Real-Time Control of a Binary Distillation Column Using HGA Fuzzy Supervisory PI Controllers

In this paper, an optimum fuzzy supervisory PI controller using hierarchical genetic algorithms is developed and implemented for controlling the top and bottom product quality of a nonlinear, multi-input multi-output binary distillation column when the disturbance enters the column in the form of the changes in feed flow rate. Two conventional PI controllers, one for the bottom product composition and another for the top product composition, are used together in the control scheme. Hierarchical genetic algorithms are used to derive the optimal number and shape of membership functions and fuzzy rules of a fuzzy supervisory system that adapts the parameters of the PI controllers. The real-time implementation results show the effectiveness of the proposed method.