Análise e implementação de estratégias de controle para mancais magnéticos em sistemas rotativo

Abstract

In this work, the mathematical modeling of a rotor–bearing system and the implementation of three advanced control strategies for the dynamic positioning of the rotating shaft are presented. The system is composed of two active magnetic bearings and a rigid rotor positioned horizontally. The bearing dynamics are described based on Maxwell’s equations, while the rotor is represented by the Newton–Euler equations of motion. From the developed formulation, a nonlinear model of the assembly is obtained, as well as a linearized representation expressed in state-space form, which is used for controller design. The adopted strategies are LQG (Linear Quadratic Gaussian) optimal control, 𝐻∞ control, and MPC (Model Predictive Control). The development is carried out using the Matlab/Simulink® software, within which the tests and validation of the controllers are performed through their application to the nonlinear model of the system. The performance of each controller is analyzed based on quantitative criteria related to transient response and steady-state behavior. The results obtained under different simulation scenarios demonstrated the superiority of the MPC in reference-tracking tests, where it achieved the best response times and the lowest integral performance indices based on the squared error. In contrast, the LQG and 𝐻∞ controllers showed superior performance in disturbance rejection tests, with the 𝐻∞ controller outperforming the other strategies on average across nearly all evaluated criteria.