基于近红外透射光谱的大豆蛋白质和粗脂肪含量快速检测  

Rapid determination of soybean protein and crude fat content by near-infrared transmittance spectroscopy

作　　者：李琳琳[1] 金华丽[1] 崔彬彬[1] 王晓君[2] 

机构地区：[1]河南工业大学粮油食品学院,河南郑州450001 [2]河南工业大学教务处,河南郑州450001

出　　处：《粮食与油脂》2014年第12期57-60,共4页Cereals & Oils

基　　金：河南省重点科技攻关项目(1121023103795)

摘　　要：该研究采用近红外光谱分析技术分别建立大豆蛋白质和粗脂肪含量的近红外检测模型,比较了不同数学处理方法和去散射校正方法及不同回归技术对模型准确性的影响,选出最佳的近红外检测模型,并通过内部验证和外部验证对模型的预测能力进行评价。结果表明:所建立的大豆蛋白质和粗脂肪含量的近红外检测模型的内部验证相关系数分别为0.947 1、0.889 0,外部验证相关系数分别为0.962 2、0.915 5。说明所建立的大豆蛋白质和粗脂肪含量的近红外检测模型具有很好的预测能力,近红外光谱技术可用于大豆成分含量的快速检测。The model of determining soybean protein content and crude fat content were built with near–infrared spectroscopy,the influences of different mathematical treatments,scattering correction methods and different regression methods on the model were discussed,thus the optimum models were selected,then the predictive ability of models were evaluated by internal validation and external validation. The results showed that the internal cross–validation correlation coefficient of NIR model of soybean protein content and crude fat content was 0.947 1 and 0.889 0 respectively,the external validation correlation coefficient of those was 0.962 2 and 0.915 5 respectively. It demonstrated that the NIR model of soybean protein content and crude fat content had good prediction ability and NIR can be used for rapid detection of soybean ingredients content.

关 键 词：大豆 蛋白质 粗脂肪 近红外光谱分析技术 

分 类 号：S565.1[农业科学—作物学]