Robust calibration model for on-line and off-line prediction of poly(ethylene terephthalate) yarn shrinkage by Raman spectroscopy

Abstract

In order to transfer a calibration model, techniques such as piecewise direct standardization (PDS) or direct standardization (DS) are used. This implies that the same subset of samples must be measured in both situations. However, when a model is transferred from an off‐line application to an on‐line application, it is not possible to remeasure the same subset of samples. In this paper, however, a method is presented which can transfer a multivariate calibration model from an off‐line application to an on‐line application without remeasuring the same subset. For this purpose a calibration model has been calculated with a subset of spectral variables. The variables, which are sensitive to the differences between the off‐line and the on‐line application, are rejected. The method is compared with the calculation of a hybrid calibration model in which off‐line and on‐line spectral data are combined into one calibration set. The methods are applied to a calibration model which is used to predict the shrinkage after heat treatment of poly(ethylene terephthalate) (PET) yarns from the Raman spectra measured both off‐line and on‐line. The Raman spectra of a calibration set have been measured off‐line and a smaller set has been measured on‐line during the production of PET yarns. It has been found that it is possible to make a universal calibration model for off‐line and on‐line predictions of PET yarn shrinkage by Raman spectroscopy. Copyright © 1999 John Wiley & Sons, Ltd.

Effect on the Partial Least-Squares Prediction of Yarn Properties Combining Raman and Infrared Measurements and Applying Wavelength Selection

“…Furtherm ore, this preference of discarding peaks is visible in previous publications. 2,7,37,38 Discarding wavelengths on peak positions also occurs for the infrared and combined spectra.…”

Calibration Transfer Methods

“…As stated before, spectral pre-processing can be of vital importance for achieving model maintenance and calibration transfer. Swierenga et al (1999) took a different pre-processing approach by selecting variables in their Raman signal, using a simulated annealing algorithm that are least infl uenced by instrumental differences while maintaining a high predictive ability. However this idea-the so-called robust multivariate calibration models in a subspace of the original full spectral domain ( Swierenga et al , 1998 )-requires samples measured on the secondary instrument, but not necessarily the samples as recorded on the primary instrument.…”

Transfer of Multivariate Calibration Models

“…28 Successful application of sensor selection in calibration is dependent on finding variables that are insensitive to extraneous variation yet sensitive to changes in the property. 34 Genetic algorithms have been employed to find a combination of variables that allow a single calibration model to be used with two separate NIR instruments. 19,[29][30][31] An example of this approach is the method developed by Mark and Workman 32 for selection of wavelengths for multivariate linear regression (MLR) models that are robust to wavelength shifts.…”