蜂蜜中还原糖组分的近红外光谱应用研究

为探索快速测定还原糖含量的方法，提出了用傅立叶变换近红外光谱技术结合偏最小二乘法(PLS)建立近红外光谱与蜂蜜还原糖含量的数学模型并进行预测。通过光谱扫描还原糖含量在61.3%～75.22%范围的蜂蜜样本，选择11992.1～7494.6cm-1波数范围、二阶导数、及10个因子数进行光谱预处理，偏最小二乘法(PLS)交叉验证。结果表明，模型的校正决定系数(Rcal)、校正均方差(RMSEE)、交叉验证决定系数(RCV)、交叉验证均方差(RMSECV)分别为99.71%、0.27%、98.44%、0.45%。用该模型对验证集样本进行预测并统计分析，表明预测值与测定值无显著差异。因此，用该方法快速准确定量分析大批蜂蜜中的还原糖含量具有重要意义。

Prediction Analysis of Reducing Sugar Content in Honey Using Fourier Transform Near-infrared Spectroscopy

For exploring a rapid determination of reducing sugars, a correlation model of near-infrared spectrum and reducing sugar content in honey was established and the reducing sugar contents in honey samples were predicted by using Fourier transform near-infrared spectroscopy (FT-NIR) combined with partial least squares (PLS). The spectra of honey sample with reducing sugars content between 61.3% and 75.22% were scanned and the 11992.1 cm-1 to 7494.6 cm-1 spectrum was selected. Through preprocess of second derivative and 10 factors, the model was established by using PLS. The results of cross validation showed that the model of the calibration coefficient of determination (Rcal), root mean standard error of estimation (RMSEE), validation coefficient of determination (RCV), root mean standard error cross validation (RMSECV) are 99.71 %, 0.27 %, 98.44 % and 0.45 %, respectively. The model was used to verify samples and the statistical results showed that there is no significant difference between the predictive and chemical values. Altogether, this method is fast and accurate for quantitative analysis of reducing sugars content in mass honey samples.