Introducing an orthogonal-triangular decomposition algorithm and its application in multivariate calibration

Abstract

An accurate and stable algorithm, Orthogonal-Triangular Decomposition Algorithm (OTDA) is for the first time presented in the field of chemistry. The basic principle of OT Multivariate Calibration Algorithm (OTMCA) is investigated. Complexes of sodium tripolyphosphate (STTP), sodium sulfate (SS) and sodium carbonate (SC), which in IR spectrometry present serious overlapping absorption peaks, are analyzed by OTMCA. We report that OTDA avoids matrix inverting, reduces the orders of matrices and speeds up the operation of matrices. So, it offers bright prospects in chemometrics. OTDA is implemented by an orthogonal and an upper triangular matrix decomposition process which is robust and convenient. Here, it is applied to overcome the problem of overlapping spectra by the different components contained in detergent washing powders.