摘要
风味是鉴别汾酒品牌最重要的标准之一。针对传统品酒师品评的主观性及现阶段光谱仪、色谱仪等大型仪器检测汾酒的局限性,实验采用表面声波型电子鼻zNose^(TM)对6种不同品牌的汾酒酒样进行指纹图谱采集。通过比较指纹图谱差异,对酒样特征峰进行提取,利用主成分分析和判别因子分析对数据进行分析,并用贝叶斯判别函数验证其准确率,采用概率神经网络建立了识别模型。结果表明,主成分分析和判别因子分析都能对不同品牌汾酒进行区分,且判别因子分析法的区分效果优于主成分分析法,建立的概率神经网络模型其识别率达到100%。研究发现,表面声波型电子鼻zNose^(TM)对不同品牌汾酒具有较好的鉴别和分类能力。
Flavor is one of the most important criteria to identify Fenjiu brand.In view of the subjectivity of traditional tasters and the limitations of current instruments such as spectrometer and chromatograph,the Surface Acoustic Wave zNose^TM was used to collect fingerprints of six different brands of Fenjiu.By comparing the difference of the fingerprint,characteristic peaks were extracted.The extracted feature data were analyzed by principal component analysis and discriminant factor analysis,and its accuracy was verified by Bayes discriminant function,and probabilistic neural network were utilized to establish the identification models.The results showed that the principal component analysis and discriminant factor analysis could distinguish different brands of Fenju,and the latter method was better than the former method.The probabilistic neural network model was established and the recognition rate was up to 100%.The study suggested that the Surface Acoustic Wave zNose^TM had good identification and classification ability for different brands of Fenjiu
作者
邓霞
李臻峰
王辉
宋飞虎
张鑫
DENG Xia;LI Zhen-feng;WANG Hui;SONG Fei-hu;ZHANG Xin(School of Mechanical Engineering,Jiangnan University,Wuxi 214122, China;Jiangsu Key Laboratory of Advanced Food Manufacturing Equipment and Technology,Wuxi 214122, China;Technology Center of Shanxi Xinghuacun Fenjiu Factory Co. Ltd, Fenyang 032200, China)
出处
《食品与发酵工业》
CAS
CSCD
北大核心
2017年第11期207-211,共5页
Food and Fermentation Industries
基金
国家自然科学基金(515082290)
江苏省产学研联合创新资金(BY2014023-32)
江南大学基本科研青年基金项目(1072050205134580)
江苏省食品先进制造装备技术重点实验室开放课题(BM2013001)
江苏省食品先进制造装备技术重点实验室开放课题(FM-201406)
关键词
电子鼻
不同品牌汾酒
指纹图谱
概率神经网络
zNose^TM
different brands of Fenjiu
fingerprint
probabilistic neural network
作者简介
第一作者：硕士研究生;李臻峰教授为通讯作者，E—mail：474639259@qq．com
