EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A New Deep Learning Method for Classification of PQ Disturbances Using DWT–IPCA in Renewable-Integrated Microgrids

Arslan Ali Memon, Ashfaque Ahmed Hashmani, Irfan Ali Channa, Suhail Khokhar, Aslam Pervez Memon, Mohsin Ali Koondhar, Lutfi Albasha, Ezzeddine Touti

IEEE Access · 2026 · 14 · 85214-85228

DOI: 10.1109/access.2026.3699524 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Arslan Ali Memon und 7 weiteren Autor:innen. Die Arbeit nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In real-time multiple energy systems, power quality causes severe problems from generation to consumption of electrical energy. Intelligent classification of power quality disturbances (PQDs) is key step to address PQ problems in renewable-integrated microgrids. In traditional framework, process of intelligent classification of PQDs consist three stages: analyzing signals, extract and optimize key features and classify the disturbance. However, the limitations of signal analysis, coupled with the complexity and inaccuracy of manual feature selection, result in reduced classification accuracy for multiple disturbances and diminished noise immunity. Given the inherent complexity and non-stationary characteristics of power quality disturbances, this study presents a hybrid methodology that leverages Discrete Wavelet Transform (DWT) for extracting salient time-frequency features. To address the high dimensionality and potential redundancy of these features, an enhanced Principal Component Analysis (PCA) is applied, ensuring the retention of critical information while reducing computational overhead. The resulting feature set is then processed by a deep convolutional neural network (CNN), where each module integrates one-dimensional convolution, pooling, and batch normalization layers to effectively capture multi-scale features and counteract overfitting. The deep architecture, composed of multiple such layers, facilitates the autonomous learning of robust and abstract representations from extensive disturbance datasets. The proposed method is evaluated on both synthetic and real-time PQ data from the experimental microgrid setup. Results show that the proposed model outperforms existing state-of-the-art approaches in accuracy and generalization.

Zitieren

APA

Memon, A. A., Hashmani, A. A., Channa, I. A., Khokhar, S., Memon, A. P., Koondhar, M. A., Albasha, L., & Touti, E. (2026). A New Deep Learning Method for Classification of PQ Disturbances Using DWT–IPCA in Renewable-Integrated Microgrids. IEEE Access, 14, 85214-85228. https://doi.org/10.1109/access.2026.3699524

Vancouver

Memon AA, Hashmani AA, Channa IA, Khokhar S, Memon AP, Koondhar MA, et al. A New Deep Learning Method for Classification of PQ Disturbances Using DWT–IPCA in Renewable-Integrated Microgrids. IEEE Access. 2026;14:85214-85228. doi:10.1109/access.2026.3699524

BibTeX
@article{Memon2026,
  author = {Memon, Arslan Ali and Hashmani, Ashfaque Ahmed and Channa, Irfan Ali and Khokhar, Suhail and Memon, Aslam Pervez and Koondhar, Mohsin Ali and Albasha, Lutfi and Touti, Ezzeddine},
  title = {A New Deep Learning Method for Classification of PQ Disturbances Using DWT–IPCA in Renewable-Integrated Microgrids},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {85214--85228},
  doi = {10.1109/access.2026.3699524}
}
Kostenlos starten →