MilkoscanFT120对原料乳中掺入尿素快速预判的研究

利用MilkoscanFr120红外光谱仪，采用主成分分析（PCA）回归方法，以本地区具有代表性原料乳样品为基础建立标准数据模型，在原料乳中加入尿素模拟牛奶掺假。通过总结原料乳成分数据的变化规律及调整控制因素以达到预判原料乳掺假和确定掺假物质的检出限的目的。结果表明该方法具有快速、准确的特点。

Study on Quick Predicting Methods of Adulteration in Raw Milk by Milkoscan FT 120

A standard data model was established on the basis of representative sample from raw milk in Harbin city, using principal component analysis（PCA） by IR of Milkoscan FT120. It can simulate milk adulteration by adding urea in raw milk. In order to predict the aduheration in raw milk and determine the detection limit of aduherated substances, researchers summed up the changes of data of raw milk composition. Results showed that this method was quick and accurate.