Fuzzy Logic vs. Traditional Techniques for PI Controller Tuning of a Propeller-Driven Pendulum

Abstract

This paper presents a comparative study of fuzzy and traditional techniques for tuning a PI controller for a propeller-driven pendulum. Two conventional tuning methods, the Ziegler–Nichols method and the PID Tuner tool in MATLAB/Simulink, are analyzed and used for PI controller design. The operating principle of the propeller-driven pendulum is described, and a nonlinear mathematical model of the system is shown. The model is subsequently linearized around multiple operating points and implemented in MATLAB/Simulink. To improve control performance over the entire operating range, a gain-scheduling strategy is employed alongside the conventionally tuned PI controller, where controller gains are adjusted according to the current operating point. In addition, a fuzzy-logic-based PI controller is designed using a set of proposed fuzzy rules. The three control approaches are implemented and evaluated through simulation for different reference values. The results demonstrate that the fuzzy-logic-based controller and the PID Tuner-based PI controller achieve comparable performance and significantly outperform the Ziegler–Nichols-tuned PI controller in terms of overall control quality.