Fuzzy cognitive approach of a molecular distillation process

Abstract

Takagi–Sugeno fuzzy model is presented as an alternative modeling tool for the molecular distillation process of heavy liquid petroleum residues. A nonlinear phenomenological model formed by partial differential equations involving temperature and composition is regarded as a virtual working plant for the generation of data required to build the fuzzy models. Furthermore, experimental data from the molecular distillation process of an atmospheric residue 673.15 K upward (673.15 K+) at different operating conditions were used to develop the fuzzy representations. In this work, the system was simulated through the development of software in Fortran 90/95 and the numerical solution by using the finite-difference method. In the fuzzy approaches, the distillation temperature and the feed flow rate are the input variables, while the liquid interface temperature, the film thickness, the concentration profiles, and the distillate flow rate were considered as the output responses. The fuzzy models obtained were compared with the results generated from the phenomenological model, showing an excellent agreement.