ATR-THz波谱结合PLS-DA对陈化小麦快速无损鉴别研究

发布时间：2018-05-09 11:20

  本文选题：太赫兹波谱 + 衰减全反射　； 参考：《光谱学与光谱分析》2016年07期

【摘要】：采用ATP附件采集了人工控制条件老化小麦样品的太赫兹波谱数据,并在PCA综合得分基础上对数据进行分集,建立陈化小麦的快速无损鉴别PLS-DA模型,并对其进行外部验证。结果显示,针对吸收系数谱,所建模型的校正集实验组、对照组以及交互验证实验组、对照组正确率分别为84.2%,94.7%,84.2%和81.6%,外部验证集实验组、对照组正确率分别为73.7%和100.0%;针对折射率谱,所建模型的校正集实验组、对照组以及交互验证实验组、对照组正确率分别为84.2%,92%,76.3%和76.3%,外部验证集实验组、对照组正确率分别为84.2%和89.5%。研究表明,ATR-THz技术在陈化小麦无损鉴别方面具有一定的应用潜力。
[Abstract]:The terahertz spectrum data of aged wheat samples under artificial control conditions were collected by ATP attachment, and the data were differentiated on the basis of PCA scores. The PLS-DA model of rapid nondestructive identification of aged wheat was established and verified externally. The results showed that the correct rates of the experimental group, the control group and the interactive verification group were 84.2% and 81.6%, respectively, according to the absorption coefficient spectrum. The accuracy rates of the control group were 73.7% and 100.0.The correct rates of the experimental group, the control group and the interactive verification group were 84.2% and 76.3%, respectively. The accuracy rates of the external validation group and the control group were 84.2% and 89.5 respectively. The results show that ATR-THz technique has some potential application in nondestructive identification of aged wheat.
【作者单位】： 北京农业质量标准与检测技术研究中心;农业部农产品质量安全风险评估实验室(北京);爱德万测试(中国)管理有限公司;
【基金】：国家重大科学仪器设备开发专项(2012YQ14000507) 北京市农林科学院青年基金项目(QNJJ201431)资助
【分类号】：O657.3;TS210.7
，

本文编号：1865822