Design of industry-centric controller for MIMO CSTH process with enhanced disturbance rejection

This paper focuses on designing an advanced control scheme tailored for largescale industrial processes, where controllers must maintain effective performance despite significant disturbances and setpoint changes. The primary focus of the proposed RTD-A controller is on robust disturbance rejection. RTDA possesses the benefits of both conventional PID and MPC control schemes. As model-based methods face challenges in addressing increasingly complex processes, data-driven techniques have gained popularity in industrial system monitoring due to their ability to handle unknown physical models. In this work, both the first-principle and transfer function models of the CSTH system are developed using real-time data and represented as a multi-input, multi-output (MIMO) system. PID, MPC, and RTD-A controllers are then applied to regulate the temperatures of the two tanks. The performance of these controllers is carefully examined using integral performance criteria and the time domain analysis to accurately assess their dynamic behavior and control precision. The results demonstrate that the RTD-A controller exhibits superior performance in mitigating disturbances. The RTD-A control strategy exhibits outstanding performance with near-zero overshoot (0% in servo and about 0.05% in regulatory responses) and stable settling times close to 430 - 440 seconds in both tanks. Although MPC and PID controllers offer quicker responses, their greater overshoot and longer settling times establish RTD-A as the preferred method for achieving reliable, precise, and safe control in industrial processes.