Research On Evaluation Method Of Chinese Rice Wine Flavor Based On Intelligent Sensory Technologies

Chinese rice wine is a traditional brewing wine originating from China,which contains rich nutrients and is known as"liquid cake".In recent years,with the rapid development of food industry,people's requirements for food have been improved from quality assurance to flavor optimization.Therefore,enterprises pay more attention to the innovation of product flavor and the upgrading of quality.In this process,it is particularly important to realize the evaluation of products through flavor evaluation.At present,the method commonly used in flavor evaluation is artificial sensory evaluation,which can directly reflect the sensory quality of products by human sensory experience.However,this method has strong subjectivity and poor repeatability,and is vulnerable to environmental factors.Chromatographic analysis is commonly used in the detection of flavor substances,which has high detection accuracy,but it is time-consuming and laborious,and can not timely feedback the actual flavor information.In this study,the combination of GC-IMS and electronic tongue technology was used to detect the olfactory and taste flavor substances of Chinese rice wine,and the correlation analysis was made with artificial sensory evaluation to realize the intelligent evaluation of Chinese rice wine flavor.The main research contents of this thesis are:1.Artificial sensory evaluation of Chinese rice wine samplesAccording to the description of the sensory requirements of Chinese rice wine in the national standard,the scores of the sensory evaluation of Chinese rice wine were formulated.The scores of the wine samples of different ages were evaluated by the score evaluation method,and the aroma and taste scores were obtained as the flavor characteristics of each sample.Quantitative evaluation used to establish the predictive model of intelligent sensory on the flavor attribute of Chinese rice wine in the later stage.The results of artificial sensory evaluation showed that different Chinese rice wine had their own flavor characteristics and did not change linearly with wine age.It was related to the type and content of flavor substances contained in wine body.2.Study on detection method of olfactory flavor substance of Chinese rice wine based on GC-IMS combined technology(1)The volatile substances in Chinese rice wine were detected and analyzed by GC-IMS.The three-dimensional spectrum of Chinese rice wine samples was analyzed and33 characteristic ion peaks were selected by differential spectroscopy.13 kinds of flavor substances in Danyang Chinese rice wine were identified by GC-IMS Library Search,including one type of terpene,four types of alcohols and four types of Esters,one type of acid,one type of aldehyde and two types of ketones.(2)By analyzing the changing trend of characteristic peaks,the changing rule of sample characteristics was analyzed.The validity of the test results was verified by the method of wine age discrimination by selecting the characteristic variables corresponding to each characteristic peak.LDA,KNN and other methods were used to classify and discriminate the age of wine.The discriminant accuracy was 90%and 100%respectively.The qualitative analysis of volatile components in Chinese rice wine and the discriminant analysis of the age of Chinese rice wine by GC-IMS were realized.It indicated that GC-IMS technology can be effectively applied to the detection of olfactory flavor substances in Chinese rice wine.3.Research on the detection method of Chinese rice wine taste and flavor based on electronic tongue technology(1)Using the titration method in the national standard to determine the physicochemical indicators related to taste,such as total sugar,total acid,and amino acid nitrogen.Using high performance liquid chromatography to determine the free amino acid content in the sample,and classify the free amino acid in the Chinese rice wine as Four kinds of fresh,sweet,bitter and sputum for analyzed.The results showed that the total sugar,total acid,amino acid nitrogen and free amino acid in Chinese rice wine showed a certain change with the age of wine.(2)Seven multi-channel lipid membrane taste sensors were selected as arrays for electronic tongue data collection.Based on the eigenvalues of seven sensors as characteristic variables,the quantitative prediction models of total sugar,total acid,amino acid nitrogen and free amino acids of Chinese rice wine were established by PLS and SVM methods,respectively.The results show that the correlation coefficient obtained by SVM method is higher than that of PLS method.The R_c of the total sugar quantitative prediction model established by SVM method is 0.9824,the RMSECV is74.9490,the prediction set R_p is 0.9370,the RMSEP is 74.9115;the total acid training set R_c is 0.9128,RMSECV is 4.9441,prediction set R_p is 0.8553,RMSEP is 4.9567;amino acid nitrogen training set R_c is 0.9805,RMSECV is 0.6586,prediction set R_p is0.9271,RMSEP is 0.6822;bitter amino acid training set R_c is 0.9743,RMSECV The predicted set R_p was 0.9497,the RMSEP was 1.7596,the sweet amino acid training set R_c was 0.9775,the RMSECV was 1.107,the predicted set R_p was 0.9462,the RMSEP was 1.1365,the umami amino acid training set R_c was 0.9793,and the RMSECV was0.5765.The set R_p was 0.9701,the RMSEP was 0.5949;the astringent amino acid training set R_c was 0.9795,the RMSECV was 0.2834,the predicted set R_p was 0.9554,and the RMSEP was 0.2796.The model was constructed to achieve effective prediction of total sugar,total acid,amino acid nitrogen and free amino acid content of Chinese rice wine.It shows that the overall response of the electronic tongue to the Chinese rice wine sample can quantitatively predict the various taste substances.4.Research on the evaluation method of Chinese rice wine flavor based on GC-IMS combined technology and electronic tongue technologyExplore the evaluation method of Chinese rice wine flavor based on GC-IMS combined technology and electronic tongue technology.The correlation variables obtained by the two intelligent sensory techniques are respectively related to the corresponding sensory scores.The modeling result of SVM is better than the PLS method.The training set R_c of the olfactory score quantitative prediction is 0.9949,and the RMSECV is 19.3140.The set R_p was 0.7686 and the RMSEP was 19.1801;The training set Rc,RMSECV,Rp and RMSEP for quantitative prediction of taste score were 0.8798,38.6976,0.8565 and 38.6221 respectively.The modeling results show that the intelligent sense can achieve quantitative prediction of the flavor attribute of Chinese rice wine.The correlation between artificial senses,physical and chemical indicators,and intelligent senses was analyzed.Based on the correlation between intelligent senses and physical and chemical indicators,the feasibility of predicting artificial sensory scores by intelligent senses was confirmed.Then the artificial sensory score obtained by intelligent sensory prediction is corresponded to the corresponding descriptor in the artificial sensory evaluation system.Intelligent sensory evaluation of rice wine flavor was completed.In order to improve the efficiency and accuracy of flavor detection,This research combines the technology of GC-IMS and electronic tongue,with the appropriate pattern recognition method,the artificial sensory score is quantitatively predicted by using intelligent sensory system.The predicted results combined with scoring criteria can correspond to specific sensory descriptions.Closed-loop sensory evaluation is completed and intelligent sensory flavor evaluation method is established.The results show that it is feasible to judge the flavor attribute of Chinese rice wine by intelligent sensory technology.The research results can provide experimental support for the practical application of flavor regulation and product flavor control in Chinese rice wine production,and provide theoretical basis for the application of rapid non-destructive testing technology in the field of flavor.