Simulation of Pulping Process Using Neural Networks and Hybrid Model

Abstract A Neural Network Model is developed for a Kraft digester which is a fundamental stage in the pulp production. A detenninistic model is used to describe the main features of the process and will provide the data to verify the performance of the artificial neural network modeling. The paper shows that the artificial neural network is a good way to represent the process since it gives equivalent results when compared to deterministic model, and it is easier and cheaper to be developed.