Model-based analysis of an industrial PLC-controlled dead-time thermal system

Temperature control is widely used in industrial processes, where Programmable Logic Controller (PLC)-based PID controllers are preferred for their robustness and ease of implementation. However, their compact structure and parameter tuning require a good understanding of system dynamics. While PLC-based thermal control systems have been widely reported, an experimentally validated model-based control (MBC) approach integrated with a comprehensive performance evaluation of industrial controllers appears to be lacking in the literature. This study presents the experimental and theoretical investigation of a PLC-based thermal system implemented with a Siemens S7-1200 PLC and a two-degree-of-freedom (2-DoF) PID controller. The experimental thermal system comprises a PT100 sensor and a 100 W resistive heater mounted in a prismatic plexiglass chamber (20 × 13 × 9.5 cm) with an insulated bottom surface. The system output is the temperature measured by the PT100 sensor, while the heater is driven by a pulse-width modulation (PWM) signal through a solid-state relay. Various open-loop and closed-loop tests are conducted on the experimental system. The thermal process is first characterized experimentally via system identification to obtain a mathematical model from input-output data. Although the thermal plant inherently exhibits nonlinear behavior due to transport delay and heat-transfer mechanisms, an equivalent linear model with a delay time is identified to facilitate controller synthesis using classical linear control theory. The accuracy of the identified model is experimentally verified through open-loop tests using an industrial commissioning methodology based on sequential pretuning and classical fine-tuning. Then, theoretical analysis is conducted to further evaluate the performance of the auto-tuned controller in closed-loop tests. Based on the validated model, equivalent FD-PID controllers are designed using controller parameters obtained from the PLC auto-tuning procedure. The closed-loop responses obtained from the MBC simulations are compared with the experimental results. The results demonstrate that the experimental dynamic behavior of the thermal system is accurately predicted by the MBC, despite the presence of unknown internal parameters in the Siemens PLC PID_Compact block. In addition, the effect of the setpoint weighting parameters of a 2-DoF PID controller on control performance is experimentally investigated to further improve the performance achieved by the auto-tuning procedure. The closed-loop performance specifications are evaluated using normalized results to ensure a fair comparison despite different initial temperatures. Overall, the proposed MBC design approach provides an effective method for the analysis and design of PLC-controlled thermal systems across a range of operating temperatures.