摘要
以电子鼻技术结合多元统计方法建立了特级初榨橄榄油-油茶籽油二元掺伪体系的定性鉴别和定量预测模型。结果表明:主成分分析能够区分初榨橄榄油掺伪-未掺伪两类样品,BP人工神经网络对初榨橄榄油掺伪的定性鉴别效果良好,验证集样品预测正确率100%;线性判别分析能够区分掺杂不同油茶籽油含量的混合油样,且区分能力优于主成分分析,偏最小二乘法建立的初榨橄榄油掺伪定量模型可以用于油茶籽油掺杂含量的预测,真实值和模型预测值的相关系数为0.994 7。
The qualitative identification and quantitative prediction model of extra virgin olive oil adultera- ted with oil -tea camellia seed oil were established by electronic nose technology combined with multiva- riate statistical method. The results showed that principal component analysis was able to distinguish pure and adulterated virgin olive oil samples and BP artificial neural network worked well for qualitative identi- fication of virgin olive oil adulteration, and correct prediction of validation samples was 100% ;linear dis- criminant analysis could distinguish mixed samples adulterated with different contents of oil -tea camellia seed oil, and the discrimination ability was better than principal component analysis;the partial least square quantitative model of virgin olive oil adulteration could be used to predict the adulteration content of oil - tea camellia seed oil, and the correlation coefficient of real value and predicted value was 0.994 7.
出处
《中国油脂》
CAS
CSCD
北大核心
2013年第12期84-87,共4页
China Oils and Fats
关键词
初榨橄榄油
油茶籽油
电子鼻
掺伪
virgin olive oil
oil - tea camellia seed oil
electronic nose
adulteration
作者简介
钟诚（1987）,男,硕士研究生,主要从事粮油食品检验工作（E-mail）zcgogo1987@126.com.
