|
Spectral analysis using Fourier transform infrared spectroscopy (FTIR) and attenuated total reflection ( ATR ) were established sugar cane the Qing syrup , molasses Brix and Jane Rapid Method for Determination of purity and spectral optimization model , and the use of chemometric methods combination of wave number ( continuous band or discrete wavenumber combination) preferably, the stability of the model study . First sugar cane the Qing syrup , molasses FTIR / ATR spectroscopy method , then , based on the optimal single-point prediction model ( SWPB ) of all samples divided into calibration set and prediction set , so that they have similarities , which has modeling representative . Were sugar cane clear syrup , molasses ( Brix and apparent purity ) of each of the indicators meet the conditions given in 50 divided modeling different divisions to discuss the stability of the the scaling prediction model calibration set prediction set by . On this basis , establish a single point of regression , the direct PLS , Savitzky - Golay smoothing PLS , moving window PLS and interval moving window MLR model . Composition preferably wavenumber spectral analysis and the analysis of the stability of the model . Clear syrup Brix , purity and molasses Brix , Jane purity optimal model for interval moving window MLR model , the best model to predict the effect of : RMSEP, RRMSEP : 0.582,093 % , 0.697,0.81 % , 1.392,1.50 % , 1.207,3.40 % . The results show that the combination of wave number model are compared with the full - spectrum model , not only greatly reduce the complexity of the model , at the same time to obtain a better prediction accuracy , FTIR / ATR spectroscopy for cane sugar industry , the development of small dedicated analytical instruments basis .
|