近红外光谱法结合C-SVM及ν-SVM方法快速无损鉴别淀粉种类  

Non-destructive determination of starch category by using C-SVM and ν-SVM on NIR spectroscopy

作　　者：邹婷婷[1] 窦英[2] 王莹[3] 刘野[1] 段紫怡[1] 张秋晨[1] 

机构地区：[1]北京工商大学食品学院,北京100048 [2]天津科技大学理学院化学系,天津300457 [3]吉林省产品质量监督检验院,吉林长春130022

出　　处：《食品工业科技》2013年第17期317-319,共3页Science and Technology of Food Industry

摘　　要：不同厂家的红薯淀粉、马铃薯淀粉和玉米淀粉共112个样品,利用近红外光谱技术对淀粉种类进行识别。分别采用马氏距离判别法、C-支持向量机(C-SVM)、ν-支持向量机(ν-SVM)建立淀粉种类鉴别的近红外光谱模型;并对比多元散射矫正、平滑、一阶微分、二阶微分等多种预处理方法后的建模结果。结果表明:同时使用平滑、多元散射矫正、微分三种预处理方法后,C-SVM和ν-SVM分类模型的效果最佳;训练集交叉验证正确率均为98.72%,测试集正确率均达到100%。实验结果表明,该模型快速准确无损的鉴别淀粉种类是可行的。A method of starch category analysis was developed by using Near-infrared （NIR）spectroscopy.All the 112 samples were obtained from different manufacturers of sweet potato starch, potato starch and corn starch.The N IR models were established using mahalanobis distance discriminant, C-Support vector machine（C-SVM） and υ-Support vector machine （υ-SVM）.The various different pretreated methods （multiplicative scatter correction （MSC） ,smooth,first-derivation and second-derivation of spectra data were applied.The results indicated the C-SVM and υ-SVM models had most advantages simultaneously used MSC,smooth and derivation.The correct ratio of the training set and the testing set was 98.72% and lO0%.The results showed the method for simultaneous, non-destructive analysis of starch category was reliable.

关 键 词：近红外光谱技术 淀粉 马氏距离判别 支持向量机(SVM) 定性分析 

分 类 号：TS237[轻工技术与工程—粮食、油脂及植物蛋白工程]