Rapid determination of FeO content in sinter ores using DRIFT spectra and multivariate calibrations

Using diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), several multivariate calibration methods are explored for the quantitative determination of FeO content in sinter ores. The multivariate calibrations include ridge regression with Variable selection, principal component regr ession, ridge principal component regression, and partial least square regr ession with the linear and the nonlinear mapping using neural networks. Spe ctral data are preprocessed by signal correction and scaling prior to the m odeling. Cross validation is employed to obtain the optimal biasing paramet er in ridge-related regression and to obtain the optimal number of principa l components (or latent variables) in component-related modeling. We consid er the possibility of reducing the number of variables involved in models w hile maintaining the prediction power to propose a final prediction model. For the quantitative determination of FeO content in sinter ores, component related regressions on auto-scaled orthogonal signal correction are sugges ted as appropriate calibration methods. (C) 2000 Elsevier Science B.V. All rights reserved.