The invention relates to the technical field of
optical fiber event detection, and discloses a phi-OTDR event detection method,
system and device and a medium, and the method comprises the steps: obtaining the
optical fiber signal data of an original phase-sensitive
optical time domain reflectometer, and carrying out the first
processing; constructing a
deep learning model based on the residual network, training the
deep learning model by using the
training set, evaluating the performance through the
verification set, and adjusting hyper-parameters according to an
evaluation result to optimize the model; and inputting the data after the first
processing into a trained
deep learning model based on a residual network to extract high-level features of the
optical fiber signals, and performing event identification and classification according to the high-level features. According to the method, the problems that the current OTDR event detection
signal is interfered in
noise, the model is insufficient in tiny event extraction capability, and the deep learning model is slow in convergence speed and poor in generalization capability are solved, automatic and accurate identification of the optical
fiber event is realized, and the efficiency and reliability of optical
fiber fault diagnosis are greatly improved.