光谱学与光谱分析
光譜學與光譜分析
광보학여광보분석
SPECTROSCOPY AND SPECTRAL ANALYSIS
2014年
10期
2673-2678
,共6页
倪力军%钟霖%张鑫%张立国%黄士新
倪力軍%鐘霖%張鑫%張立國%黃士新
예력군%종림%장흠%장입국%황사신
近红外光谱%液态奶掺假物质判别%改进与简化的支持向量机方法%改进与简化的KNN方法
近紅外光譜%液態奶摻假物質判彆%改進與簡化的支持嚮量機方法%改進與簡化的KNN方法
근홍외광보%액태내참가물질판별%개진여간화적지지향량궤방법%개진여간화적KNN방법
Near-infrared spectroscopy%Discrimination of adulterated substances in liquid milk%Improved and simplified K-nea-rest neighbor classification algorithm%Improved and simplified of support vector machine
以287例上海及上海周边地区牧场的生鲜奶作为真奶样本集组成3个真奶样品集合，配制了526例含有糊精（或淀粉）＋三聚氰胺（或尿素、或硝酸铵）的掺假牛奶形成6个不同种类的假奶样品集合，其中糊精、淀粉在掺假奶中的含量为0.15％～0.45％；硝酸铵、尿素和三聚氰胺的含量分别为700～2100，524～1572与365.5～1096.5 mg · kg -1，以保证掺假奶中凯氏定氮法测得的蛋白含量不低于3％。所有样本的近红外光谱均经过标准正态变换（SNV）预处理。将3个真奶样品集合和6个假奶样品集合进行不同的组合并对其采用改进与简化的K最邻近结点算法（IS-KNN ）和改进与简化的支持向量机法（ν-SVM ）建立了判别糊精、淀粉、三聚氰胺、尿素、硝酸铵这5类掺假物质的近红外判别模型，探寻掺假物质的浓度与识别正确率之间的关系。结果表明IS-KNN和ν-SVM两种方法对含三聚氰胺、尿素、硝酸铵的掺假牛奶的平均判别正确率分别在49.55％～51.01％，61.78％～68.79％与68.25％～73.51％区间波动，说明在该研究的掺假物浓度范围内，很难用近红外模型良好区分不同类型伪蛋白的掺假奶；IS-KNN和ν-SVM 两种方法对含淀粉的掺假牛奶的判别正确率分别为92.33％与93.66％、对含糊精的掺假牛奶的平均判别正确率分别为77.29％与85.08％。从整体结果上来看ν-S V M法进行建模判别的结果大部分优于IS-K N N法进行建模判别的结果。对判别正确率与样品中掺假物质的含量水平分析表明近红外光谱结合非线性模式识别方法能良好地区分掺假奶中含量较高（0.15％～0.45％）的糊精和淀粉，而对含量偏低的三聚氰胺等伪蛋白的判别效果不佳，说明近红外光谱技术不适于鉴别牛奶中含量低于0.1％的掺假物质。
以287例上海及上海週邊地區牧場的生鮮奶作為真奶樣本集組成3箇真奶樣品集閤，配製瞭526例含有糊精（或澱粉）＋三聚氰胺（或尿素、或硝痠銨）的摻假牛奶形成6箇不同種類的假奶樣品集閤，其中糊精、澱粉在摻假奶中的含量為0.15％～0.45％；硝痠銨、尿素和三聚氰胺的含量分彆為700～2100，524～1572與365.5～1096.5 mg · kg -1，以保證摻假奶中凱氏定氮法測得的蛋白含量不低于3％。所有樣本的近紅外光譜均經過標準正態變換（SNV）預處理。將3箇真奶樣品集閤和6箇假奶樣品集閤進行不同的組閤併對其採用改進與簡化的K最鄰近結點算法（IS-KNN ）和改進與簡化的支持嚮量機法（ν-SVM ）建立瞭判彆糊精、澱粉、三聚氰胺、尿素、硝痠銨這5類摻假物質的近紅外判彆模型，探尋摻假物質的濃度與識彆正確率之間的關繫。結果錶明IS-KNN和ν-SVM兩種方法對含三聚氰胺、尿素、硝痠銨的摻假牛奶的平均判彆正確率分彆在49.55％～51.01％，61.78％～68.79％與68.25％～73.51％區間波動，說明在該研究的摻假物濃度範圍內，很難用近紅外模型良好區分不同類型偽蛋白的摻假奶；IS-KNN和ν-SVM 兩種方法對含澱粉的摻假牛奶的判彆正確率分彆為92.33％與93.66％、對含糊精的摻假牛奶的平均判彆正確率分彆為77.29％與85.08％。從整體結果上來看ν-S V M法進行建模判彆的結果大部分優于IS-K N N法進行建模判彆的結果。對判彆正確率與樣品中摻假物質的含量水平分析錶明近紅外光譜結閤非線性模式識彆方法能良好地區分摻假奶中含量較高（0.15％～0.45％）的糊精和澱粉，而對含量偏低的三聚氰胺等偽蛋白的判彆效果不佳，說明近紅外光譜技術不適于鑒彆牛奶中含量低于0.1％的摻假物質。
이287례상해급상해주변지구목장적생선내작위진내양본집조성3개진내양품집합，배제료526례함유호정（혹정분）＋삼취청알（혹뇨소、혹초산안）적참가우내형성6개불동충류적가내양품집합，기중호정、정분재참가내중적함량위0.15％～0.45％；초산안、뇨소화삼취청알적함량분별위700～2100，524～1572여365.5～1096.5 mg · kg -1，이보증참가내중개씨정담법측득적단백함량불저우3％。소유양본적근홍외광보균경과표준정태변환（SNV）예처리。장3개진내양품집합화6개가내양품집합진행불동적조합병대기채용개진여간화적K최린근결점산법（IS-KNN ）화개진여간화적지지향량궤법（ν-SVM ）건립료판별호정、정분、삼취청알、뇨소、초산안저5류참가물질적근홍외판별모형，탐심참가물질적농도여식별정학솔지간적관계。결과표명IS-KNN화ν-SVM량충방법대함삼취청알、뇨소、초산안적참가우내적평균판별정학솔분별재49.55％～51.01％，61.78％～68.79％여68.25％～73.51％구간파동，설명재해연구적참가물농도범위내，흔난용근홍외모형량호구분불동류형위단백적참가내；IS-KNN화ν-SVM 량충방법대함정분적참가우내적판별정학솔분별위92.33％여93.66％、대함호정적참가우내적평균판별정학솔분별위77.29％여85.08％。종정체결과상래간ν-S V M법진행건모판별적결과대부분우우IS-K N N법진행건모판별적결과。대판별정학솔여양품중참가물질적함량수평분석표명근홍외광보결합비선성모식식별방법능량호지구분참가내중함량교고（0.15％～0.45％）적호정화정분，이대함량편저적삼취청알등위단백적판별효과불가，설명근홍외광보기술불괄우감별우내중함량저우0.1％적참가물질。
In the present work ,two hundred and eighty seven raw milks collected from pastures in Shanghai and surrounding ar-eas of Shanghai were used as true milk samples and divided into three true milk sets .Five hundred and twenty six adulterated milk samples ,which contained dextrin (or starch) mixed with melamine (or urea ,or ammonium nitrate) ,were prepared as six different adulterated milk sets .The concentrations of these adulterants in the adulterated milks were designed to be 0.15% ~0.45% (starch or dextrin) ,700~2 100 mg · kg -1 (ammonium nitrate) ,524~1 572 mg · kg -1 (urea) ,and 365.5~1 096.5 mg · kg -1 (melamine) to guarantee the protein content of adulterated milks detected by Kjeldahl method not lower than 3% .All the near infrared spectra (NIR) of the samples should have a pretreatment of normal variable transformation (SNV) before they were used to build discriminating models .The three true milk sets and six adulterated milk sets were combined in different ways in order to build NIR models for discriminating different kinds of adulterants (i .e .,dextrin ,starch ,melamine ,urea and ammo-nium nitrate) based on simplified K-nearest neighbor classification algorithm (IS-KNN) and an improved and simplified of sup-port vector machine (ν-SVM ) method .The relationship between mass concentration of the adulterants and the rate of correct discrimination was also investigated .The results show that the average discrimination accuracy of IS-KNN andν-SVM for identi-fying melamine ,urea and ammonium nitrate were in the region of 49.55% to 51.01% ,61.78% to 68.79% and 68.25% to 73.51% ,respectively .Therefore within the concentration regions designed in this study ,it is difficult to distinguish different kinds of pseudo proteins by NIR spectroscopy .However ,the average accuracy of IS-KNN andν-SVM for identifying starch and dextrin are 92.33% and 93.66% ,77.29% and 85.08% ,respectively .Most discrimination results of ν-SVM are better than those of IS-KNN .The correlative analysis between the discrimination accuracy rate and the content levels of the adulterants indi-cated that near infrared spectroscopy combined with non-linear pattern recognition methods can distinguish dextrin and starch in milks with higher concentration levels (>0.15% ) ,but do not work well on identifying the adulterants with lower concentrations such as melamine (365.5 to 1 096.5 mg · kg -1 ) ,urea (524 to 1 572 mg · kg -1 ) ,ammonium nitrate (700 to 2 100 mg · kg -1 ) . Therefore near Infrared Spectroscopy is not suitable for identifying the adulterants with concentrations are below 0.1% .
