近红外光谱结合线性判别快速无损鉴别婴幼儿和中老年奶粉.PDFVIP

《乳业科学与技术》 2010 年第3 期 （总第142 期） 125 近红外光谱结合线性判别快速无损鉴别婴幼儿和中老年奶粉 唐玉莲 中南大学化学化工学院中药现代化研究中心，湖南长沙 410083 摘 要：应用近红外光谱分析技术(NIRS)并结合线性判别，对中老年和婴幼儿2种不同年龄阶段奶粉品 种进行鉴别。对奶粉样本的原始光谱进行一阶微分处理后，采用Kennard-Stone法对80个样本进行挑选， 选出60个作为训练集，剩余的20个作为预测集。采用留一交叉验证评价算法，选取最优主成分数2建立 模型，并对测试集进行预测，正确率为100 %。说明提出的方法具有很好的分类和鉴别作用，为婴幼儿 奶粉的品种快速无损鉴别提供了一种新方法。 关键词：近红外光谱；线性判别；奶粉品种；鉴别 中图分类号: TS252.7 文献标识码: A 文章编号: 1671-5187(2010)03-0125-04 Fast Discrimination of Infant and Middle-aged Milk Powder Using Near Infrared Spectra Infrared Spectra with Linear Discriminant Analysis Tang Yulian Research Center of Modernization of Chinese Medicines, College of Chemistry and Chemical Engineering, Central South University, Changsha of Hunan 410083, China Abstract: In this paper, a new method for discrimination of varieties of infant milk powder and middle-aged milk powder by means of near infrared spectroscopy (NIRS) technique combined with linear discriminant analysis (LDA) was developed. After first derivative of the original spectral of all milk powder, 60 training set(total 80 samples) was selected using Kennard-Stone method, model ，and the remainder samples formed the prediction set. 2 was the best principal component with leave-one-out cross-validation algorithm. With the founded model, the recognition accuracy of 100% was achieved. This model is reliable and practicable. So the present paper could offer a new approach to the fast discrimination of varieties of milk powder. Key words: NIRS, linear discriminant Aanalysis, milk powder, discrimination 奶粉是日常生活必不可少的营养品。奶粉中 别奶粉已成为维护食品安全中的一项重要任务。 的营养成分比例按照不同的摄食人群搭配，奶粉 线性判别是有监督模式识别中常用的判别方 的质量好坏直接影响消费人群的生命健康。同时 法，Titterington认为在很多分类问题中，线性判别