基于人工神经网络模拟啤酒酿造过程中糖度及乙醇浓度的变化  

Prediction of sugar density and alcohol content during beer fermentation based on artificial neural network

作　　者：黄奕雯[1] 戴玉杰[1] 钟成[1] 李清亮[1] 贾士儒[1] 郝俊光 

机构地区：[1]工业发酵微生物教育部重点实验室,天津科技大学,天津300457 [2]啤酒生物发酵工程国家重点实验室(筹),山东青岛266061

出　　处：《中国酿造》2013年第1期25-28,共4页China Brewing

基　　金：国家重点基础研究发展计划‘973计划’(No.2010CB735706);啤酒生物发酵工程国家重点实验室开放基金(No.K2012006)

摘　　要：建立BP神经网络模型模拟啤酒酿造过程中糖度变化和乙醇浓度变化。将啤酒酿造过程中的发酵温度、麦汁浓度、接种量及发酵时间作为输入数据,将糖度变化和乙醇浓度的变化作为输出数据,运用BP神经网络建立啤酒酿造过程的模型。使用此模型模拟了主酵温度8℃、麦汁浓度11°P、接种量为2×107个/mL时糖度变化和乙醇浓度变化,结果糖度预测的均方根误差为2.66%,乙醇浓度预测的均方根误差为14.60%。结果表明,使用此模型能够准确预测啤酒酿造过程糖度变化和乙醇浓度的变化。The back-propagation （BP） neural network was used to predict sugar density and alcohol content during beer fermentation. A BP neural net- work model of beer fermentation was established using fermentation temperature, sugar density of wort, inoculum and fermentation time as input val- ues, and sugar density and alcohol content during beer fermentation as output values. After the model was trained, the sugar density and alcohol con- tent were predicted for the beer fermentation conducted at 8~C with 1 l^P wort and an inoculum of 2~107cells/ml. The root mean square error of pre- diction of sugar density and alcohol content were 2.66% and 14.60%, respectively. The results showed that the model could be applied for the predic- tion of sugar density and alcohol content during beer fermentation.

关 键 词：糖度 乙醇浓度 BP神经网络 

分 类 号：TS262.5[轻工技术与工程—发酵工程]