基于BP神经网络模型的梨醋发酵产酸量的预测  

Prediction of total acid from pear vinegar by artificial neural network

作　　者：丁城[1] 刘璐[1] 胡勇[1] 

机构地区：[1]湖北工业发酵协同创新中心,湖北省食品发酵工程技术研究中心,教育部发酵工程重点实验室中心,湖北工业大学轻工学部,武汉430068

出　　处：《食品科技》2017年第2期273-276,共4页Food Science and Technology

基　　金：湖北省教育厅科学技术研究计划青年人才项目(Q20151412);湖北省自然科学基金项目(2015CFB678);湖北工业大学博士启动基金项目(Grant No.BSQD10036)

摘　　要：研究不同发酵条件(发酵温度、乙醇体积分数、接种量和梨的添加量)对梨醋总酸含量的影响,并且建立BP神经网络预测模型。神经网络模型能够根据发酵条件,准确预测总酸含量,总酸相关系数为0.986。在梨醋发酵中,发酵温度对总酸含量影响最大。与传统的模型比较,BP神经网络误差小,构建简单,拟合性高。In this research, the effect of different pretreatments（fermentation temperature, volume fractions of ethanol, inoculating amount and the dosage of pear） on total acid content of pear vinegar was investigated, and artificial neural network were applied for modeling of these parameters during fermentation. According to the fermentation conditions, BP artificial neural network model provided an accurate prediction method for total acid content of pear vinegar, moreover, total acid content with correlation coefficient of 0.986. In the pear vinegar fermentation, the temperature greatly affected the content of total acid. The results showed that the accuracy and goodness of modeling of the the content of total acid predicted by ANN were better than those predicted by the traditional modes.

关 键 词：梨醋 BP神经网络 总酸 发酵 

分 类 号：TS264.22[轻工技术与工程—发酵工程]