Design Algorithm for Unbalance Compensation for Magnetic Rotor–Bearing System

Unbalance-induced vibrations remain a critical challenge in high-speed rotor systems supported by active magnetic bearings (AMBs). This study presents an optimization-based procedure for systematically designing the parameters of an Unbalance Force Rejection Controller (UFRC). The UFRC is capable of suppressing synchronous currents caused by rotor mass unbalance and resonance, thereby improving energy efficiency and reducing vibration-induced noise. The proposed method requires only four inputs: the system's sensitivity function, minimum and maximum operating speeds, and the allowable peak sensitivity. The proposed method is optimization-based, ensuring that the designed parameters are optimal with respect to the objectives and constraints. Furthermore, the method is nonparametric, it does not require a full system model and system parameters, making it highly suitable for industrial applications. The optimized UFRC parameters were experimentally validated on the shaft of a magnetically levitated chiller system. Results show that, after activating the UFRC, the magnitude of the control current was reduced by 67% compared to the case without compensation and approached the same level as static levitation, showing the effectiveness of the optimized parameters.