基于近红外光谱技术的白茶3种典型感官滋味特征属性定量评价模型研究

Quantitative Evaluation of Taste-characteristic Attributes in Three White Tea Based on Near Infrared Spectroscopy

针对不同产地、不同类型及不同年份的白茶的感官滋味品质特征、滋味特征属性进行深入剖析及基于近红外光谱技术的滋味属性预测评价研究。研究显示甘甜度与醇爽度是白茶滋味品质特征表现的必要评价指标,相同原料及前期工艺条件下陈化度提高可显著提升滋味甘甜度及醇爽度。利用NIRS-PLS法对滋味特征总分、甘甜度、醇爽度及陈化度分别建模,模型定标集校正相关系数R_c为0.837~0.987,预测相关系数R_p为0.748~0.888,均具有较高预测能力,表明利用近红外光谱技术能够实现白茶感官滋味特征属性的快速预测评价。

In this study, tea taste-quality evaluations and attribution analysis were performed in white teas of different producing areas, tea types and storage times. The potential of near infrared spectroscopy（NIRS） in determining taste-attributes of white tea was also evaluated. The results showed that sweet, mellow and brisk tastes were the main aspects of taste-quality evaluation in white teas. Increasing of post-oxidation degrees could significantly enhance the sweet, mellow and brisk tastes of white tea with the same leave materials and processing technology. Near infrared spectroscopy（NIRS） was used to develop four models correlating spectral data to general taste-characteristics and its individual attributes（sweet, mellow, brisk tastes and post-oxidation characteristics）. The four optimized models based on near infrared spectroscopy-partial least squares（NIRS-PLS） had considerable potential for taste-quality evaluation with correlation coefficient of calibration（ Rc） ranging from 0.837 to 0.987 and that of prediction（Rp） ranging from 0.748 to 0.888, which suggested that fast taste-attribute evaluation of white tea could be achieved by NIRS-PLS.