近红外光谱技术快速检测露酒指标的研究

Study on Rapid Detection of Dew Wine Index by Near Infrared Spectroscopy

为实现露酒理化指标的快速检测,结合露酒检测的实际需要和相关检测标准的要求,运用近红外光谱技术,建立了露酒酒精度和总糖的快速检测模型。结果表明:通过筛选不同光谱预处理方法和波数范围,运用偏最小二乘法(PLS)建立模型,使用内部交叉验证得到最优模型,酒精度和总糖指标模型的相关系数(R2)值分别为0.9984和0.9934,预测模型线性良好,交叉验证均方根(RMSECV)分别为0.258和2.24,预测误差较小。使用盲样对露酒各指标的模型进行外部验证,酒精度和总糖的相对标准偏差分别为0.36%和2.01%,符合生产检测的需求,可运用于露酒的日常检测分析。

In order to realize the rapid detection of physical and chemical indexes of liqueur,combined with the actual needs of liqueur detection and the requirements of relevant detection standards,the near-infrared spectroscopy technology was used to establish the rapid detection model of liqueur precision and total sugar.The results showed that:by screening different spectral pretreatment methods and wave number range,the partial least squares(PLS)method was used to establish the model,and the internal cross validation was used to obtain the optimal model.The correlation coefficient(R2)values of alcohol and total sugar index models were 0.9984 and 0.9934,respectively.The prediction model had good linearity,and the root mean square of cross validation(RMSECV)was 0.258 and 2.24,respectively,with small prediction error.The relative standard deviations of alcohol content and total sugar were 0.36%and 2.01%,respectively,which met the needs of production detection and could be applied to the daily detection and analysis of dew wine.