采用闽南、闽北和广东乌龙茶为样本,进行了基于化学成分、因子分析和判别分析的茶样品质评定和产品判别模型研究。应用因子分析得出累积变异量为85.041%的7个公共因子,再以公共因子为自变量,采用同步判别分析法对茶样进行产地、品种和级别鉴别,建立判别函数;交叉验证和回代样本验证结果显示,产地和品种的判别模型正确区别度较高,产地的回代和交叉验证总正判率分别为97.6%和94.0%,品种的为86.9%和75.0%。并确立7个乌龙茶品质因子得分函数和1个综合评价函数,得到判别效果良好的产地判别函数3条、品种判别函数6条。

The model of quality evaluation and products discrimination of Oolong tea based on chemical composition, factor and discriminant analysis were studied with Guangdong, southern and northern Fujian Oolong tea as the samples. Seven factors with cumulative variance of 85.041% were obtained along with factor analysis, set the factors as the independent variable, and discriminating classification of the samples area,variety and grade use discriminant analysis. Then the discrimination functions were established. Results of cross-validation and resubstitution of testing samples demonstrated the correct rates of discrimination model of area and variety was satisfactory. Recognition accuracy of the resubstitution and cross-validation of area were 97.6% and 94.0%, which of variety was 86.9% and 75.0%. Seven score function of quality factor and one comprehensive evaluation function were established for Oolong tea, and three discrimination functions for producing area and six of variety were obtained and the discrimination results were satisfactory.