Control of Dividing-Wall Columns via Fuzzy Logic

Because of their potential energy savings, dividing-wall columns (DWCs) have been viewed as an interesting alternative to the use of sequences based on conventional distillation columns. The energy savings, however, must be supported by suitable controllability properties for a proper operation. In this study we use a design methodology to optimize the gains of the Takagi-Sugeno fuzzy controller and show its implementation for the control of a DWC. Comparison with optimal proportional-integral-derivative (PID) controllers was carried out to further analyze the performance of the proposed controller under load disturbances in feed composition and set point changes. The case studies considered here are the seperation of pentane-hexane-heptane and the industrially relevant separation of benzene-toluene-xylene. The results show that the fuzzy controller provides a suitable option for DWC control and that it produces improved control performance with respect to the conventional PID controller.