Distributed vibration fiber based method and system for monitoring cable passage from external damage

Abstract

The present application relates to the technical field of cable monitoring, and particularly relates to a cable passage anti-external damage monitoring method and system based on distributed vibration optical fibers, which comprises: collecting multi-source heterogeneous data in subareas, calculating section risk scalars, and generating multi-dimensional regional risk prior constraint factors; configuring differentiated acquisition control parameters for each spatial detection unit, combining a pulse digital coding phase optical time domain reflection architecture to acquire multi-dimensional state data for the spatial detection unit; calculating a multi-dimensional state Euclidean distance to generate a global state distance matrix, and calculating an external damage index to screen potential risk units; analyzing disturbance propagation trajectories, filtering out interference signals, and outputting noise reduction trajectory features; calculating temperature deviations through a temperature prediction neural network, fusing the external damage index, calculating a comprehensive risk index, and outputting a risk warning result. The present application establishes a subarea differentiated acquisition and multi-parameter neighborhood fusion and judgment mechanism, and solves the technical problem of high false and missed alarm rates.

Description

Technical Field

[0001] This invention relates to the field of cable monitoring technology, and in particular to a method and system for monitoring cable channels against external damage based on distributed vibration optical fibers. Background Technology

[0002] Urban cable tunnels, cable trenches, and underground optical fiber cable linear channels are critical, concealed infrastructure for urban energy and information transmission. These pipelines are laid over long distances, cover wide areas, and are exposed to complex surrounding conditions. They are constantly subjected to municipal construction, surface loads, geological changes, and complex natural environments, making external damage a major cause of pipeline failures. Currently, the monitoring of external damage to underground linear channels in China generally adopts a combined operation and maintenance model of traditional fiber optic vibration monitoring, fixed-point video monitoring, and manual inspection. This overall monitoring system has many technical shortcomings.

[0003] Existing monitoring systems mostly adopt uniform acquisition thresholds and fixed sampling frequencies across the entire pipeline, failing to implement zoned control based on differences in pipeline location risks. This results in insufficient monitoring resolution for high-risk sections with dense construction activity, and excessive redundant acquisition in low-risk sections such as green belts and suburban areas, leading to unreasonable allocation of system computing resources. Furthermore, traditional vibration monitoring relies solely on single-point vibration amplitude thresholds to determine anomalies, failing to utilize the spatial correlation characteristics of pipeline neighborhoods. This makes it unable to identify weak, gradual geological subsidence disturbances and highly susceptible to environmental factors such as wind and rain, vegetation movement, and small animal contact, resulting in frequent false alarms and missed alarms.

[0004] Furthermore, existing monitoring technologies primarily collect single physical quantities such as vibration and temperature independently, lacking a multi-parameter fusion and analysis mechanism, and thus failing to distinguish between actual construction and excavation disturbances and natural environmental interference. Traditional equipment lacks disturbance trajectory analysis and noise adaptive suppression capabilities, cannot quantify disturbance propagation patterns, and suffers from limited risk identification dimensions and assessment accuracy. Overall, existing monitoring technologies suffer from insufficient differentiated perception capabilities, poor anti-interference capabilities, and limited risk assessment dimensions, making it difficult to meet the refined, comprehensive, and highly reliable external damage prevention monitoring and maintenance needs of long-distance underground cable tunnels. Summary of the Invention

[0005] This invention addresses the technical problems of insufficient monitoring accuracy in high-risk areas, large data redundancy in low-risk areas, and inability to effectively distinguish between environmental interference and real external disturbances, resulting in high false alarm rates and poor accuracy in risk assessment by establishing a zone-differentiated data collection and multi-parameter neighborhood fusion judgment mechanism based on spatial risk prior constraints.

[0006] The technical solution proposed in this invention is: a method for monitoring external damage to cable channels based on distributed vibration optical fibers, the method comprising: The fiber optic line is divided into several spatial detection units, and multi-source heterogeneous data are collected respectively. The segment risk scalar corresponding to each spatial detection unit is calculated by combining historical data, and the corresponding multi-dimensional regional risk prior constraint factor is generated based on the segment risk scalar. Based on multi-dimensional regional risk prior constraint factors, differentiated acquisition control parameters are configured for each spatial detection unit, and multi-dimensional state data is acquired by dividing the spatial detection units into spatial detection units using a pulse digital coding phase light temporal reflectance architecture. Based on multidimensional state data, a multidimensional state vector is constructed for each spatial detection unit. The multidimensional state Euclidean distance between any two spatial detection units is calculated to generate a global state distance matrix. The local outlier factor algorithm is used to calculate the external failure index and screen potential risk units. Vibration data of different pipeline points under continuous time sequence of potential risk units are extracted, disturbance propagation trajectory is analyzed, dynamic threshold is generated to filter out interference signals in combination with real-time environmental data, and noise-reduced trajectory features are output. Based on noise reduction trajectory characteristics, section risk scalars, and real-time cable temperature, temperature deviation is calculated through a temperature prediction neural network, and the external damage index is fused to calculate a comprehensive risk index, outputting early warning results for pipeline external force damage risks.

[0007] Preferably, the process for obtaining the multi-dimensional regional risk prior constraint factors is as follows: Collect multi-source basic data and historical disturbance sample data of optical fiber lines. The multi-source basic data includes spatial geographic data, regional construction registration data and real-time environmental monitoring parameters. After normalizing the multi-source basic data of the optical fiber line and the historical disturbance sample data, feature fusion is completed, and the segment risk scalar corresponding to each spatial detection unit is solved by weighted mapping operation. The parameters representing the regional risk level, historical disturbance probability, and environmental disturbance intensity in the segment risk scalar are encapsulated into a unified parameter set to generate multi-dimensional regional risk prior constraint factors for corresponding spatial detection units. Each multi-dimensional regional risk prior constraint factor is independent of the others.

[0008] Preferably, the process of obtaining the multidimensional state data is as follows: Based on multi-dimensional regional risk prior constraint factors, differentiated acquisition control parameters including vibration trigger sensitivity threshold, time series data parsing period, and signal feature extraction depth are calculated. The differentiated acquisition control parameters are imported into the pulse digital code phase light time domain reflectance sensing architecture, and the sensing fiber is partitioned into spatial detection units for synchronous signal acquisition. By relying on the vibration triggering sensitivity threshold to remove weak noise below the reference amplitude, a fixed-length continuous time-series data stream is extracted according to the time-series data parsing period. The data stream is extracted based on the signal characteristics and the pipeline state variables are extracted simultaneously. The pipeline state variables are collected and encapsulated to generate multi-dimensional state data.

[0009] Preferably, the process of obtaining the global state distance matrix is ​​as follows: Based on multidimensional state data, a multidimensional state vector corresponding to the dimensions of pipeline state quantities is established for each spatial detection unit. Traverse all spatial detection units, arbitrarily select two sets of multidimensional state vectors from different spatial detection units, and calculate the Euclidean distance between the two sets of multidimensional state vectors. The Euclidean distances calculated pairwise are arranged in rows and columns according to the cell number and uniformly standardized into a two-dimensional matrix format to generate the global state distance matrix.

[0010] Preferably, the process for obtaining the potential risk unit is as follows: All Euclidean distance values ​​in the global state distance matrix are imported into the local outlier factor algorithm, and the external failure index is calculated unit by unit based on the overall state distribution law of the entire pipeline section. An external failure judgment threshold is pre-set. The external failure index of each spatial detection unit is compared with the external failure judgment threshold one by one. All spatial detection units with external failure index values ​​greater than the external failure judgment threshold are selected and marked as potential risk units. The continuous time-series vibration data corresponding to the potential risk units are extracted.

[0011] Preferably, the process of obtaining the standardized noise reduction trajectory features is as follows: Extract the multi-point vibration amplitude change sequence from the continuous temporal vibration data corresponding to the potential risk unit, fit and generate the spatial propagation trajectory of the disturbance, and extract three types of trajectory quantitative indicators: disturbance movement direction, energy diffusion rate, and disturbance duration. By introducing real-time environmental monitoring parameters and historical disturbance sample sets for continuous iterative calculation, the dynamic suppression threshold is dynamically updated to adapt to the current working conditions. The disturbance trajectory features are matched and verified with the dynamic suppression threshold. Risk-free disturbance signals that match natural wind and rain or small animal activities are removed. After filtering out environmental interference, the data format is standardized and regularized to generate standardized noise-reduced trajectory features.

[0012] Preferably, the calculation process of the comprehensive risk index is as follows: Based on noise reduction trajectory characteristics, section risk scalar, real-time cable temperature and external failure index of each unit, a temperature prediction neural network is built to obtain the theoretical predicted temperature of the cable, and the absolute value of the difference between the real-time cable temperature and the theoretical predicted temperature is calculated to obtain the cable temperature deviation. A multi-parameter weighted fusion model is constructed, which performs weighted coupling operations on temperature deviation, noise reduction trajectory characteristics, segment risk scalar, and external failure index to obtain a comprehensive risk index.

[0013] Preferably, the process for obtaining the pipeline external force damage risk warning result is as follows: Pre-set alarm thresholds, compare the comprehensive risk index with the alarm thresholds, and when the comprehensive risk index is greater than the alarm threshold, generate a cable external force damage warning signal that is bound to the corresponding spatial detection unit number, section risk level, disturbance trajectory type and cable temperature rise deviation. It outputs early warning signals of cable damage caused by external forces, simultaneously initiates joint response procedures, locates risk points based on pipeline spatial geographic data, retrieves front-end image equipment for verification, and sends local audible and visual alarms and SMS messages to operation and maintenance terminals.

[0014] The present invention also provides a cable channel external damage prevention monitoring system based on distributed vibration optical fiber, the system being used to execute the aforementioned cable channel external damage prevention monitoring method based on distributed vibration optical fiber.

[0015] The present invention also provides a computer-readable storage medium storing a computer program, which is executed by a processor to implement the aforementioned method for monitoring cable channels against external damage based on distributed vibration optical fiber.

[0016] The beneficial effects of this invention are: 1. By spatially partitioning cable channels according to geographical conditions, construction risks, and personnel activities, and generating prior constraint factors based on the risk weights of each section, adaptive and differentiated configurations of the acquisition sensitivity, analysis cycle, and feature extraction depth of each monitoring unit are achieved. This completely changes the traditional mode of uniform parameter acquisition across the entire line, enabling more detailed monitoring of high-risk construction sections, reducing invalid sampling in low-risk idle sections, effectively balancing the system's computing load, and significantly improving the targeting and adaptability of long-distance pipeline monitoring across the entire line.

[0017] 2. By relying on the global state distance matrix, spatial neighborhood working conditions are compared and identified. Combined with dynamic environmental threshold iteration and trajectory feature matching, environmental interference disturbances are eliminated, overcoming the shortcomings of traditional single-point threshold judgment which is easily affected by wind, rain, and small animal disturbances. By quantifying the direction of disturbance propagation, diffusion rate, and duration, the system can effectively distinguish between real external disturbances and natural disturbances, significantly reducing the probability of false alarms and missed alarms in the monitoring system.

[0018] 3. By integrating multiple parameters such as the inherent risk weight of the section, vibration outlier index, disturbance noise reduction trajectory characteristics, and cable temperature deviation, a weighted coupled analysis is conducted, abandoning the traditional single vibration feature analysis method. Through neural network temperature prediction correction and multi-parameter joint evaluation, the comprehensive external damage risk level of pipelines can be accurately quantified, effectively improving the identification accuracy and risk assessment reliability of hidden dangers such as underground cable construction excavation, mechanical compaction, and geological subsidence. Attached Figure Description

[0019] Figure 1 The flowchart shows a method and system for monitoring cable channels against external damage based on distributed vibration optical fiber. Figure 2 A flowchart of the partitioned differential parameter configuration and synchronous acquisition process for a cable channel external damage prevention monitoring method and system based on distributed vibration optical fiber; Figure 3 A flowchart for anomaly screening and noise reduction trajectory feature generation in a cable channel external damage monitoring method and system based on distributed vibration optical fiber; Figure 4 This is a flowchart of a multi-parameter fusion analysis process for a cable channel external damage monitoring method and system based on distributed vibration optical fiber. Detailed Implementation

[0020] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.

[0021] It is understood that the term "a" should be understood as "at least one" or "one or more," that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0022] like Figure 1 As shown, a method for monitoring external damage to cable channels based on distributed vibration optical fiber is described. The method includes: The fiber optic line is divided into several spatial detection units, and multi-source heterogeneous data are collected for each unit. Historical data is combined to calculate the segment risk scalar for each spatial detection unit, and a corresponding multi-dimensional regional risk prior constraint factor is generated based on the segment risk scalar. Differential acquisition control parameters are configured for each spatial detection unit based on the multi-dimensional regional risk prior constraint factor. Multi-dimensional state data is collected for each spatial detection unit using a pulse digital code phase light time-domain reflectometry architecture. Based on the multi-dimensional state data, a multi-dimensional state vector is constructed for each spatial detection unit. The multi-dimensional state Euclidean distance between any two spatial detection units is calculated to generate a global state distance matrix. A local outlier factor algorithm is used to calculate the external damage index and screen potential risk units. Vibration data from different pipeline locations under continuous time series of potential risk units are extracted, and the disturbance propagation trajectory is analyzed. A dynamic threshold is generated to filter interference signals based on real-time environmental data, and noise-reduced trajectory features are output. Based on the noise-reduced trajectory features, segment risk scalar, and real-time cable temperature, temperature deviation is calculated using a temperature prediction neural network. The external damage index is then fused to calculate a comprehensive risk index, and the pipeline external force damage risk warning result is output.

[0023] This embodiment is applied to long-distance linear pipeline channels such as urban cable tunnels, cable trenches, and underground communication optical cables. The field sensing carrier is a pulse digitally encoded phase light time-domain reflectometry distributed vibration optical fiber. The entire monitoring equipment is divided into four layers, and data transmission between devices is completed via optical fiber and industrial network cables.

[0024] The first layer is the sensing and acquisition unit, which consists of a combination of detection optical cables and temperature measurement optical cables. This unit is laid along the entire length of the cable trench sidewalls and tunnel top, simultaneously collecting three types of physical signals: vibration deformation around the pipeline, optical cable tension, and tunnel temperature. One end of the optical cable is fused to the optical port of the pulse digital coded phase optical time domain reflectometer host, and all raw optical signals generated on-site are transmitted to the demodulation host for further processing.

[0025] The second layer is the signal demodulation unit, the main body of which is a pulse digital coded phase optical time-domain reflectometry host. The host is responsible for transmitting coded probe light pulses, receiving reflected light from the optical cable and performing photoelectric conversion. It then filters, intercepts, and extracts features from the raw signals according to pre-defined pipeline sections, thereby obtaining multi-dimensional status data. The device's optical port connects downwards to the field sensor optical cable to acquire the raw optical signal, and its network port transmits the partitioned status data to the backend intelligent management platform via a network cable.

[0026] The third layer is the data storage unit, which is a server hard drive. The hard drive stores pipeline GIS ledgers, municipal construction records, original disturbance signals from previous years, historical environmental temperature and humidity data, and all intermediate processing data from each monitoring unit. The storage device and the intelligent management platform's industrial control host are connected to the same local area network. When the platform calculates risk weights or updates filtering thresholds, it can directly retrieve the historical data stored on the hard drive.

[0027] The fourth layer is the control and assessment unit, with the core equipment being the intelligent management platform's industrial control host. The host performs pipeline area risk weight calculations, configures parameters for each section, performs state distance matrix calculations across the entire line, solves for external damage indices, analyzes vibration trajectories, measures cable temperature, and determines the comprehensive risk index. Simultaneously, it issues audible and visual alarms and maintenance SMS linkage commands. The host receives real-time pipeline status data uploaded by the demodulation host, reads various historical records stored on the hard drive, and after determining the existence of potential external damage, outputs alarm commands to the on-site audible and visual alarms and the maintenance SMS module.

[0028] The pulse digital coded phase optical time domain reflectometer uses a 10km single-channel fiber optic demodulator. The equipment is equipped with a single FC / APC fiber optic interface and two gigabit adaptive network ports, with a signal sampling rate of 250MS / s. The overall operating temperature range is controlled between -15℃ and 55℃. The basic data acquisition interval is uniformly set to 100 milliseconds, dynamically adjusted based on the risk level of the pipeline section. For high-risk sections with construction records in the past year, the acquisition interval is shortened to 50 milliseconds, while for low-risk sections such as green belts and idle pipelines in the wild, the acquisition interval is extended to 200 milliseconds. The equipment simultaneously acquires four types of data along the optical cable: vibration amplitude, optical phase offset, vibration frequency, and optical cable deformation tension.

[0029] The detection and sensing optical cable laid on site is a 4-core single-mode communication sensing optical cable. During construction, it is laid in a straight line along the top of the cable channel, and the minimum bending radius of the optical cable at the bend is not less than 10 times the outer diameter of the optical cable. The optical cable passively senses the mechanical disturbance transmitted by the soil and pipelines throughout the entire process, and converts the vibration into an optical phase change signal, which is continuously transmitted to the demodulation host.

[0030] The temperature-measuring optical cable is integrated inside the sensing optical cable, with a fixed temperature acquisition interval of 1 second, simultaneously acquiring the real-time temperature of the cable sheath of each monitoring unit and the ambient temperature inside the channel.

[0031] The site is equipped with an external environmental data acquisition terminal, which is equipped with temperature, humidity, wind speed and rainfall sensors. The standard deployment is one unit installed every 200 meters of cable channel. The device uploads a set of environmental parameters to the backend every 30 seconds. The collected data includes real-time wind speed, rainfall and ambient temperature and humidity. The collected data is synchronously pushed to the intelligent management platform for dynamic adjustment of alarm suppression thresholds.

[0032] The sensor optical cables laid on-site are uniformly divided into lengths of 10 meters. The boundaries of each monitoring unit are determined by combining static data retained from the three types of cables. The signal acquisition parameters within each section are adaptively adjusted according to the risk level. The specific zoning criteria are as follows: The first is the pipeline geographical boundary standard, which uses natural pipeline dividing nodes such as cable inspection wells, tunnel expansion joints, and optical cable junction boxes as unit dividing points. The joint locations are divided into independent monitoring units to avoid optical signal attenuation and monitoring data distortion caused by cross-joint data collection. The second is the construction risk classification standard. The earthwork construction filing ledger of the housing and construction and municipal departments in the jurisdiction is retrieved. Pipeline sections with records of road reconstruction, pipeline excavation and earthwork loading construction in the past twelve months are designated as high-risk units; pipelines with no earthwork operations for a long time and idle pipelines under green belts are designated as low-risk units. The third is the classification standard for personnel activities. Substation entrances and exits, cable tunnel maintenance openings, and pipelines under urban sidewalks are designated as medium-risk units; wild woodlands and enclosed unmanned underground corridors are designated as low-risk units.

[0033] Once all zoning work is completed, each pipeline section with a fixed length and fixed boundaries becomes an independent spatial detection unit. Risk weight calculations are performed on all units to generate multi-dimensional regional risk prior constraint factors for the corresponding sections.

[0034] Furthermore, the process of obtaining the multi-dimensional regional risk prior constraint factors is as follows: Multi-source basic data and historical disturbance sample data of optical fiber lines are collected. The multi-source basic data includes spatial geographic data, regional construction registration data, and real-time environmental monitoring parameters. After normalizing the multi-source basic data and historical disturbance sample data of optical fiber lines, feature fusion is performed. The segment risk scalar corresponding to each spatial detection unit is solved by weighted mapping operation. The parameters representing the regional risk level, historical disturbance probability, and environmental disturbance intensity in the segment risk scalar are encapsulated into a unified parameter group to generate multi-dimensional regional risk prior constraint factors for the corresponding spatial detection unit. Each multi-dimensional regional risk prior constraint factor is independent of the others.

[0035] like Figure 2 As shown, the sensing fiber optic lines laid along the cable channel are divided into continuous and non-overlapping independent spatial monitoring units according to the zoning standard. The baseline monitoring length of each unit is fixed at 10 meters, and the entire pipeline forms a continuous segmented monitoring interval.

[0036] The spatial geographic data of the pipelines is obtained from the as-built GIS drawings of the power operation and maintenance unit. On-site operation and maintenance personnel are arranged to check the data unit by unit along the cable channel, recording the start and end coordinates, burial depth, and channel type of each spatial detection unit. After verification, the data is exported to an Excel spreadsheet and stored on the backend server. Regional construction registration data is collected monthly by operation and maintenance personnel from the local housing and construction and municipal engineering management departments, collecting road excavation, pipeline renovation, and earthwork construction filing documents, extracting the scope of construction sections, construction period, and type of construction machinery, and simultaneously entering the data into the system ledger. Real-time environmental monitoring parameters are automatically collected by temperature, humidity, wind speed, and rainfall sensors deployed every 200 meters along the channel. The devices upload a set of on-site environmental values ​​every 30 seconds, and the data is stored in the management platform database in real time via the local area network.

[0037] Historical disturbance sample data is compiled and archived based on the platform's on-site monitoring records over the past 36 months. Maintenance personnel export the raw vibration time series data of the entire line monthly, distinguishing four types of disturbance signals: vehicle traffic, wind and rain swaying, small animal contact, and mechanical excavation. Each time series data segment is labeled with the disturbance type, occurrence section, and duration, and uniformly stored in an independent historical disturbance sample library as a historical reference for risk calculation.

[0038] The geographical coordinates, construction frequency, wind speed and rainfall, and various disturbance amplitude data collected on-site have different dimensions. Before calculation, a unified linear normalization process was performed to map all parameter values ​​to the interval between 0 and 1. The normalization calculation formula adopted uniformly is as follows: ; In the formula Here are the normalized parameters, and x is the original collected value. This is the minimum value for this type of parameter across the entire line. This is the maximum value for this type of parameter across the entire line.

[0039] After all dimensional parameters are normalized, multi-source feature fusion is completed using a fixed-dimensional sequential splicing method, and an 8-dimensional fusion feature vector is independently constructed for each spatial detection unit. The vector consists of four main categories of features arranged sequentially, each category containing two sub-quantitative indicators: the first dimension is the pipeline geographic normalized feature, which includes the risk of the laying location and the risk of the pipeline burial depth, representing the inherent risks of the pipeline laying location, burial depth, and channel type; the second dimension is the construction registration normalized feature, which includes the annual construction frequency and the hazard level of construction machinery, representing the construction frequency, construction scale, and construction machinery level in the past 12 months; the third dimension is the real-time environment normalized feature, which includes the intensity of wind and rain disturbance and the amplitude of temperature and humidity fluctuations, integrating the comprehensive disturbance intensity of wind speed, rainfall, and environmental temperature and humidity; the fourth dimension is the historical disturbance time series normalized feature, which includes the proportion of disturbance duration and the monthly disturbance frequency, calculated from the cumulative duration and frequency of various disturbance signals in this unit within 36 months.

[0040] set up For the segment risk scalar of the i-th independent spatial detection unit, the global segment risk scalar is obtained by combining all units. Generate the formula for solving the risk scalar of a single segment: ; In the formula: The inherent geographical risk score for the i-th unit is assigned based on whether the pipeline is buried on a main road or sidewalk. The historical construction risk score for the i-th unit is calculated based on the construction frequency over the past 12 months. The real-time environmental disturbance score for the i-th unit is calculated from the on-site wind speed and rainfall intensity. , , Fixed weighting coefficients are assigned to the three dimensions, which are manually calibrated by the operation and maintenance unit based on years of pipeline external failure statistics to meet the requirements. + + =1.

[0041] By individually substituting the formula into each independent spatial detection unit along the entire line, the risk scalar of the corresponding unit section is obtained, identifying construction sections where mechanical excavation has been most frequent in the past. The values ​​are too large, resulting in an overall high weight value in the final matrix. This is particularly true for green belts and unused pipeline sections without pedestrian access. , The values ​​are all relatively low, resulting in a relatively small overall value in the corresponding weight matrix.

[0042] Calculated by all spatial detection units along the entire line The values ​​are sorted uniformly from smallest to largest, and then three intervals are evenly divided according to the value range: low-risk interval [0, 1 / 3], medium-risk interval (1 / 3, 2 / 3], and high-risk interval (2 / 3, 1). Let the minimum weight of the entire line after sorting be denoted as . The maximum value is Perform a quadratic linear normalization scaling on the current unit weights: ; In the formula: This represents the risk level value for the i-th unit area, with a value falling within the range of 0-1. The larger, The higher the number, the higher the corresponding risk level.

[0043] Historical construction risk component It is a weight matrix The components, the probability of disturbance occurrence are... Using this as the core basis, and making slight adjustments based on the proportion of historical disturbance sample durations, the formula is modified as follows: ; In the formula: This is the cumulative duration of disturbance signals in the archived samples of this unit over 36 months; This refers to the total monitoring duration of 36 months for this unit. This represents the probability of a segment disturbance occurring.

[0044] directly with As a base, a small-scale fitting correction is performed by combining real-time meteorological normalized values: ; In the formula: The real-time environmental correction coefficient is obtained by combining wind speed, rainfall, temperature, and humidity, with a value range of 0.8-1.2. This represents the final environmental disturbance intensity. Under calm and clear conditions... Approaching 0.8, under severe working conditions of heavy rain and strong winds The closer the value is to 1.2, the more severe the environment. The higher the value.

[0045] The above-mentioned reliance The converted regional risk level values Probability of segment disturbance Strong environmental disturbance Three independent quantitative parameters are encapsulated into a one-dimensional structured parameter group according to a fixed field order, forming the multi-dimensional regional risk prior constraint factor of this spatial detection unit.

[0046] Furthermore, the process of obtaining multidimensional state data is as follows: Based on multi-dimensional regional risk prior constraint factors, differentiated acquisition control parameters including vibration trigger sensitivity threshold, time series data parsing period, and signal feature extraction depth are calculated. These differentiated acquisition control parameters are then imported into a pulse digital code phase optical time domain reflectance sensing architecture, and synchronous signal acquisition is performed on the sensing fiber according to spatial detection unit partitions. By relying on the vibration trigger sensitivity threshold to remove weak noise below the reference amplitude, a fixed-length continuous time series data stream is extracted according to the time series data parsing period. The data stream is then parsed layer by layer according to the signal feature extraction depth, and pipeline state variables are extracted synchronously. The pipeline state variables are then collected and encapsulated to generate multi-dimensional state data.

[0047] Based on the risk level values ​​of the unit area The specific process for determining the vibration trigger sensitivity threshold is as follows: High-risk areas: The vibration trigger sensitivity threshold setting range is 0.02V to 0.05V. Slight mechanical scratches or small vibrations caused by digging with a small pickaxe can trigger signal acquisition. Medium-risk areas: The vibration trigger sensitivity threshold setting range is 0.05V to 0.15V, which only captures obvious disturbances such as people stepping on things and small equipment bumping into things, while filtering out normal light wind disturbances. Low-risk zone: 0 The vibration trigger sensitivity threshold is set in the range of 0.15V to 0.20V, which significantly raises the judgment threshold. The shaking of grass and trees in the green belt, airflow vibration, and natural micro-deformation of the underground soil will be directly filtered out, leaving only strong disturbance signals such as heavy rolling and manual excavation.

[0048] Probability of segment disturbance The value range is [0,1], calculated based on the construction components of the weight matrix and the historical disturbance duration. The collection period is defined according to three fixed boundary points: High-disturbance section: A 50ms time series analysis period was adopted to shorten the sampling interval and fully record the time series changes of instantaneous excavation impact; Medium disturbance section: A 100ms basic analysis cycle is adopted as the standard duration for routine monitoring across the entire line; Low-disturbance section: By adopting a 200ms parsing cycle, the sampling interval is lengthened, the amount of useless time-series data stored is reduced, and the computational load of the demodulation host is reduced.

[0049] Environmental disturbance intensity The value range is [0,1], which is obtained by combining the environmental components of the weight matrix with real-time temperature, humidity, wind speed, and rainfall correction. It is divided into three analytical levels, and the boundaries of the corresponding levels are clear and fixed. Deep parsing levels: The three-layer analysis was enabled to fully extract all pipeline status quantities, including vibration change, optical cable tension change, optical phase offset, vibration frequency change, and real-time temperature of the cable channel. Intermediate parsing level: Two-layer analysis is enabled to extract four types of pipeline status quantities: vibration change, optical cable tension change, vibration frequency change, and real-time temperature of cable channel. Basic parsing levels: One layer of analysis is enabled, extracting only two core basic state quantities: vibration change and real-time temperature of the cable channel.

[0050] After the three sets of acquisition parameters are mapped and solved unit by unit, they are bound and solidified with the corresponding spatial detection unit number to form a unique and differentiated acquisition parameter set, which is then sent out and stored in the local register cache of the pulse digital coded phase light time domain reflectance host.

[0051] The pulse digital coded phase optical time domain reflectance sensing architecture of this system is an integrated hardware system. The core functional modules include a narrow linewidth laser, a pulse code modulation module, an acousto-optic frequency shifting module, an optical fiber coupler, a balanced photodetector, and a 250MS / s high-speed signal acquisition card.

[0052] During the parameter configuration phase, the background control industrial host sends the differentiated acquisition parameters of each spatial detection unit to the host high-speed signal acquisition card in batches through the gigabit Ethernet TCP communication protocol, covering the default general parameters of the device, and completing the solidification of the partition parameters of acquisition threshold, parsing period and extraction depth. After the device returns a loading completion response signal, it officially enters the synchronous acquisition state.

[0053] After the on-site data acquisition begins, a narrow-linewidth laser outputs a stable, continuous single-mode laser beam. A pulse-code modulation module generates encrypted timing probe pulses according to preset encoding rules. These pulses are then processed by an acousto-optic frequency shift module to achieve a fixed frequency shift, and finally coupled into the entire-line-laid sensing optical cable via an optical fiber coupler. The optical pulses propagate along the optical cable axis. Deformation and vibration along the cable line cause phase changes in the light, and the corresponding backscattered Rayleigh signal is transmitted back along the original path to the optical fiber coupler. After being spun out, the signal is input to a balanced photodetector.

[0054] The photodetector accurately converts the light signals, which change in phase and intensity, into analog voltage signals, which are then transmitted to a high-speed signal acquisition card for analog-to-digital conversion. Based on locally cached 10-meter unit spatial scales and partition parameters, the acquisition card performs parallel, partitioned, and synchronous acquisition across the entire optical cable. Each spatial detection unit independently calls its own configured acquisition parameters, ensuring that the acquisition processes of each section do not interfere with each other and are completely synchronized in timing. This avoids the problems of cross-section parameter mixing and timing misalignment, achieving differentiated and precise signal acquisition across the entire line.

[0055] After the acquisition card completes the analog-to-digital conversion and outputs the raw electrical signal, it compares the real-time signal voltage amplitude with the vibration trigger sensitivity threshold bound to the current unit point by point. For underlying steady-state noise such as soil micro-vibration, air flow, and slight vegetation swaying with amplitudes below the threshold, it directly performs zeroing and rejection processing; only the effective disturbance waveform with amplitudes exceeding the threshold is retained to complete the raw signal noise reduction preprocessing and prevent invalid noise data from participating in subsequent calculations.

[0056] Using a fixed sliding window length for the pre-configured time-series data parsing cycle, the effective disturbance waveform after noise reduction is continuously and without intervals segmented, with the duration of each segment strictly matching the parsing cycle of the corresponding unit. After segmentation, each set of time-series data is bound to the corresponding spatial unit number and the start and end timestamps of the acquisition, and temporarily cached in the host memory to ensure accurate correspondence between the data and the monitoring segment.

[0057] Based on the three-level analytical depth matching the environmental disturbance intensity of each unit, the time-series waveform is decomposed layer by layer, and the pipeline operating status parameters are extracted in stages: Level 1 Basic Analysis: Adapted to low-risk, low-disturbance sections, extracting only two core state quantities: vibration change and real-time temperature of the cable channel; Level 2 Intermediate Analysis: Adapted to medium-risk conventional sections, it supplements the extraction of optical cable tension variation and vibration frequency variation on top of basic features; Level 3 In-Depth Analysis: Adapted to high-risk construction disturbance sections, it fully extracts five types of pipeline status parameters: vibration change, optical cable tension change, optical phase offset, vibration frequency change, and real-time temperature of the cable channel.

[0058] After completing the hierarchical feature extraction, all state variables obtained from the current unit are uniformly organized and sorted, and three unique identifiers—unit number, collection timestamp, and segment risk level—are simultaneously attached. This data is then converted into a system-recognizable structured array format, completing the data standardization and encapsulation. The resulting single-unit multidimensional state data, which serves as the monitoring data for the corresponding spatial detection unit, is uploaded in batches to the backend management and analysis unit via the local area network.

[0059] Furthermore, the process of obtaining the global state distance matrix is ​​as follows: Based on multidimensional state data, a multidimensional state vector corresponding to the pipeline state quantity is established for each spatial detection unit. All spatial detection units are traversed, and two sets of multidimensional state vectors from different spatial detection units are arbitrarily selected to calculate the Euclidean distance between the two sets of multidimensional state vectors. The Euclidean distances calculated pairwise are arranged in rows and columns according to the unit number and uniformly standardized into a two-dimensional matrix format to generate a global state distance matrix.

[0060] like Figure 3 As shown, the multidimensional state data of each spatial detection unit is completed and normalized into a five-dimensional standardized state vector. The five dimensions of the vector correspond to five types of pipeline physical state quantities: the first dimension is the vibration change, the second dimension is the optical cable tension change, the third dimension is the optical phase offset, the fourth dimension is the vibration frequency change, and the fifth dimension is the real-time temperature of the cable channel.

[0061] For the missing feature dimensions of low- and medium-risk units, the system does not collect additional data, but only fills in the gaps using the long-term steady-state mean of the current unit, ensuring that the state vector dimension of all spatial detection units across the entire line is uniformly five-dimensional. Subsequently, the original five-dimensional state values ​​are uniformly subjected to global linear normalization, mapped to the 0-1 interval, eliminating differences in different physical dimensions, and finally forming the standardized five-dimensional state vector corresponding to the i-th spatial detection unit. .

[0062] The system traverses all numbered spatial detection units along the entire line, performs pairwise matching on any two independent detection units, does not perform self-matching or duplicate / missed matching, and performs Euclidean distance calculation on the standardized five-dimensional state vector of each matching unit.

[0063] Let any two different spatial detection units be the i-th unit and the j-th unit, and their corresponding state vectors be: ; ; The formula for calculating the Euclidean distance between the two is: ; In the formula: The normalized state value of the k-th dimension of the i-th unit; The normalized state value of the k-th dimension of the j-th unit; The difference in operating status between the two pipeline units.

[0064] The larger the calculated distance value, the more significant the difference in the overall operating conditions of the two pipeline segments in terms of vibration, tension, phase, frequency, and temperature. Conversely, the smaller the distance value, the more consistent the operating conditions of the two segments are, indicating they are in a normal and stable range. The system completes the distance calculation for all pairwise unit combinations along the entire pipeline, and all distance results are temporarily stored according to the paired unit number.

[0065] Based on the on-site pipeline layout, all spatial detection units have been assigned unique, consecutive numbers according to their physical arrangement. The system uses the unit number as the row and column index for a two-dimensional regularized arrangement: the matrix row number corresponds to the i-th detection unit, the matrix column number corresponds to the j-th detection unit, and the matrix position (i,j) is filled with the Euclidean distance calculated from the corresponding pairing. .

[0066] The diagonal positions of the matrix are for unit self-comparison and are uniformly set to 0; the off-diagonal positions are filled with the state difference distances between each pair of cross-segment along the entire line. Finally, the matrix is ​​regularized to form a two-dimensional square matrix with an order equal to the total number of detection units along the entire line, which is the global state distance matrix.

[0067] Furthermore, the process of obtaining potential risk units is as follows: All Euclidean distance values ​​within the global state distance matrix are imported into the local outlier factor algorithm. Based on the overall state distribution pattern of the entire pipeline section, the external failure index is calculated unit by unit. An external failure judgment threshold is pre-set, and the external failure index of each spatial detection unit is compared with the external failure judgment threshold one by one. All spatial detection units with external failure index values ​​greater than the external failure judgment threshold are selected and marked as potential risk units. The continuous time-series vibration data corresponding to the potential risk units are extracted.

[0068] The pairwise Euclidean distance values ​​of all cells in the global state distance matrix are imported into the local outlier factor algorithm. Based on the overall distribution characteristics of the normal operation status of the entire pipeline section, the algorithm statistically analyzes the state density and dispersion deviation of the surrounding neighboring sections for each cell.

[0069] First, the k nearest neighbor distance of the i-th spatial detection unit is set. In engineering, the value k is fixed at 8, that is, the 8 nearest surrounding pipeline sections are selected as normal reference samples for each unit, and the reachability distance of the i-th unit is calculated: ; In the formula: The k-th distance from neighboring cell j; Let k be the reachable distance of cell i relative to its neighboring cell j.

[0070] The local reachability density is calculated based on the reachability distance, which characterizes the degree of clustering of samples around the current cell: ; In the formula: Let i be the set of k nearest neighbor units; This represents the local reachability density of the current unit. The higher the density value, the more closely the unit's operating state conforms to the normal distribution across the entire line.

[0071] Finally, the element external failure index (local outlier factor) is calculated: ; This is the final normalized external failure index of the i-th unit.

[0072] When the external damage index is greater than 1, it indicates that the local reachability density of the unit is lower than that of the surrounding sections, the state deviates from the normal distribution of the whole line, and there is an abnormal disturbance trend; the larger the index value, the more significant the deviation of the working condition, and the higher the risk of external damage.

[0073] Based on the sample data of actual external damage accidents to cable channels in the region over the past five years and samples of daily environmental interference, the system in this embodiment sets the external damage judgment threshold to 1.2.

[0074] The algorithm performs numerical comparisons of the external damage index for each of all spatial detection units along the entire line: For satisfying The spatial detection unit determines that the overall state of the current section deviates from the normal range, indicating a tendency for external force disturbance, and marks it uniformly as a potential risk unit; for those that meet the requirements... The spatial detection unit determined that the operating status of the section conformed to the normal distribution of the entire line, with no abnormal deviations, and classified it as a normal unit.

[0075] After completing the risk assessment of all units, the system classifies and distributes the monitoring data based on the marking results. For all marked potential risk units, the platform backend accurately locates the cached raw acquisition data packets based on the unit number, and extracts the continuous time-series vibration raw data within the complete acquisition cycle of the corresponding unit, including the millisecond vibration amplitude time-series sequence, disturbance timestamp, and spatial location number. This part of the high-frequency time-series data is cached separately in high-speed computing memory.

[0076] For low-risk units that are deemed normal, their high-frequency time-series data will no longer be loaded for subsequent complex calculations. Only the basic status statistics of the unit will be retained. The original time-series data will be transferred to a low-speed archive hard drive in the background for lightweight storage and will only be used for later ledger tracing. This will effectively reduce the platform's real-time computing power consumption and ensure the efficient operation of the risk assessment process.

[0077] Furthermore, the process of obtaining the standardized denoising trajectory features is as follows: Multi-point vibration amplitude change sequences are extracted from continuous temporal vibration data corresponding to potential risk units, and disturbance spatial propagation trajectories are fitted to generate them. Three types of trajectory quantification indicators are extracted: disturbance movement direction, energy diffusion rate, and disturbance duration. Real-time environmental monitoring parameters and historical disturbance sample sets are introduced for continuous iterative calculation to dynamically update the dynamic suppression threshold adapted to the current working conditions. The disturbance trajectory features are matched and verified with the dynamic suppression threshold. Risk-free disturbance signals matching natural wind and rain and small animal activities are removed. After filtering out environmental interference, the data format is standardized and regularized to generate standardized noise-reduced trajectory features.

[0078] The system locates the optical cable spatial mileage intervals of each potential risk unit, batch reads the raw time-series monitoring data of all optical fiber monitoring points within this interval, sorts them by millisecond-level timestamps, and organizes them into a continuous and complete multi-point vibration amplitude change sequence. A three-dimensional spatiotemporal disturbance data matrix is ​​constructed with the data acquisition timestamp as the horizontal axis of time, the normalized vibration amplitude of each point as the vertical axis of intensity, and the actual optical cable laying mileage as the vertical axis of space. For the discretely distributed monitoring point data, a second-order least squares polynomial algorithm is used to perform continuous curve fitting, eliminating single-point data fluctuation deviations, and fitting the global spatial propagation trajectory curve of this disturbance event, realistically reproducing the diffusion and change law of the disturbance along the cable channel. Based on the standardized trajectory curves that have been fitted, three types of core trajectory quantification parameters are extracted one by one: First, the direction of disturbance movement. By comparing the spatial mileage position offset of the optical cable corresponding to the vibration peak at different time points, the overall propagation direction of the disturbance is determined, and three types of propagation modes are distinguished: unilateral advancement, bidirectional diffusion, and fixed-point disturbance. Second, the vibration energy diffusion rate. Using 100ms as a fixed statistical time unit, the spatial coverage expansion distance of the high-energy vibration range per unit time is calculated to quantify the speed of disturbance diffusion. Third, the duration of disturbance. The start time when the disturbance first exceeds the steady-state baseline and the end time when the disturbance energy decays back to the normal vibration range of the channel are accurately recorded. The time difference between the two is counted as the complete duration of disturbance. The three types of parameters are integrated to form the original disturbance trajectory feature group.

[0079] Then, the dynamic suppression threshold is updated online iteratively. The system retrieves real-time monitoring data of wind speed, rainfall, and ambient temperature and humidity along the channel, and simultaneously retrieves characteristic parameters of environmental disturbance samples from a 36-month historical disturbance sample database. First, the original measured wind speed and rainfall values ​​are mapped and converted in segments to establish a unified dimensionless correction conversion function: the original measured wind speed unit is m / s, and the segmented threshold is divided as follows: when wind speed < 2 m / s =0, 2m / s ≤ wind speed < 8m / s =0.1-0.4 linear interpolation, when wind speed ≥8m / s =0.4; Original rainfall measurement unit: mm / h, no rainfall. =0, light rain 1-5mm / h corresponding =0.1-0.2 linear interpolation, corresponding to moderate to heavy rain of 5-20 mm / h. =0.2-0.5 linear interpolation, fixed for heavy rain ≥20mm / h =0.5, after piecewise mapping , A dimensionless correction value of 0-0.5 is uniformly applied. A linear iterative correction formula is used to update the dynamic suppression threshold in real time. The iterative formula is as follows: In the formula: This is a real-time dynamic suppression threshold; Set a basic suppression threshold for the system; This is the wind speed correction amount; the higher the wind speed, the greater the correction lift. To compensate for rainfall, the threshold is automatically raised under heavy rain conditions, and reset to zero under sunny, windless, and rainless conditions, with the threshold falling back to the base value, ensuring that the interference filtering standard matches the actual situation on site under different environmental conditions.

[0080] Finally, trajectory feature matching verification, interference removal, and data normalization are performed to generate standardized denoised trajectory features. The system precisely compares the three core parameters—vibration amplitude, energy diffusion rate, and disturbance duration—from the fitted original disturbance trajectory feature set with the dynamic suppression threshold corresponding to the current operating condition dimension by dimension. Simultaneously, it performs similarity matching calculations with a historical environmental interference feature template library. A fixed matching threshold of 85% is set. If the matching degree between the current disturbance trajectory feature and risk-free interference templates such as wind and rain swaying, small animal contact, and vegetation swaying is ≥85%, the disturbance is determined to be an environmental interference signal, and the current time-series disturbance data is completely removed and not included in subsequent risk calculations. For disturbance events with a matching degree <85%, they are determined to be valid external force disturbance signals and are retained. After interference screening and removal, the retained valid disturbance time-series data are uniformly normalized to finally generate standardized denoised trajectory features.

[0081] Furthermore, the calculation process for the comprehensive risk index is as follows: Based on noise reduction trajectory characteristics, section risk scalar, real-time cable temperature, and external damage index of each unit, a temperature prediction neural network is built to obtain the theoretical predicted temperature of the cable. The absolute value of the difference between the real-time cable temperature and the theoretical predicted temperature is calculated to obtain the cable temperature deviation. A multi-parameter weighted fusion model is constructed to perform weighted coupling operations on temperature deviation, noise reduction trajectory characteristics, section risk scalar, and external damage index to obtain a comprehensive risk index.

[0082] like Figure 4 As shown, this system constructs a three-layer fully connected temperature prediction neural network. The specific network configuration and calculation process are as follows.

[0083] The temperature prediction neural network has a three-layer fully connected structure: an input layer, a hidden layer, and an output layer. The input layer contains three fixed input features: the real-time load current of the cable, the cable's historical average load over the past hour, and the baseline temperature of the channel environment. The hidden layer has 16 neurons, using the ReLU activation function to fit the nonlinear temperature rise correlation. The output layer has one neuron, outputting the theoretically predicted cable temperature. The network training dataset is taken from 36 consecutive months of measured operation samples of cable channels in this region, including actual temperature rise data of cables under different loads, ambient temperatures, and seasonal conditions. After training, the network weight parameters are fixed and are not updated during online monitoring.

[0084] During the online monitoring phase, the system normalizes the three types of input parameters collected in real time and imports them into the trained neural network. The theoretical predicted cable temperature under the current operating conditions is then calculated through forward propagation. Based on this, the synchronously collected real-time cable temperature is retrieved, and the cable temperature deviation is calculated using the absolute value of the difference method. The calculation formula is as follows: In the formula: For cable temperature deviation; Provides real-time temperature measurement data for the cable; The theoretical temperature prediction for the cable is output by the neural network. The larger the temperature deviation value, the more significant the deviation of the actual temperature rise of the cable from the theoretical steady-state temperature rise, indicating a potential risk of abnormal heating of the cable due to external disturbances.

[0085] This system adopts a four-layer structured multi-parameter weighted fusion model, using four types of parameters as inputs: standardized noise reduction trajectory characteristics, section risk scalar score, unit external failure index, and cable temperature deviation, to achieve multi-physical quantity and multi-dimensional risk coupling assessment.

[0086] The model has a four-layer progressive computational structure: the first layer is the data normalization layer, which maps four different types of input parameters to the 0-1 interval; the second layer is the single-parameter weight assignment layer, which uses engineering statistical calibration to fix the weights and separately weights the risk characteristics of each dimension; the third layer is the cross-coupling layer, which realizes the mutual correction coupling of spatial inherent risk, vibration disturbance risk, and abnormal temperature rise risk; the fourth layer is the output normalization layer, which outputs the comprehensive risk index in the 0-1 interval.

[0087] The model's fixed input includes four core parameters: the first is the comprehensive feature value of the noise-reduced trajectory, representing the actual damage characteristics of the disturbance event; the second is the scalar score of the section risk, representing the inherent basic risk of the section; the third is the unit external damage index, representing the degree of deviation of the pipeline's operating status; and the fourth is the cable temperature deviation, representing the degree of abnormal cable temperature rise. Based on the statistical results of on-site external damage accidents over the past five years, fixed weighting coefficients are manually calibrated, with a weight sum of 1. In this embodiment, the trajectory feature weight is 0.4, the spatial risk weight is 0.25, the external damage index weight is 0.2, and the temperature deviation weight is 0.15.

[0088] The comprehensive risk index is calculated using a fixed weighted coupling formula: In the formula: This is a comprehensive risk index.

[0089] Furthermore, the process for obtaining the early warning results of pipeline external force damage risk is as follows: An alarm threshold is preset, and the comprehensive risk index is compared with the alarm threshold. When the comprehensive risk index is greater than the alarm threshold, a cable external force damage warning signal is generated, which is bound to the corresponding spatial detection unit number, section risk level, disturbance trajectory type, and cable temperature rise deviation. The cable external force damage warning signal is output to the outside, and the linkage response process is initiated simultaneously. The risk point is located based on the pipeline spatial geographic data, the front-end image equipment is retrieved for verification, and local audible and visual alarms and SMS pushes are sent to the operation and maintenance terminal.

[0090] Based on the cable channel damage accident log of the past five years in the jurisdiction, historical valid alarm samples and environmental interference false alarm samples, the alarm threshold is calibrated. In this embodiment, the warning threshold is 0.6.

[0091] In addition, based on the comprehensive risk index range, three levels of risk hazard are classified, with the following classification standards: Low-risk risks: Overall risk index 0.6–0.75; Medium-risk risk: Comprehensive risk index 0.75–0.88; High-risk hazards: Comprehensive risk index 0.88~1.00.

[0092] The platform compares the real-time comprehensive risk index of each spatial detection unit across the entire line with the preset alarm threshold, and performs unit-by-unit judgment and screening.

[0093] The system determines that there is a real external force damage hazard in the section only when the comprehensive risk index of the spatial detection unit is greater than 0.6, and immediately triggers the early warning mechanism to automatically generate a standardized cable external force damage early warning message.

[0094] The early warning message is fixedly bound to five core fields: First, it is bound to the unique number of the current abnormal space monitoring unit, accurately locating the 10-meter monitoring interval; Second, it matches the risk level of the corresponding section, marking low, medium, and high hazard levels; Third, it enters the noise reduction trajectory feature classification results corresponding to this disturbance, clearly distinguishing four types of disturbances: mechanical excavation disturbance, mechanical compaction disturbance, personnel intrusion disturbance, and ground settlement disturbance; Fourth, it is bound to the measured value of the cable temperature deviation corresponding to this assessment, indicating whether it is accompanied by abnormal cable temperature rise; Fifth, it carries the basic geographic information of the unit, such as precise geographic coordinates, channel type, and laying depth.

[0095] After the early warning message is generated, the platform immediately initiates the preset linkage response logic, executing four response actions simultaneously and in parallel: First, precise geographic location analysis. The platform retrieves pipeline spatial geographic ledger data, matches the corresponding GIS coordinates according to the early warning unit number, automatically highlights and locks the risk section on the system's electronic map, selects the precise range of potential hazards, and marks the pipeline burial depth, channel type, and surrounding construction environment, providing accurate location information for on-site operation and maintenance verification.

[0096] Second, front-end video verification is linked. The system automatically retrieves high-definition cameras and infrared thermal imaging devices deployed near the risk points, instantly pulls real-time monitoring images, and displays them in a pop-up window on the platform's operation interface. This enables visual verification of disturbances, distinguishes between actual construction damage and temporary environmental anomalies, and assists maintenance personnel in quickly identifying the authenticity of potential hazards.

[0097] Third, local equipment audible and visual alarm prompts. After receiving the early warning command, the computer room monitoring host immediately activates the built-in audible and visual alarm device, reminding on-site personnel to promptly address potential hazards through fixed-frequency buzzing and flashing alarm indicator lights.

[0098] Fourth, mobile maintenance SMS push. The platform automatically edits structured alarm SMS messages, including the location of the risk area, unit number, disturbance type, risk level, abnormal temperature, and time of occurrence of the hazard, and accurately pushes them to the mobile terminals of the corresponding maintenance personnel in the area, ensuring that field staff receive hazard information and carry out on-site verification and protection work as soon as possible.

[0099] After each warning is triggered, the system automatically corresponds to the current risk unit and batch archives all intermediate calculation data of the entire monitoring chain. The archived content includes: multi-dimensional regional risk prior constraint factors, standardized multi-dimensional state data, global state distance matrix, unit external failure index, noise reduction trajectory feature parameters, and comprehensive risk index.

[0100] The processes described above with reference to the flowcharts in the embodiments disclosed in this invention can be implemented as computer software programs. The embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the functions defined in the methods of this application. It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wire segments, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless segments, wire segments, optical fibers, RF, etc., or any suitable combination thereof.

[0101] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0102] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functions and structural principles of the present invention have been shown and explained in the embodiments. Without departing from the stated principles, the implementation of the present invention may have any variations or modifications.

Claims

1. A method for monitoring external damage to cable channels based on distributed vibration optical fibers, characterized in that, The method includes: The fiber optic line is divided into several spatial detection units, and multi-source heterogeneous data are collected respectively. The segment risk scalar corresponding to each spatial detection unit is calculated by combining historical data, and the corresponding multi-dimensional regional risk prior constraint factor is generated based on the segment risk scalar. Based on multi-dimensional regional risk prior constraint factors, differentiated acquisition control parameters are configured for each spatial detection unit, and multi-dimensional state data is acquired by dividing the spatial detection units into spatial detection units using a pulse digital coding phase light temporal reflectance architecture. Based on multidimensional state data, a multidimensional state vector is constructed for each spatial detection unit. The multidimensional state Euclidean distance between any two spatial detection units is calculated to generate a global state distance matrix. The local outlier factor algorithm is used to calculate the external failure index and screen potential risk units. Vibration data of different pipeline points under continuous time sequence of potential risk units are extracted, disturbance propagation trajectory is analyzed, dynamic threshold is generated to filter out interference signals in combination with real-time environmental data, and noise-reduced trajectory features are output. Based on noise reduction trajectory characteristics, section risk scalars, and real-time cable temperature, temperature deviation is calculated through a temperature prediction neural network, and the external damage index is fused to calculate a comprehensive risk index, outputting early warning results for pipeline external force damage risks.

2. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 1, characterized in that, The process of obtaining the multi-dimensional regional risk prior constraint factors is as follows: Collect multi-source basic data and historical disturbance sample data of optical fiber lines. The multi-source basic data includes spatial geographic data, regional construction registration data and real-time environmental monitoring parameters. After normalizing the multi-source basic data of the optical fiber line and the historical disturbance sample data, feature fusion is completed, and the segment risk scalar corresponding to each spatial detection unit is solved by weighted mapping operation. The parameters representing the regional risk level, historical disturbance probability, and environmental disturbance intensity in the segment risk scalar are encapsulated into a unified parameter set to generate multi-dimensional regional risk prior constraint factors for corresponding spatial detection units. Each multi-dimensional regional risk prior constraint factor is independent of the others.

3. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 2, characterized in that, The process of obtaining the multidimensional state data is as follows: Based on multi-dimensional regional risk prior constraint factors, differentiated acquisition control parameters including vibration trigger sensitivity threshold, time series data parsing period, and signal feature extraction depth are calculated. The differentiated acquisition control parameters are imported into the pulse digital code phase light time domain reflectance sensing architecture, and the sensing fiber is partitioned into spatial detection units for synchronous signal acquisition. By relying on the vibration triggering sensitivity threshold to remove weak noise below the reference amplitude, a fixed-length continuous time-series data stream is extracted according to the time-series data parsing period. The data stream is extracted based on the signal characteristics and the pipeline state variables are extracted simultaneously. The pipeline state variables are collected and encapsulated to generate multi-dimensional state data.

4. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 3, characterized in that, The process of obtaining the global state distance matrix is ​​as follows: Based on multidimensional state data, a multidimensional state vector corresponding to the dimensions of pipeline state quantities is established for each spatial detection unit. Traverse all spatial detection units, arbitrarily select two sets of multidimensional state vectors from different spatial detection units, and calculate the Euclidean distance between the two sets of multidimensional state vectors. The Euclidean distances calculated pairwise are arranged in rows and columns according to the cell number and uniformly standardized into a two-dimensional matrix format to generate the global state distance matrix.

5. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 4, characterized in that, The process for obtaining the potential risk units is as follows: All Euclidean distance values ​​in the global state distance matrix are imported into the local outlier factor algorithm, and the external failure index is calculated unit by unit based on the overall state distribution law of the entire pipeline section. An external failure judgment threshold is pre-set. The external failure index of each spatial detection unit is compared with the external failure judgment threshold one by one. All spatial detection units with external failure index values ​​greater than the external failure judgment threshold are selected and marked as potential risk units. The continuous time-series vibration data corresponding to the potential risk units are extracted.

6. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 5, characterized in that, The process of obtaining the standardized noise reduction trajectory features is as follows: Extract the multi-point vibration amplitude change sequence from the continuous temporal vibration data corresponding to the potential risk unit, fit and generate the spatial propagation trajectory of the disturbance, and extract three types of trajectory quantitative indicators: disturbance movement direction, energy diffusion rate, and disturbance duration. By introducing real-time environmental monitoring parameters and historical disturbance sample sets for continuous iterative calculation, the dynamic suppression threshold is dynamically updated to adapt to the current working conditions. The disturbance trajectory features are matched and verified with the dynamic suppression threshold. Risk-free disturbance signals that match natural wind and rain or small animal activities are removed. After filtering out environmental interference, the data format is standardized and regularized to generate standardized noise-reduced trajectory features.

7. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 6, characterized in that, The calculation process for the comprehensive risk index is as follows: Based on noise reduction trajectory characteristics, section risk scalar, real-time cable temperature and external failure index of each unit, a temperature prediction neural network is built to obtain the theoretical predicted temperature of the cable, and the absolute value of the difference between the real-time cable temperature and the theoretical predicted temperature is calculated to obtain the cable temperature deviation. A multi-parameter weighted fusion model is constructed, which performs weighted coupling operations on temperature deviation, noise reduction trajectory characteristics, segment risk scalar, and external failure index to obtain a comprehensive risk index.

8. The method for monitoring external damage to cable channels based on distributed vibration optical fiber according to claim 7, characterized in that, The process of obtaining the pipeline external force damage risk warning result is as follows: Pre-set alarm thresholds, compare the comprehensive risk index with the alarm thresholds, and when the comprehensive risk index is greater than the alarm threshold, generate a cable external force damage warning signal that is bound to the corresponding spatial detection unit number, section risk level, disturbance trajectory type and cable temperature rise deviation. The system outputs early warning signals for cable damage caused by external forces, simultaneously initiates joint response procedures, locates risk points based on pipeline spatial geographic data, retrieves front-end image equipment for verification, and sends local audible and visual alarms and SMS messages to maintenance terminals.

9. A cable channel external damage prevention monitoring system based on distributed vibration optical fiber, characterized in that, The system is used to perform the cable channel external damage monitoring method based on distributed vibration optical fiber as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is executed by a processor to implement the cable channel external damage monitoring method based on distributed vibration optical fiber as described in any one of claims 1-8.

Images