ROBUST PI/PID CONTROL IN CYBER-PHYSICAL SYSTEM-BASED MINI BATCH DISTILLATION COLUMN

• Main Author: Sekar Ayu, Wirenda
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/41814
• Institution: Institut Teknologi Bandung

Interconnected system between computation and physical process (Cyber-Physical System) has been widely used in industrial processes. In CPS-based industrial process, sensors, controllers, and actuators are connected into a communication network. The communication network may introduce delay time uncertainties due to shared resources and load traffic in the network. Furthermore, the non-linear time-varying characteristic of batch distillation column may causes another uncertainties to take into account in control system design. Modeling error and delay process uncertainty is introduced due to linierized system approximation that unmodeled high-frequency dynamics. The dynamic uncertainty on both I/O channel are also introduced to the system uncertainty. In this research, robust PI/PID controller using AMIGO method with appropriate weighting function is designed to guarantee robust stability and robust performance spesification of batch distillation column. The impact of system uncertainties to closed-loop system performances is also investigated. MATLAB/Simulink simulation is used to validate the method before its implementation in CPS-based distillation column. Based on simulation and implementation, the proposed robust PI/PID controller gives better performances on integral error, distillate mixture ratio, and robustness criteria compared to conventional PI/PID controller. The results show that robust PI/PID can compensate the effect of system uncertainties to its performances.