Application of a Hybrid Deep Model based on Harmonic Distortion for High-Impedance Fault Detection

Distribution systems face significant challenges in dealing with high impedance faults (HIF) due to the unpredictable, asymmetric, and nonlinear characteristics of these faults. Identifying HIFs in power networks is critical to prevent device damage and human safety. Most conventional detection and classification HIF methods, use a vast dataset and mathematical transformations, but intrinsic waveform characteristics are not considered. In this paper, a novel framework including a bidirectional long short-term memory (BiLSTM) algorithm based on convolution neural networks (CNN), incorporating harmonic distortion (HD) is proposed for HIF detection and classification along with the influence of background noise. The simulation results indicate the efficiency of the proposed method, which can identify HIFs. The evaluation results indicate that the proposed method outperforms baselines regarding precision and accuracy.