河北大学学报(自然科学版) ›› 2016, Vol. 36 ›› Issue (4): 374-379.DOI: 10.3969/j.issn.1000-1565.2016.04.008

• • 上一篇    下一篇

矿物元素结合簇类独立软模式法对冬枣产地判别模型的优化

夏立娅,李小亭,李晓杨,张晓瑜,尹洁璇   

  • 收稿日期:2016-03-01 出版日期:2016-07-25 发布日期:2016-07-25
  • 通讯作者: 李小亭(1958—),女,河北保定人,河北大学教授,博士生导师,主要从事光学分析及应用研究.E-mail:lxt@hbu.com
  • 作者简介:夏立娅(1978—),女,河北临西人,河北大学副教授,博士,主要从事食品品质评定及原产地鉴别研究. E-mail:xialiya@126.com
  • 基金资助:
    国家自然科学基金资助项目(31501447);河北省自然科学基金资助项目(B2013201235);河北大学自然科学研究项目(2014-02)

Optimization of traceability model of Ziziphus jujuba geographic origin by multi-element analysis combined SIMCA

XIA Liya,LI Xiaoting,LI Xiaoyang,ZHANG Xiaoyu,YIN Jiexuan   

  1. College of Quality and Technical Supervision, Hebei University, Baoding 071002, China
  • Received:2016-03-01 Online:2016-07-25 Published:2016-07-25

摘要: 为了提高冬枣产地鉴别方法的准确性,测定了69个冬枣样本中10种元素的含量,并对数据进行了差异性分析、主成分分析和簇类独立软模式分析(SIMCA).结果表明,冬枣中Mg、B、Mn、Fe、Zn等元素在不同产地间存在显著差异,利用主成分分析可以看出不同产地样本有较好的聚类趋势.在前4个主成分中,Fe、B、Mn、Zn和K元素的载荷值较高,是重要的产地识别元素.利用SIMCA建立的产地判别模型,置信水平为5%时对验证集样本判别结果最好,识别率为100%,拒绝率为78.95%.研究结果证实了农产品中多元素分析结合SICMA法可以有效用于原产地的鉴别.

关键词: 冬枣, 元素, 主成分分析, 簇类独立软模式

Abstract: In order to improve the accuracy of identification method of geographical origin,the contents of 10 elements in 69 Ziziphus jujubas were measured,and the data were analyzed by difference analysis,principal component analysis(PCA)and soft independent modeling of class analogies(SIMCA).The results showed that there were significant differences in the contents of Mg,B,Mn,Fe,Zn and other elements from different origin places.The result of PCA showed that the samples from different origins had clustering trend.In the first four principal components,the Fe,B,Mn,Zn and K elements had higher loading values,they were considered important origin elements.In the SICMA,with the 5% significance level,the identified result was best with 100% recognition rate and 78.95% rejection rate.This study confirmed that multi-element analysis with SICMA is on effective method to determine geographical origin of agricultural products.

Key words: Ziziphus jujuba, element, principal component analysis, soft independent modeling of class analogies

中图分类号: