摘要
本研究对来自于云南不同厂家的55个样品进行感官审评与理化分析,将滋味分为鲜醇型、浓醇型、醇厚型及其他型四种类型,并以理化分析数据为基础,进行主成分分析(PCA),筛选滋味评价因子,然后对各滋味类型分别建立贝叶斯(Bayes)预测模型。通过预测模型进行回判,各组模型正确率分别为100%、100%、75%和100%,回判结果良好,从而达到量化滋味类型的目的。
55 samples of Kungfu black tea from different manufacturers in Yunnan province were collected in this study for physical and chemical analysis and Sensory evaluation,with four taste types estimating in the fresh mellow,mellow concentration,mellow and others. Then Principal Component Analysis(PCA)were based on the data of physical and chemical analysis. Bayes prediction model was established by different flavors. By the model,the accuracy of each group model was desired effect of 100%,100%,75% and 100%. As a result,the purpose quantifying flavor type was achieved in this study.
作者
熊一帆
安会敏
欧行畅
张杨波
黄建安
刘仲华
李适
李文雄
江鸿健
XIONG Yi-fan;AN Hui-min;OU Xing-chang;ZHANG Yang-bo;HUANG Jian-an;LIU Zhong-hua;LI Shi;LI Wen-xiong;JIANG Hong-jian(Key Laboratory of Tea Science,Ministry of Education,Hunan Agricultural University,Changsha 410128,China;National Research Center of Engineering Technology for Botanical Functional Ingredients,Changsha 410128,China;Lincang City,Yunnan Province Biological Industry Development Office,Lincang 677000,China)
出处
《食品工业科技》
CAS
北大核心
2019年第24期251-255,269,共6页
Science and Technology of Food Industry
关键词
滇红
感官审评
理化分析
滋味类型
贝叶斯模型
Yunnan province black tea
sensory evaluation
flavor type
Bayes model