IMPLEMENTATION OF ADVANCED PROCESS CONTROL FOR FLOW CONTROL APPLICATION

In the process control industry in recent years, majority of control loops is using Proportional-Integral-Derivative (PID) controllers due to its structure that's easy to regulate the process output. The most difficult process variable (PV) to control is flow due to the property changes in fluid and fast response. Therefore when disturbance occurs, the system is becoming dynamic process. Thus, to reduce the waste, the conventional PID controller is not the best option. However, model predictive control (MPC) has a success story in the process industry and manufacturing of its high performance in controlling a system with multivariable control simultaneously. Therefore, this paper builds on the last three previous final year student that work on the comparison between PID, fuzzy logic and artificial intelligence controller. This work provides an overview of the basic concept of an MPC technology handling flow response. Based on the research done by previous students, Adaptive Neuro-Fuzzy Inference System (ANFIS) has the best control performance compared to Fuzzy Logic Control (FLC) and PID. Therefore this work also will illustrate the performance of MPC in terms of its stability and handling robustness compared to the conventional PID Controller. A general MPC control algorithm is developed using MATLAB/Simulink Toolboxes. After that, the designed controllers will be implemented in PcA SimExpert Mobile Pilot Plant SE231B-21-Flow Control and Calibration Process Unit. The Data Acquisition (DAQ) card is used for interfacing between hardware and software. The result will be analyzed to see the comparison of PID and MPC performance in flow control application. Based on the experiment results, MPC shows better performance compared to PID Controller.