拉曼光谱快速监测荔枝酒发酵过程酒精度的研究

Rapid monitoring of alcohol content in fermentation process of litchi wine by Raman spectroscopy

酒精度是荔枝酒发酵过程及成品酒质量控制的关键因素,该研究使用拉曼光谱仪快速采集样本信号,并对拉曼光谱数据进行2阶导预处理后,分别采用偏最小二乘法(PLS)和向量角偏最小二乘法(VAPLS)建模,同时对模型效果进行对比分析。结果表明,VAPLS酒精度预测模型效果最好,该模型预测值与真实值的相关系数(R^(2))为0.9931,均方根误差(RMSE)为0.2856,验证集预测相对误差为-5.0%~1.6%,优于PLS建模方法。所建分析方法简便快速,能满足生产中荔枝酒酒精度的快速检测精度要求,并可以拓展实现酒类发酵过程中多个性能指标的同时分析。

Alcohol content is a key factor in the fermentation process of litchi wine and the quality control of finished wine.The sample signals were quickly collected using Raman spectra,based on the samples after 2 derivative Raman spectra data pretreatment,partial least square(PLS)and vector angle partial least square(VAPLS)were used respectively to build the model and the modeling effect was analyzed and compared.The results showed that VAPLS alcohol content prediction model was the optimal,the correlation coefficient(R2)of the predicted value and the real value was 0.9931,the root mean square error(RMSE)was 0.2856,the validation set prediction relative error was-5.0%-1.6%,better than PLS modeling method.The established analytical method was simple and convenient,could meet the requirement of rapid detection of alcohol content of litchi wine in production,and could be extended to realize the simultaneous analysis of multiple performance indicators in the production process of fermented alcoholic drink.