МҰНАЙДЫ БАСТАПҚЫ ӨҢДЕУ ҚОНДЫРҒЫСЫНДА ТҰРАҚТЫ БЕНЗИН ӨНДІРУ ПРОЦЕСІН АЙҚЫНСЫЗДЫҚТА МОДЕЛЬДЕРІ НЕГІЗІНДЕ ОҢТАЙЛАНДЫРУ ТӘСІЛІ

This article examines the problems of optimizing the production process of stable gasoline based on models of the object, which is the target product of a primary oil refining installation, and proposes approaches to their effective solution. Currently, solving the problem of optimizing the main processes at a primary oil processing plant in the oil refining industry is one of the most pressing scientific and production problems, as it ensures the efficiency of further deep processing of the resulting petroleum products. In this regard, this study formulates a mathematical formulation of the problem of optimizing the production process of stable gasoline - the target product of the primary oil refining installation of the Atyrau Oil Refinery, characterized by uncertainty, and proposes a heuristic approach for its effective solution in a fuzzy environment. As a result of the research, mathematical models of complex objects characterized by ambiguity of some parameters, such as the stabilization column of a primary oil treatment plant, were developed, and on their basis a heuristic approach was developed that effectively optimizes the operation of the object in conditions of uncertainty. Models that determine the volumes of gasoline and gas from the stabilization column depending on the input operating parameters of the column are determined on the basis of experimental statistical data and the systematic application of the approach of sequential sequential connection of regressors and the least squares method. Fuzzy models that evaluate vaguely described quality indicators of stable gasoline are synthesized on the basis of expert assessment methods, fuzzy sets and modified methods of sequential inclusion of regressors, least squares. Based on the obtained models of the stabilization column, a heuristic algorithm has been developed for the effective optimization of the production process of stable gasoline under conditions of fuzzy conditions, based on a modification and combination of the principles of Pareto optimality and the ideal point for working in a fuzzy environment. [ABSTRACT FROM AUTHOR]