欢迎访问《茶叶科学》,今天是

茶叶科学 ›› 2009, Vol. 29 ›› Issue (1): 34-40.doi: 10.13305/j.cnki.jts.2009.1.006

• • 上一篇    下一篇

基于杠杆率校正的PLS-DA法对正半岩武夷岩茶的识别研究

周健1, 成浩1,*, 王丽鸳1, 薛凤仁2, 何孝延2, 徐锦斌3, 贺巍4   

  1. 1. 中国农业科学院茶叶研究所国家茶树改良中心,浙江 杭州 310008;
    2. 福建省农业厅种植业管理局,福建 福州 350003;
    3. 福建省南平市农业局,福建 南平 353000;
    4. 南京农业大学,江苏 南京 210095
  • 收稿日期:2008-09-05 修回日期:2008-11-03 出版日期:2009-02-15 发布日期:2019-09-06
  • 通讯作者: *
  • 作者简介:周健(1979— ),男,浙江平湖人,助理研究员,硕士,从事茶树生物技术方面的研究工作。
  • 基金资助:
    国家科技基础条件平台项目(2005DKA1002)和浙江省自然科学基金(Y305096)资助

Recognition of Wuyi Zheng Yan Tea with PLS-DA Based on Leverage Correction Using near Infrared Spectra

ZHOU Jian1, CHENG Hao1,*, WANG Li-yuan1, XUE Feng-ren2, HE Xiao-yan2, XU Jin-bin3, HE Wei4   

  1. 1. National Center for Tea Improvement, Tea Research Institute, Chinese Academy of Agriculture Science, Hangzhou 310008 China;
    2. department for management of growing industry of Fujian Agriculture Bureau, Fuzhou 350003, China;
    3. Nanping Municipal Bureau of Agriculture, Nanping 353000, China;
    4. Nanjing Agricultural University, Nanjing 210095, China
  • Received:2008-09-05 Revised:2008-11-03 Online:2009-02-15 Published:2019-09-06

摘要: 武夷岩茶属于闽北乌龙茶,是福建省的特有茶类,也是我国传统名茶中的精品,尤其以武夷中心地带所产的正岩茶品质最佳。本文采用近红外光谱技术,利用杠杆率校正(leverage correction)结合偏最小二乘法(Partial least Squares Discriminant Analysis, PLS-DA)的分析方法,进行了武夷岩茶中的正岩茶与半岩茶的分类识别研究。结果表明不仅定标集42个样本的识别正确率为100%,对于未参与模型建立的验证集20个样本的识别正确率也达到了100%。初步证明了利用近红外技术结合PLS-DA分析方法对正岩茶与半岩茶进行分类识别的可行性。

关键词: 乌龙茶, 武夷岩茶, 近红外光谱, 识别

Abstract: Wuyi Yan tea was one of Olong tea which was produced only in north of Fujian province. It was a famous tea in China and the quality of Zheng Yan tea which was manufactured of tea leaves planted in the centre area of Wuyi mountain was best in Wuyi Yan tea. To recognize Zheng Yan tea , Partial least Squares Discriminant Analysis(PLS-DA)combined with leverage correction was used to discriminate Zheng Yan tea and Ban yan tea based on near-infrared spectra in this paper. The result showed that Zheng Yan tea could be recognized by PLS-DA method. 100% was correctly recognized in training set(42 samples), test set(20 samples) separately. The result proved the feasibility of discriminant recognition of Zheng Yan tea and Ban Yan tea.

Key words: olong tea, Wuyi Yan tea, near-infrared spectra, recognition

中图分类号: