为了找出一种快速、准确鉴别黄酒酒龄和产地的方法，本研究采用超快速气相色谱电子鼻采集不同酒龄和产地黄酒样品的嗅觉指纹信息，采用主成分分析对嗅觉指纹信息进行降维处理，并采用判别因子分类法建立黄酒酒龄和产地鉴别模型。经主成分分析降维后的1年陈和3年陈基酒样品点落在相近的区域内，7年陈和9年陈基酒样品点落在相近的区域内，陈酿时间5年是基酒风味变化的一个转折点；判别因子分类法所建黄酒酒龄鉴别模型的鉴别正确率为100%。主成分分析降维后，绍兴黄酒与上海黄酒、浙江嘉善黄酒和江苏南通黄酒差异明显，进一步采用判别因子分类，4个产地黄酒的鉴别正确率均为100%。研究表明，超快速气相色谱电子鼻结合化学计量学方法建立的模型可以较好地应用于黄酒酒龄和产地的鉴别。

This study was performed to establish a method that can quickly and accurately identify the age and origin of Chinese rice wine. A nultra-fast gas chromatography-based electronic nose (E-nose) was used to collect olfactory fingerprints of wine samples from different age and origins. Dimension reduction of the olfactory fingerprints was performed using principal component analysis (PCA) followed by discriminant factor analysis (DFA), to establish a model for the purpose of this study. After dimension reduction by PCA, the points of one-year-old and three-year-old base wine samples fell in similar regions, as did the points of seven-year-old and nine-year-old base wine samples; the results indicated that the fifth year was a turning point in the aging process of base wine. The correct identification rate model built using DFA for rice wine was 100%. After the dimension reduction by PCA, obvious differences were observed between rice wine samples from Shaoxing and Shanghai/Jiashan/Nantong. The DFA models for the identification of geographical origin showed that the correct identification rates were 100%. The results demonstrate that the model built using an E-nose in combination with chemometric methods could be used for the identification of age and geographical origin of Chinese rice wine.

国家自然科学基金资助项目（21105065）；全国优秀博士学位论文作者专项资金资助项目（201059）