Spectral Pattern Recognition of Mixed Alcohols by Means of the Method Based on Judging the Subspace Coincidence
  
View Full Text    Download reader
DOI:10.3969/j.issn.1004-4957.年份.月份
KeyWord:Raman spectra  multicomponent qualitative detection  subspace angle  chemometrics
  
AuthorInstitution
黄培贤,姚志湘,粟晖,孙阔 广西工学院生物与化学工程系
Hits: 3858
Download times: 1210
Abstract:
      Based on the deficiency of qualitative detection of traditional characteristic peak comparison,a new method of the spectral pattern recognition based on judging the subspace coincidence was developed for the qualitative detection of nine mixed alcohols solution with different proportions.The spectra of mixed alcohol were treated as a vector in the method.First,the angle of raman spectra of mixed alcohol and the gather of standards were calculated.Then the least number of subspace,which contains the standard samples composed of the mixed alcohol undetermined,was selected according to the change of the subspace angle.Finally,it is successful to complete the qualitative recognition of all nine mixed alcohols of different proportions,and the detection rate of the method is 100%.With advantages of accurancy,simplicity,sensitivity and time saving,the method was suitable for the qualitative detection of multicomponent mixed system.
Close