高等学校化学学报 ›› 2004, Vol. 25 ›› Issue (6): 1023.

• 研究论文 • 上一篇    下一篇

独立组分分析在红外光谱分析中的应用

毕贤, 李通化, 吴亮   

  1. 同济大学化学系, 上海 200092
  • 收稿日期:2003-05-08 出版日期:2004-06-24 发布日期:2004-06-24
  • 通讯作者: 李通化(1948年出生),男,教授,博士生导师,从事化学计量学研究.E-mail:lith@tongji.edu.cn E-mail:lith@tongji.edu.cn
  • 基金资助:

    国家自然科学基金(批准号:20275026)资助

Application of Independent Component Analysis to the IR spectra Analysis

BI Xian, LI Tong-Hua, WU Liang   

  1. Department of Chemistry, Tongji University, Shanghai 200092, China
  • Received:2003-05-08 Online:2004-06-24 Published:2004-06-24

摘要: 针对红外光谱的黑色体系分析,提出了一种基于独立组分分析(ICA)的红外光谱定性分析方法.其主要优点在于可从混合光谱中分离出独立组分的光谱,且这种分离是盲源分离,混合物的组分预先是未知的.对ICA在正己醇、丙酮和正丁醇的混合中红外光谱分离中的应用进行了研究,验证了体系的独立性,并对ICA算法做了一定的改进,讨论了各种预处理方法对其分离结果的误差.

关键词: 独立组分分析, ICA, IR谱, 应用

Abstract: A new method for separating the IR spectra based on the Independent Component Analysis(ICA) is presented. The main advantage of this method is that it can be used to separate the spectra of the independent components from the mixed spectra. The separation is a blind source separation,since the independent components in the mixtures was unknown. ICA was used to process the mixed system of {n-hexyl} alcohol,acetone and n-butyl alcohol in this study. We also discussed the effect of various preprocessing methods on the result of ICA and developed the ICA algorithm by adding PCA in front of ICA.

Key words: Independent Component Analysis, ICA, IR, Application

中图分类号: 

TrendMD: