基于PLS-LDA和拉曼光谱快速定性识别食用植物油

吴静珠; 石瑞杰; 陈岩; 刘翠玲; 徐云

doi:10.13386/j.issn1002-0306.2014.06.002

基于PLS-LDA和拉曼光谱快速定性识别食用植物油

1. 北京工商大学计算机与信息工程学院
2. 中国农业大学信息与电气工程学院

基金项目:

北京市自然科学基金面上项目(4132008)；北京教委重点项目(KZ201310011012)；北京市属高等学校人才强教项目；

详细信息

作者简介:
吴静珠 (1979-) , 女, 博士, 副教授, 主要从事基于分子光谱技术的农产品及食品检测方面的研究。;

中图分类号: O657.3;TS227
计量
- 文章访问数: 208
- HTML全文浏览量: 24
- PDF下载量: 555
出版历程
- 收稿日期: 2013-08-13

Rapid qualitative identification method of edible vegetable oil based on PLS-LDA and Raman

School of Computer and Information Engineering, Beijing Technology and Business University
College of Information and Electrical Engineering, China Agricultural University

摘要

摘要: 以6种食用油共计23个样本为分析对象,采用偏最小二乘线性判别分析法（PLS-LDA）和拉曼光谱进行单一种类（橄榄油、花生油和玉米油）食用油快速定性检测,通过自适应迭代惩罚最小二乘法（airPLS）对拉曼信号进行背景扣除,以及蒙特卡洛无信息变量消除法筛选波长变量,不但有效减少了波长点数,降低了建模运算量,而且提高了单一种类食用油的识别率,使得总体识别率均高于90%,并在此基础上进一步提出了采用PLS-LDA进行多种类食用油识别的检测流程。实验结果表明PLS-LDA在食用油定性识别检测中具有较好的应用前景和可行性,该方法也可为定性检测食品及农产品品质提供借鉴。
- 偏最小二乘线性判别分析法 /
- 拉曼光谱 /
- 食用植物油 /
- 蒙特卡洛无信息变量消除法
Abstract: This paper choose 6 kinds of edible vegetable oils for a total of 23 samples as a typical tested object.Partial Least Squares-Linear Discriminant Analysis (PLS-LDA) method was employed to quickly identify a certain kind of edible vegetable oil (olive oil, peanut oil and corn oil) based on Raman. Raman backgrounds were subtracted by adaptive iteratively reweighted Penalized Least Squares (airPLS) method and wavelength variables were selected by Monte Carlo Uninformative Variable Elimination (MCUVE) method. The above spectra preprocessing not only effectively reduced the wavelength points and modeling computation, but also improved the general recognition rates higher than 90%, respectively. The process of identifying different kinds of edible oil using PLS-LDA method was suggested further on above basis. The experimental results showed that the PLS-LDA method had good application prospects and feasibility to identify edible oil species. This method provided a reference for processing the similar problems in food and agricultural products quality detection.
- Partial Least Squares-Linear Discriminant Analysis (PLS-LDA) /
- Raman /
- edible vegetable oil /
- Monte Carlo Uninformative Variable Elimination (MCUVE)

HTML全文

参考文献(15)

[1]	褚小立.化学计量学方法与分子光谱分析技术[M].北京:化学工业出版社, 2011.
[2]	陈健, 肖凯军, 林福兰.拉曼光谱在食品分析中的应用[J].食品科学, 2007, 28 (12) :554-558.
[3]	施玉珍, 陈志春, 林贤福.拉曼光谱与红外光谱无损检测技术新进展[J].分析化学, 2005, 33 (2) :272-276.
[4]	Xiaofang Zhang, Xiaohua Qi, Mingqiang Zou, et al.Rapid Authentication of Olive Oil by Raman Spectroscopy Using Principal Component Analysis[J].Analytical Letters, 2011, 44 (12) :2209-2220.
[5]	PIETER SAMYN, DIETER VAN NIEUWKERKE, GUSTAAF SCHOUKENS, et al.Quality and statistical classification of brazilian vegetable oils using mid-infrared and Raman spectroscopy[J].Applied Spectroscopy, 2012, 66 (5) :552-562.
[6]	Stewart Francis Graham, Simon Anthony Haughey, Robert Marc Ervin, et al.The application of near-infrared (NIR) and Raman spectroscopy to detect adulteration of oil used in animal feed production[J].Food Chemistry, 2012, 132:1614-1619.
[7]	周秀军, 戴连奎, 李晟.基于拉曼光谱的食用植物油快速鉴别[J].光谱学与光谱分析, 2012, 32 (7) :1829-1833.
[8]	房承宣, 李建华, 梁逸曾.拉曼光谱结合背景扣除化学计量学方法用于汽油中MTBE含量的快速测定研究[J].分析测试学报, 2012, 31 (5) :541-545.
[9]	Z-M Zhang, S Chen, Y-Z Liang.Baseline correction using adaptive iteratively reweighted penalized least squares[J].Analyst, 2010, 135 (5) :1138-1146.
[10]	陈珊.拉曼背景扣除算法及其应用研究[D].长沙:中南大学, 2011.
[11]	Li H-D, Zeng M-M, Tan B-B, et al.Recipe for revealing informative metabolites based on model population analysis[J].metabolomics, 2011, doi: 10.1007/s11306-010-0213-z.
[12]	袁大林, 梁逸曾, 许青松.QSAR/QSPR模型中的蒙特卡罗交叉效验评价[J].计算机与应用化学, 2006, 23 (6) :569-573.
[13]	Li H-D, Y-Z Liang, Q-S Xu, et al.Model population analysis for variable selection[J].J Chemometr, 2009, 24:418-423.
[14]	Roman M Balabin, Sergey V Smirnov.Variable selection in near-infrared spectroscopy:Benchmarking of feature selection methods on biodiesel data[J].Analytica Chimica Acta, 2011, 692:63-72.
[15]	Wensheng Cai, Yankun Li, Xueguang Shao.A variable selection method based on uninformative variable elimination formultivariate calibration of near-infrared spectra[J].Chemometrics and Intelligent Laboratory Systems, 2008, 90:188-194.