【作者】 杨仁杰； 杨延荣； 董桂梅； 杜艳红； 赵玉连；

【Author】 YANG Ren-jie;YANG Yan-rong;DONG Gui-mei;DU Yan-hong;ZHAO Yu-lian;College of Engineering and Technology,Tianjin Agricultural University;

【机构】 天津农学院工程技术学院；

【摘要】 为进一步探索基于二维红外相关谱在判别掺杂牛奶方面应用的可行性与潜力,将二维红外相关谱与核隐变量正交投影算法相结合,发展了一种快速、高效的掺杂牛奶判别方法。采集64个合格牛奶样品,并分别配置浓度范围均为0.01⁰.3 g·L^-1掺杂尿素牛奶、三聚氰胺牛奶、四环素牛奶和葡萄糖牛奶各16个。采集所有样品在900¹ 700cm^-1红外光谱,并以牛奶中掺杂物浓度为外扰,构建二维红外相关谱。直接采用二维红外相关谱矩阵（128×201×201）建立掺杂牛奶与纯牛奶的核隐变量正交投影判别模型。所建的模型对预测集中未知样品的判别正确率为97.5%。更多还原

【Abstract】 The aim of the current study was to further investigate the potential application of two-dimensional（2D） infrared correlation spectra in classifying adulterated milk.A rapid,cost-effective and high-throughput discriminate method of adulterated milk was proposed by combing 2D infrared correlation spectroscopy with kernel orthogonal projection to latent structure（KOPLS）.64 pure milk samples and 64 adulterated milk samples,including 16 urea-tainted milk,16 melamine-tainted milk,16tetracycline-tainted milk,and 16 glucose-tainted milk,were prepared,respectively.The concentration ranges of adulterants all were 0.01⁰.3 g·L^-1.The one-dimensional infrared spectra of 128 samples were recorded in the spectral region 900¹ 700cm^-1.2D infrared correlation spectra under the perturbation of adulterant concentration were calculated.The calibration KOPLS discriminate model was constructed using 2D infrared correlation spectra of adulterated milk and pure milk.The recognition rate of unknown samples was 97.5%in prediction set using K-OPLS model.更多还原