Detecting Arcing Faults in Switchgear by Using Deep Learning Techniques

Switchgear and control gear are susceptible to arc problems that arise from slowly developing defects such as partial discharge, arcing, heating due faulty connections. These issues can now be detected monitored using modern technology. This study aims explore the effectiveness of deep learning techniques, specifically 1D-CNN model, LSTM 1D-CNN-LSTM in detecting arcing switchgear. The hybrid model was preferred for fault detection switchgear because its superior performance both time frequency domains, allowing analysis generated sound wave during an event. To investigate algorithms, experiments were conducted locate faults switchgear, domain analyses conducted. proved most effective differentiating between non-arcing situations training, validation, testing stages. Time (TDA) showed high success rates 99%, 100%, 98.4% 1D-CNN; LSTM; 100% distinguishing cases respective phases. Furthermore, (FDA) also demonstrated accuracy 95.8% Therefore, it concluded developed particularly effectively recognize providing efficient method monitoring systems.