Potential of a Portable Electronic Nose for Control Quality of Moroccan Traditional Fresh Cheeses

摘要

An investigation was made to assess the ability of a portable electronic nose to monitor freshness of Moroccan traditional fresh cheeses in real-time and stored at 4℃. Six commercially available metal oxide gas sensors jointly with suitable pattern recognition methods are used to analyze cow, goat cheeses during 30 days of storage. The sensors array coupled with features extraction and pattern recognition methods namely Principal Component Analysis (PCA) and Discriminant Factor Analysis (DFA) can be trained to classify cheeses produced from bovine milk and caprine milk mixed in ratios of 100:0; 10:90; 25:75; 50:50; 0:100 (bovine:caprine) in order to detect the fraud sign, in the first storage day. On the other hand, the data coming from the response of the sensors have been elaborated by PCA and Support Vectors Machines (SVMs) in order to obtain a classification of the data clusters related to ageing time of cow and goat cheeses. PCA results show that cheese samples could be grouped into three categories (fresh, medium and aged), which corresponded to an increasing number of days that cheeses had spent under cold storage. A SVM based classifier and the one-against-one voting strategy were implemented, which achieved very good accuracy in the identification of the three categories predicted by PCA. Hence, the developed portable electronic nose technology was capable to objectively characterize cheese aroma profiles at different days of storage and to detect irregularities in fresh cheeses during the early stage.