Robust Fault Classification with High Accuracy and Low Latency for Wind-Integrated Transmission Lines Using LSTM

Wind-integrated power systems face unique challenges due to the intermittent nature of wind energy and the risk of faults. Precise and timely fault classification is crucial for maintaining system stability and minimizing outages. This research proposes a novel approach to fault classification, utilizing Long Short-Term Memory (LSTM) networks. LSTM's capacity to model sequential data and mitigate noise contributes to its suitability for achieving accurate and efficient fault classification. The proposed method utilizes current and voltage waveforms as input features, followed by LSTM training and testing. Experimental results demonstrate superior performance compared to other popular deep learning-based methods, achieving a fault classification accuracy of 99.20%.