Comparative Study of Adaptive Mixed Fuzzy Controller and Adaptive Traditional Fuzzy Controller for Fluid Catalytic Cracking Unit Using Real-Coded Genetic Algorithm

A Fluid Catalytic Cracking Unit (FCCU) is always considered to be a benchmark problem in the petroleum refinery industry. The process is interactive, with each process depending on the other. In this study, the control of the reactor and regenerator temperature in the FCCU involves the utilization of Adaptive Traditional Fuzzy (ATF) logic controllers and Adaptive Mixed Fuzzy (AMF) logic controllers. The purpose of designing the AMF controller is to offer solutions for minimizing the coupling effect resulting from the interaction of the FCCU. The development of this controller involves a combination of ATF and Coupling Fuzzy (CF) controller elements across different degrees of freedom. To optimize the scale factors for the ATF and CF control schemes, a Real-coded Genetic Algorithm (RGA) is employed. The results show that the RGA-tuned AMF controller tracks the setpoint better than the ATF controller, with less interaction effect.