以市售的10种花椒油为研究对象,采用电子鼻技术结合主成分分析(principal component analysis,PCA)、聚类分析(cluster analysis,CA)、线性判别分析(linear discriminant analysis,LDA)和判别因子分析(discriminant factor analysis,DFA)对不同氧化程度的花椒油进行定性判别分析,并建立相应的花椒油氧化快速判别模型。利用所建模型对40组验证样品进行判别,将判别结果与国标法所测过氧化值进行比较分析,结果表明,CA、PCA、LDA和DFA的判别正确率分别为95.0%、97.5% 、100%和97.5%,说明基于电子鼻技术能够对花椒油氧化进行判别,且LDA法判别效果较好。
The commercial samples of 10 kinds of pepper oil were qualitatively analyzed of different oxidation degrees by electronic nose technology combined with principal component analysis (PCA), cluster analysis (CA), linear discriminant analysis (LDA) and discriminant factor analysis (DFA). A model of rapid discrimination of pepper oil with different oxidation degrees was established using PCA, CA, LDA and DFA, respectively. Then 40 groups of samples were tested by the models. The results were compared with GB peroxide values method. The correct recognition rates of CA, PCA, LDA and DFA were 95.0%, 97.5%, 100% % and 97.5%, respectively. The results showed that electronic nose technology was able to determine the oxidation degree of pepper oil, and LDA analysis was a better method.