts in tea were measured by near infrared spectroscopy （NIR）, and the calibrationmodel of above- mentioned components was established by partial least square （PLS） and Multi Linear Regression（MLR）. The prediction model was evaluated by correlation coefficient, Interaction validation of mean square error and pre-diction mean square error. The correlation coefficient of the model of water, total free amino acids, caffeine, catechins, teapolyphenols and water extract in tea was 0. 971 5,0. 910 7,0. 970 6,0. 922 5,0. 964 1 and 0. 903 7 respectively, and theRMSEP was 0. 574,0. 418,0. 279,0. 959,0. 735 and 0. 28 respectively. The method is simple ,accurate, it is more suitablefor rapid analysis of major constituents in Guizhou tea.

关 键 词：茶叶 近红外光谱 预测模型 定量分析 

分 类 号：TS272.7[农业科学—茶叶生产加工]