Design of controllers for the rotary inverted pendulum

Abstract

The rotary inverted pendulum is a typical nonlinear control system widely studied in the field of automatic control, and is
commonly used to test and evaluate various control algorithms. Due to its inherent instability and strong nonlinearity, controlling this system presents significant challenges, requiring effective control strategies to maintain the pendulum in the upright
equilibrium position. Numerous control methods have been applied to this system, including classical algorithms such as PID,
state feedback, and Linear Quadratic Regulator (LQR), as well as modern approaches like fuzzy control and model predictive
control (MPC). This study focuses on the design, simulation, and performance comparison of two primary controllers: LQR
and fuzzy PID. The control algorithms are validated through MATLAB Simulink simulations and real-time experiments using
the Quanser QUBE-Servo 2 module. Evaluation criteria include stability, adaptability, and sensitivity to disturbances. The
results show that the LQR controller delivers high performance with minimal oscillations and a short settling time, while the
fuzzy PID controller demonstrates better adaptability to varying environmental conditions and unexpected disturbances.