Electronic tongue as a rapid analytical tool was used to assess the quality of green tea in this work. "Biluochen" green tea was studied. The total taste score value for the brewed tea infusion was attained by chemical evaluation method of green tea taste which was considered as the reference measurement. Firstly, the effects of production dates on taste quality of green tea and sensor signals of green tea infusion were analyzed using one-way analysis of variance (One-Way ANOVA). Subsequently, partial least squares and least squares support vector machines methods were contradistinctive used to establish the relationship between sensor signals and the total taste score value of tea infusion. Results showed that production dates had significant effects on green tea taste quality and sensor signals. The established least squares support vector machines model got better predictive effect with correlation coefficient of 0.906 and root mean square error of prediction of 4.077 when 4 principal components were included. This method will provide a basic for the intelligent evaluation of tea quality.%该文研究利用电子舌技术快速评价绿茶的滋味品质.试验以“碧螺春”绿茶为研究对象,以绿茶滋味化学鉴定法作为绿茶滋味品质的评价方法,获得的滋味总得分值作为电子舌评价模型的参考测量值.在数据分析过程中,首先对不同生产日期的碧螺春茶汤滋味总得分值和各传感器响应值进行单因素方差分析；然后对比采用偏最小二乘法和最小二乘支持向量机建立电子舌传感器响应值与滋味总得分值之间的相关模型.结果显示不同生产日期对绿茶滋味品质及各传感器响应信号都具有极显著影响；当采用4个主成分时,建立的最小二乘支持向量机模型最优.用独立样本检验模型精度,模型预测值与参考值的相关系数为0.906,预测集均方根误差为4.077.研究结果可为茶叶品质智能化评价提供参考.
展开▼
