Method and system for vibration signal pattern recognition of distributed optical fiber sensor

Abstract

The invention discloses a vibration signal pattern recognition method and system of a distributed optical fiber sensor. The original back Rayleigh scattering signal is obtained through a signal acquisition terminal to form a time-space two-dimensional matrix; when vibration occurs, a signal including the vibration signal is obtained. The time-space two-dimensional matrix is ​​mapped to the corresponding gray level co-occurrence matrix, and edge detection is performed to obtain the gradient matrix; the time accumulation of the gradient matrix can obtain the output of the vibration signal space-amplitude curve, and the vibration position can be obtained from the gradient matrix. The vibration time-amplitude signal at the location; the signal is subjected to feature extraction and classifier classification and identification. The present invention can accurately locate the vibration signal in a complex and high-noise environment, at the same time classify and identify the vibration signal, and has a high classification accuracy.

Description

technical field [0001] The invention relates to the fields of perimeter security and energy pipeline intrusion detection, in particular to a vibration signal pattern recognition method and system of a distributed optical fiber sensor. Background technique [0002] Optical fiber sensing technology uses optical fiber as the sensing medium. It has been widely used in industry, municipal transportation, aviation due to its strong anti-electromagnetic interference ability, intrinsic safety, convenient laying, good concealment, and the ability to work in harsh conditions. aerospace and military. Among them, distributed optical fiber sensing has been used in many occasions such as perimeter security, real-time detection of energy transmission pipelines, submarine optical cables, and smart grids due to its advantages of distributed and long-distance measurement. [0003] The distributed optical fiber vibration sensor based on the principle of phase-sensitive optical time domain ref...

AI Technical Summary

Problems solved by technology

[0006] The technical problem to be solved by the present invention is: the sensor needs to effectively extract the vibration positioning point and obtain the time-amplitude signal about the vibration in a complex and high-noise environment; due to the impact, randomness, and complex frequency components of the vibration signal and other characteristics, often appearing as non-stationary signals, so choosing a reasonable feature extraction method and classifier can ensure high classification accuracy

Images