Abstract
In practice, when developing models of many production facilities, problems often arise related to the fuzziness of some of the initial information, affecting important indicators of the quality of the facility’s operation. The purpose of the study is to develop a systematic method for synthesizing statistical, fuzzy and linguistic models of complex objects in conditions of shortage and fuzziness of initial information. Then, using the proposed method, various models of the atmospheric unit of a primary oil refining plant are developed. At the same time, statistical models are developed on the basis of traditional methods. With crisp input and operating parameters and fuzzy output parameters, based on the proposed method, fuzzy models of the atmospheric unit are developed that determine the quality of the manufactured products. And when the input, operating and output parameters of the object are fuzzy, there have been developed linguistic models that evaluate the qualities of the target product of the atmospheric unit based on expert assessment methods, logical rules of conditional inference and the proposed method. The developed linguistic models in Fuzzy Logic Toolbox make it possible to evaluate the quality of gasoline from an atmospheric unit depending on the content of chloride salts and the mass fraction of sulfur in the raw material. The advantages of the proposed modeling method compared to known ones are shown.