基于电子舌技术定量分析黄酒理化指标  

Quantitative Analysis on Physical and Chemical Indexes of Rice Wines Based on Electronic Tongue

作　　者：汤海青[1] 吴维儿 王晓龙 TANG Haiqing;WU Weier;WANG Xiaolong(Department of Food Science,Zhejiang Pharmaceutical College,Ningbo,Zhejiang 315100,China;Technology Center,Ningbo Customs District,Ningbo,Zhejiang 315012,China)

机构地区：[1]浙江医药高等专科学校食品学院,浙江宁波315100 [2]宁波海关技术中心,浙江宁波315012

出　　处：《农产品加工》2020年第12期66-69,共4页Farm Products Processing

基　　金：浙江省教育厅一般科研项目(Y201534608);浙江省大学生科技创新活动计划暨新苗人才计划项目(2015R435006);浙江省大学生科技创新活动计划暨新苗人才计划项目(2016R435006)。

摘　　要：为快速有效地检测黄酒理化指标,使用电子舌对黄酒的滋味特征进行识别,结合理化检测手段,对黄酒样品分别建立偏最小二乘(PLS)和多元线性回归(MLR)的定量预测模型。结果表明,应用PLS和MLR对传感器信号与国标方法检测结果进行拟合,MLR模型的pH值、总酸、酒精度和氨基酸态氮的RPD值分别为3.8,2.9,2.9,2.8,R2均接近0.9,RMSEC和RMSEP比值均在0.8~1.2,建立的模型效果良好,可进行准确的定标和预测;MLR模型在准确性和稳定性方面优于PLS模型,更适合所用黄酒样本集的定标和预测。研究结果为应用电子舌对黄酒理化指标进行快速定量分析提供了理论和实践基础。In order to detect the physical and chemical indexes of rice wine quickly and effectively,the taste characteristics of rice wine were identified by electronic tongue,and the quantitative prediction models of partial least squares(PLS)and multiple linear regression(MLR)were established for rice wines samples respectively.The results showed that the RPD values of pH,total acid,alcohol and amino acid nitrogen of MLR model were 3.8,2.9,2.9,2.8,respectively,and the R2 was close to 0.9,and the ratios of RMSEC and RMSEP were between 0.8~1.2.The established model had good effect and could be accurately calibrated and predicted.The stability and accuracy of MLR model was better than the PLS moclel,which was more suitable for the calibration and prediction of the sample set of rice wines.The results of this study provided a theoretical and practical basis for the rapid quantitative analysis of the physical and chemical indexes of rice wine with electronic tongue.

关 键 词：黄酒 电子舌 偏最小二乘(PLS) 多元线性回归(MLR) 

分 类 号：TS261.7[轻工技术与工程—发酵工程]