Vibration Compensation Control of BPMSM With Dead-Time Effect Based on Adaptive Neural Network Band-Pass Filter

In order to solve the problem of rotor vibration for the bearingless permanent magnet synchronous motor, a vibration compensation control method based on adaptive neural network band-pass filter (ANNBPF) is proposed. First, the mechanism of rotor vibration is obtained by theoretical analysis. Second, the ANNBPF is used to extract the components related to the frequency of the unbalanced vibration and the dead-time (DT) vibration to realize the vibration compensation control. Additionally, the DT effect is compensated by detecting the current polarity to improve the compensation effect. Finally, a vibration compensation control system based on ANNBPF and current polarity detection is constructed.