ABSTRACT
A distributed fiber optic sensor based on dual-Michelson interferometers for disturbance localization and pattern recognition is proposed. The system obtains the phase difference of each of the two interferometers using a passive demodulating algorithm based on a 3×3 coupler. Two correlation signals with disturbance position information are obtained by delaying and subtracting the phase difference signals through which the disturbance location can be obtained. This method has the same frequency response over the whole sensing path, and there is no localization blind spot. The pattern recognition method of this sensing system is to obtain the spectral signal by fast Fourier transform of the demodulated interferometer phase information and input it as a feature vector into a one-dimensional convolutional neural network to verify the correct rate of pattern recognition for four behaviors: stepping, shearing, sweeping, and shaking. The total transmission distance of the system can reach 100 km; the location errors are within ±35m, and the correct rate of four pattern recognitions is higher than 97%. The sensing system has good polarization stability, which can ensure the stability of long-term operation and has a broad application prospect in long-distance perimeter security.
ABSTRACT
A distributed optic fiber perimeter security system is proved to be an effective strategy for the security monitoring of some vital targets, such as power plants, power substations and telecommunication base stations. However, this method can hardly distinguish different categories of the intrusion behavior and is easily mis-triggered by different kinds of environmental interference. To distinguish different intrusion patterns and different interference events effectively, a vibration pattern recognition algorithm is proposed and demonstrated based on the merged Sagnac interferometer structure. The method consists of two parts: the pre-processing algorithm and the multi-layer perceptron neural networks (MLP-NNs). The pre-processing algorithm is applied to retrieve and extract the vibration signal from the captured source signal, and the MLP-NN is used to realize pattern recognition from each type of input. Typically, a high-dimensional vector group which contains hundreds of orders of vibration signal's power frequency is obtained to cover as many signalized features as possible. Moreover, results of the experiment deployed on a 10 kilometer long perimeter fence in the transformer substation show that the proposed classification-based model achieves 97.6% classification accuracy in the test. Through multiple comparison tests, the proposed model gives a solid performance in the subsequent integrated evaluation to classify each intrusion pattern.
