A new simulation-driven method for diagnosing mixed eccentricity in PMSMs using stray magnetic fields

Existing data-driven methods for mixed eccentricity (ME) fault diagnosis in permanent magnet synchronous motors (PMSMs) face the challenge of limited large-scale labeled datasets. This paper proposes a novel simulation-driven approach for diagnosing mixed eccentricity faults in permanent magnet synchronous motors (PMSMs) using stray magnetic field data. By leveraging a finite element model and radial basis function neural network, our method achieves accurate predictions of eccentricity parameters. The results demonstrate the effectiveness of our approach, providing a non-intrusive and simulation-based solution for fault diagnosis in PMSMs. This study contributes to the growing field of predictive maintenance and Industry 4.0, enabling more efficient and reliable motor operation.