Dynamic Matrix Control of a Reactive Distillation Process for Biodiesel Production

Abstract

This study has been carried out to demonstrate the control of a reactive distillation process in which the production of biodiesel was taken as the case study using an advanced control method, which is known as dynamic matrix control. The control was accomplished by employing the transfer function model of the reactive distillation process developed, using the System Identification Toolbox of MATLAB, from the dynamic data generated when the prototype plant of the process was simulated with the aid of ChemCAD process simulator. The results obtained from the dynamic matrix control were compared with those of a proportional-integral-derivative (PID) control system tuned with Ziegler-Nichols and Cohen-Coon methods, and it was discovered that the dynamic matrix control was able to perform best among the three (dynamic matrix control method, PID tuned with Ziegler-Nichols method and PID tuned with Cohen-Coon method) because it (the dynamic matrix control) was able to make the biodiesel mole fraction response not to exceed the maximum limit value of 1 in addition to having the lowest sum of absolute errors (SAE) and sum of squared errors (SSE) from the control systems that were simulated.