The study is devoted to multicriteria optimization of the operation of the reactor unit intended for reforming of oil fractions. The optimization is based on the earlier developed mathematical model, which can take into account changes in the temperature and molar consumption of the reaction mixture during chemical transformations, the composition of the feedstock being processed, and the circulation quality and ratio of the hydrogen-containing gas. The following criteria are used in such optimization: octane number, yield of the target product (reformate), content of the sum of aromatic hydrocarbons, and content of benzene. The optimization was implemented using Pareto approximation with the genetic algorithm. Temperatures at the reactor inlets served as the control parameters. As a result of optimization, the sum content of aromatic hydrocarbons decreases from 56 to 45 wt.%, and the octane number measured by the research method (RON) decreases from 92.7 to 90.7 points. The three-criteria optimization is also considered; it provides a decrease in the benzene content from 4 to 3.08 wt.% with a decrease in RON from 92.7 to 91.8 points. In both cases, the decrease in the octane number is admissible, taking into account the fulfilled requirements to the limiting content of benzene and aromatic hydrocarbons in the reformate.
Optimizing an Industrial Scale Naphtha Catalytic Reforming Plant Using a Hybrid Artificial Neural Network and Genetic Algorithm Technique
