Modeling and Control of Magnetic Bearings With Nonlinear Magnetization (Simulation vs. Experiment)

This paper is a continuation of the work presented in (Gerami,,et al., 2015), which explored the use of the Lur’e formulation to achieve increased dynamic load capacity in active magnetic bearings (AMBs). The increase in the load capacity was achieved by developing a nonlinear model that enabled the system to operate in the nonlinear region of the magnetization curve. While the previous work focused on the theoretical development of the modeling and control approach, this paper verifies the results by experiments conducted on a rocking beam test rig. The experiment demonstrates that the proposed modeling and control approach significantly improves the transient response and disturbance rejection capabilities of the magnetic bearing system compared with present industrial practice. Based on the excellent agreement between simulation and experimental results, it is shown that existing industrial AMBs could be modified to be significantly more resilient to unknown external disturbances, and therefore result in fewer shutdowns and less associated damage to backup bearings. In addition, for new AMB applications, smaller AMBs could be designed using this method which would be able to handle the same worst-case disturbance forces as larger AMB systems based on standard modeling and control methods.