MODEL OF MULTIVARIATE CALIBRATION IN ORDER TO DETERMINE PARAMETERS OF DISPERSED PARTICLES FROM SMALL-ANGLE SCATTERING SPECTRA

摘要

A modular least-squares method is used for multivariate calibration in order to determine the parameters of a suspension of dispersed particles from small-angle light scattering spectra. Prior to using this method, transfer to new variables is accomplished using a discrete Fourier transformation. The mean square error for predicting the size of particles and their polydispersion using this method does not exceed 30%.