(561d) Iterative Learning Control of Reactive Distillation Process

Recently, the use of batch distillation with chemical reaction is common practice in the chemical industries. If one of the reaction products has a lower boiling point than other products and reactants, it can be taken in higher conversion when reaction and separation take place simultaneously. And it can provide a significant reduction of capital investment.

The constraints of the distillation and reaction are different each other. So there are many problems when distillation and reaction take place in same area. Especially, in reactive batch distillation process, the operating conditions (like a composition and temperature) are varying with times. For this reason, the optimal reflux trajectory should be required to satisfy constraints of product.

The ILC(Iterative Learning Control) is a control system that can be improved by learning from previous executions (trials, iteration, passes) with the same task multiple time, even if the dynamic characteristics are not known. The ILC can be applied to batch distillation because it is performed in iterative same multiple batch. Then it is possible the control error goes to zero perfectly as repeating process of each batch on the assumption that the same disturbance is entered. In this paper, the process is designed to trace optimal reflux trajectory with using ILC in reactive batch distillation.