Fault Diagnosis Techniques for Permanent Magnet Synchronous Machines

Permanent Magnet Synchronous Machines (PMSMs) are integral to electric vehicles, renewable-energy installations and aerospace applications, valued for their high efficiency and power density. They are, however, prone to electrical faults such as stator inter-turn short circuits, magnetic faults including partial or uniform demagnetisation of rotor magnets, and mechanical defects like bearing failure and eccentricity. Early detection of these faults is crucial to prevent downtime, reduce maintenance costs and extend machine life. Fault diagnosis techniques can be grouped into model-based, signal-processing and data-driven approaches. Model-based methods employ observers or Kalman filters on mathematical machine models to generate residuals that indicate anomalies. Signal-processing strategies analyse stator currents and vibration signals in the time, frequency or time–frequency domains—using Fourier transforms, wavelet packet decomposition or motor current signature analysis—to extract sensitive diagnostic features. Data-driven techniques harness machine learning and deep learning algorithms, including convolutional neural networks, transfer learning and classifiers such as k-nearest neighbours or multilayer perceptrons, to recognise fault patterns across varying operational conditions. Recent hybrid schemes that fuse electrical and mechanical signals further enhance robustness. Together, these advances promise more accurate, faster and cost-effective monitoring solutions that underpin the reliability of PMSM-driven systems worldwide.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Fault Diagnosis Techniques for Permanent Magnet Synchronous Machines publication trend

The graph below shows the total number of articles in fault diagnosis techniques for permanent magnet synchronous machines across all publications each year (not limited to Nature Index journals).

Technical terms

Permanent Magnet Synchronous Machine (PMSM): An electric motor design that uses permanent magnets on the rotor to create a constant magnetic field and operates synchronously with the rotating magnetic field of the stator.

Model-based diagnostics: Techniques that employ mathematical representations of machine dynamics and observers or filters to generate residuals used for fault detection.

Time–frequency analysis: Signal-processing methods, such as the short-time Fourier transform and wavelet transforms, that decompose signals to reveal temporal variations of spectral content.

Motor Current Signature Analysis (MCSA): A diagnostic approach that examines frequency components of stator current to detect characteristic fault harmonics.

Convolutional Neural Network (CNN): A deep learning architecture that applies convolutional layers to extract hierarchical features from input data for classification tasks.

k-Nearest Neighbours (kNN): A non-parametric machine learning classifier that assigns a class to a sample based on the majority class of its nearest neighbours in feature space.

References

1. Faults and Diagnosis Methods of Permanent Magnet Synchronous Motors: A Review. Applied Sciences (2019).
2. Detection and Identification of Demagnetization and Bearing Faults in PMSM Using Transfer Learning-Based VGG. Energies (2020).
3. Fault Detection of Stator Inter-Turn Short-Circuit in PMSM on Stator Current and Vibration Signal. Applied Sciences (2018).
4. Motor Current Signature Analysis-Based Permanent Magnet Synchronous Motor Demagnetization Characterization and Detection. Machines (2020).
5. Fault Diagnosis of PMSG Stator Inter-Turn Fault Using Extended Kalman Filter and Unscented Kalman Filter. Energies (2020).
6. On-line Detection and Classification of PMSM Stator Winding Faults Based on Stator Current Symmetrical Components Analysis and the KNN Algorithm. Electronics (2021).
7. Demagnetization Fault Diagnosis of Permanent Magnet Synchronous Motors Based on Stator Current Signal Processing and Machine Learning Algorithms. Sensors (2023).
8. Fault Detection and Diagnosis of the Electric Motor Drive and Battery System of Electric Vehicles. Machines (2023).