ERROR ESTIMATION IN THE QUANTITATIVE SPECTROCHEMICAL ANALYSIS OF MULTICOMPONENT MIXTURE WITH RANDOMLY VARYING SPECTRAL BASELINE

Abstract

The errors of quantitative analysis of multicomponent mixtures using absorption spectra perturbed by randomly varying baseline could be significantly reduced by the introduction of the noise covariance matrix in least squares estimation. The effectiveness of the proposed method is demonstrated on the example of Excedrin drug analysis.

Keywords

• quantitative spectrochemical analysis
• direct calibration
• errors
• randomly varying base line