CLASSIC AND GENERATIVE INTELLIGENT CONTROL APPLIED TO INDUSTRIAL MIXERS, WITH IMPROVEMENTS IN QUALITY, MANAGEMENT, SAFETY, AMONG OTHERS

• DOI: https://doi.org/10.22533/at.ed.3175825011014

• Palavras-chave: Fuzzy Cognitive Maps, Hebbian Learning, Process Control, Fuzzy Logic, Artificial Neural Network.

• Keywords: Fuzzy Cognitive Maps, Hebbian Learning, Process Control, Fuzzy Logic, Artificial Neural Network.

• Abstract: This study presents the application of intelligent control techniques to an industrial mixing process. The proposed controller designed based on a Hebbian adaptation of the Fuzzy Cognitive Map (FCM) learning mechanism, which results in a Dynamic Fuzzy Cognitive Map (DFCM) model. The research develops and validates this DFCM using Hebbian learning algorithms to improve adaptability and robustness in nonlinear industrial systems. To ensure reliability, a classical fuzzy controller and a standard Proportional-Integral-Derivative (PID) controller implemented as benchmarks to validate the simulation results of the DFCM-based control for the industrial mixer. Extensive simulation experiments conducted to compare the performance of the controllers. The results demonstrate that the proposed DFCM provides superior performance in adaptability and robustness compared to the benchmarks, while also suggesting low computational complexity for practical implementation.