This paper proposes a long distance pipeline threat event identification method based on Gramian Angular Field (GAF) and deep learning networks for phase sensitive Rayleigh distributed fiber optic vibration system. Eleven types of threat events of vibration signals along a 28.9 km natural pipeline are collected and converted to two-dimensional images using GAF. The performance of five deep learning networks: GoogLeNet, VGG, AlexNet, ShuffleNet, and DenseNet are compared for events classification. The results show that DenseNet achieved the highest classification accuracy within the eleven types of threat events. Moreover, the false alarm rates of manual digging and mechanical excavation among the eleven events were 0.31% and 0.18%, respectively. The proposed method is suitable for complicated pipeline safety monitoring niche and offers fast and accurate recognition solution for long distance pipeline safety monitoring.

A terahertz photonic crystal fiber with two sensing channels was designed. Graphene coated on the micro-grooves in the cladding was used as plasma material to introduce tunability. The dispersion relation, mode coupling, and sensing characteristics of the fiber were studied using the finite element method. Ultrahigh sensitivity of 2.014 THz/RIU and 0.734 GHz/°C were obtained for analytes with refractive index in the range of 1.33 to 1.4 and environment temperature in the range of 10–60 °C, respectively. Refractive index resolution can reach the order of 10⁻⁵. The dual parameter simultaneous detection, dynamic tunable characteristics, and working in the low-frequency range of terahertz enable the designed photonic crystal fiber to have application prospects in the field of biosensing.