Linear and Nonlinear Feature Extraction for Transformer Partial Discharge Severity Classification: A Comparative Study Using Artificial Neural Networks.
Saved in:
| Title: | Linear and Nonlinear Feature Extraction for Transformer Partial Discharge Severity Classification: A Comparative Study Using Artificial Neural Networks. |
|---|---|
| Authors: | Thobejane, Lucas1 (AUTHOR), Thango, Bonginkosi A.1 (AUTHOR) bonginkosit@uj.ac.za |
| Source: | Energies (19961073). Jun2026, Vol. 19 Issue 11, p2642. 29p. |
| Subject Terms: | *Feature extraction, *Partial discharges, *Nonlinear analysis, *Dimensional reduction algorithms, *Autoencoders, *Artificial neural networks, *Classification, *Principal components analysis |
| Abstract: | Accurate classification of transformer partial discharge (PD) severity is essential for insulation diagnostics yet remains challenging due to nonlinear feature relationships and class imbalance. This study evaluates whether feature extraction improves PD severity classification and compares the effectiveness of linear and nonlinear extraction methods. A dataset of 294 samples was categorized into four IEC-aligned severity classes. Two raw measurements (discharge magnitude and applied voltage) were expanded into a 15-dimensional feature space. Principal Component Analysis (PCA) and a bottleneck Autoencoder (AE) were used for linear and nonlinear feature extraction, respectively. Extracted features were classified using an identical Multilayer Perceptron (MLP). Both feature extraction methods improved classification performance over raw and full-feature baselines (96.6%). PCA+ANN achieved 100.0% accuracy (k = 9), while AE+ANN achieved 98.3% (k = 8). The AE misclassified one minority "Normal" sample due to poor latent boundary representation. Reconstruction analysis showed the highest error for the Normal class, reflecting imbalance-driven optimization bias. Feature extraction enhances PD severity classification, with linear PCA outperforming nonlinear AE in this near-linearly separable dataset. PCA's deterministic projection preserves minority class boundaries more effectively, whereas AE performance is limited by class imbalance. These findings suggest that nonlinear methods provide advantages only in more complex feature spaces. [ABSTRACT FROM AUTHOR] |
| Database: | Energy & Power Source |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Text: Availability: 1 |
|---|---|
| Header | DbId: enr DbLabel: Energy & Power Source An: 194588030 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Linear and Nonlinear Feature Extraction for Transformer Partial Discharge Severity Classification: A Comparative Study Using Artificial Neural Networks. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Thobejane%2C+Lucas%22">Thobejane, Lucas</searchLink><relatesTo>1</relatesTo> (AUTHOR)<br /><searchLink fieldCode="AR" term="%22Thango%2C+Bonginkosi+A%2E%22">Thango, Bonginkosi A.</searchLink><relatesTo>1</relatesTo> (AUTHOR)<i> bonginkosit@uj.ac.za</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22Energies+%2819961073%29%22">Energies (19961073)</searchLink>. Jun2026, Vol. 19 Issue 11, p2642. 29p. – Name: Subject Label: Subject Terms Group: Su Data: *<searchLink fieldCode="DE" term="%22Feature+extraction%22">Feature extraction</searchLink><br />*<searchLink fieldCode="DE" term="%22Partial+discharges%22">Partial discharges</searchLink><br />*<searchLink fieldCode="DE" term="%22Nonlinear+analysis%22">Nonlinear analysis</searchLink><br />*<searchLink fieldCode="DE" term="%22Dimensional+reduction+algorithms%22">Dimensional reduction algorithms</searchLink><br />*<searchLink fieldCode="DE" term="%22Autoencoders%22">Autoencoders</searchLink><br />*<searchLink fieldCode="DE" term="%22Artificial+neural+networks%22">Artificial neural networks</searchLink><br />*<searchLink fieldCode="DE" term="%22Classification%22">Classification</searchLink><br />*<searchLink fieldCode="DE" term="%22Principal+components+analysis%22">Principal components analysis</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Accurate classification of transformer partial discharge (PD) severity is essential for insulation diagnostics yet remains challenging due to nonlinear feature relationships and class imbalance. This study evaluates whether feature extraction improves PD severity classification and compares the effectiveness of linear and nonlinear extraction methods. A dataset of 294 samples was categorized into four IEC-aligned severity classes. Two raw measurements (discharge magnitude and applied voltage) were expanded into a 15-dimensional feature space. Principal Component Analysis (PCA) and a bottleneck Autoencoder (AE) were used for linear and nonlinear feature extraction, respectively. Extracted features were classified using an identical Multilayer Perceptron (MLP). Both feature extraction methods improved classification performance over raw and full-feature baselines (96.6%). PCA+ANN achieved 100.0% accuracy (k = 9), while AE+ANN achieved 98.3% (k = 8). The AE misclassified one minority "Normal" sample due to poor latent boundary representation. Reconstruction analysis showed the highest error for the Normal class, reflecting imbalance-driven optimization bias. Feature extraction enhances PD severity classification, with linear PCA outperforming nonlinear AE in this near-linearly separable dataset. PCA's deterministic projection preserves minority class boundaries more effectively, whereas AE performance is limited by class imbalance. These findings suggest that nonlinear methods provide advantages only in more complex feature spaces. [ABSTRACT FROM AUTHOR] |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=enr&AN=194588030 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.3390/en19112642 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 29 StartPage: 2642 Subjects: – SubjectFull: Feature extraction Type: general – SubjectFull: Partial discharges Type: general – SubjectFull: Nonlinear analysis Type: general – SubjectFull: Dimensional reduction algorithms Type: general – SubjectFull: Autoencoders Type: general – SubjectFull: Artificial neural networks Type: general – SubjectFull: Classification Type: general – SubjectFull: Principal components analysis Type: general Titles: – TitleFull: Linear and Nonlinear Feature Extraction for Transformer Partial Discharge Severity Classification: A Comparative Study Using Artificial Neural Networks. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Thobejane, Lucas – PersonEntity: Name: NameFull: Thango, Bonginkosi A. IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 06 Text: Jun2026 Type: published Y: 2026 Identifiers: – Type: issn-print Value: 19961073 Numbering: – Type: volume Value: 19 – Type: issue Value: 11 Titles: – TitleFull: Energies (19961073) Type: main |
| ResultId | 1 |
