A deep transfer learning approach with a compact convolutional neural network architecture for robust high impedance fault detection in energy distribution grids | Synapse
March 3, 2026
A deep transfer learning approach with a compact convolutional neural network architecture for robust high impedance fault detection in energy distribution grids
Key Points
High impedance fault detection accuracy increased significantly using deep transfer learning techniques, with notable improvements over traditional methods.
A 95% detection rate was achieved in simulated energy distribution grid scenarios during extensive testing.
Application of a compact convolutional neural network architecture allows efficient processing and swift identification of faults.
Results support the potential for real-time implementations in modern energy distribution systems, enhancing reliability.