Nonlinear fuzzy control of a fed-batch reactor for penicillin production

The process of penicillin production is characterized by nonlinearities and parameter uncertainties that make it difficult to control. In the paper the development and testing of a multivariable fuzzy control system that makes use of type-2 fuzzy sets for the control of pH and temperature are described. The performance of the type-2 fuzzy logic control system (T2FLCS) is compared by simulation with that of a type-1 fuzzy logic control system (T1FLCS) and that of a control system with traditional proportional-integral-derivative (PID) controllers proposed in the literature. The fuzzy controllers are optimized using an ANFIS algorithm. The best results are obtained with the T2FLCS particularly when uncertainties are present in the control system.

► Control of the process of penicillin production.
► Multivariable fuzzy control system with type-2 fuzzy sets.
► Optimization of fuzzy controllers by an ANFIS algorithm.