Advanced Control for Double Acting Pneumatic Valve Positioning System

This study investigates the control of the nozzle flapper system in the pneumatic positioning for valves, where the Proportional-Integral-Derivative known as PID is the most used controller. Enhancing performance remains challenging, with questions about whether the limits stem from software, electronic control systems, mechanical design, or environmental factors. Hence, four advanced control methods are used, namely, genetic-algorithms based PID tuning (GA-PID), fuzzy logic (FL), PID fuzzy logic (FPID), and adaptive neuro-fuzzy inference systems (ANFIS). Controller’s synthesis aims to enhance the accuracy of pneumatic valve positioners, specifically focusing on position control of a nonlinear double-acting butterfly valve, to get the fastest response time with the lowest overshoot. Based on practical tests and theoretical modelling review, the whole system including pneumatic, mechanical, and electrical components has a nonlinear behaviour in addition to position related motion hysteresis, which makes practical frequency system identification methods not applicable. After mathematical modelling and approximations, a linear second-order model with an integrator structure was selected. For the system identification this structure is used to get the parameters of the proposed linear model around the middle position as an operating point. The listed control techniques are employed to determine the optimal controller. A comparative study assessed their performance in reducing response time and minimizing overshoot in pneumatic positioning systems. Additionally, a novel software and hardware design scheme for a digital valve positioner was developed, integrating these control methods on the STM32 microcontroller. Experimental results demonstrate that the use of ANFIS approach shows high performance, offering a more efficient solution in term of precise control in nonlinear pneumatic valve positioning systems.