Model Predictive Control for Active Magnetic Bearings

Active magnetic bearings (AMBs) have lots of advantages than conventional mechanical bearings so that they are being applied to various industrial rotating machineries. However, AMB system is a complex and nonlinear open-loop unstable mechatronic system. For most AMBs, the bias current of actuator is needed which means the input constraint. In this paper, a model predictive control (MPC) technique is developed for eight-pole AMB system with input constraints to improve the ability of resisting external disturbances, e.g. sensor noise and external force. And stability analysis is given and shown that this method enlarging the region of attraction. Finally, the performance of the proposed control method is verified via the simulations.