近红外光谱在食醋品牌和贮藏年份鉴别中的应用研究

近红外光谱用于食醋品牌和食醋贮藏年份的鉴别。144个食醋样品用于食醋品牌的鉴别,其中114个为"恒顺"牌食醋,其余30个为山西"水塔"和天津"独流"品牌食醋;114个"恒顺"牌食醋用于食醋贮藏年份的鉴别;采用kennard-stone(KS)方法将样品划分为校正集和验证集,主成分分析(principal component analysis,PCA)和偏最小二乘判别分析(partial least squares-discriminant analysis,PLS-DA)用于食醋品牌和年份的鉴别;competitive adaptive reweighted sampling(CARS)方法用于食醋特征变量的筛选。结果表明,样品经PLS-DA方法后,食醋品牌的校正识别率为98.96%,验证识别率为95.83%;食醋年份的校正识别率为97.37%,验证识别率为100%;食醋光谱经CARS方法优选波长后结合PLS-DA方法进行判别,所有的识别率都为100%;近红外光谱结合CARS-PLS-DA方法可以用于食醋品牌和贮藏年份的鉴别.为食醋原产地保护和年份鉴别提供一种快速的参考方法。

Research on identification of vinegar brand and storing-year by near-infrared spectroscopy

Near-infrared spectroscopy （NIRS） was used to identify vinegar brand and storing-year in 144 vinegar samples, among which, 114 samples were ＂Hengshun＂ brand, and 30 samples belonged to ＂Shuita＂ and ＂Duliu＂. Kennard-stone method was adopted for the partitions of calibration and validation sets. The principal component analysis （PCA） and partial least squares- discrimination analysis （PLS-DA） were used for identification of vinegar brands and storage year. Competitive adaptive re-weighted sampling （CARS） was used for the selection of the feature variables. The results showed that the recognition rate of brands for calibration set was 98.96%, and 95.83% for validation set; the recognition rate of storing-year for calibration set was 97.37%, and 100% for validation set. When CARS-PLS-DA method was used for identification, 100% recognition rates were obtained for both brands and storage year in calibration and validation sets. It can offer a fast reference method for the country origin protection and storing-year identification of vinegar.