An electronic nose was used to evaluate the bitterness and astringency of green tea, and the possible application of the sensor was assessed for the evaluation of different tasting green tea samples. Three different grades of green tea were measured with the electronic nose and electronic tongue. The sensor array of the E-nose was optimized by correlation analysis. The relationship between the signal of the optimized sensor array and the bitterness and astringency of green tea was developed using multiple linear regression (MLR), partial least squares regression (PLSR), and back propagation neural network (BPNN). BPNN is a multilayer feedforward neural network trained by an error propagation algorithm. The results showed that the BPNN model possessed good ability to predict the bitterness and astringency of green tea, with high correlation coefficients (R = 0.98 for bitterness and R = 0.96 for astringency) and relatively lower root mean square errors (RMSE) (0.25 for bitterness and 0.32 for astringency) for the calibration set. The R value is 0.92 and 0.87, and the RMSE is 0.34 and 0.55, for bitterness and astringency, respectively, of the prediction set. These results indicate that the electronic nose could be used as a feasible and reliable method to evaluate the taste of green tea. These results can provide a theoretical reference for rapid detection of the bitter and astringent taste of green tea using volatile odor information.

The taste of Xinyang Maojian tea was tested using an electronic tongue to conduct a correlation analysis between the taste and quality of tea. It was determined that the astringent aftertaste had a significant relationship with the quality of tea (p < 0.01). The aftertaste properties of tea were evaluated before and after adding bitter, astringent, and rich (fresh aftertaste) aftertastes to six types of tea, and compared by principal component analysis (PCA) and linear discriminant analysis (LDA). PCA and LDA results were affected by the aftertaste attributes of tea soup. After the addition of aftertaste, the classification effect of PCA was reduced, and that of LDA was improved. After introducing aftertaste, LDA had the best effect on tea differentiation. Finally, a support vector machine (SVM) model was established for tea grade evaluation based on the tea aftertaste and basic taste attributes. The values of the SVM penalty function c and kernel function search radius g were optimized by the grid search and particle swarm optimization (PSO) methods, respectively, to compare the classification performance characteristics of the two optimization methods. The grid search method was used to analyze the global classification effect, which had a heavy calculation burden and was unable to achieve further optimization near the optimum point. The PSO method had iterative characteristics and it was easier to determine the optimal point. The classification accuracies of the training and test sets of the grid search method were 86.11 and 87.5%, respectively, and those of the training and test sets of the PSO method both reached 100%, indicating the superiority of the PSO method over the grid search method. In conclusion, the effectiveness of an electronic tongue for distinguishing the quality of Xinyang Maojian tea was demonstrated.