基于近红外光谱技术的白酒酒醅在线监测研究

为实现白酒生产过程中出窖酒醅智能化配粮，该文开发了一套在线近红外光谱监测系统，实时在线检测出窖酒醅配粮前的水分、淀粉及酸度，实时检测结果转化为4~20 mA电控输入信号，解决酒醅智能配粮问题。通过将在线近红外分析系统与台式近红外分析仪相结合，在四川宜宾六尺巷酒厂开展酒醅在线监测与配粮智能化应用。结果表明，台式仪器检测水分、淀粉、酸度的平均误差分别为-0.25%、0.38%和0.29 mmol/10 g，检测水分、淀粉、酸度的预测标准差(SEP)分别为0.60%、0.75%和0.18 mmol/10 g；在线仪器检测酒醅水分、淀粉、酸度的平均误差分别为-0.75%、0.48%、0.29 mmol/10 g，检测水分、淀粉、酸度的SEP分别为0.66%、0.97%、0.22 mmol/10 g。相比台式近红外分析结果，在线结果的平均误差及ＳＥＰ均有所放大，但在线分析的准确度能满足酒醅在线配粮控制的精度要求。

Online Monitoring of Fermented Grains Parameters for Chinese Liquor Brewing Based on Near Infrared Spectroscopy

In order to realize the intelligent grain preparation of fermented grains in the process of liquor production,an online near infrared(NIR) spectroscopy system was developed in this paper for the real time detection of the water,starch and acid contents of fermented grains.The online signal was converted as 4-20 mA current for the intelligent allocation of grains in the process of Chinese liquor brewing.This real time detection system was carried out at liuchixiang Chinese liquor brewing plant in Yibin,Sichuan province.The average errors for Lab NIR analysis of water,starch,acid in fermented grains were -0.25%,0.38% and 0.29 mmol/10 g,while the standard errors of prediction(SEP) of water,starch and acid were 0.60%,0.75% and 0.18 mmol/10 g,respectively.The average errors for online NIR quantitative analysis of water,starch,acid in fermented grains were -0.75%,0.48% and 0.29 mmol/10 g,while the SEP of water,starch and acid were 0.66%,0.97%and 0.22 mmol/10 g,respectively.Compared with the results of Lab NIR analysis,the average errors and standard deviations of this online analysis were amplified,but the precision of online NIR analysis could be qualified for grains addition in the automation of the whole production line.