市售酱油理化特性分析和模式识别研究  

Physicochemical Properties and Pattern Recognition of Commercial Soy Sauce

作　　者：于寒松[1,2] 李松[1,2] 王玉华[1,2] 朴春红[1,2] 刘俊梅[1,2] 任大勇[1,2] 代伟长[1,2] 胡耀辉[1,2] 

机构地区：[1]吉林农业大学食品科学与工程学院,长春130118 [2]国家大豆产业技术研发中心加工研究室,长春130118

出　　处：《中国调味品》2015年第4期50-55,共6页China Condiment

基　　金："十二五"国家支撑项目(2012BAD34B03)子课题(2012BAD34B03-03);现代农业产业技术体系建设专项资金项目(CARS-04)

摘　　要：检测了25种市售酱油的7项理化指标,根据检测结果分析了酱油的理化特性,并进行酱油的模式识别研究,结果将25种酱油聚类为品质较好、品质一般和老抽酱油3个子类,依据7个理化指标建立了3个主成分模型,并得到了酱油的2个典型判别方程:D1=-1.465X1+0.426X2-2.706X3+0.254X4+0.055X5+0.516X6+0.129X7-4.486,D2=1.804X1+0.468X2-2.197X3+0.088X4+0.024X5+0.046X6-0.024X7-7.325,总判别正确率为100%,为酱油品质的控制和改进提供了理论基础和参考方法。Seven Physicochemical indexes are determined in 25 kinds of soy sauce. The physicochemical properties are analyzed depend on the detection results. Then 25 kinds of soy sauce is classified into three groups： better quality soy sauce, general quality soy sauce and dark soy sauce by pattern recognition method. Three principal component models are gotten according to seven physicochemical indexes, and two typical discrimination equations of soy sauce are written as D1 = -1. 465X1 ＋0. 426X2 -2. 706X3 nu 0. 254X＋0. 055X6 ＋0. 516X6＋0. 129X7 -4. 486 and D2 = 1. 804X1 ＋0. 468X2 -2. 197X3 ＋0. 088X＋0. 024Xs ＋ 0. 046X6-0. 024X7 -7. 325, with the total accurate rate of 100M. The equation could be used as a tool and useful method for quality control and improvement of soy sauce.

关 键 词：酱油 理化特性 主成分分析 聚类分析 判别分析 

分 类 号：TS264.21[轻工技术与工程—发酵工程]