鸭蛋清腥味评价模型的建立及其在酵母脱腥中的应用  

Establishment of Duck Egg White Fishy Smell Evaluation Model and Its Application in Yeast Deodorization

作　　者：黎海涛[1] 张婞 郭燕 陈晓燕 张小涛 侯焘 LI Haitao;ZHANG Xing;GUO Yan;CHEN Xiaoyan;ZHANG Xiaotao;HOU Tao(Zhejiang Pharmaceutical College,Ningbo 315100,China;College of Food Science and Technology,Huazhong Agricultural University,Key Laboratory of Environmental Food Science,Ministry of Education,Wuhan 430070,China)

机构地区：[1]浙江医药高等专科学校,浙江宁波315100 [2]华中农业大学食品科学技术学院,环境食品学教育部重点实验室,湖北武汉430070

出　　处：《食品科技》2021年第12期93-99,共7页Food Science and Technology

基　　金：2020年浙江省基础公益基金项目(LGC208070002)。

摘　　要：为降低鸭蛋清的腥味,基于电子鼻对不同腥味浓度鸭蛋清的响应结果,建立了腥味评价模型,并利用模型对酵母发酵脱除鸭蛋清腥味的工艺进行优化。结果表明,多元线性回归分析(MLR)、偏最小二乘回归分析(PLS)以及BP神经网络法均能建立较好的鸭蛋清腥味定量预测模型,三者建模集的决定系数R;和均方根误差RMSE分别为:0.97681、0.97526、0.99033和1.086、1.1656、0.7055。采用MLR得出的拟合方程对不同酵母添加量、反应温度和反应时间下的酵母脱腥试验条件进行优化,结合BP神经网络模型的验证实验,得出最佳脱腥条件为酵母添加量0.9 g/100 mL、反应温度40 ℃、反应时间2.5 h。In order to decrease the fishy smell of duck egg white,the fishy smell evaluation model was established based on the response values of different concentrations of duck egg white by the electronic nose.Then,duck egg white deodorization test by yeast fermentation was optimized by using this model.The results showed that the multiple linear regression (MLR),partial least-squares (PLS) regression and BP neural network could simulate great quantitative prediction models of duck eggs fishy smell,and the coefficient of determination R;and root mean square error RMSE of the three prediction models were:0.97681,0.97526,0.99033 and 1.086,1.1656,0.7055,respectively.The fitting equation of MLR was used in yeast fermentation deodorization test to optimize the experimental conditions of different yeast addition amount,reaction temperature and reaction time.Combined with validation experiments of BP neural network model,it was concluded that the optimum condition was 0.9 g/100 mL yeast,40 ℃ and 2.5 h.

关 键 词：鸭蛋清 腥味 模型预测 酵母 脱腥 

分 类 号：TS253.9[轻工技术与工程—农产品加工及贮藏工程]