基于二维相关近红外光谱的白酒酒龄鉴别

目的:杜绝以低龄酒冒充高龄酒,实现对白酒酒龄的判别。方法:利用近红外光谱仪采集白酒样品的近红外透射光谱,以酒龄作为外扰,构建各白酒样品的同步和异步二维相关谱;在此基础上,提取每个样品的自相关谱,并结合马氏距离法建立白酒酒龄判别模型。结果:1400~1800 nm波段范围内包含白酒样品随酒龄变化的有效信息;模型校正集判别正确率为93.3%,预测集判别正确率为92.0%。结论:二维相关谱结合马氏距离法能够对白酒酒龄进行有效鉴别。

Identification of the age of Baijiu based on two-dimensional correlation near infrared spectroscopy

Objective: To avoid young liquor as old liquor and realize the discrimination of Baijiu age. Methods: The near infrared transmission spectra of Baijiu samples were collected by near infrared spectrometer, and the synchronous and asynchronous two-dimensional correlation spectra of Baijiu samples were constructed with the wine age as the external disturbance. On this basis, the autocorrelation spectrum of each sample was extracted, and the age discrimination model of Baijiu was established by combining Markov distance method. Results: The wave band from 1400 nm to 1800 nm contained the effective information about the changes of Baijiu samples on age. The discrimination accuracy of model correction set was 93.3%, and that of prediction set was 92.0%. Conclusion: Two dimensional correlation spectrum combined with Mahalanobis distance method can effectively identify the age of Baijiu.