摘要
提出了用欧马距离评价混合体系指纹图谱的方法,并结合主成分分析计算了6种香精标准样品和2种掺兑样品、6种香精标准样品和8种无水乙醇稀释样品GC/MS指纹图谱的欧马距离。结果表明,欧马距离能较好区分6种香精标准样品与2种香精的混合物、6种香精标准样品和8种乙醇稀释样品。欧马距离算法结合主成分分析可用于香精样品指纹图谱识别。
A method for evaluating the fingerprint chromatographs of a mixed system by Euclidean- Mahalanobis distance was put forward. The Euclidean-Mahalanobis distances of GC/MS fingerprint chromatographs of six standard samples and two mixture samples, and six standard samples and eight absolute alcohol-diluted samples were calculated by combining with principal component analysis. The results showed that the standard samples were well discriminated from mixture or diluted samples by their Euclidean- Mahalanobis distances. It suggested that Euclidean-Mahalanobis distance combining with principal component analysis could be used to recognize flavor fingerprint chromatographs.
出处
《烟草科技》
EI
CAS
北大核心
2011年第2期52-57,共6页
Tobacco Science & Technology
关键词
指纹图谱
模式识别
距离
算法
香精香料
Fingerprint chromatograph
Pattern recognition
Distance
Algorithm
Flavor
