Fuzzy supervisory PI controllers for a binary distillation column using hierarchical genetic algorithms

This paper presents the optimum fuzzy supervisory PI controllers of a binary distillation column. A hierarchical genetic algorithm (HGA) is used to define the optimal number and shape of membership functions, and fuzzy rules. The optimum fuzzy system then adapts the parameters, K_p and K_I, of the PI controller. Two optimum fuzzy supervisory PI controllers are developed to maintain the top and bottom product composition due to changes in feed flow rate. Simulation results are compared with the PI controllers using minimum ISTSE method.