Identification of Fuzzy Models Applied to a Three-Phase Catalytic Hydrogenation Reactor

Abstract

This paper proposes the identification of Fuzzy models applied to a three-phase catalytic reactor in which is carried out the o-cresol hydrogenation for obtainment of 2-methyl-cyclohexanol. Through the deterministic model that represents the three-phase catalytic reactor are obtained the data to build the Fuzzy models. The identification of models is an important step for the control design since the Fuzzy model can be presented as an alternative to the model of the Model Predictive Controllers. Two possible control configurations for this reactor is to maintain the exit concentration of o-cresol in the liquid phase at the set point manipulating the reactor feed temperature and to maintain the reaction medium temperature at the exit of the reactor at the set point manipulating the reactor feed temperature. Considering the different order of magnitude of the variables of the first control configuration, in order to procedure with the identification of the model, the codification of such variables is proposed as a first step. The results show a very good approximation of the Fuzzy model and the deterministic model confirming that this Fuzzy model can be used in substitution to the rigorous model of the reactor in control applications. In future works, the Fuzzy model can be used, for instance, as an alternative to the convolution model of the Dynamic Matrix Control, configuring in a Fuzzy-DMC Hybrid Controller.