Fuzzy Logic Control Techniques for compensation of Stiction in a Pneumatic control valve

Abstract

Stiction in the pneumatic control valve is one of the biggest reasons for creating oscillations in the process variable (PV) and corresponding controller output (OP) in a process plant. Oscillation in the control loops diminishes economical benefits of the process plant because of quickened wear and tear of the valve stem, lower quality products, and higher rejection rate. For industrial scenarios, the loss in performance due to stiction is a serious problem, which ought to be resolved. In the present work, the oscillatory behavior due to the sticky pneumatic control valve in a laboratory scale pressure control unit has been resolved by utilizing fuzzy logic control techniques and these have been addressed as; newline 1. Fuzzy proportional integral (PI) controller. newline2. Fuzzy integral plus conventional proportional plus derivative (Fuzzy I+PD) controller. newline3. Fuzzy gain scheduling of proportional plus integral plus derivative (FGS PID) controller. newline4. Fuzzy gain scheduling of integral minus proportional minus derivative (FGS I-P-D) newline controller. newlineThe fuzzy PI is a simple-structure controller, which damps out the oscillations in PV and OP. The fuzzy I+PD controller is developed by cascading a fuzzy integral controller with a conventional proportional plus derivative controller. FGS PID and FGS I-P-D are adaptive controllers that tune the gain parameters of conventional PID and I-P-D controllers in real time to manage the stiction in the pneumatic control valve. newlineThe performance of above-mentioned fuzzy controllers has been compared to conventional PI, PID, and I-P-D controllers for setpoint tracking capability, robustness, and external disturbance rejection at different operating points. The experimental results reveal that the proposed fuzzy controllers outperform the conventional controllers in all aspects of the investigation performed and suppress oscillations in the PV and OP efficiently. Further, the comparison among fuzzy controllers has been done and it has been observed that the performance of the adaptive FGS I-P-D controller has been superior at all the operating points. These controllers have been implemented using Matlab/Simulink (2011b) software, having fourth-order Runge-Kutta ODE solver with sampling rate of 0.01 seconds. The uniqueness of the proposed fuzzy control techniques is that they do not require any additional compensating component and provide a standalone solution to mitigate stiction newline newline