老陈醋陈酿过程中不挥发酸和氨基酸态氮的近红外光谱技术分析

应用近红外光谱高效、绿色、无损的检测技术，结合改良的偏最小二乘法（MPLS）建立了不挥发酸、氨基酸态氮（AA-N）的近红外定量分析模型。收集了多个具有代表性的样品为定标样品，通过不同的数学处理和散射校正方法，筛选出预测不挥发酸、AA-N含量的较优模型。结果表明，采用原始光谱和原始散射方法获得的模型效果最好。不挥发酸模型和AA-N模型定标相关系数（RSQ）为0.986 0、0.974 0，定标标准误差（SEC）为0.084 8、0.013 6，内部交叉验证相关系数（1-VR）为0.979 5、0.960 1，内部交叉验证标准差（SECV）为0.091 7、0.014 4。经外部验证，不挥发酸、AA-N的预测值与化学值均有较高的相关性（＞0.900 0），平均误差分别为0.072 g/100 mL和0.020 g/100 mL，在实际应用中可获得良好的预测效果，表明利用该技术快速评价陈酿期半成品的质量是可行的。

Analysis of non-volatile acid and amino acid nitrogen in aging process of aged vinegar by near infrared spectroscopy

Using efficient,green,non-destructive near-infrared spectroscopy technology,combined with modified partial least squares(MPLS),a near-infrared quantitative analysis model of non-volatile acids and amino acid nitrogen(AA-N)was established.Multiple representative samples were collected as calibration samples,and a better model for predicting the content of non-volatile acids and AA-N was selected through different mathematical processing and scattering correction methods.The results showed that the model obtained by the original spectrum and the original scattering method had the optimal effect.The regression squared(RSQ)of non-volatile acid model and AA-N model were 0.9860 and 0.9740,the standard error of the calibration(SEC)were 0.0848 and 0.0136,the regression squared of the cross validation(1-VR)were 0.9795 and 0.9601,and the standard error of cross-validation(SECV)were 0.0917 and 0.0144.After external verification,the predicted values of non-volatile acids and AA-N had a high correlation with chemical values(>0.9000).The average error was 0.072 g/100 ml and 0.020 g/100 ml,respectively.Good prediction effect could be obtained in practical application,and it showed that it was feasible to use the technology to quickly evaluate the quality of semi-finished products during the aging period.