关键词: 羊奶, 蛋白质掺假, 电子鼻, 化学计量法

Abstract: Goat milk adulterated with proteins was qualitatively discriminated and quantitatively analyzed using electronic nose combined with chemometric methods. Milk samples adulterated with different proteins were detected by electronic nose and then the response values were analyzed qualitatively by principal component analysis (PCA), and linear discriminant analysis (LDA) and quantitatively by linear regression analysis, Fisher discriminant analysis (FDA) and K nearest neighbor (KNN) analysis. The results showed that PCA and LDA were able to distinguish different adulterants. The determination coefficient of linear regression analysis was 84.5%, indicating high reliability of the regression equation. The accuracy of the original classification by FDA reached 100.0%, and the accuracy of cross validation was 98.2%, indicating good predication performance. The classification accuracy of the training set by KNN analysis was 95.1%, and the classification accuracy of the validation set was 97.1%, indicating good prediction performance of the model. All of these results showed that it is feasible to apply electronic nose technology in recognition of protein adulteration in goat milk.

Key words: goat milk, protein adulteration, electronic nose, chemometric methods