Abstract

This paper discusses a Kappa-number estimator using different modelling approaches in continuous cooking. Alkaline, total dissolved solids and lignin content are measured on-line from several circulations in a continuous digester. ABB's Cooking Liquor Analyser (CLA 2000) and an on-line Kappa-number measurement were used in this research. With these measurements it is possible to estimate the blow-line Kappa-number long before the end of the cook and to gain a better understanding on the dependence of the Kappa-number on these variables. The developed models will be used in cooking control to reduce the Kappa-number variation.

Artificial neural networks, fuzzy logic, partial least squares method and linguistic equations are used in building a model to map input-output relationships of the measurements.