AI-DRIVEN FAULT DETECTION AND AUTONOMOUS SELF-HEALING IN SMART GRIDS
Abstract
The modernization of electrical power systems into smart grids has led to significant improvements in efficiency, reliability, and renewable integration. However, smart grids’ increasing complexity and dynamic operation pose critical challenges in fault detection, diagnosis, and restoration. Traditional protection systems often rely on fixed thresholds and manual interventions, resulting in delayed fault handling and prolonged outages. This paper proposes an integrated artificial intelligence (AI) framework that combines deep learning with reinforcement learning-based autonomous self-healing control for real-time fault detection, classification, and localization. The deep learning models exploit time-series sensor data to accurately identify and classify various fault types, while the reinforcement learning agent optimizes switching operations to isolate faults and restore power without human intervention. The framework is validated on IEEE 33-bus and 69-bus test systems, achieving fault detection accuracy above 95% and localization errors below 5% of line length. Compared with traditional methods [1,2], which typically achieve 85%–90% accuracy and require manual fault isolation, the proposed system reduces outage durations by up to 40%, demonstrating substantial improvements in operational efficiency. This research lays the groundwork for scalable AI-driven fault management solutions that are adaptable to evolving smart grids.
Keywords:
smart grid, fault detection, deep learning, reinforcement learning, self-healing, power system automationDownloads
Published
DOI:
https://doi.org/10.5281/zenodo.16894058Issue
Section
How to Cite
License
Copyright (c) 2025 Modu Abba Gana, Muhammad Musa
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
References
Zhang, Y., Li, H., & Wang, J. (2023). Deep learning-based fault diagnosis for smart grids. IEEE Transactions on Power Systems, 38(2), 1234–1245.
Liu, X., Y. Chen, and M. Zhao. 2022. LSTM-based fault classification in distribution systems. Renewable Energy 185: 445–456.
Fang, X., Misra, S., Xue, G., Yang, D., 2012. Smart grid —New and improved power grid: A survey IEEE Communications Surveys and Tutorials, 14(4), 944–980.
Momoh, J. A. (2012). Smart grid design for efficient and flexible power networks operation and control. IEEE/PES Power Systems Conference, 1–8.
Amin, M., and Wollenberg, B. (2005). Toward a smart grid: power delivery for the 21st century. IEEE Power & Energy Magazine, 3(5), 34–41.
Gonen, T. (1986). Electric Power Distribution System Engineering. McGraw-Hill, New York.
Das, J. C. (2010). Modern Techniques of Power System Analysis. McGraw-Hill.
Pal, B. C., et al. (2010). Challenges in Distribution System State Estimation with High Penetration of Distributed Generation. IEEE Power and Energy Society General Meeting, 1–6.
Sharma, A., Singh, R., & Kumar, P. (2021). Deep CNN for fault diagnosis in smart grids Electric Power Systems Research, Vol. 193, pp. 107058.
Singh, S., & Chatterjee, K. (2020). LSTM networks for fault detection in power systems. IEEE Access, 8, 120345–120355.
Mao, C., Wang, X., & Sun, Y. (2022). Reinforcement learning-based self-healing in distributed networks IEEE Transactions on Smart Grid, 13(1), 345–356.
Chen, Z., Li, J., & Wu, Q. (2023). RL agents for autonomous network reconfiguration in smart grids. IEEE Trans. Power Syst., 38(1), 567–577.
Li, J., Zhang, Y., & Chen, L. (2022). AI methods for fault classification in power systems. Renewable and Sustainable Energy Reviews 155: 111925.
Wang, H., & Zhou, X. (2024). RL-based self-healing control strategies for smart grids. Electric Power Systems Research, vol. 215, pp. 108739.
