Precise Fault Distance Prediction in Distributed Generators-Integrated Transmission Lines Using CNN-LSTM Ensemble Networks

This is the authors' abstract. We don't add key points for this paper.

Precise fault distance estimation is critical for effective transmission system protection, especially in modern power systems with integrated distributed generators (DGs). As the global demand for energy continues to rise, the adoption of renewable energy sources and their integration into the grid has significantly increased. However, the presence of various DGs complicates fault distance estimation due to the diverse power flows and interactions between DGs and existing grid infrastructure. The integration of large-scale DGs introduces additional challenges in fault detection and location, as the dynamic behavior of these generators affects the fault characteristics. This paper focuses on precise fault distance estimation within the IEEE-9 bus system, which includes two distinct DGs. The proposed solution leverages an ensemble deep learning approach, combining convolutional neural networks (CNNs) and long short-term memory (LSTM) networks. This hybrid CNN-LSTM model takes advantage of CNN’s ability to extract spatial features and LSTM’s strength in capturing temporal dependencies, making it well-suited for fault distance prediction. The model’s performance is evaluated using mean absolute error (MAE), giving 0.6387 average MAE for the studied system. Comparative analysis with existing deep learning models and validated through OPAL-RT 4510 real-time simulator shows that the proposed model delivers superior performance, demonstrating its effectiveness in handling the complexities introduced by DG integration.