<![CDATA[Level and flow control systems are widely used in industrial processes, where poor control performance may cause process instability, energy inefficiency, and reduced product quality. This study aims to design and compare the performance of Model Predictive Control and Fuzzy-PID in a cascade level-flow control system based on models obtained through system identification. An experimental simulation approach was conducted, including system identification, process modeling, controller design, and performance evaluation. Black-box modeling was performed using bump test data processed in the MATLAB System Identification Toolbox, producing level and flow models with average fit estimations of 87.54% and 85.19%, respectively. The identified models were used to develop both controllers. MPC was designed with a prediction horizon of 40, a control horizon of 3, and a sampling time of 0.01 s, while Fuzzy-PID employed a fuzzy inference system to tune PID parameters based on error and change of error. Simulation results show that both controllers achieved the desired setpoint. However, MPC provided superior performance with no overshoot, a settling time of approximately 8 s, and a steady-state error of 0.0001. In contrast, Fuzzy-PID produced 24% overshoot, a settling time of 28 s, and oscillatory responses under disturbances. These results indicate that MPC is more effective and reliable for cascade level-flow control applications in industrial systems.]]>
Copyrights © 2026
