Interpretable Time-Frequency Features for the Classification of Power Quality Disturbances
Overview & Synopsis
Power Quality (PQ) disturbances in power transmission systems affect manufacturing processes, cause equipment malfunctions, and lead to economic losses. This paper proposes ten new interpretable features, grounded in both mathematical and application perspectives, for the classification of PQ disturbances, including voltage sags, swells, interruptions, harmonics, and combined disturbances. The proposed features are derived from different linear time-frequency transforms, enabling a comparative analysis of their sensitivity with respect to the chosen representation. To evaluate their effectiveness, a machine learning model is trained on these features using synthetic signals and tested on more realistic data generated from the IEEE 9-bus system. The results demonstrate high classification accuracy even under low SNR conditions, reaching up to 99.4% at 5 dB when combining the optimized time-frequency representation with a Random Forest classifier, and achieving near-perfect classification under ideal conditions (SNR = 40 dB). In addition, experiments on the IEEE 9-bus system highlight the model's ability to generalize in more physically consistent scenarios. A comparison with state-of-the-art studies demonstrates the competitiveness of the proposed features in this paper with respect to existing approaches.
hal.science)
← / → for pages • + / - for zoom • 0 to fit width
Powered by Mozilla PDF.js v3.11
Document Information
| Document Title | Interpretable Time-Frequency Features for the Classification of Power Quality Disturbances |
|---|---|
| Author / Organization | Gandorj Darambazar, Ali Moukadem, Bruno Colicchio, Patrice Wira |
| Subject & Discipline | Multidisciplinary Academic Open Research |
| Categories | Science & Mathematics |
| Publication Year | 2026 |
| License / Rights | Open Access / Creative Commons |
| Resource Verification | Listed PDF (Verification Pending) |
| Original Host Website |
HAL Open Science
(hal.science)
|
| File Format | application/pdf (Portable Document Format) |
| Language | EN |
| Catalog ID | DOC-256064 |
| Original Source Link | Visit Source Page on hal.science » |