| Strong aroma type is one of the basic flavor of Chinese liquors. It has a rich, harmonious aroma and mellow taste. The sales of strong aroma type liquor has been in a leading position in Chinese liquor market. In this paper, the three-dimensional fluorescence spectra of strong aroma type liquor were measured and the fluorescent characteristic were analyzed. The discrimination of different bands and different years of strong aroma type liquors were realized by using three-dimensional fluorescence spectroscopy technology.The fluorescent characteristic about four different brands as well as a brand in different years was analyzed in this paper. The results show that the outline of three-dimensional fluorescence spectroscopy and the fluorescent characteristic are the same between the same bands of liquors, while one or two bands were different. The fluorescent characteristic are different in different bands liquors, but there are also similarities. Moreover, with the increase of the year, the fluorescence peak wavelength in 336-366nm is first blueshifted then redshifted. The fluorescence intensity of the peak around 432nm increases gradually with the increase of the year. In addition, three-dimensional fluorescence spectroscopy about Fen-lavor, Sesame-flavor and soybean-flavor liquors were also measured. Compared with the fluorescent characteristic of three strong aroma type liquor, the fluorescent characteristic of Fen-lavor liquor is similar with, as will as the Sesame-flavor liquor is different.In order to maintain the stability of the liquor market and protect the interests of consumers, a method for discrimination of different bands liquor with strong aroma type based on three-dimensional fluorescence spectrum technology was developed in this paper. Firstly, the three-dimensional fluorescence spectra of seven different brands liquor were measured. The data preprocessing methods of partial derivative and wavelet transform were carried out. Then the new data matrix was analyzed by principal component analysis (PCA) and the principal components were extracted to be used as the inputs of support vector machine (SVM). Finally the prediction model was constructed. The effect of three different spectral data after processing on the model is compared:original data, the first-order and second-order partial derivatives on the spectral data. The results show that the three-dimensional fluorescence spectra with the pretreatment of second-order partial derivatives coupled with PCA and SVM can make a good performance on the brands identification of strong aroma type liquors. The accuracy of the established model and prediction accuracy were 98.98% and 100%, respectively.In order to ensure the safe and reliable of year liquors and improve integrity of the wineries. The discrimination of different year liquors were realized by using three dimensional fluorescence spectroscopy combined with non-negative matrix factorization (NMF) in this paper. Taking two bands of Chinese strong liquor, for example,10 years,20 years,30 years liquors were chosen and mixed in different proportions to obtain a weighted years of 15 years,20 years and 25 years, respectively. Then three-dimensional fluorescence spectra of all mixed year liquor samples were measured by the fluorescence spectrometer. Non-negative matrix factorization of multiplicative iteration method was used on three-dimensional fluorescence spectral data, and the basis matrix and encoding matrix were obtained. Finally, the encoding matrix was used as the inputs of support vector machine and the prediction model about the mixed year liquor was constructed. The results obtained compared with the principal component analysis, and the results show that three-dimensional fluorescence spectroscopy combined with non-negative matrix can make a good performance on the discrimination of mixed year liquor brands, the accuracy of the established model and prediction accuracy were up to 100% and 100%, respectively. |
PDF Full Text Request