EUVIMEDEUVIMED entdecken
Systematische ÜbersichtOpen Access · CC BY 4.0

A Systematic Review on Integrating Artificial Intelligence for Enhanced Fault Detection in Power Transmission Systems: A Smart Grid Approach

Nijat Taghizad, Pietro Lio

Computing&AI Connect · 2025

DOI: 10.69709/caic.2024.103229 ↗

Kurz gefasst

Systematische Übersicht in der Zeitschrift Computing&AI Connect, 2025. Verfasst von Nijat Taghizad und 1 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 24 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Modern electrical systems rely on sensors and relays for fault detection in three-phase transmission lines and distribution transformers, but these devices often face time complexity issues and false alarms. In this study, the fault detection accuracy is compared in models studied in 2023 and 2024 following PRISMA guidelines. The objectives were to identify fault types, utilize machine learning models to assess their predictive efficacy, and establish accuracy levels. To explore this further, a systematic literature review was performed based on AI accuracy in fault detection from scholarly articles published in reputable journals. The inclusion criteria required journals published between 2023 and 2024 that tested AI systems for three-phase transmission lines and distribution transformers, while sources older than 2023 were excluded. The selected journals used both simulated and real databases to assess AI-based fault detection accuracy. A total of 12 sources were searched, with two selected for comparative analysis based on their relevance to the study’s objectives. The risk of bias was assessed using the Robvis method. Findings were presented using narratives, graphs, and tables, and the results were synthesized through comparative data analysis. The Novel Glass Box-Based Model was ranked as the most accurate fault detection model (99% accuracy), followed by Convolutional Neural Network (98%), Gated Recurrent Unit (92%), Random Forest (90%), Logistic Regression (74%), and Support Vector Classifier (63%). Both selected studies compared older fault detection systems with AI-based models, demonstrating the superior accuracy of modern AI approaches. However, the study was limited by reliance on only two sources, which introduced potential bias in the exclusion criteria. The findings suggest that AI developers should aim for 99% accuracy in fault detection systems to meet industry requirements.

Zitieren

APA

Taghizad, N., & Lio, P. (2025). A Systematic Review on Integrating Artificial Intelligence for Enhanced Fault Detection in Power Transmission Systems: A Smart Grid Approach. Computing&AI Connect. https://doi.org/10.69709/caic.2024.103229

Vancouver

Taghizad N, Lio P. A Systematic Review on Integrating Artificial Intelligence for Enhanced Fault Detection in Power Transmission Systems: A Smart Grid Approach. Computing&AI Connect. 2025. doi:10.69709/caic.2024.103229

BibTeX
@article{Taghizad2025,
  author = {Taghizad, Nijat and Lio, Pietro},
  title = {A Systematic Review on Integrating Artificial Intelligence for Enhanced Fault Detection in Power Transmission Systems: A Smart Grid Approach},
  journal = {Computing&AI Connect},
  year = {2025},
  doi = {10.69709/caic.2024.103229}
}
Kostenlos starten →