近红外技术识别掺假原奶的可行性研究  

The feasibility study of discriminating adulteration raw milk by near infrared technology

作　　者：钟珍珍[1] 张立国[1] 张鑫[2] 顾欣[2] 倪力军[1] 

机构地区：[1]华东理工大学化学与分子工程学院,上海200237 [2]上海市动物疫病预防控制中心兽药检测所,上海201103

出　　处：《计算机与应用化学》2010年第12期1691-1693,共3页Computers and Applied Chemistry

摘　　要：为探索近红外技术识别掺假原奶的可行性,本文从上海市以及周边地区各牧场采集两批共162个原奶样本,将冷藏保鲜的原奶在40℃下超声20min后配制相同数目、糊精水溶液含量为1～20%的含三聚氰胺或尿素的掺假牛奶,采用漫反射模式采集其近红外光谱。用SNV、MSC及一阶导数对光谱进行预处理、分别以IS-KNN、KNN以及PLS-DA法建立掺假原奶的近红外判别模型,轮流留1/4样品作为检验集统计平均预测正确率。结果表明IS-KNN法可取得良好效果,该法对掺假原奶的判别正确率在90%以上,对原奶的判别正确率在81%以上,对含三聚氰胺假奶的判别正确率要高于含尿素假奶,该法对不区分假奶中伪蛋白种类所得真、假牛奶的识别正确率仍可达85%与91%以上。本文结果表明采用合适的模式识别方法及光谱预处理方法可实现对掺假原奶的良好识别。样本的采集地、采集时间、样本数目等因素对原奶近红外判别模型的影响尚不明确,该技术的实际应用有待进一步深入研究和完善。In order to explore the feasibility of discriminating adulteration raw milk by the NIR technology,two groups of raw milk,including 162 samples,were gathered from the pastures of Shanghai and surrounding areas.Every raw milk sample,which was kept fresh in cold,was put in an ultrasonic cleaner for 20min at 40℃ to homogenize it.The same number of counterfeit milk by adding 1%～20% dextrin water solution that contained melamine or urea were prepared after homogenizing.The near infrared（NIR） spectra of the raw and adulterate milk samples were collected on diffuse reflection mode.Multiplicative scatter correction（MSC）,standard normal variate（SNV） and first derivative were applied to pre-process the spectra.The NIR model for discriminating adulteration of raw milk was built based on IS-KNN,KNN and PLS-DA methods,respectively.A quarter of samples were selected as validation sets by turns and the average of four times prediction accuracy ratio was used to evaluate the performance of the NIR discrimination models.The results demonstrate that the method of IS-KNN is more suitable for discriminating raw milk samples.The prediction accuracy ratio of this method to discriminating adulterate milk was higher than 90% and 81% to raw milk.The prediction accuracy of this method to discriminate counterfeit milk containing melamine was higher than those containing urea.For the adulteration milks added different pseudo-proteins,the method still gives prediction accuracy ratio with higher than 91%.So it is possible to discriminate adulteration milks well based on NIR spectra by means of suitable chemometric method and spectra pretreatment method.The influence of collecting zone,collecting time and number of samples etc.factors on the NIR model for discriminating adulteration milks is as yet unknown.Thus the practical application of this technology should be further investigated and improved.

关 键 词：生鲜奶 掺假奶 近红外 三聚氰胺 尿素 

分 类 号：TQ015.9[化学工程] TP391.9[自动化与计算机技术—计算机应用技术]