Discrimination and Classification of Plastic Steel Windows by FTIR in Combination with Chemometrics

42 samples of plastic steel windows were collected from 7 manufacturing companies, and the samples were divided into 7 groups according to their brands. The samples (9, 17, 21 and 32) used as samples to be distinguished. The remainder 38 samples were used for study on the method of their identification. The above mentioned samples were analyzed by FTIR to obtain their spectra. The spectral informations were then standardized by the Z-method and studied by the method of system cluster analysis and the Spss software of statistic analysis. As shown by the results, the 38 samples were classified into 27 categories. On this basis, the 4 samples to be distinguished were classified by selecting the typical discriminant analysis, and a high rate of classification accuracy of 94.6% was attained.