粳稻储藏食用品质近红外光谱快速测定技术研究  

Study on Determination of Storage Eating Quality of Rice by Near-infrared Spectroscopy

作　　者：林家永[1] 范维燕[1] 陆晖[1] 

机构地区：[1]国家粮食局科学研究院,北京100037

出　　处：《中国粮油学报》2011年第9期110-114,共5页Journal of the Chinese Cereals and Oils Association

基　　金：国家科技支撑计划(2006BAD08B07)

摘　　要：采用近红外光谱(NIRS)分析技术和化学计量方法建立了粳稻储藏品尝评分值的近红外分析模型,并对模型进行了预测准确性评价;在建立定标模型的过程中,探讨了不同光谱散射、数学等优化处理对定标模型的影响。结果表明:修正偏最小二乘法(MPLS)是建立粳稻储藏品尝评分值定标模型的最佳回归方法,所建立颗粒状和粉末状样品模型的定标相关系数(RSQ)分别为0.927 4和0.923 0,定标标准偏差(SEC)分别为2.347 9和2.539 1。定标模型具有较好的预测准确性。The predicted mathematic model of eating score of rice by the technique of near - infrared spectrosco- py （NIRS） was established. The result showed that the calibration model developed by the partial least square （MPLS） regression was the best one. The coefficient of correlation （RSQ） of granular samples and powdery samples were 0.927 and 0.9230 respectively; the standard errors of calibration （SEC） were 2. 3479 and 2. 5391 respectively. This NIRS method can be applied to predict the storage eating quality of rice.

关 键 词：稻谷 储藏食用品质 近红外光谱(NIRS) 定标模型 

分 类 号：TS227[轻工技术与工程—粮食、油脂及植物蛋白工程]