Research on Power Grid Outbreak Identification Technology Based on Distributed Optical Fiber Sensing

With continuous expansion of power grid scale, the power grid security is facing unprecedented challenges, especially the increasingly accidents caused by external forces, which pose a serious threat to the stable operation of the power system. To overcome the limitations of traditional monitoring methods, this study innovatively integrates the Distributed Fiber Optic Vibration Sensing System (DVS) with the Long Short-Term Memory - Convolutional Neural Network (LSTM-CNN) algorithm model, developing a new type of detection technology for external damage to the power grid. It utilizes the distributed measurement characteristics of DVS, enabling real-time monitoring and accurate capture of weak signals generated by external damage. At the same time, combined with the data processing and pattern recognition capabilities of the LSTM-CNN model, it achieves high-precision detection of external damage events to the power grid, with a recognition accuracy rate of over 98%. The research results indicate that this detection technology, which combines DVS with LSTM-CNN, not only significantly improves the detection efficiency of external damage events but also effectively reduces the rates of missed alarms and false alarms, which is of great significance for enhancing the security protection capability of the power grid.