Inferential Feedforward Control of a Distillation Column

An inferential feedforward control strategy is developed and applied to a simulated distillation column. In this control strategy, the effects of disturbances on the primary process variables (top and bottom compositions) are inferred from uncontrolled secondary process variables (tray temperatures) which can be easily measured. The proposed strategies are particularly useful when disturbances cannot be measured easily or economically. Robustness of the inferential feedforward controllers and the selection of appropriate secondary measurements are discussed. Nonlinear dynamic simulation results demonstrate the superior performance of this control strategy and verify the robustness analysis.