Static Eccentricity Detection in Permanent Magnet Synchronous Machines based on Embedded Linear Hall Sensors

Manufacturing and assembly errors, long-term operation, and other factors may induce rotor static eccentricity (SE) in a permanent magnet synchronous motor (PMSM), which is a key drivetrain component in electric vehicles and rail transport systems. Further, SE generates an unbalanced magnetic pull that degrades motor efficiency and operational longevity, posing reliability risks in electrified transportation applications. Hence, proactive eccentricity detection is essential to ensure functional safety and operational availability. Traditional SE diagnosis methods are normally based on current or vibration signals, susceptible to misdiagnosis due to interference from load variations or other mechanical faults. While magnetic flux density (MFD)- based SE diagnostic methods benefit from signal stability, their previous implementations relied solely on MFD amplitude variation. This approach is susceptible to various disturbances, e.g., Hall sensor misplacement, sensor gain drift, temperature variations, and such a susceptibility frequently leads to misdiagnosis on SE. To overcome these limitations, this paper proposes a SE diagnosis method using embedded linear Hall sensors for in-situ flux monitoring. Firstly, the amplitude-phase dual modulation effect of SE on MFD at antipodal Hall sensing points is revealed. Then, an initial calibration approach to enhance the diagnostic accuracy is put forward by decoupling interference sources and compensating for component-level errors. Thirdly, an amplitude-phase dual diagnostic indicator is proposed, and a combined online monitoring and offline quantification framework is implemented, which enables simultaneous quantification of SE magnitude and directional orientation. Finally, the effectiveness of the proposed method is validated through finite-element analysis and experimental tests.