Fuzzy Delphi method for identifying key parameters in the optimization of intelligent control systems for oil refining processes

Oil refining involves a series of intricate, nonlinear, and highly dynamic processes that require accurate control to maintain product quality, ensure operational safety, and achieve energy efficiency. As global energy demand grows and refineries aim for improved performance, the demand for intelligent control systems capable of managing uncertainties, external disturbances, and complex multi-variable interactions has increased considerably. Control artificial intelligence-based approaches, like fuzzy logic systems, neural network models, and hybrid intelligent frameworks are increasingly investigated to enhance refinery efficiency and adaptability. A major challenge in developing and optimizing such intelligent control systems is determining the critical parameters that most strongly affect system behaviour. Conventional parameter selection methods frequently rely on deterministic assumptions or limited experimental calibration, which may fail to capture the operational uncertainties characteristic of real refinery conditions. Moreover, this selection process often depends on expert insights typically expressed in qualitative or subjective terms, which traditional quantitative approaches are not well equipped to interpret. The Fuzzy Delphi Method (FDM) presents a powerful means of addressing this issue by merging expert consensus techniques with fuzzy set theory. FDM supports the integration of diverse expert opinions while accounting for ambiguity and uncertainty, making it particularly suitable for identifying parameters within complex industrial systems where both qualitative judgment and quantitative assessment are important. Despite these strengths, the use of FDM for optimizing intelligent control systems in oil refining has not yet been sufficiently investigated. Accordingly, this study employs the fuzzy Delphi technique to systematically determine the key parameters that influence the optimization of intelligent control systems in oil refining processes. By combining expert knowledge through fuzzy evaluation, the proposed methodology establishes a structured and trustworthy foundation for developing more resilient, efficient, and adaptable control strategies.