基于高效液相色谱的普洱晒青毛茶指纹图谱识别方法

为了对不同产地的晒青毛茶进行鉴别，研究了晒青毛茶高效液相色谱指纹图谱的不同识别方法。运用相关系数、夹角余弦和重叠率3种方法分别计算指纹图谱的相似度，以数字化指纹图谱为基础，对样品进行了系统聚类，分析了18个茶叶样品的主成分，并以前2个主成分作二维散点图。结果表明，3种方法均能准确地体现指纹图谱的相似程度，使毛茶与绿茶得到良好的分离；茶叶内含成分复杂，主成分较多，7个主成分累计贡献率为88.61%，系统聚类和二维排序散布图能够区分晒青毛茶和绿茶，具有简便、直观的特点。4种识别模式均能较好地对指纹图谱进行识别，为普洱茶原料鉴别和质量控制提供新的试验依据。

Fingerprint identification method of Pu'er raw tea based on high performance liquid chromatography profiles

Different recognition methods of the fingerprint for solar dried Pu'er raw tea was investigated based on their high performance liquid chromatography (HPLC) profiles, in order to identify Pu'er raw tea from different places. Similarity of the fingerprint was calculated by comparing their HPLC profiles using three methods, namely correlation coefficient, angle cosine and overlap rate, respectively. Based on the digitalized fingerprint, tea samples were systematically clustered. The main components of 18 tea samples were analyzed, in which the former two main components were employed to plot a two-dimensional scatter map. The results showed that each of the three employed methods could reflect the similarity of the fingerprint accurately, by which solar dried Pu'er raw tea could be well identified from green tea. Although the component of tea sample is complicated and the main component is rich, the selected seven main components contributed 88.61% of the total. Solar dried Pu'er raw tea samples could be distinguished obviously from green tea samples by using method of either systematic clustering or two-dimensional map. The recognition modes described could not only provide an identification method for fingerprints, but also provide an experimental foundation for quality control of raw material for Pu'er tea.