A Hybrid Method for the Development of Mathematical Models of a Chemical Engineering System in Ambiguous Conditions

• Published in: Mathematical models and computer simulations Vol. 10; no. 6; pp. 748 - 758
• Main Authors: Orazbayev, B. B., Ospanov, E. A., Orazbayeva, K. N., Kurmangazieva, L. T.
• Format: Journal Article
• Language: English
• Published: Moscow Pleiades Publishing 01.11.2018; Springer Nature B.V
• Subjects: Catalysis; Chemical engineering; Construction; Mathematical analysis; Mathematical Modeling and Industrial Mathematics; Mathematical models; Mathematics; Mathematics and Statistics; Optimization; Organic chemistry; Petroleum refineries; Refineries; Reforming; Simulation and Modeling; chemical engineering systems; reforming block; theory of fuzzy sets; membership function; mathematical models
• ISSN: 2070-0482; 2070-0490
• DOI: 10.1134/S2070048219010125

A hybrid method is proposed for the development of a structured set (complex) of mathematical models for a complicated chemical engineering system (CES) of an oil refinery in ambiguous conditions based on a different type of information. Based on the research results for each CES element, the collected information, and the selection criterion, a mathematical model of a CES element is constructed, and then the developed models are combined into a single model to model the chemical engineering system as a whole. The developed method is successfully implemented when constructing a set of models for the main units of the reforming block of the catalytic reforming unit of the Atyrau oil refinery. The results of modeling based on the proposed method are compared with the well-known results and the experimental data from the LG unit of the Atyrau oil refinery. The statement of the problem of fuzzy optimization based on the models for CES operation regimes is formalized and an algorithm of its solution is proposed. The structure of a computer-aided system for modeling and optimizing an oil refinery’s CES operation regimes under the conditions of multicriteriality and fuzzy initial information is created.