Approaches for Modeling and Optimization of a Cellulose Production Process

Cellulose is a ground material for paper production. Reaching good cellulose quality will result in high-class paper. Strong quality demands determine the production. My work was to model the production process and to look for optimization approaches. A lot of chemical reactions take place during the phase of a cellulose production and many parameters influence the production process. Several difficulties we had to master. One of them was the small data set we have gotten from the paper mill. Another one was attributed to the little knowledge of the chemical process that describes the cellulose production. The thesis is structured into the following parts. At first the modeling and optimization methods followed by a comparison are demonstrated. Nearest neighbor, RBF nets, hyperplane and LOESS are used as modeling methods. Simulated annealing and genetic algorithms are discussed as optimization tools. Afterwards I show the principle of cellulose production. Finally I discuss the practical application at the paper mill.