Application of Artificial Neural Network in Detecting, Locating, Classifying And Isolating AC Faults in Grid-Integrated Microgrid

In recent years, the integration of artificial intelligence (AI) techniques in power systems has gained significant traction, particularly for enhancing the reliability and efficiency of fault management processes. This paper presents an innovative approach for fault detection, location, and isolation on the AC side of power systems using Artificial Neural Networks (ANN). Traditional methods for fault management often encounter challenges such as delayed response times, low accuracy, and the inability to adapt to complex and dynamic system conditions. The proposed ANN-based methodology addresses these limitations by leveraging the pattern recognition and learning capabilities of neural networks. The research encompasses the development of a robust ANN model trained with a comprehensive dataset comprising various fault scenarios, including short circuits, open circuits, and ground faults. The dataset also incorporates variations in load conditions and system configurations to ensure the model's adaptability and accuracy. The performance of the ANN model is evaluated through extensive simulations conducted on a representative power system model, demonstrating its ability to accurately detect, locate, and isolate faults with minimal latency. Results indicate that the ANN approach outperforms conventional fault management techniques in terms of detection speed and precision, significantly reducing the downtime and operational risks associated with faults. Additionally, the proposed method offers the flexibility to be integrated into existing power system infrastructures with minimal modifications.