基于电子鼻山茶油芝麻油掺假的检测研究

本文采用电子鼻系统对山茶油、芝麻油的掺假（大豆油）作了检测.通过对传感器信号进行方差分析可知,三种油脂的传感器响应有显著差异.主成分分析（PCA）对山茶油与大豆油及其混合物检测效果较差,对芝麻油、大豆油及两者混合物取得了较好的检测效果;而线性判别式分析（LDA）对山茶油和芝麻油的掺假都有较好的检测效果,并优于PCA方法.运用BP神经网络拟对混合油脂进行定量预测,对山茶油掺假的定量预测效果较差,对芝麻油掺假的预测效果略好于山茶油,但最大绝对误差已达0.134,还不能取得较为准确的结果.

Detection of Adulteration in Camellia Seed Oil and Sesame Oil Using an Electronic Nose

An electronic nose was used for detection of soy bean oil adulteration in camellia seed oil and sesame oil. The results of multivariate analysis of variance show that the sensors signals of different kinds of oil are different with each other significantly. Principal component analysis (PCA) can not be used to detect the adulteration of camellia seed oil, but can be used in detection of adulteration in sesame oil. Linear discriminant analysis (LDA) is more effective than PCA, which can be used in adulteration detection of both camellia seed oil and sesame oil. The BP model has been used to detect the percentage of adulteration in camellia seed oil and sesame oil. The results show that, except in certain particular cases, it is difficult to obtain good calibration models for the quantification of the percentage of adulteration. Although the prediction is more precise in sesame oil than in camellia seed oil, the maximum absolute error is 0.134.