基于EGCG,ECG和咖啡碱含量的乌龙茶产地鉴定

doi:10.13305/j.cnki.jts.2018.03.003

摘要/Abstract

摘要： 提出一种量化判定乌龙茶产地的方法。共收集主要乌龙茶产区闽南、闽北、广东和台湾等地的代表性乌龙茶样品130个,用高效液相色谱方法检测其没食子酸、儿茶素、咖啡碱和茶氨酸等理化成分的含量。用遗传算法和连续投影算法筛选出重要的化合物,基于这些化合物指标分别用支持向量机、反向传播人工神经网络及随机森林模型对闽南、闽北、广东和台湾4个产区的乌龙茶进行分类和预测。结果表明,用遗传算法筛选的3个化合物（咖啡碱、EGCG和ECG）结合反向传播人工神经网络模型BPNN能够实现对4个产区乌龙茶的高效判别,且训练集和预测集的判别率分别为97.13%和98.34%。该研究结果能为乌龙茶产地的鉴别提供科学依据。

关键词: 乌龙茶, 产地, 鉴定, 化学成分

Abstract: A quantitative method to discriminate the geographical origins of Oolong teas was proposed to promote the fair of tea trade. A total of 130 Oolong tea samples were collected across China, and the chemical compositions including gallic acid, catechins, caffeine and theanine were quantified by high performance liquid chromatography. Genetic algorithm and successive projections algorithm were applied to identify important compounds, and then support vector machine, back propagation artificial neural networks and random forest models were used to classify and predict Oolong tea samples from Minnan, Minbei, Guangdong and Taiwan based on the selected compounds. The overall results indicated that compounds selected by genetic algorithm (caffeine, EGCG and ECG) combined with back propagation artificial neural networks could achieve a high efficiency in identifying Oolong tea samples from four origins, and the total identification rate in the training and prediction sets were 97.13% and 98.34%. The results provided scientific credibility to identify Oolong tea origins.

Key words: Oolong tea, geographical origins, identification, chemical compositions

中图分类号:

曹琼, 苏欢, 宛晓春, 宁井铭. 基于EGCG,ECG和咖啡碱含量的乌龙茶产地鉴定[J]. 茶叶科学, 2018, 38(3): 237-243. doi: 10.13305/j.cnki.jts.2018.03.003.

CAO Qiong, SU Huan, WAN Xiaochun, NING Jingming. Identification of the Geographical Origins of Oolong Tea Based on EGCG, ECG and Caffeine Contents[J]. Journal of Tea Science, 2018, 38(3): 237-243. doi: 10.13305/j.cnki.jts.2018.03.003.

参考文献 21

[1]	梅宇, 王智超. 2016年全国乌龙茶产销形势调研报告[J]. 广东茶业, 2017(Z1): 1-8.
[2]	蔡烈伟, 许勇泉, 周炎花, 等. 不同产区乌龙茶感官品质与茶汤化学成分分析[J]. 福建茶叶, 2016, 38(11): 17-19.
[3]	孙威江, 董青华, 周卫龙, 等. 乌龙茶品质评定与产品判别研究[J]. 茶叶科学, 2011, 31(4): 305-312.
[4]	Lin J, Zhang P, Pan Z, et al.Discrimination of Oolong tea (Camellia sinensis) varieties based on feature extraction and selection from aromatic profiles analysed by HS-SPME/ GC-MS[J]. Food Chemistry, 2013, 141(1): 259-265.
[5]	ISO 1573. Tea-Determination of loss in mass at 103 degrees CSO 1573. Tea-Determination of loss in mass at 103 degrees C[S]. Technical Committee ISO/TC 34: Agricultural Food Products, 1980.
[6]	ISO 14502-2. Determination of substances characteristic of green and black tea—Part 2: content of catechins in green tea—Method using high performance liquid chromatography, MODSO 14502-2. Determination of substances characteristic of green and black tea—Part 2: content of catechins in green tea—Method using high performance liquid chromatography, MOD[S]. Technical Committee ISO/TC 34, Food Products, Subcommittee SC 8, Tea, 2005.
[7]	ISO 10727. Tea and instant tea in solid form—Determination of caffeine content—Method using high-performance liquid chromatographySO 10727. Tea and instant tea in solid form—Determination of caffeine content—Method using high-performance liquid chromatography[S]. Technical Committee ISO/TC 34, Food Products, Subcommittee SC 8, Tea, 2002.
[8]	ISO 19563. Determination of theanine in tea and instant tea in solid form using high performance liquid chromatographySO 19563. Determination of theanine in tea and instant tea in solid form using high performance liquid chromatography[S]. Technical Committee ISO/TC 34, Food Products, Subcommittee SC 8, Tea, 2017.
[9]	Wiegand P, Pell R, Comas E.Simultaneous variable selection and outlier detection using a robust genetic algorithm[J]. Chemometrics & Intelligent Laboratory Systems, 2009, 98(2): 108-114.
[10]	Leardi R, Seasholtz M B, Pell R J.Variable selection for multivariate calibration using a genetic algorithm: prediction of additive concentrations in polymer films from Fourier transform-infrared spectral data[J]. Analytica Chimica Acta, 2002, 461(2): 189-200.
[11]	Soares S, Gomes A A.The successive projections algorithm[J]. Trac Trends in Analytical Chemistry, 2013, 42(42): 84-97.
[12]	陈斌, 孟祥龙, 王豪. 连续投影算法在近红外光谱校正模型优化中的应用[J]. 分析测试学报, 2007(1): 66-69.
[13]	刘华煜. 基于支持向量机的机器学习研究[D]. 大庆: 大庆石油学院, 2005.
[14]	Zhao J, Chen Q, Huang X, et al.Qualitative identification of tea categories by near infrared spectroscopy and support vector machine[J]. Journal of Pharmaceutical and Biomedical Analysis, 2006, 41(4): 1198-1204.
[15]	Chen Q, Zhao J, Fang C H, et al.Feasibility study on identification of green, black and Oolong teas using near-infrared reflectance spectroscopy based on support vector machine (SVM)[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2007, 66(3): 568-574.
[16]	欧文娟, 孟耀勇, 张小燕, 等. 紫外可见吸收光谱结合主成分-反向传播人工神经网络鉴别真假蜂蜜[J]. 分析化学, 2011, 39(7): 1104-1108.
[17]	王丽琼, 范琦, 易珍奎, 等. HPLC指纹图谱结合反向传播人工神经网络和判别分析鉴定不同的麻黄药材[J]. 西南师范大学学报(自然科学版), 2012, 37(5): 73-77.
[18]	李欣海. 随机森林模型在分类与回归分析中的应用[J]. 应用昆虫学报, 2013, 50(4): 1190-1197.
[19]	马玥, 姜琦刚, 孟治国, 等. 基于随机森林算法的农耕区土地利用分类研究[J]. 农业机械学报, 2016, 47(1): 297-303.
[20]	詹曙, 姚尧, 高贺. 基于随机森林的脑磁共振图像分类[J]. 电子测量与仪器学报, 2013, 27(11): 1067-1072.
[21]	方匡南, 吴见彬, 朱建平, 等. 随机森林方法研究综述[J]. 统计与信息论坛, 2011, 26(3): 32-38.