摘要
应用近红外光谱(NIRS)技术建立了大米食味品质分析与种类快速鉴别的方法。提取了102份粉碎后大米样品的近红外光谱,采用偏最小二乘法(PLS)建立了大米水分、蛋白质和直链淀粉定量分析模型,对模型预测结果的准确性进行了评价。预测模型的内部交叉验证决定系数(R2)分别为:0.992、0.9792和0.9736;内部交叉验证标准差(RMSECV)分别为:0.141、0.201和0.209;模型外部验证决定系数(R2)分别为0.9861、0.912和0.9373;外部验证标准差(RMSEP)分别为0.179、0.206和0.243。通过计算样品的近红外光谱图之间的欧氏距离来反映不同样品间的差异,对不同属性和不同产地的大米进行了定性聚类分析,种类识别准确率达到100%。结果表明,NIRS分析技术可以用于对大米品质和种类的快速无损检测。
The methods for rice eating quality analysis and category identification of rice were established based on the near infrared spectroscopy（NIRS）.A total number of 102 rice samples from different origins with different category were collected,and the crushed rice samples were applied for near infrared spectra collection.Quantitative analysis models of rice moisture,protein and amylose were developed by partial least square（PLS）.The accuracy of the prediction result was evaluated.Internal cross-validation decided coefficient（R2） of prediction model was 0.992,0.9792 and 0.9736,respectively.Internal cross-validation RMSECV was 0.141,0.201and 0.209,respectively.External validation determination coefficient（R2） of model was 0.9861,0.912 and 0.9373,respectively.External validation RMSEP was 0.179,0.206 and 0.243,respectively.The differences among samples could be tested by calculation of the euclidean distance between near infrared spectra of samples.The differences of category of rice samples were evaluated by cluster analysis.The accurate for the category identification reached to 100%.Results showed that NIRS,a rapid and non-destructive analytical technique,can be used for rice quality and category analysis.
出处
《食品工业科技》
CAS
CSCD
北大核心
2012年第3期322-325,共4页
Science and Technology of Food Industry
基金
安徽省优秀青年基金(08040106804)
关键词
近红外光谱
大米
偏最小二乘法
品质分析
聚类分析
种类鉴别
near infrared reflectance spectra（NIRS）
rice
partial least squares algorithm（PLS）
quality analysis
clustering analysis
category identification
