Recognition of a Phase-Sensitivity OTDR Sensing System Based on Morphologic Feature Extraction

Abstract

This paper proposes a novel feature extraction method for intrusion event recognition within a phase-sensitive optical time-domain reflectometer (Φ-OTDR) sensing system. Feature extraction of time domain signals in these systems is time-consuming and may lead to inaccuracies due to noise disturbances. The recognition accuracy and speed of current systems cannot meet the requirements of Φ-OTDR online vibration monitoring systems. In the method proposed in this paper, the time-space domain signal is used for feature extraction instead of the time domain signal. Feature vectors are obtained from morphologic features of time-space domain signals. A scatter matrix is calculated for the feature selection. Experiments show that the feature extraction method proposed in this paper can greatly improve recognition accuracies, with a lower computation time than traditional methods, i.e., a recognition accuracy of 97.8% can be achieved with a recognition time of below 1 s, making it is very suitable for Φ-OTDR system online vibration monitoring.

Long-range distributed optical fiber hot-wire anemometer based on chirped-pulse ΦOTDR

“…For example, when working in intrusion detection via ΦOTDR, a further analysis of the measurements (beyond simple thresholding of the ΦOTDR data) is frequently performed. Typically, pattern recognition and machine learning strategies are applied [45,46]. Such techniques generally work much better with a high-fidelity input, and keeping the sampling rate as high as possible.…”

Machine Learning Methods for Pipeline Surveillance Systems Based on Distributed Acoustic Sensing: A Review

“…Although some works have been presented in this direction [3,[12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27], there is still, many times, a lack of understanding of how a reliable and efficient system can be built, and what is the adequate and rigorous methodology to apply in the system evaluation processes. In addition, to the best of our knowledge, no thorough written report that summarizes the main characteristics of these systems exists in the literature applied to DAS+PRS technology.…”

“…In others, there were no details of the system description [30], there were no details on the classification strategy [26], or not enough details of the experimental procedure (training and testing conditions, recording protocol, etc.) were given [19,21,26,43].…”

“…Some recent works present significant improvements over those previously reported, by generating recording environments with longer optical fibers (in the range of tens of kilometers), such as [45] (17 km), [21] (20 km), [27] (24 km), [46] (50 km), and [26] (220 km), thus approaching the idea of a more realistic environment. Also, their experimental procedure is more rigorous [45,46], even applying cross-validation (CV) techniques [21].…”

“…Also, their experimental procedure is more rigorous [45,46], even applying cross-validation (CV) techniques [21]. However, some of them generate all the measurements from a single position [27,46], hence biasing the system to recognize the position instead of the real event, which we demonstrated in [23] that was a major issue when facing realistic environments.…”