CN-122027020-A - Intelligent counting analysis system based on optical fiber vibration sensing

CN122027020ACN 122027020 ACN122027020 ACN 122027020ACN-122027020-A

Abstract

The invention discloses a digital intelligent analysis system based on optical fiber vibration sensing, and relates to the technical field of optical fiber intelligent maintenance. The digital intelligent analysis system based on the optical fiber vibration sensing comprises an acquisition layer and a data layer, wherein the acquisition layer is used for acquiring optical fiber vibration signals and related data, the acquisition layer comprises a vibration monitor, an OTDR tester, an electronic tag reader-writer, GPS positioning equipment and image acquisition equipment, and the data layer is used for processing and storing the acquired data and comprises a signal preprocessing module, a feature extraction module, a vibration source identification model, an optical cable resource database and an early warning rule base. Based on optical fiber vibration sensing, the deep learning and multi-source data are fused, 8 vibration sources such as construction, damage and the like can be accurately identified, positioning error is less than or equal to 2.5 meters, and identification rate is more than 91%. The optical cable health can be early warned and evaluated in real time, the fault handling time is shortened to 45 minutes, the maintenance cost is reduced by 38%, and high-efficiency support is provided for intelligent operation and maintenance of the power optical cable.

Inventors

LIU HAO
LIU CHENG
SHEN ZHEN
Sui Lianzhu
XIE XIAOWEI
WANG SHAOPENG
SHI RAN
LI GUOXIN

Assignees

内蒙古电力(集团)有限责任公司呼和浩特供电分公司

Dates

Publication Date: 20260512
Application Date: 20251208

Claims (6)

1. A digital intelligent analysis system based on optical fiber vibration sensing, comprising: the acquisition layer is used for acquiring optical fiber vibration signals and related data and comprises a vibration monitor, an OTDR tester, an electronic tag reader-writer, GPS positioning equipment and image acquisition equipment; The data layer is used for processing and storing the acquired data and comprises a signal preprocessing module, a characteristic extraction module, a vibration source identification model, an optical cable resource database and an early warning rule base; The service layer is used for providing core business logic processing, including resource management service, vibration analysis service, early warning decision service and quality assessment service; the application layer is used for realizing man-machine interaction and external interfaces, and comprises a Web management interface, a mobile APP, a central control large screen and a third party system interface; the acquisition layer transmits the acquired data to the data layer, the data layer transmits the data to the service layer after processing the data, and the service layer transmits the processing result to the application layer for displaying and interaction.
2. The digital intelligent analysis system based on optical fiber vibration sensing as claimed in claim 1, wherein the vibration source identification model adopts a CNN-LSTM hybrid model, and the specific steps comprise: carrying out windowing and framing treatment on the acquired vibration signals, wherein a window function is a Hamming window, and the expression is as follows: Extracting multi-characteristic parameters of signals by utilizing a local characteristic scale decomposition method, fitting an extreme point envelope curve by means of cubic B spline function interpolation, wherein the basis functions of the cubic B spline function are as follows: The extracted features are input into a one-dimensional convolutional neural network for training and recognition, and the calculation formula of the convolutional layer is as follows: Wherein, the For the layer 1 jth feature map, For the number of feature maps of layer l-1, K is the convolution kernel size, For the convolution kernel weights, In order for the offset to be a function of, To activate the function.
3. The system for digital intelligence analysis based on fiber vibration sensing as set forth in claim 2, wherein the vibration source identification model further comprises an LSTM layer for processing time sequence characteristics, and the calculation formula of the LSTM unit is as follows: forgetting the door: an input door: Cell status update: cell status: Output door: Hidden state: Wherein, the As a function of the sigmoid, For the element-level multiplication to be performed, For the input at the time of t, The state is hidden for the moment t, The state of the cells at the time t, As a matrix of weights, the weight matrix, Is a bias vector.
4. The digital intelligent analysis system based on fiber vibration sensing of claim 1, wherein the signal preprocessing module comprises a phase demodulation process for vibration signals, and based on a phi-OTDR technology, external vibration is sensed by detecting phase change of Rayleigh scattered light in the fiber, and a calculation formula of the phase change is as follows: is the change in length of the optical fiber.
5. The system for intelligent analysis of optical fiber vibration sensing according to claim 1, wherein the resource management service comprises checking and managing the optical fiber resources, and the specific steps comprise: positioning the optical cable node by vibration knocking to obtain the optical length and the disc information of the node; reading NFC/RFID label information of the node, and acquiring an electronic identifier of the node; Acquiring longitude and latitude information of the node through GPS positioning equipment; acquiring a node site image through image acquisition equipment; acquiring fiber core quality data through an OTDR tester; And integrating the data and updating the optical cable resource database.
6. The digital intelligent analysis system based on optical fiber vibration sensing as set forth in claim 1, wherein the early warning decision service comprises a three-stage early warning mechanism, and generates an alarm, a main alarm and an emergency alarm three-stage early warning according to the vibration type and the intensity, wherein the calculation formula of early warning level judgment is as follows: Wherein R is an early warning grade, S is vibration intensity, T is vibration duration, D is the distance between a vibration point and an optical cable, I is an optical cable importance coefficient, Is a weight coefficient.

Description

Intelligent counting analysis system based on optical fiber vibration sensing Technical Field The invention relates to the technical field of intelligent maintenance of optical fibers, in particular to a digital intelligent analysis system based on optical fiber vibration sensing. Background In the current infrastructure construction and operation field, the safety monitoring requirement for various pipelines and facilities is urgent. Conventional monitoring approaches have a number of limitations. For example, in the aspect of monitoring the power optical cable, manual inspection is mostly relied on in the past, so that the efficiency is low, fine abnormal conditions are difficult to detect in real time, and the severe requirements of the power system on stability and reliability cannot be met. Even if a part of the sensor is used for monitoring, the sensor is often arranged sparsely and is easily interfered by the environment, so that the monitoring result is inaccurate and the early warning has hysteresis. With the rise of optical fiber sensing technology, the distributed optical fiber vibration sensing technology brings new ideas for solving the problems. The technology is based on the elasto-optical effect of the optical fiber, when external vibration acts on the sensing optical fiber, the refractive index and the length of the optical fiber can generate tiny changes, so that the light intensity and the phase of a transmission signal in the optical fiber are changed, and vibration detection and positioning can be realized by detecting the changes. However, existing monitoring systems based on fiber optic vibration sensing still face a number of challenges. On the one hand, in a complex environment, various interference sources exist, vibration signals generated by different interference sources are mutually interwoven, so that effective vibration signals (such as construction excavation, artificial destruction and other threatening vibrations) and noise interference are extremely difficult to accurately identify and distinguish, and the false alarm rate are high. For example, in urban areas, the environments such as vehicle running, wind blowing and the like vibrate frequently, so that vibration signals which are threatening to pipelines are easily covered, and the reliability of a monitoring system is affected. On the other hand, existing systems suffer from deficiencies in data processing and analysis capabilities. In the face of massive optical fiber vibration data, the processing and the depth analysis cannot be performed quickly and efficiently, accurate assessment and prediction of potential risks are difficult to achieve, and powerful support cannot be provided for operation and maintenance decisions. For example, in long-distance oil and gas pipeline monitoring, because the pipeline is complicated and changeable along the line environment, the generated vibration data volume is huge, and the traditional system is difficult to extract key information from the pipeline rapidly, so that potential safety hazards can be found in time. Disclosure of Invention Aiming at the defects in the prior art, the invention provides a digital intelligent analysis system based on optical fiber vibration sensing, which comprises the following components: the acquisition layer is used for acquiring optical fiber vibration signals and related data and comprises a vibration monitor, an OTDR tester, an electronic tag reader-writer, GPS positioning equipment and image acquisition equipment; The data layer is used for processing and storing the acquired data and comprises a signal preprocessing module, a characteristic extraction module, a vibration source identification model, an optical cable resource database and an early warning rule base; The service layer is used for providing core business logic processing, including resource management service, vibration analysis service, early warning decision service and quality assessment service; the application layer is used for realizing man-machine interaction and external interfaces, and comprises a Web management interface, a mobile APP, a central control large screen and a third party system interface; the acquisition layer transmits the acquired data to the data layer, the data layer processes the data and transmits the processed data to the service layer, and the service layer transmits the processing result to the application layer for display and interaction; preferably, the vibration source identification model adopts a CNN-LSTM hybrid model, and the specific steps include: carrying out windowing and framing treatment on the acquired vibration signals, wherein a window function is a Hamming window, and the expression is as follows: The method comprises the steps of obtaining a frame of a signal, wherein N is a sample index, N is a window length, extracting time-frequency domain characteristics of the signal after framing, wherein the time-frequency domain characte