Control Strategies for the Quadruple Tank System: A Comparative Study of PID and MPC

This is the authors' abstract. We don't add key points for this paper.

Control engineering faces unique challenges when dealing with interconnected tank systems, particularly those with multiple tanks. Precise and robust control is challenging due to the interdependence of water levels between tanks, leading to potential inefficiencies within the system. To address this issue, Quadruple Tank Systems (QTS) is used to study interconnected tank systems and develop effective control approaches. The QTS, comprised of four interconnected tanks, necessitates precise control of water levels in its lower tanks. To assess controller efficacy within the QTS, two case studies were tested under three control scenarios: Open Loop system, Proportional-Integral-Derivate (PID) Controller, and Model Predictive Controller (MPC). The tuning parameters for PID and MPC is determined heuristically, and the study found that MPC outperformed PID with a significant settling time of 169.38s during case 1 and 165.11s during case 2. The steady-state error has been eliminated completely, ensuring that the water level can be accurately regulated and sustained at the intended setpoint of 8cm.