CN-121984580-A - Method, system, equipment and medium for detecting optical fiber loss event classification

CN121984580ACN 121984580 ACN121984580 ACN 121984580ACN-121984580-A

Abstract

The embodiment of the invention provides a method, a system, equipment and a medium for detecting optical fiber loss event classification, belonging to the technical field of optical fiber line detection. The method comprises the steps of obtaining multiband echo data of an optical fiber to be tested under a preset wavelength set, constructing a cross-band alignment relation based on the multiband echo data, executing amplitude comparability processing to generate multiband comparability echo data, detecting event candidate bands based on the multiband comparability echo data, extracting corresponding multiband feature sets, and outputting event position and event category results based on the multiband feature sets. According to the scheme, through cross-band distance alignment and amplitude comparability processing, the spatial consistent expression of the multi-band echoes at the same physical event is realized, so that the positioning accuracy of the optical fiber event and the reliability of event type judgment are improved.

Inventors

ZHOU QIANG
ZHANG QI
Zheng Sigang
XIAO YAO
ZHANG DENGKE
Fan Yunru

Assignees

天府绛溪实验室

Dates

Publication Date: 20260505
Application Date: 20260113

Claims (10)

1. A method for classifying and detecting fiber loss events, the method comprising: acquiring multiband echo data of an optical fiber to be tested under a preset wavelength set; Constructing a cross-band alignment relation based on the multi-band echo data, and executing amplitude comparability processing to generate multi-band comparability echo data; detecting event candidate bands based on the multiband comparable echo data and extracting a corresponding multiband feature set; and outputting event positions and event category results based on the multiband feature set.
2. The method for classifying and detecting optical fiber loss events according to claim 1, wherein acquiring multiband echo data of an optical fiber to be measured at a preset wavelength set comprises: Based on a preset wavelength set, controlling a multiband detection light source to sequentially inject detection light with corresponding wavelength into an optical fiber to be detected or according to a preset time sequence so as to form a multiband detection process under the same measurement link condition; In the process of injecting detection light of each wave band, respectively collecting a back scattering signal and a reflection signal generated by propagation along an optical fiber to be detected, and marking the collected back scattering signal and reflection signal according to corresponding wave bands to form echo sequence data of each wave band independently; and uniformly mapping the echo sequence data of each wave band to a distance dimension to generate multi-wave band echo data taking the optical fiber distance as an independent variable and the echo amplitude as a dependent variable.
3. The method of claim 1, wherein constructing a cross-band alignment relationship based on the multi-band echo data and performing an amplitude comparison process to generate multi-band comparable echo data comprises: selecting a preset reference wave band from the multi-wave band echo data, and identifying echo positions corresponding to preset reference events in the echo data of each wave band, wherein the echo positions are used for representing the distance corresponding relation of the same physical position under different wave bands; constructing a cross-band distance alignment relation based on the corresponding relation between the echo position of the reference band and the corresponding echo positions in other bands, and performing distance axis correction on echo data of non-reference bands according to the cross-band distance alignment relation to obtain multi-band echo data after distance alignment; And based on the reference amplitude information corresponding to each wave band, performing amplitude normalization processing on the multi-band echo data after the distance alignment to generate multi-band comparable echo data.
4. The method of claim 3, wherein performing amplitude normalization processing on the multi-band echo data after the distance alignment based on the reference amplitude information corresponding to each band to generate multi-band comparable echo data, comprises: For each wave band, acquiring reference amplitude information corresponding to echo data of the wave band, wherein the reference amplitude information is used for representing the reference amplitude level of the wave band under the current measurement condition; determining amplitude normalization parameters of corresponding wave bands based on the reference amplitude information, wherein the amplitude normalization parameters are used for eliminating amplitude deviation caused by transmission power difference or link response difference between different wave bands; And respectively executing amplitude normalization processing on the multi-band echo data after the distance alignment according to the amplitude normalization parameters, so that the echo data of different bands have comparability under the unified amplitude reference, and multi-band comparable echo data are generated.
5. The method of claim 1, wherein detecting event candidate segments and extracting corresponding multiband feature sets based on the multiband comparable echo data comprises: selecting a preset detection wave band from the multiband comparable echo data as an event detection reference, performing change analysis on the multiband comparable echo data along a distance dimension, identifying an echo change interval meeting a preset change criterion or a preset reflection criterion, and generating a corresponding event candidate segment; performing cross-band mapping on the event candidate segments in the multi-band comparable echo data, and determining echo intervals corresponding to the event candidate segments in each band; And extracting multiband characteristic parameters for representing the event in the echo intervals of the corresponding wave bands aiming at each event candidate wave band, and aggregating the multiband characteristic parameters to form a corresponding multiband characteristic set.
6. The method for classifying and detecting optical fiber loss event according to claim 5, wherein the preset change criteria are: Comparing the echo data with adjacent sampling points along the distance dimension to obtain an adjacent differential sequence, wherein each differential value of the adjacent differential sequence represents the difference value of the echo amplitude value at the adjacent distance position; when absolute values of differential values of the adjacent differential sequences in the number of preset continuous sampling points are all larger than a preset change threshold, determining a corresponding distance range as an echo change interval; the preset reflection criteria are: Identifying a local peak in echo data, and setting preset background windows on two sides of the peak position of the local peak to obtain a background amplitude reference, wherein the background amplitude reference is obtained by echo amplitude statistics in the preset background window; And when the difference value between the peak value amplitude of the local peak and the background amplitude reference is larger than a preset reflection threshold value and the distance span corresponding to the local peak at the preset amplitude proportion is smaller than a preset peak width threshold value, determining the distance range corresponding to the local peak as an echo change interval.
7. The method of claim 1, wherein outputting event location and event category results based on the multiband feature set comprises: correspondingly reading a multiband feature set of each event candidate band, and constructing a distinguishing feature vector for representing the event characteristic based on the multiband feature set, wherein the distinguishing feature vector is formed by combining a plurality of feature parameters in the multiband feature set; Comparing the distinguishing feature vector with the preset event category feature description one by one, and determining the event category corresponding to the event candidate segment according to the matching degree between the distinguishing feature vector and each event category feature description; And determining the event position corresponding to the event candidate segment based on the position range of the event candidate segment in the multi-band comparable echo data after the distance alignment, correlating the event position with the event category, and outputting a corresponding event position and event category result.
8. A fiber loss event classification detection system, the system comprising: the acquisition unit is used for acquiring multiband echo data of the optical fiber to be detected under a preset wavelength set; The processing unit is used for constructing a cross-band alignment relation based on the multi-band echo data and executing amplitude comparability processing to generate multi-band comparability echo data; a feature extraction unit for detecting an event candidate band based on the multiband comparable echo data and extracting a corresponding multiband feature set; And the output unit is used for outputting event positions and event category results based on the multiband feature set.
9. An electronic device, comprising: One or more processors; Storage means having stored thereon one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the fiber loss event classification detection method of any of claims 1-7.
10. A computer readable storage medium having instructions stored thereon, which when run on a computer causes the computer to perform the fiber optic loss event classification detection method of any of claims 1-7.

Description

Method, system, equipment and medium for detecting optical fiber loss event classification Technical Field The invention relates to the technical field of optical fiber line detection, in particular to a method, a system, equipment and a medium for detecting optical fiber loss event classification. Background When the optical fiber circuit runs for a long time in communication and monitoring applications, loss events such as fiber breakage, bending, abnormal connection and the like are easily generated due to construction, environmental change or external force action, and detection and analysis on fiber loss distribution and event positions are required. The existing optical fiber detection method is mostly based on an optical time domain reflection principle, and realizes optical fiber loss measurement and fault positioning by analyzing a backward scattering signal and a reflected signal generated by propagation along an optical fiber. In practical application, a single-wavelength optical fiber detection mode has limitation in event type distinction, and different types of events may be similar in echo characteristics and difficult to distinguish stably. In order to acquire more detection information, a multiband detection mode is introduced in part of the technical scheme, but the problems of inconsistent distance axes and incomparable amplitude scales are commonly existed among echo data in different wave bands, so that events in the same physical position are difficult to directly correspond to each other in different wave bands, and the reliability of event detection and classification is affected. Therefore, in the multi-band optical fiber detection scene, how to effectively align and process the amplitude of echo data in different bands, and realize event candidate segment detection and event category judgment on the basis of the alignment and the amplitude, and further research and improvement are still needed. Disclosure of Invention The embodiment of the invention aims to provide a method, a system, equipment and a medium for detecting optical fiber loss event classification, which are used for at least solving the problems of alignment and comparability of multi-band echo data. In order to achieve the above purpose, the first aspect of the invention provides an optical fiber loss event classification detection method, which comprises the steps of obtaining multi-band echo data of an optical fiber to be detected under a preset wavelength set, constructing a cross-band alignment relation based on the multi-band echo data and executing amplitude comparability processing to generate multi-band comparability echo data, detecting event candidate bands based on the multi-band comparability echo data and extracting corresponding multi-band feature sets, and outputting event positions and event category results based on the multi-band feature sets. The method comprises the steps of acquiring multiband echo data of an optical fiber to be detected under a preset wavelength set, controlling a multiband detection light source to inject detection light with corresponding wavelengths into the optical fiber to be detected sequentially or according to a preset time sequence based on the preset wavelength set so as to form a multiband detection process under the same measurement link condition, respectively acquiring backscattering signals and reflection signals generated by propagation along the optical fiber to be detected in the detection light injection process of each waveband, marking the acquired backscattering signals and reflection signals according to corresponding wavebands to form independent echo sequence data of each waveband, and uniformly mapping the echo sequence data of each waveband to a distance dimension to generate multiband echo data taking the optical fiber distance as an independent variable and the echo amplitude as a dependent variable. The method comprises the steps of selecting preset reference wave bands from multi-wave band echo data, identifying echo positions corresponding to preset reference events in the multi-wave band echo data, representing distance corresponding relations of the same physical position under different wave bands, constructing a cross-wave band distance alignment relation based on the corresponding relations between the echo positions of the reference wave bands and the corresponding echo positions in other wave bands, carrying out distance axis correction on echo data of non-reference wave bands according to the cross-wave band distance alignment relation to obtain multi-wave band echo data after distance alignment, and carrying out amplitude normalization processing on the multi-wave band echo data after distance alignment based on reference amplitude information corresponding to each wave band to generate multi-wave band comparable echo data. Optionally, amplitude normalization processing is performed on the multi-band echo data after the distance alignment