基于近红外光谱的藏区酥油脂肪和蛋白质含量快速检测分析

Rapid determination of fat and protein contents in ghee using near-infrared spectroscopy

为了建立藏区酥油中脂肪和蛋白质含量快速检测方法,以藏区不同产地酥油为研究对象,采用组合区间偏最小二乘法优选出脂肪和蛋白质的红外光谱特征波段,在此基础上分别对几种光谱预处理方法进行比较挑选出最佳预处理方法,最后建立脂肪和蛋白质的定量分析模型并对模型进行外部验证。结果表明:在脂肪和蛋白质的特征波段内采用SNV+数据归一化+二阶导数+S-G滤波法对光谱预处理后建立脂肪和蛋白质定量模型,模型相关系数(R)分别为0.994和0.997,交叉验证均方差(RMSECV)分别为4.09%和0.286%。对验证样本的预测值和实测值进行配对t检验,两种测量值之间没有显著性差异(P>0.05),表明建立的脂肪和蛋白质的近红外定量模型具有良好的预测能力,基于近红外光谱快速测定藏区酥油中脂肪和蛋白质的含量是可行的。

Rapid determination method of fat and protein contents in ghee was established using near - in frared spectroscopy with partial least squares. The ghee was collected from different regions in Tibet. The quantitative analysis model of fat and protein in those ghee samples was built with SNV, second deriva- tive, data normalization and S - G filter in the characteristic wavelength band using synergy interval par- tial least squares. Finally, the validation samples were employed to verify the model. The results showed that the correlation coefficients of the quantitative models of fat and protein were 0.994 and 0.997 respec- tively, and RMSECV were 4.09% and 0.286% respectively. The t test indicated that there was no sig- nificant difference （P 〉 0.05 ） between predicated value and measured value of validation samples. The established near- infrared quantitative model of fat and protein had good prediction ability, and it was feasible to rapidly determine the fat and protein contents in ghee using near- infrared spectroscopy.