食品科学 ›› 2018, Vol. 39 ›› Issue (8): 231-236.doi: 10.7506/spkx1002-6630-201808036

• 安全检测 • 上一篇    下一篇

基于傅里叶变换近红外光谱法鉴别五常大米

钱丽丽1,宋雪健1,张东杰1,*,左锋1,鹿保鑫1,迟晓星1,赵海燕2   

  1. (1.黑龙江八一农垦大学食品学院,黑龙江?大庆 163319;2.青岛农业大学食品科学与工程学院,山东?青岛 266109)
  • 出版日期:2018-04-25 发布日期:2018-04-17
  • 基金资助:
    国家自然科学基金面上项目(81673170);黑龙江八一农垦大学校内培育课题重点项目(XA2016-04;XZR2016-07);黑龙江省农垦总局“十三五”重点科技攻关项目(HNK135-06-06)

Authentication of Wuchang Rice Based on Fourier Transform Near Infrared Spectroscopy

QIAN Lili1, SONG Xuejian1, ZHANG Dongjie1,*, ZUO Feng1, LU Baoxin1, CHI Xiaoxing1, ZHAO Haiyan2   

  1. (1. College of Food Science, Heilongjiang Bayi Agricultural University, Daqing 163319, China;2. College of Food Science and Engineering, Qingdao Agricultural University, Qingdao 266109, China)
  • Online:2018-04-25 Published:2018-04-17

摘要: 为保护五常大米的品牌效益,利用傅里叶变换近红外光谱法(Fourier transform near infrared spectroscopy,FTNIS)结合化学计量学分析法对来自五常地区和非五常地区共计191?份大米样品进行产地鉴别研究,结果表明:利用因子化法结合FTNIS技术所建立的鉴别分析模型对五常地区大米的鉴别正确率为100%,非五常地域的大米鉴别正确率为97.50%;利用聚类分析模型判别,对五常大米的判别正确率为100%,非五常大米判别正确率为100%。利用偏最小二乘法结合FTNIS建立的定量分析模型对五常大米判别正确率为95.83%,非五常大米判别正确率为95.00%。通过以上研究结果可知FTNIS法对五常大米产地鉴别是可行的。

关键词: 大米, 产地鉴别, 定性分析, 聚类分析, 定量分析

Abstract: In order to protect the brand benefits of Wuchang rice, a total of 191 rice samples collected form Wuchang county and other places were identified by Fourier transform near infrared spectroscopy (FTNIS) combined with chemometrics. The results indicated that Wuchang rice and non-Wuchang rice were identified by factor analysis combined with FTNIS with 100% and 97.50% accuracy, by cluster analysis model with 100% and 100% accuracy, and by the quantitative analysis model developed using partial least squares (PLS) based on FTNIS spectra with 95.83% and 95.00% accuracy, respectively. It can be concluded that FTNIS combined with chemometrics is feasible to identify the geographical origin of Wuchang rice.

Key words: rice, geographical origin identification, qualitative analysis, cluster analysis, quantitative analysis

中图分类号: