Abstract
In the context of the digitalization of the oil refining industry, the need for intelligent methods for ana-lyzing technological processes characterized by incomplete and fuzzy information is increasing. This study addresses the problem of expert evaluation for a benzene production unit, where experts can provide only linguistic rather than quantitative assessments of the influence of input parameters on output indicators. Based on fuzzy set theory and artificial intelligence, an intelligent method of expert evaluation in a fuzzy environment (I-EE-FE) is proposed, integrating fuzzy logical inference, machine learning, and elements of the Delphi method. The proposed method enables the automatic formation of term sets and membership functions, verification of expert opinion consistency, and the construction of linguistic and fuzzy models. Using the benzene production unit as an example, it is demonstrated that the application of I-EE-FE makes it possible to develop hybrid models of the “expert knowledge + AI” type, suitable for decision-support systems and digital twins, thereby enhancing the accuracy and justification of technological process control.