DATA DRIVEN ADAPTIVE CONTROL SYSTEM FOR CYBERPHYSICAL DISTILLATION COLUMN

• Main Author: Made Aswin Nahrendra, I
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/38933
• Institution: Institut Teknologi Bandung

A mini distillation column is used to separate two different mixtures from its’ liquid phase by using separation principle based on the different boiling point. However, a distillation column posed high complexity and difficult to be controlled to achieve a desirable performance. Non-linearity and time varying characteristics are the main problems which are caused by the variation of the raw material’s quantity and composition. To compensate such condition, one of the simplest approach is to regularly tune and maintain the controller to adapt with the input’s variations, however, it would be not efficient if implemented manually, i.e. by using a human operator. Therefore, a data driven adaptive control scheme which enabled the controller to identify the system’s model based on previous measurements data to tune the controller was proposed in this thesis research. There are three major steps for the controller, i.e. model estimation, controller Pengaturan using metaheuristics algorithm, and control signal computation. The results achieved in this research show the ability of the proposed adaptive controller to change its control gain to with subject to the system’s condition. The controller was implemented on a Cyber- Physical Distillation Column (CPDC) system which was connected to a database server and able to perform data mining, storage, and processing for the physical system to achieve a desirable performance.