Research On The Formation Factors And Quantitative Evaluation Of The Umami And Sweet Flavor Of Keemun Black Tea

Umami and sweetness are important indicators of keemun black tea’s taste.The garden environment of tea trees,tenderness of fresh leaf,drying method and processing process will affect the formation of tea taste.In this paper,four important factors affecting the umami and sweet taste of keemun black tea were studied: region,drying method,grade and processing process.The umami and sweet values of samples were evaluated by equivalent quantitative sensory evaluation and electronic tongue,and 17 compounds in tea were detected by physicochemical detection combined with chemometrics to explore the umami and sweetness of keemun black tea.The results provide a scientific explanation for further understanding of keemun black tea,the formation of sweetness and substances.The main results are as follows:(1)Combined with the equivalent quantitative umami and sweet taste score and17 compounds to explore the samples of different regions,drying methods,different grades and processing processes,it is found that D3 has the best umami and sweet taste and high content of free amino acids in different regions,the content of tea polyphenols was lower,the ratio of phenol to ammonia was the lowest,and the content of soluble sugar was higher.Among the samples of drying method,Z3 had the best umami and sweet performance,with high contents of free amino acids,soluble sugar,TFDG,TFs and low ratio of phenol to ammonia.In three different grades of samples,the umami taste decreased with the decrease of grade,the sweetness of N2 and N3 was significantly higher than that of N1,the free amino acids decreased significantly with the decrease of grade,the soluble sugar increased with the decrease of grade,and the contents of TFs and caffeine decreased with the decrease of grade.In the processing samples,the umami taste increased slightly after withering,decreased significantly after rolling,and the sweetness increased gradually with the processing,and its change was closely related to free amino acids,TF1,TCs,caffeine and soluble sugar.(2)The principal component analysis of 17 chemical components among samples of various factors shows that the total interpretation variation of the first two principal components of samples in different regions is 77.4%,and the separation effect of D2 and D3 is good.However,the separation effect of D1 and D4 is not significant;the total variation of the first two principal components of different drying samples is 82.9%.The separation effect of five drying methods is significant.The total variation of the first two principal components of different grades of samples is73.5%.The separation effect of three grades is significant.The total variation of the first two principal components of the processing samples is 90.6%.The separation effect of R3 and R4 is good.R1 and R2 have partial overlap and the separation effect is not significant.(3)A support vector machine model is constructed by combining the umami(sweet)taste values of the sample electronic tongue with 16 flavor substances.The results show that the training set of PCA-SVM umami taste model: Rc=0.9999,RMSECV=0.0754,prediction set Rp=0.9951,RMSEP=0.6527;PCA-SVM sweetness model,Rc=0.9999,RMSEC=0.0004,prediction set Rp=0.9603,RMSEP=0.3549;The fitting accuracy and prediction ability of umami(sweet)taste PCA-SVM training model are better than SVM model,so the stability,accuracy and prediction ability of PCA-SVM model are the best.