GA tuned LQR controller for active vibration control of a rotor bearing system using piezoelectric actuators

Abstract

This paper proposes a vibration identification method based on a Genetic Algorithm (GA) to improve the reliability of rotating machinery by effectively controlling rotor vibrations. The classical active control approach, known as the Linear Quadratic Regulator (LQR) control algorithm, is influential in this regard but heavily depends on specific parameters. However, using a predetermined set of parameters may not suffice to address the diverse vibration control needs of different rotating systems. To address this issue, an active vibration control system employing flexible piezoelectric patches attached to the external surface of the rotating shaft, functioning as actuators, is designed. These patches are individually supplied with actuation voltages. Through a numerical example, the effectiveness of the proposed system is validated. The numerical results demonstrate a significant reduction of the amplitude of the unbalance response, with a reduction of 79% at the first critical speed during a steady-state condition and 75% during run-up. The numerical findings underscore the effectiveness of the proposed control strategy (LQR-GA), highlighting its notable dynamic impact compared to the conventional LQR algorithm with fixed parameters.