Control for Quality of the Separated Products of a Butylene-Butane Distillation Column Based on High Order Iterative Learning Control

This paper presents the design of an advanced control system for product quality of a butylene-butane distillation column (BBDC) based on High Order Iteration Learning Control (HOILC). The multicomponent distillation process was adapted to a pseudo-binary distillation process for the development of automatic control systems for the product's quality. Starting from this assumption, dynamic models of the process were obtained for the input-output channels reflux flowrate – concentration of the pseudo-light component and bottom product flowrate – concentration of the pseudo-heavy component. To enhance the control accuracy of product quality, this paper adopts a high-order iterative learning algorithm to optimize the accuracy of the product quality control system for the butene-butane distillation tower. Compared with conventional iterative learning, a high-order iterative learning algorithm contains more degree of freedom, which significantly improves the algorithm performance, speeds up the convergence process, reduces the required number of batches, and ultimately achieves the goal of reducing the number of trays and improving efficiency. Simulation and modeling of a product quality control system for a butene-butane distillation column in Simulink, and its validation through MATLAB coding and Simulink, have demonstrated that High-order ILC effectively improves the performance and cost of the product quality control system for this type of distillation column.