Research on fault identification methods for transmission line arc faults based on acoustic signals

The transmission line is a crucial component of the power system responsible for electricity transmission, and fault detection is vital for ensuring the safe and reliable operation of the power grid. When an arc fault occurs in the transmission line, it generates specific acoustic signals that are associated with the occurrence of the fault. This paper proposes a diagnostic method for transmission line faults based on acoustic signals from arcing. By establishing an experimental platform for fault arcing, we collected the acoustic signals generated during arc faults and conducted feature extraction and analysis of these signals. To address the issue of indistinct feature vectors, we optimized the extracted Mel-frequency cepstral coefficients using Principal Component Analysis (PCA), and subsequently built a fault recognition model based on Long Short-Term Memory (LSTM) networks. The research results indicate that the proposed feature vectors effectively reflect fault characteristics and can achieve efficient identification of line faults, which is significant for the stable operation of the power grid.