Journal of Hebei University (Natural Science Edition) ›› 2019, Vol. 39 ›› Issue (1): 56-62.DOI: 10.3969/j.issn.1000-1565.2019.01.0010

Previous Articles     Next Articles

Geographic origin identification of Ziziphus Jujuba based on anion characteristic analysis

XIA Liya1,2, SHEN Shigang1, LI Chao2, LIU Xiaohui3, LI Yunsi2, LI Jiaojiao2   

  1. 1. College of Chemistry and Environmental Science, Hebei University, Baoding 071002, China; 2. College of Quality and Technical Supervision, Hebei University, Baoding 071002, China; 3. Baoding Entry-exit Inspection and Quarantine, Baoding 071002, China
  • Received:2018-10-01 Online:2019-01-25 Published:2019-01-25

Abstract: In order to investigate the possibility of anions on the identification of the geographic origin, the contents of anions in Ziziphus Jujuba and its soil from Huanghua, Zhanhua and Dali were determined by ion chromatography. The data were analyzed by variance, partial correlation analysis, stepwise discriminant analysis(SDA)and radial basis function artificial neural network analysis(RBF-ANN). The results showed that F-, Cl-, NO-2, PO3-4, SO2-4 and C2O2-4 were significant differences between the three places, and they significant correlated with soil. In the SDA, the discriminant equation was established by the six anions with strong discriminating ability. The analysis of RBF-ANN further confirmed that the accuracy of model constructed by the six anions was higher than that of all anions. Br- interfered the model and reduced the accuracy of identification. This study confirmed that the anion binding multivariate statistical- DOI:10.3969/j.issn.1000-1565.2019.01.0010基于阴离子特征分析的冬枣产地鉴别夏立娅1,2,申世刚1,李超2, 刘晓慧3,李运思2, 李姣姣2(1.河北大学 化学与环境科学学院,河北 保定 071002; 2.河北大学 质量技术监督学院,河北 保定 071002; 3. 保定出入境检验检疫局,河北 保定 071002)摘 要:为了考察阴离子用于农产品产地鉴别的可能性,利用离子色谱测定了黄骅、沾化和大荔冬枣及土壤中多种阴离子的含量,并对数据进行了方差分析、偏相关分析、逐步判别分析(SDA)和径向基人工神经网络分析(RBF-ANN). 结果表明,冬枣中F-、Cl-、NO-2、PO3-4、SO2-4和C2O2-4在不同产地间存在显著性差异,与土壤中阴离子具有显著相关性.逐步判别分析中,上述6种阴离子的判别能力较强,所建判别方程可以准确地鉴别冬枣的产地.RBF-ANN的分析进一步证实了,6种阴离子所建模型的准确率高于全部阴离子的分析结果,Br-对产地鉴别有一定的干扰作用,鉴别准确度降低.研究结果证实了阴离子结合统计学算法可以建立有效的冬枣产地鉴别模型,选择合适的产地因子是提高产地鉴别模型准确度的关键步骤.关键词:冬枣;阴离子;偏相关分析;逐步判别;径向基人工神经网络中图分类号:O657.3 文献标志码:A 文章编号:1000-1565(2019)01-0056-07Geographic origin identification of Ziziphus Jujuba based on anion characteristic analysisXIA Liya1,2, SHEN Shigang1, LI Chao2, LIU Xiaohui3, LI Yunsi2, LI Jiaojiao2(1. College of Chemistry and Environmental Science, Hebei University, Baoding 071002, China;2. College of Quality and Technical Supervision, Hebei University, Baoding 071002, China;3. Baoding Entry-exit Inspection and Quarantine, Baoding 071002, China)Abstract:In order to investigate the possibility of anions on the identification of the geographic origin, the contents of anions in Ziziphus Jujuba and its soil from Huanghua, Zhanhua and Dali were determined by ion chromatography. The data were analyzed by variance, partial correlation analysis, stepwise discriminant analysis(SDA)and radial basis function artificial neural network analysis(RBF-ANN). The results showed that F-, Cl-, NO-2, PO3-4, SO2-4 and C2O2-4 were significant differences between the three places, and they significant correlated with soil. In the SDA, the discriminant equation was established by the six anions with strong discriminating ability. The analysis of RBF-ANN further confirmed that the accuracy of model constructed by the six anions was higher than that of all anions. Br- interfered the model and reduced the accuracy of identification. This study confirmed that the anion binding multivariate statistical- 收稿日期:2018-10-01 基金项目: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);河北大学实验室开发项目(sy201674) 第一作者:夏立娅(1978—),女,河北邢台人,河北大学教授,博士,主要从事原产地鉴别技术研究.E-mail: xialy@hbu.edu.cn 通信作者:申世刚(1964—),男,河北大学教授,博士生导师,主要从事分析检测技术研究. E-mail: shensg@hbu.edu.cn第1期夏立娅等:基于阴离子特征分析的冬枣产地鉴别methods were effective method to determine geographical origin of Jujube, and the selection of suitable anions was the key step to provide the desired accuracy.

Key words: Ziziphus jujuba, anion, partial correlation analysis, stepwise discriminant analysis, radial basis function artificial neural network

CLC Number: