摘要
采用近红外光谱技术与偏最小二乘法(PLS)相结合的方式对核桃油中菜籽油、大豆油的现象进行定量分析。结果显示,当核桃油中菜籽油时,采用最小-最大归一化进行光谱预处理,菜籽油定量模型校正相关系数RC及校正标准误差RMSEC和交叉验证模型的相关系数RCV和交叉均方根RMSECV分别为99.97%,0.532%和99.96%,0.562%;当核桃油中大豆油时,采用一阶导数+矢量归一化进行光谱预处理,大豆油定量模型RC及RMSEC和RCV及RMSECV分别为99.99%,0.336%和99.99%,0.352%;当核桃油中2种油(菜籽油和大豆油)时,采用二阶导数进行光谱预处理,菜籽油定量模型的RC及RMSEC和RCV和RMSECV为99.99%,0.220%和99.98%,0.313%,大豆油定量模型的RC及RMSEC和RCV和RMSECV为99.99%,0.210%和99.97%,0.386%;当核桃油中菜籽油、大豆油和玉米油时,分别以菜籽油、大豆油和玉米油为研究对象,同时分别用一阶导数+多元散射校正、二阶导数、一阶导数方法进行光谱预处理,可以得到菜籽油定量模型的RC及RMSEC和RCV和RMSECV为99.93%,0.347%和99.90%,0.401%,大豆油定量模型的RC及RMSEC和RCV和RMSECV为99.87%,0.524%和99.80%,0.615%,玉米油定量模型的RC及RMSEC和RCV和RMSECV为99.94%,0.371%和99.89%,0.451%。经过重复性试验和外部试验验证可知,近红外光谱分析技术可以快速准确地检测核桃油中掺入1种油、2种油和3种油的含量。
In this study,near infrared spectroscopy（ NIR） combined with partial least squares（ PLS） was used to rapidly quantify the authentication of walnut oil. Walnut oils adulterated with different concentrations of rapeseed oil,soybean oil and corn oil were distinguished. The results showed that the spectra of walnut oils adulterated with rapeseed oil were subjected to Min / Max normalization and the relative correlation（ RC） and Root Mean Square Error of Calibration（ RMSEC） were 99. 97% and 0. 532%,and the relative correlation（ RCV） and Root Mean Square Error of Cross Validation（ RMSECV） for cross validation were 99. 96% and 0. 562%. The spectra of walnut oils adulterated with soybean oil were subjected to First Derivative and Vector Normalization and the RCand RMSEC were99. 99% and 0. 336%,and the RCVand RMSECV for cross validation were 99. 99% and 0. 352%. The spectra of walnut oils adulterated with rapeseed oil and soybean oil were subjected to Second Derivative and the RC,RMSEC,RCVand RMSECV of repressed oil were 99. 99%,0. 220%,99. 98% and 0. 313%; and the RC,RMSEC,RCVand RMSECV of soybean oil were 99. 99%,0. 210%,99. 97% and 0. 386%. The spectra of walnut oils adulterated with rapeseed oil,soybean oil and corn oil were subjected to First Derivative and Multiple Scattering Correction,Second Derivative,First Derivative and the RC,RMSEC,RCVand RMSECV of 99. 99%,0. 210%,99. 97% and 0. 386%.The spectra of walnut oils adulterated with rapeseed oil,soybean oil and corn oil were subjected to First Derivative and Multiple Scattering Correction,Second Derivative,First Derivative and the RC,RMSEC,RCVand RMSECV of rapeseed oil were 99. 93%,0. 347%,99. 90% and 0. 401%; and the RC,RMSEC,RCVand RMSECV of soybean oil were 99. 87%,0. 524%,99. 80% and 0. 615%; the RC,RMSEC,RCVand RMSECV of soybean oil were 99. 94%,0. 371%,99. 89% and 0. 451%. In this paper,NIR method played a good role in quantification,which offered a new approach to rapid quantification of pure and adulterated walnut oil.
出处
《中国粮油学报》
EI
CAS
CSCD
北大核心
2015年第12期106-113,共8页
Journal of the Chinese Cereals and Oils Association
基金
陕西省科技统筹创新工程计划(2012KTCG01-09)
西北农林科技大学大学生创新创业训练计划(2201210712178)
关键词
近红外光谱
偏最小二乘法
鉴别
核桃油
near infrared spectroscopy
PLS
quantification
walnut oil
作者简介
彭星星，女，1989年出生，硕士，果蔬贮藏与加工
通讯作者：徐怀德，男，1964年出生，教授，饮料、果品蔬菜贮藏与加工、天然产物提取
