Sensory Evaluation Of Wine Using Electronic Nose And Electronic Tongue

Wine sensory evaluation mainly depend on the chemical techniques,such as High Performance Liquid Chromatography or Gas Chromatography-Mass Spectrometer etc.,and traditional human sensory evaluation.Because the complexity of wine aroma composition,the human evaluation is still unsurpassed to appraise its quality.However,the main limitations of sensory analysis are related to the non-reproducibility of the evaluation and to the individual variation of tasters likely linked to anatomical,psychological,and cultural aspects.At chemical analysis respect,the techniques of GC-MS and HPLC currently provide robust analytical methods that are extensively used in enology laboratories to identify and quantify volatile aroma compounds,polyphenol,amino acids etc.However,these methods are usually labor-intensive and require sophisticated,expensive equipment and highly skilled personnel.Therefore,many studies focuse on finding a time-consumed and objective wine analysis method.Electronic nose(EN)and electronic tongue(ET)are thought to emerge as third approach to wine analysis in a wide range of applications and a good supplemen for chemical techniques and human beings.And they offer a new way for objective wine sensor analysis.In this study,EN and ET were used to detect the wine samples that fermented in different wineries of China and with different yeasts and different corks and different grape varieties.Combining with various pattern recognition analysis,the aim of this study is to evaluate the capacity of wine identification and objective sensory evaluation for electronic nose and electronic tongue,and to establish the wine region identification model of electronic nose and the model for wine sensory evaluation of electronic tongue,which are of great importance for intelligent sensory system using in wine industry.(1)In this study,EN has been used to detect different wine samples,by the loadings analysis,principal component analysis(PCA)and linear discriminant analysis(LDA)to evaluate EN abilities of recognition and volatile aroma compounds sensory analysis.The results come out that based on PCA and LDA,the EN could discern the different wines well and could evaluate the aroma of wine in a certain degree.(2)In order to establish the model for wine region recognition,the wine samples were divided into two groups of the training set and testing set,and the software WinMuster was used to establish the wine identification model for different regions of China.Combining the wines of testing set with four algorithms of Euclid,Mahalanobis,DFA and Correlation to detect the blind wines identification capability of the model.The results shows that the identification model also could differentiate the testing set wines.(3)ET has been used to detect the sourness,saltness,astringency,bitterness,umami,umami richness,the bitter aftertaste and the astringent aftertaste of different wine samples.And analysis the data obtained by ET to explore and study the ability of wine sensory evaluation of ET.At the same time,chemical parameters were analysed to verify the ability of electronic tongue of wine sensory evaluation.The research results show that there is a good correlation of the wine sensory evaluation on acidity between ET and pH of chemical analysis.The radar chart and bubble graph of ET resuslts can describe the differences of all wine samples well.(4)And the principal component analysis is used to build the model for wine sensory evaluation by the software SPSS19.0 and the data obtained by ET.And the model was used to evaluate wine samples,comparing the sensory evaluation resulsts of the model with the expert tasters tasting results.The study results show that there is a good consistency between the results of human tasting and the PCA wine sensory evaluation model.The study provides a new way and some reference for more objective wine sensory evaluation.