^1HNMR-偏最小二乘法-判别法分析椰奶的质量

Analysis of coconut milk based on ^1HNMR -PLS -DA

目的建立一种基于核磁共振-模式识别技术来鉴别椰奶质量的新方法。方法利用1HNMR测定椰奶样品,采集信息并转化成数据矩阵;利用模式识别方法中的PLS-DA法对数据进行分析,并建立鉴别模型。结果 5个厂家的椰奶样品均能获得有效的鉴别,含有添加物的椰奶样品也能有效地与正常椰奶样品鉴别开来。结论所用方法可靠、快速,能有效地鉴别椰奶的品质。

OBJECTIVE To establish a new quality control method of coconut milk based on ^1HNMR - PLS - DA. METHODS The ^1HNMR was used for obtaining the data which was transformed into data matrix. And the partial least squares - discriminate analy- sis （PLS -DA） was used to analyze the data and establish the identification model. RESULTS The coconut milk samples from 5 manufacturers were effectively distinguished by ^1HNMR - PLS - DA methods. The samples of coconut milk and the samples containing additives were effectively distinguished by ^1HNMR - PLS - DA methods. CONCLUSION The method of ^1HNMR - PLS - DA is reliable, fast and effective identification for coconut milk.