Abstract:
The structures and model parameters of the hydrotreating reactor of a catalytic
reforming unit were identified based on available information of various types. The structural
identification of the developed reactor models was carried out on the basis of the ideas of the
sequential switching of regressors method, and the parameters were identified using the least
squares method modified for working in a fuzzy environment. The developed models have the
structure of nonlinear regression models (for output, i.e. hydrogenated) and fuzzy multiple
regression equations (for qualitative indicators of hydrogenated). When developing the models
with the source information, we used available statistical and experimental data describing the
operation of the reactor and fuzzy information representing the knowledge, experience and
judgment of experts about the hydrotreatment process. Based on the simulation of the
hydrotreatment reactor, a graph of the temperature dependence of the hydrogenate yield of the
reactor is constructed. A linguistic model has been constructed that estimates the dependence
of the temperature of the hydrotreating process on the quality of the feedstock. This model
allows to determine the optimal temperature value in a fuzzy environment depending on the
quality of the raw material being cleaned.
