摘要
应用主成分分析(Principal component analysis,PCA)和聚类分析法(Cluster analysis,CA)对9种(27个)常见食用植物油及100个餐饮废油的低场核磁共振(Low-field nuclear magnetic resonance,LF-NMR)(T2)弛豫特性数据进行分析。结果表明:在正常食用油种类区分方面,主成分分析的效果较优,9种食用油在主成分分布图上按种类正确分组,边界清晰。而在正常食用油与餐饮废油的区分方面,聚类分析效果较优,引入30个待测样本后,聚类分析(127个样品,欧式距离=5)的正确率为94.49%,分析误判率为5.51%,分组效果良好。LF-NMR结合化学模式识别可实现对油脂种类及餐饮废弃油脂的鉴别。
To establish an effective analysis method for evaluating the quality of edible oil is of great significance to ensure the safety of edible oil market.Principal component analysis（PCA）and cluster analysis（CA）were used to analyze the low field nuclear magnetic resonance（LF-NMR） T2 relaxation characteristics of 9 kinds of normal edible oil and 100 catering waste oil samples.The results indicated that good classification of refined edible oil according to their vegetable types could be achieved by PCA,and the distributions of different vegetable oils on the PCA plot have clear boundaries.While for the discrimination of authentic vegetable oil and the catering waste oil,good identification results could be achieved by CA（Euclidean distance=5）.After the introduction of 30 testing samples,the overall correct classification rate was still as high as 94.49%,and the misjudgment rate was only 5.51%.Therefore,LF-NMR combined with chemometrics method is feasible for rapid classification of edible vegetable oils and discrimination of catering waste oils.
作者
毛锐
王欣
史然
MAO Rui WANG Xin SHI Ran(School of Medical Instrument and Food Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China China Agricultural University, Beijing 100083, China)
出处
《分析测试学报》
CAS
CSCD
北大核心
2017年第3期372-376,共5页
Journal of Instrumental Analysis
基金
国家自然科学基金项目(NSFC31201365)
上海市科委重点攻关项目(11142200403)
上海市教委科研创新项目(11YZ109)
