Improvement of the piecewise direct standardisation procedure for the transfer of NIR spectra for multivariate calibration

Standardisation of near-infrared spectrometric instruments using the piecewise direct standardisation method is studied. Two important steps of this method, namely the selection of the subset samples and the determination of the local optimal number of principal components, can yield artefacts in the transferred spectra, if they are not carefully performed. Three different algorithms are tested on two different data sets in order to study the subset selection, and a method based on the prediction of an independent test sample is proposed in order to detect the artefacts due to bad local rank determinations. Improvements of the standardisation procedures were obtained, particularly when low numbers of standardisation samples were included in the subset.