Study on discrimination of different grades of rapeseed oil based on flavor fingerprint

Tong Chen,Chunyou Liu,Bin Chen,Yongchun Huang
DOI: https://doi.org/10.1556/1326.2020.00817
2021-04-28
Acta Chromatographica
Abstract:Abstract In this work, Gas chromatograph-Mass Spectrometry (GC-MS) combined with solid phase micro-extraction technology was used to analyze the difference of volatile organic compounds (VOCs) in rapeseed oil of different grades, and the relationship between changes of VOCs and refining process were also investigated in order to construct a non-linear model, which could realize rapid and accurate discrimination of different grade rapeseed oils. 124 rapeseed oil samples with different grades were collected and analyzed by GC-MS technology and 55 VOCs were identified and selected as variables to characterize the internal quality information of rapeseed oils. Then, principal component analysis (PCA) method was used to extract useful features and reduce data dimensionality, and finally a discriminant model was built using linear discriminant analysis (LDA) algorithm. The correct recognition rate of sample set was close to 94.59%. The results showed that the proposed method is promising in discriminating different grades of vegetable oils. Besides, it provides a theoretical basis for studying the relationship between VOCs composition and vegetable oil quality.
chemistry, analytical
What problem does this paper attempt to address?