Multifrequency Vibration Suppression of Magnetic Bearing Systems Applied Variable Step-size Automatic Learning Control

Active Magnetic Bearing (AMB) has been widely applied in occasions acquire high-speed rotating, high-precision machining, and so on. The AMB can provide active control force to offset centrifugal force and other disturbance to suppress vibration while rotating. This paper proposed a strategy to compensate vibration in AMB system based on Variable Step-size Automatic Learning Control (VALC). Firstly, a two degrees of freedom model of radial AMB is established. Phase error between of centrifugal force and synchronous vibration is analyzed. Then, principle and process of VALC is introduced. A generalized notch filter is applied to obtain synchronous and harmonic vibration. Next, simulations are carried out based on AMB control system applied VALC. The results of simulations show the feasibility of the compensating strategy. Experiments to reduce synchronous and harmonic vibration based on VALC are carried out on a test rig. Spectrums of vibration signals show that vibration is reduced by about 90%, which demonstrates the effectiveness of proposed strategy for vibration compensation.