Experimental investigations on fuzzy logic for process control

The present work is concerned with the design and experimental testing of fuzzy algorithms for the on-line control of a few processes: temperature of a methyl methacrylate batch polymerization reactor, evaporating and condensing temperatures of a refrigeration system and the overhead composition of a batch distillation column. PID-fuzzy algorithms were developed and compared to conventional PID controllers, proving to be more suitable and reliable for the polymerization process. For the batch distillation column, fuzzy control reduced the batch time and the energy consumption. The PID-fuzzy also outperformed the conventional PID in terms of energy savings for the refrigeration system.