We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
偏最小二乘方法在不同白酒香型判别分析中的应用.
- Authors
魏泉增; 范江涛; 刘嘉玲; 宋丽君; 李瑞; 谢玉锋
- Abstract
In order to build a method for Baijiu (Chinese liquor) flavor types identification, using DB-WAX as chromatographic column, the samples were analyzed after adding internal standard by GC method. The results showed that correlation coefficient of volatile compounds was R²>0.999. The detection limit was less than 5.11 mg/L. The relative standard deviation (RSD) of precision test was less than 4.11%. The average recovery rates of samples spiked was 80.20%-97.93%. Partial least square (PLS) method was used to modeling and predicting the data of different flavor liquor. The samples of different flavor liquor could be separated in the PLS score diagram, the model was verified by using the data of training set and verification set, and the results showed that the prediction accuracy was high. The obtained root mean square error (RMSEE) of correction and root mean square error of prediction (RMSEP) values also showed the PLS model had strong predictability for discriminating the Baijiu flavor types. The method had high accuracy in determining Baijiu flavor analysis.
- Subjects
STANDARD deviations; FLAVOR; FORECASTING; LEAST squares; STATISTICAL correlation; DETECTION limit; FOOD aroma
- Publication
China Brewing, 2020, Vol 39, Issue 10, p183
- ISSN
0254-5071
- Publication type
Article
- DOI
10.11882/j.issn.0254-5071.2020.10.034