Abstract:

In order to effectively detect and identify their volatile flavor, pork samples were detected by electronic nose and
solid-phase micro-extraction gas chromatography-mass spectrometry (HS-SPME-GC-MS). In view of the similarity between
the two detection methods, hierarchical clustering was used to analyze the detection data. A linear regression model was
established to obtain the mapping relationship between the electronic nose detection data and the volatile flavor compounds
in pork for the classification of the electronic nose data. Consistent results were obtained from the clustering analysis of the
electronic nose detection data and volatile components of pork. The regression equation showed that a linear relationship
existed between the two detection data. Sample recognition accuracy reached 90%. This study demonstrated the feasibility
of using electronic nose to detect the difference in volatile flavor compounds of pork.

Key words: electronic nose, solid phase micro-extraction, statistical space mapping, pattern recognition