Comparative study of PID and Fuzzy control in thermal and hydraulic systems.

• Grant number: 26/18221-3
• Support Opportunities: Research Grants - Regular Innovation Grant - AIR
• Start date: October 01, 2026
• End date: March 31, 2028
• Field of knowledge: Engineering - Production Engineering
• Principal Investigator: Waldemar Bonventi Júnior
• Grantee: Waldemar Bonventi Júnior
• Host Institution: Faculdade de Tecnologia de Sorocaba (FATEC Sorocaba). Centro Paula Souza (CEETEPS). Sorocaba , SP, Brazil
• City of the host institution: Sorocaba

Abstract

Temperature control in furnaces and ovens, as well as level control in tanks with variable inlet and outlet flow rates, are essential in industrial processes that transform raw materials into finished products. Process control is fundamental in the chemical, food, and metallurgical industries. The PID controller is widely used due to its simplicity and efficiency, yet it faces limitations in non-linear systems. Fuzzy logic emerges as a robust alternative capable of handling uncertainties and environmental variations (Costa, 2016; Sinzato & Moers, 2025). Furthermore, Fuzzy modeling incorporates expert knowledge using linguistic expressions, such as "If the temperature is high, then decrease furnace power" or "If the tank level is high, then decrease the inlet flow rate." A comparative study of PID and Fuzzy control in thermal systems contributes to practical training in automation and to Industry 4.0 applications by incorporating cloud-based information exchange (IoT - Internet of Things). The PID (Proportional-Integral-Derivative) controller is the most widely used control algorithm in the industry, found in approximately 84% of industrial applications (Song, 2016). Its popularity stems from its ease of implementation and robustness across various processes. PID combines three actions: proportional, which reacts to the instantaneous error; integral, which eliminates steady-state error; and derivative, which anticipates error trends, thereby improving stability (Aström & Hägglund, 2004). Proper tuning of the $K_p$, $K_i$, and $K_d$ parameters is essential to ensure satisfactory performance. Fuzzy logic, proposed by Zadeh (1965), allows for the handling of uncertainties and linguistic variables such as "low," "medium," and "high." Unlike PID, which relies on precise mathematical models, Fuzzy logic uses heuristic rules to determine the control action. This approach is particularly useful for nonlinear systems or those that are difficult to model mathematically (Gomide & Gudwin, 1994). Fuzzy control has been successfully applied in industrial processes, offering greater robustness and flexibility (Lee, 1990). Thus, according to Silva (2025), while PID is effective for simple systems, Fuzzy control demonstrates greater robustness against noise and disturbances. Scilab/Xcos and Octave are free tools that enable the simulation of both controllers (Silva, 2025; GitHub - Schuenck, 2024). This project aims to compare the performance of PID and fuzzy controllers in thermal and hydraulic systems. The comparative study of PID and Fuzzy control in these systems contributes to practical training in automation and to Industry 4.0 applications, including cloud-based information exchange (IoT - Internet of Things). The objective is to analyze which controller-PID or Fuzzy-performs better in thermal or hydraulic systems, considering ease of implementation and clarity of parameterization. Both control methods will be applied to the system. The study expects to identify their respective advantages and limitations, contribute to research on industrial system automation, and evaluate their feasibility and applicability in real-world production systems. (AU)