A distributed fiber optic synchronous vibration monitoring system and method

Abstract

This invention belongs to the field of security monitoring technology. It discloses a distributed optical fiber synchronous vibration monitoring system and method, comprising: acquiring the parameter requirements of the target monitoring scene; calling a preset scene parameter mapping table for scene adaptation configuration; generating scene parameter packages and synchronous trigger signals; filtering vibration signals of corresponding frequencies; intercepting candidate target signal segments by combining energy attainment with temperature and polarization state abrupt changes; encapsulating these segments into synchronous spatiotemporal data packets; accessing a synchronous data pool to bind three-dimensional data; generating a set of suspicious events; decoupling concurrent events to obtain real events; and generating a list of high-confidence events; the first reporting node broadcasting an inquiry signal containing event fingerprints through dual communication channels; neighboring nodes responding to lock the smallest geographical grid segment; generating grid-level relative positions and location confidence; classifying alarm levels; executing differentiated operation and maintenance linkages; archiving end-to-end data; periodically optimizing end-to-end parameters; and forming a closed loop.

Description

Technical Field

[0001] This invention relates to the field of security monitoring technology, and more specifically, to a distributed optical fiber synchronous vibration monitoring system and method. Background Technology

[0002] With the intelligent upgrading of security protection needs in key areas such as oil and gas pipelines and perimeter security, the demand for accurate identification, rapid location, and efficient handling of abnormal events such as leaks and cuts has surged. However, traditional monitoring relies heavily on single-dimensional sensing and fixed parameter configurations, making it difficult to adapt to complex environmental fluctuations. Furthermore, the fragmented nature of the data collection, verification, location, and handling processes leads to numerous missed and false alarms, large location deviations, and delayed responses, failing to meet the requirements of "early detection, accurate location, and rapid handling."

[0003] Existing solutions still have shortcomings: for example, they only focus on monitoring a single signal such as vibration or temperature, lack multimodal collaborative verification, and have a high rate of missed detection in edge scenarios; they rely on static parameters, resulting in poor scenario adaptability and a high false alarm rate; they lack methods for decoupling concurrent events, making it easy to confuse real events with interference; the single-node positioning architecture has poor fault tolerance, and node failure directly leads to positioning failure; and they lack full-process data archiving and feedback optimization, which makes it impossible to continuously improve performance and results in a high problem reproducibility rate. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a distributed optical fiber synchronous vibration monitoring system, comprising:

[0005] Scene adaptation unit: Obtains the parameter requirements of the target monitoring scene, calls the preset scene parameter mapping table to perform scene adaptation configuration, and generates scene parameter package and synchronous trigger signal;

[0006] Optical domain acquisition unit: Based on scene parameter package and synchronous trigger signal, it filters vibration signals of corresponding frequency, and by combining energy reaching the standard with the coordinated abrupt change of temperature and polarization state, it intercepts candidate target signal segments; it marks the candidate target signal segments with event fingerprints and scene tags, and encapsulates them into synchronous spatiotemporal data packets;

[0007] Event decoupling unit: Based on synchronous spatiotemporal data packets, it accesses the synchronous data pool and binds 3D data. It generates a set of suspicious events through three-level judgment and edge scene complementation rules. Then, it obtains real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generates a list of high-confidence events.

[0008] The positioning calibration unit, based on a list of high-confidence events, broadcasts an inquiry signal containing event fingerprints through dual communication channels. Neighboring nodes respond and lock the smallest geographic grid segment. Then, it calls the static benchmark library and dynamic anchor point library for calibration, generating grid-level relative position and positioning confidence.

[0009] Incident handling unit: Based on grid-level relative location and location reliability, classify alarm levels and execute differentiated operation and maintenance linkage; archive full-link data, and regularly optimize full-link parameters to form a closed loop.

[0010] Furthermore, the generation method of the scene parameter package and the synchronization trigger signal includes:

[0011] The target monitoring scene information is acquired and parsed into structured parameter requirements. The preset scene parameter mapping table is retrieved to match the corresponding benchmark parameters, forming a scene parameter package.

[0012] The initial configuration is performed based on the scene parameter package, and a synchronization trigger signal is generated and distributed to the entire link.

[0013] Furthermore, the method for intercepting the candidate target signal segment includes:

[0014] Based on the scene parameter package and the synchronous trigger signal, laser pulses are emitted and backscattered signals from optical fibers are received. The vibration signal corresponding to the target frequency is obtained by directional frequency filtering.

[0015] The energy value of the vibration signal is collected in real time and compared with the reference parameters. The energy qualified signal is selected by screening. At the same time, the temperature and polarization state data of the corresponding optical fiber position are collected. The temperature and polarization coordinated abrupt change qualified signal is determined by baseline comparison.

[0016] For vibration signals that meet both criteria, segments are extracted according to preset rules, and after quality verification, they are marked as candidate target signal segments.

[0017] Furthermore, the method of encapsulating the event fingerprint and scene markers for labeling candidate target signal segments into a synchronous spatiotemporal data packet includes:

[0018] For candidate target signal segments, an event fingerprint is generated by matching the scene parameter package with the actual frequency characteristics of the signal, and at the same time, the scene type identifier is bound as a scene marker.

[0019] The tagged candidate target signal segments are embedded with synchronization timestamps, fiber optic geoidentifiers, and signal quality information, and then encapsulated into synchronization spatiotemporal data packets.

[0020] Furthermore, the generation method of the suspicious event set includes:

[0021] Based on synchronous spatiotemporal data packets, candidate target signal segments are accessed into a pre-set synchronous data pool, and temperature and polarization state data with the same geographic identifier and time sequence are retrieved to form a three-dimensional data group.

[0022] Then, high-confidence candidates are marked using a three-level decision rule:

[0023] First, a dynamic baseline is constructed based on vibration data during event-free periods to verify the vibration amplitude. If the amplitude meets the standard, the event is judged as a primary triggering pass. Then, the coordinated abrupt change in temperature or polarization state is verified. If the amplitude meets the standard, the event is judged as a suspicious event. Finally, the consistency of vibration, temperature and polarization state three modes is verified for suspicious events. If the consistency meets the standard, the event is marked as a high-confidence candidate.

[0024] Based on high-confidence candidates and combined with edge scene complement rules, a set of suspicious events with quality labels is generated.

[0025] Furthermore, the method of obtaining real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generating a list of high-confidence events, includes:

[0026] The suspicious event set is grouped by geographic grid, and multi-dimensional completion verification is performed on suspicious events at the edge of the group. If they pass the verification, they are upgraded to candidates and included in the group.

[0027] Duplicate events are eliminated by clustering events within a group based on their spatiotemporal correlation; then, modal differentiation is performed by combining event fingerprints and physical quantity characteristics to filter out real events and determine their types.

[0028] From the real events categorized into groups, events with high confidence levels and clear types are selected and labeled as standardized event types to form a list of high-confidence events.

[0029] Furthermore, the first reporting node broadcasts an inquiry signal containing an event fingerprint through dual communication channels, and neighboring nodes respond by locking the smallest geographic grid segment in the following ways:

[0030] Select the node closest to the event and with the best signal strength from the distributed monitoring nodes as the first reporting node;

[0031] The first reporting node generates an inquiry signal containing event fingerprints based on a list of high-confidence events, and broadcasts it to upstream and downstream neighboring nodes in parallel through primary and backup dual communication channels. Each neighboring node retrieves local data for matching and returns valid response information.

[0032] The first reporting node summarizes the valid response information and filters valid response nodes according to matching degree and response time; then, by combining the geographical coordinates of the first reporting node and valid response nodes, the smallest geographical grid segment where the event is located is located.

[0033] Furthermore, the generation methods for the grid-level relative positions and location confidence include:

[0034] Using the locked smallest geographic grid segment as the search range, the built-in static benchmark library and dynamic anchor point library are called to match all surrounding static benchmark points and historical event anchor points in the same scene and environment as two types of benchmark points, thereby obtaining the grid-level relative position of the event relative to the benchmark point.

[0035] The location reliability of the relative position is obtained by evaluating the confidence level through a confidence level assessment mechanism.

[0036] Furthermore, the method of classifying alarm levels and executing differentiated operation and maintenance linkages based on grid-level relative location and location confidence includes:

[0037] Based on grid-level relative location, different alarm levels are divided according to event type and location reliability, and alarms are pushed through multiple channels.

[0038] Based on the alarm level, joint maintenance personnel will carry out differentiated maintenance coordination and track the maintenance progress in real time.

[0039] After the operation and maintenance is completed, submit and verify the handling result form. If the verification is successful, mark the event as closed and notify the operation and maintenance personnel.

[0040] Archive end-to-end data using event IDs as indexes, and regularly generate event analysis reports to optimize end-to-end parameters, forming a closed loop.

[0041] Furthermore, a distributed optical fiber synchronous vibration monitoring method includes:

[0042] S1: Obtain the parameter requirements of the target monitoring scene, call the preset scene parameter mapping table to perform scene adaptation configuration, and generate scene parameter package and synchronous trigger signal;

[0043] S2: Based on the scene parameter package and the synchronous trigger signal, the corresponding frequency vibration signal is screened. By combining the energy threshold with the coordinated abrupt change in temperature and polarization state, candidate target signal segments are extracted. Event fingerprints and scene markers are added to the candidate target signal segments, and they are encapsulated into a synchronous spatiotemporal data package.

[0044] S3: Based on synchronous spatiotemporal data packets, access the synchronous data pool to bind 3D data, generate a set of suspicious events through three-level judgment and edge scene complementation rules; then obtain real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generate a list of high-confidence events;

[0045] S4: Based on a list of high-confidence events, the first reporting node broadcasts an inquiry signal containing event fingerprints through dual communication channels. Neighboring nodes respond and lock the smallest geographic grid segment. Then, the static benchmark library and dynamic anchor point library are called for calibration to generate grid-level relative position and location confidence.

[0046] S5: Based on grid-level relative location and fixed location reliability, classify alarm levels and execute differentiated operation and maintenance linkage; archive full-link data, and regularly optimize full-link parameters to form a closed loop.

[0047] The technical effects and advantages of the distributed optical fiber synchronous vibration monitoring system and method of the present invention are as follows:

[0048] This invention addresses existing shortcomings through end-to-end collaborative design. First, it analyzes scene features to generate dynamic parameter packages and synchronizes them across the entire chain, replacing static configuration and adapting to different scenes and environmental changes, thus reducing false alarms from the source. Second, after collecting and encapsulating vibration signals, it retrieves temperature and polarization state data to construct a three-dimensional data set. Combined with three-level decision-making and edge verification, it solves the problem of single-dimensional missed detection and improves recognition accuracy.

[0049] Then, spatial clustering is used to remove duplicates and modality differentiation to filter real events, thus solving the problem of concurrent confusion; the first report and neighbor node triangulation are linked, and the benchmark library is used for calibration, and fault tolerance is provided when a node fails, thereby improving the positioning accuracy and reliability; next, alarms are issued through multiple channels according to the event level, and emergency events are automatically linked for monitoring and dispatching, and the handling progress is tracked in real time to ensure response efficiency; after verifying the handling results, the loop is closed, and the entire chain of data is archived by event ID;

[0050] Finally, parameters are iteratively optimized based on archived data to achieve a closed loop of "monitoring-handling-optimization" and improve system robustness.

[0051] This solution provides precise control throughout the entire process, is adapted to critical protection scenarios, significantly improves identification, location, and response capabilities, and reduces labor costs. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a distributed optical fiber synchronous vibration monitoring system according to the present invention;

[0053] Figure 2 This is a schematic diagram illustrating the process of generating a suspicious event set in a distributed optical fiber synchronous vibration monitoring system according to the present invention.

[0054] Figure 3 This is a schematic diagram of a distributed optical fiber synchronous vibration monitoring method according to the present invention. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example 1

[0057] Please see Figure 1 and Figure 2 As shown in the figure, the distributed optical fiber synchronous vibration monitoring system described in this embodiment includes:

[0058] Scene adaptation unit: Obtains the parameter requirements of the target monitoring scene, calls the preset scene parameter mapping table to perform scene adaptation configuration, and generates scene parameter package and synchronous trigger signal.

[0059] It should be noted that this is used to address the problems of traditional fixed parameters having weak adaptability to different scenarios and complex environments, and high false alarm rates.

[0060] Optical domain acquisition unit: Based on scene parameter package and synchronous trigger signal, it filters vibration signals of corresponding frequency, and intercepts candidate target signal segments by combining energy reaching the standard with the coordinated abrupt change of temperature and polarization state; it marks the candidate target signal segments with event fingerprint and scene mark, and encapsulates them into synchronous spatiotemporal data packets.

[0061] Event decoupling unit: Based on synchronous spatiotemporal data packets, it accesses the synchronous data pool and binds 3D data. It generates a set of suspicious events through three-level judgment and edge scene complementation rules. Then, it obtains real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generates a list of high-confidence events.

[0062] It should be noted that this technology is used to address the issues of single-dimensional sensing failing to detect edge scenarios and concurrent events being easily confused with interference signals.

[0063] Positioning and calibration unit: Based on a list of high-confidence events, the first reporting node broadcasts an inquiry signal containing event fingerprints through dual communication channels, and neighboring nodes respond to lock the smallest geographic grid segment; then, it calls the static benchmark library and dynamic anchor point library for calibration to generate grid-level relative position and positioning confidence.

[0064] It should be noted that this technology is used to address the problem of low positioning accuracy in single nodes and positioning failure caused by the lack of a fault-tolerant mechanism.

[0065] Incident handling unit: Based on grid-level relative location and location reliability, classify alarm levels and execute differentiated operation and maintenance linkage; archive full-link data, and regularly optimize full-link parameters to form a closed loop.

[0066] It should be noted that this is used to address issues such as delayed response times and the lack of a full-process data archiving and continuous optimization mechanism.

[0067] The methods for generating scene parameter packages and synchronization trigger signals include:

[0068] The operation and maintenance personnel can obtain target monitoring scenario information through the local industrial control computer interface or remote operation and maintenance platform, including: monitoring objects (such as long-distance pipelines, industrial park perimeters), monitoring range, core monitoring targets (such as pipeline leaks, construction disturbances), environmental level (such as normal, high noise, extreme temperature change), special monitoring points (such as "a certain chainage is a key section"), and operation and maintenance response time (such as "early warning within 10 minutes").

[0069] Verify the standardization of the acquired target monitoring scene information. For example, the monitoring range must be a positive integer ≥1. If it is not a numerical value or <1, ​​a pop-up window will prompt "Please enter a monitoring range of 1km or more". At least one core monitoring target must be selected. If it is not selected, it cannot be submitted. The naming of special monitoring points must follow the "station number + section" rule (such as "KXX-KXX"). All input anomalies are recorded in the verification log (including operator, time, and anomaly type).

[0070] For the target monitoring scene information that has passed the verification, the built-in parsing rule library that matches the scene type, core requirements and parameter dimensions is called to parse it into a structured parameter requirement document (JSON format), which includes core parameter requirements such as frequency range, positioning accuracy, decision threshold and anchor point density. If there is a high noise environment, the adaptation requirements such as "energy dynamic threshold increased by 20%" and "decision time window shortened to 0.8s" are automatically added to the parameter requirements.

[0071] The example's parsing rule base includes:

[0072] Monitoring object: long-distance pipeline; core monitoring target: pipeline leakage; core parameter requirements after analysis: vibration frequency range 150Hz-200Hz, positioning accuracy ±2m, decision threshold (e.g., synchronization signal deviation threshold 10ns), anchor point density 0.5 / km;

[0073] Retrieve the preset scene parameter mapping table stored in the local database, and match the baseline parameters in the table according to the parsed structured parameter requirements;

[0074] The core fields of the scene parameter mapping table include: scene type, vibration frequency range, energy dynamic threshold, anchor point density for positioning, and synchronization signal deviation threshold; for example:

[0075] Scenario type: Conventional pipeline monitoring; Vibration frequency min: 150Hz; Vibration frequency max: 200Hz; Energy dynamic threshold -30dBm; Positioning anchor density: 0.5 / km; Synchronization signal deviation threshold 10ns;

[0076] It should be noted that maintenance personnel are allowed to make fine adjustments to the matched benchmark parameters within ±10% based on the actual site conditions (such as high-noise sections). For example, the energy threshold can be adjusted from -30dBm to -27dBm. The "reason for fine adjustment" must be filled in for the fine adjustment operation (such as "the noise of the pump group on site is high"). All fine adjustment records are synchronously stored in the parameter change log.

[0077] Verify whether the reference parameters after matching and fine-tuning are within the hardware range of the system module (e.g., the vibration frequency range should be within the 10-500Hz range of the optical domain acquisition module). If they are outside the range, prompt "Parameters exceed hardware range, it is recommended to adjust to XX-XX" and automatically roll back to the reference parameter matching stage.

[0078] Integrate the verified baseline parameters to form a scenario parameter package;

[0079] Based on the scene parameter package, it is encapsulated into standardized configuration instructions and issued to modules (e.g., the optical domain acquisition module issues vibration frequency range and energy threshold parameters, and the distributed positioning module issues anchor point density and positioning accuracy parameters). Each module completes the initialization configuration according to the configuration instructions. After the instructions are issued, wait for the module's "configuration successful" receipt. If no receipt is received within a timeout (e.g., 10 seconds), the instructions are reissued. If there are 3 consecutive timeouts, a "module communication abnormality" alarm is triggered.

[0080] After all modules are successfully initialized, the optical frequency comb module generates a nanosecond-level synchronization trigger signal with the corresponding linewidth (e.g., <10kHz) and frequency stability (±1ppm) according to the synchronization signal deviation threshold requirements in the scene parameter package. The signal is then distributed to the entire link modules, such as optical domain acquisition and electrical domain verification, through a dedicated optical communication channel.

[0081] Methods for capturing candidate target signal segments include:

[0082] Based on the scene parameter package and the synchronous trigger signal, the corresponding reference parameters (vibration frequency range and energy dynamic threshold) are loaded into the optical domain acquisition module. The laser pulse is emitted by the onboard φ-OTDR laser emitting device according to the timing of the synchronous trigger signal, and the fiber backscatter signal is received by the receiving end. After being focused by the lens of the receiving end, the fiber backscatter signal is divided into multiple signals, one of which is used for frequency filtering (the rest are reserved).

[0083] Based on the applied vibration frequency range, the scattered signal is directionally filtered by the onboard optical filtering device (e.g., filtering 150-200Hz signals in a pipeline scenario) to obtain the vibration signal at the corresponding target frequency; during the filtering process, the timing is kept aligned with the synchronous trigger signal to ensure that the timestamp of the filtered signal is consistent with the synchronization reference.

[0084] The energy value of the filtered vibration signal is collected in real time and compared with the dynamic energy threshold. If the energy value exceeds the threshold for three consecutive sampling periods (default sampling rate 1kHz, period 1ms), it is determined that the energy meets the standard and is used as the energy compliant signal. Vibration signals that do not meet the standard are marked as low energy signals and temporarily stored in the buffer (automatically discarded after 5s).

[0085] Simultaneously, temperature and polarization state data at the corresponding fiber location are collected and baseline comparison is performed: if the temperature change is ≥ ±1℃ and the duration is > 100ms, or the polarization state deviates from the baseline by ≥ ±0.1rad and the time difference with the energy standard is < 50ms, it is determined that the temperature-polarization coordinated change meets the standard, and the temperature-polarization coordinated change meets the standard signal. Signals that do not meet the standard are marked as having no coordinated change and are temporarily stored for 5s before being discarded.

[0086] For vibration signals that meet both criteria, segments are extracted according to preset rules: starting from the moment the energy meets the criteria for the first time, 0.5 seconds are extracted forward and 2 seconds are extracted backward to form a complete target signal segment from 0.5 seconds before triggering to 2 seconds after triggering.

[0087] The captured target signal segments undergo quality verification, with key indicators including: signal-to-noise ratio ≥20dB (to avoid noise interference); signal amplitude fluctuation range ≤±10% (to ensure signal stability); timestamp continuity (no jumps, no repetitions); and qualified signals are marked as candidate target signal segments.

[0088] Signals that fail quality verification (e.g., signal-to-noise ratio < 20dB) are marked as substandard and discarded, and the non-compliant indicators are recorded (e.g., "signal-to-noise ratio 18dB < 20dB"). If 10 consecutive signals fail, a signal quality anomaly alarm is triggered, prompting the user to check the fiber optic link or adjust the screening parameters.

[0089] Methods for labeling candidate target signal segments with event fingerprints and scene tags, and encapsulating them into synchronous spatiotemporal data packets include:

[0090] Verify the completeness of the core attributes of the candidate target signal segment: timestamp continuity (no jumps, no repetitions, deviation from the synchronous trigger signal ≤10ns); signal length meets the standard; quality label is valid (including signal-to-noise ratio and stability indicators); if the verification fails, mark the signal as incomplete and discard it, and record the verification log (including signal ID and non-compliance items) simultaneously.

[0091] The vibration frequency range is extracted from the combined scene parameter package. Combined with the actual frequency characteristics of the signal, the event fingerprint is generated by matching the frequency range + feature code format (e.g., 150-200Hz corresponds to "L01", which represents suspected pipeline leakage. The feature code and the frequency and event type mapping in the scene parameter package need to correspond).

[0092] Simultaneously, the scene type identifier (e.g., "P001" for "pipeline monitoring") is extracted from the scene parameter package and used as a scene marker to bind to the candidate target signal segment. During the binding process, the logical consistency between the scene marker and the event fingerprint is ensured (e.g., the "P001" scene is not allowed to have an event fingerprint specific to the "B001" scene). If they are inconsistent, a marker conflict alarm is triggered. All marking operations strictly follow the synchronous trigger signal timing to ensure that the timestamps of the event fingerprint, scene marker, and signal segment are completely aligned to avoid misalignment between the marker and the signal.

[0093] Nanosecond-level synchronization timestamps, fiber optic geographic identifiers (in the format of "station number + distance", such as "K120+350" representing 350m after station number K120) and signal quality information (signal-to-noise ratio, amplitude fluctuation range, etc.) are embedded into the marked candidate target signal segments. The packets are encapsulated into a synchronization spatiotemporal data packet using a three-segment structure of "header-body-tail" (packet header (64 bits): containing event fingerprint, scene marker, synchronization timestamp, geographic identifier; packet body (variable length): original candidate target signal segment data; packet tail (32 bits): containing signal quality label, MD5 checksum (used for data integrity verification)).

[0094] The methods for generating suspicious event sets include:

[0095] The synchronous spatiotemporal data packet is decompressed to extract candidate target signal segments (i.e., vibration signals) and embedded synchronization timestamps and geographic identification information.

[0096] The extracted data is connected to a pre-set synchronous data pool (containing a pre-set data cache queue (capacity ≥ 1000 synchronous spatiotemporal data packets), supporting real-time writing and associated retrieval of three types of data: vibration, temperature, and polarization state. The data pool is indexed by "fiber optic geographic identifier + synchronization timestamp" to ensure rapid matching of data at the same location and time sequence). Temperature data and polarization state data with the same geographic identifier and time sequence as the current synchronous spatiotemporal data packet are retrieved and bound together with candidate target signal segments to form a three-dimensional data group. The associated fields include geographic identifier, unified synchronization timestamp, and scene type.

[0097] It should be noted that the core of the candidate target signal segment comes from the synchronous spatiotemporal data packet. This data packet encapsulates the target vibration signal segment (including synchronization timestamp and geographic identifier) ​​that has been pre-screened and marked in the optical domain. After receiving it, the vibration signal is first decompressed and extracted, and then stored in the synchronous data pool as the core basis for binding.

[0098] Temperature data is independently acquired by the distributed temperature sensor (DTS) and written to the synchronization data pool in real time. It is not included in the synchronization spatiotemporal data package (the data package only contains vibration signal related information and does not contain complete temperature time series data); polarization state data is independently acquired by the polarization state detection device and written to the synchronization data pool in real time. It is also not included in the synchronization spatiotemporal data package and needs to be matched by "geographic identifier + timestamp" through the synchronization data pool.

[0099] The synchronous data pool is the central hub for the aggregation and correlation of multimodal data.

[0100] The vibration signal in the synchronous spatiotemporal data packet is the anchor point data that triggers the binding (only the target vibration signal that has passed the optical domain pre-screening will initiate the subsequent multimodal binding).

[0101] Temperature and polarization state data are real-time background data, which are continuously collected by a dedicated device hardware module and stored in a synchronous data pool, waiting to be matched with the target vibration signal.

[0102] The binding logic is as follows: using the geographic identifier and synchronization timestamp in the synchronization spatiotemporal data packet as an index, temperature and polarization state data of the same location and time sequence are retrieved in the synchronization data pool. The three are combined to form a three-dimensional data group of "vibration-temperature-polarization state" to ensure the spatiotemporal consistency of the data.

[0103] The purpose of this design is that temperature and polarization state are time-series continuous data, and the synchronous spatiotemporal data packet only encapsulates candidate target signal fragments without carrying complete time-series data. This can greatly reduce the size of the data packet, and when needed, the corresponding time-series data can be retrieved from the synchronous data pool.

[0104] The time axes of the three types of data in the three-dimensional dataset are aligned, and then confidence levels are graded using a three-level decision rule:

[0105] First-level judgment: Based on vibration data during the event-free period (nearly 10 minutes), construct a dynamic baseline (baseline value = historical data mean + 2 standard deviations).

[0106] The vibration amplitude of candidate target signal segments in the three-dimensional data set is verified. If the amplitude of three consecutive sampling points (sampling rate 1kHz, period 1ms) exceeds the dynamic baseline, it is determined that the primary trigger has passed and a primary event record (including timestamp, geographic identifier, and vibration amplitude) is generated. If it fails, the vibration is marked as not meeting the standard and the three-dimensional data set is temporarily stored for 5 seconds and then automatically discarded.

[0107] Level 2 Decision: For events that pass the initial trigger, a 1-second verification time window is initiated to verify the synergy between temperature and polarization state. The requirements for temperature data are: a sudden change of ≥±1℃ exists within the time window, and the time difference between the sudden change and the peak vibration time is ≤100ms. The requirements for polarization state data are: a baseline deviation of ≥±0.1rad exists within the time window, and the time difference between the deviation and the peak vibration time is ≤50ms. If either the temperature data requirement or the polarization state data requirement is met, the event is judged as a suspicious event, and the event record is updated as a suspicious event entry (adding temperature and polarization state sudden change data). If neither condition is met, it is marked as having no synergistic sudden change, the data set is discarded, and the decision log is recorded.

[0108] Level 3 Judgment: For suspicious events, a three-mode consistency check is performed on vibration signals, temperature abrupt changes, and polarization state deviations: Vibration signals: duration > 0.3s (excluding instantaneous interference), and amplitude stable (fluctuation range ≤ ±15%); Temperature abrupt changes: the direction of the abrupt change matches the event fingerprint (e.g., "pipeline leak" fingerprint corresponds to temperature drop, "mechanical cutting" corresponds to temperature rise); Polarization state deviation: the overlap between the deviation duration and the vibration duration is ≥ 80%.

[0109] If all three conditions are met, the candidate is marked as high-confidence; if not, it remains in the "suspicious event" state and is not removed (to be further decoupled later).

[0110] The edge scene complementation rule is defined as follows: For edge scenes with "weak vibration (e.g., amplitude = dynamic baseline × 1.1-1.2) + strong temperature deviation (e.g., temperature ±2℃ or polarization state ±0.15rad)" (e.g., pipeline micro-leakage), the verification time window is automatically extended from 1s to 2s to reduce the risk of missing low-amplitude events; if the high confidence condition is still not met after the extension, it is marked as an edge suspicious event.

[0111] All high-confidence candidates, suspicious events, and marginally suspicious events are sorted in ascending order by geographic identifier + timestamp to generate a structured set of suspicious events. Each entry includes: event ID (geographic identifier + timestamp), event type (high-confidence candidate, suspicious, marginally suspicious), 3D data group summary, quality label, event fingerprint, and scene marker.

[0112] Methods for generating a high-confidence event list include: decoupling concurrent events through event fingerprint clustering and modality differentiation to obtain real events; and using these methods.

[0113] Suspicious event sets are grouped by geographic grid identifiers (geographic grids are divided into 1km×1km, associated with location information in the scene parameter package), and suspicious events at the edge of the group are bound to the historical event database of the corresponding grid (event data of the same grid in the past 30 days, including event type and related data thresholds); if a grid has no historical data, the threshold of edge events of the same type of scene in the scene parameter package is automatically called as a reference (e.g., the vibration amplitude threshold of micro-leakage in pipeline scene is lowered by 10%).

[0114] For edge-suspicious events after binding, multi-dimensional completion verification is performed: First, the time window is extended from 1-2 seconds to 5 seconds to verify the weak and continuous stability of the vibration signal (amplitude fluctuation ≤ ±20%); second, multi-node cross-verification is performed by retrieving synchronous data from two adjacent monitoring nodes upstream and downstream of the corresponding geographic grid. If there is a coordinated signal of weak vibration + temperature deviation change between adjacent nodes, the cross-verification is deemed successful; third, dynamic threshold adaptation is performed by adjusting the judgment threshold according to the environmental level (e.g., high noise, normal) (e.g., the temperature deviation change threshold in high noise environment is increased to ±1.5℃). If all three verification steps are successful, the suspicious event is upgraded to a candidate; if they fail, the edge-suspicious label is retained and an explanation of the verification failure is attached. Events that have been upgraded to candidates after passing the verification are included in the group of events.

[0115] Concurrent event decoupling includes spatiotemporal clustering and modality differentiation;

[0116] Spatiotemporal clustering involves clustering events within a group based on their spatiotemporal correlation and removing duplicate events. Specifically, clustering rules are first set: events within the same geographic grid, with a timestamp deviation ≤ 5s and an event interval ≤ 50m, are grouped into the same cluster. Then, feature fusion is performed within the cluster: the mean vibration frequency, the mean peak temperature change, and the mean polarization state change of events within the cluster are calculated, and abnormal events whose features deviate from the cluster mean by ±30% are removed (judged as interference). Finally, trajectory processing is performed: if events within a cluster show a continuous movement trend in geographic location after being sorted by timestamp (e.g., movement ≤ 10m within 5s), they are merged into moving target trajectory events, and the trajectory direction and speed are marked; if multiple events overlap at a fixed location, the one with the clearest feature is retained as the core event, and the rest are marked as duplicate events and removed.

[0117] Modal differentiation: After removing duplicate events within a group, modal differentiation is performed by combining event fingerprints and physical quantity characteristics to distinguish event types and filter out real events. Specifically:

[0118] First, fingerprint matching is performed: the event fingerprints of clustered events (e.g., 150-200Hz corresponds to the suspected leakage fingerprint L01) are extracted and matched with the fingerprint-event type mapping in the scene parameter package to preliminarily determine the event type; then, physical quantity cross-validation is performed: further distinction is made based on the direction of temperature change and vibration duration, such as "suspected leakage" needing to meet "temperature drop of more than ±1℃ and vibration duration of 0.5-2s"; finally, interference is removed: events with ambiguous fingerprint matching and contradictory physical quantity characteristics (e.g., "leakage fingerprint but temperature rise") are marked as interference events, and after being labeled with interference, they are temporarily stored (not included in high-confidence events).

[0119] From the remaining real events after removing interfering events within each group, events with qualified confidence and clearly defined types are selected. These events must meet the following criteria: increased confidence after spatial clustering (e.g., marginal events are upgraded to candidates after cross-validation); clear modal differentiation results (unique event type; physical quantity features and fingerprints are perfectly matched); and qualified 3D data quality (vibration signal-to-noise ratio ≥25dB; no abnormal jumps in temperature and polarization state data). Suspicious events and interfering events that do not meet the criteria are retained in the local cache (for 24 hours, for manual review to determine whether to reuse them).

[0120] For the selected high-confidence events, they are labeled with a standardized type according to a preset type (such as "pipeline leak (L01)" and "perimeter cutting (C02)"). The event ID, geographic grid identifier, synchronization timestamp and event fingerprint are associated to form a "high-confidence event list".

[0121] The first reporting node broadcasts an inquiry signal containing the event fingerprint through dual communication channels. Neighboring nodes respond and lock onto the smallest geographic grid segment in the following ways:

[0122] The system receives a list of high-confidence events through a distributed node communication network. Centered on the geographic grid corresponding to each high-confidence event, it selects the node from the distributed monitoring nodes that is closest to the event location, has the highest signal reception strength, and meets the historical positioning accuracy standard (error ≤ ±3m) as the first reporting node. Simultaneously, it marks the first reporting node and notifies it to initiate the collaborative process.

[0123] The first reporting node generates an inquiry signaling based on the list of high-confidence events, which includes: the first reporting node ID, the synchronization timestamp, the signaling validity period, and the core characteristics of the high-confidence event (event fingerprint, geographic grid identifier, vibration frequency range).

[0124] The query signal is broadcast in parallel to upstream and downstream neighboring nodes through primary and backup dual communication channels. The primary channel is a dedicated fiber optic network (bandwidth ≥ 100Mbps, packet loss rate < 0.1%), covering 5 nodes upstream and downstream of the event location (a total of 10 neighboring nodes). The backup channel is a dedicated wireless network (4G, 5G industrial modules, latency ≤ 100ms), with the same coverage as the primary channel, used as a backup in case of primary channel failure. The broadcast includes a response timeout (e.g., 500ms), requiring neighboring nodes not to respond within the timeout period; otherwise, the broadcast is considered invalid.

[0125] After receiving the query signaling, each neighboring node retrieves local data using the event fingerprint and timestamp in the query signaling. If there is a record of the same event that is in the same period and the timestamp deviation meets the expectation (e.g., ≤500μs) and the event fingerprint matching degree meets the expectation (e.g., ≥90%, meaning that more than 90% of the event fingerprint is consistent), then it is determined to be a valid response and a valid response signaling is generated; if there is no matching record or the signaling is invalid, it is discarded directly and no response is returned.

[0126] The effective response signaling includes: its own node ID, event matching degree value, geographical coordinates of the locally monitored event (calibrated based on the node's preset location), and response timestamp; it is returned to the first reporting node through the original receiving channel to ensure that the response information of the primary and backup channels is transmitted independently and without interference.

[0127] The first reporting node summarizes all valid response information and filters valid response nodes based on the conditions of event matching degree ≥90% and response time ≤500ms, removing invalid nodes that do not meet the conditions of event matching degree or response event. Then, weights are assigned to valid nodes, with the weight coefficient being larger (range 0.8-1.2) for nodes that are closer to the event center and have higher historical positioning accuracy.

[0128] Based on the geographic coordinates of the first reporting node and the effective response nodes, and combined with the divided geographic grid, a triangulation method is used to construct the connections between nodes. If there are ≥2 effective response nodes, the grid segment corresponding to the intersection of each connection is taken as the minimum grid set (e.g., "grid G120-G121 at the intersection of the lines connecting node A-K120 and node B-K122"). If there is only 1 effective response node, the 3 consecutive grids between the first reporting node and the effective response node are taken as the minimum grid set. The accuracy of the range is optimized by node weight (because nodes that are close and have high positioning accuracy have higher weights). If there are no effective response nodes, the single node is marked and the 5 adjacent grids covered by the first reporting node are locked. The range of upstream and downstream neighboring nodes is expanded, the query signal is rebroadcast, and the static reference call density is increased to supplement the reference anchor points for grid positioning. If there is still no response after expanding the range, the historical positioning error data of the single node is used for correction to obtain the minimum grid set.

[0129] The smallest grid set is taken as the smallest geographic grid segment where the event is located.

[0130] Methods for generating grid-level relative positions and location-based confidence include:

[0131] Using the locked minimum geographic grid segment as the search range, the built-in static benchmark library and dynamic anchor point library are called to match all surrounding static benchmark points and historical event anchor points in the same scene and environment as two types of benchmark points, including static benchmark points and dynamic anchor points.

[0132] The static baseline database stores the precise coordinates of fixed facilities such as manholes, markers, and valves, indexed by geographic grid partitions; the dynamic anchor point database stores the locations of manually confirmed high-confidence events, along with event type, positioning accuracy, and environmental labels (such as "rainy day" or "high noise").

[0133] Matching rules: Static reference points retrieve the coordinates of all static reference points within the smallest geographic grid segment and its two adjacent grids; dynamic anchor points are selected from historical event locations with the same scene and environment label, and whose event fingerprints and vibration characteristics are similar (similarity ≥ 90%) and included in the reference.

[0134] First, select 2-5 static reference points that are closest to the smallest geographic grid segment. Then, select the static reference point with the highest coordinate accuracy as the origin. Calculate the relative distance between the center of the smallest geographic grid segment and the origin (e.g., "32m downstream of origin 5"). Combine this with the relative deviation correction of the dynamic anchor point (e.g., the average deviation between historical events and the static reference is 0.3m, and the current result is corrected in reverse) to obtain the grid-level relative position of grid identifier + relative position.

[0135] The location reliability of the relative position is obtained by evaluating the confidence level through a confidence level assessment mechanism;

[0136] The confidence assessment mechanism is as follows: based on the base score of the number of benchmark matching points (≥2 static benchmark points + 1 dynamic anchor point, then the base score is 80 points), combined with environmental adaptability (add 20 points if there is no extreme environment, deduct 20 points if there is), the total score corresponds to the location reliability (e.g., 100 points → 100%).

[0137] Based on grid-level relative location and location reliability, alarm levels are classified, and differentiated operation and maintenance linkage methods are implemented, including:

[0138] Based on grid-level relative location, and combined with event type and location reliability, alarm levels are divided into three levels. The trigger condition for a Level 1 alarm is a high-confidence event + a high-risk type (the risk type is preset in advance by expert experience, such as leakage or cutting); the trigger condition for a Level 2 alarm is a non-high-confidence event + a normal risk type (such as construction disturbance); and the trigger condition for a Level 3 alarm is a non-high-confidence level + a low-risk type (such as personnel loitering).

[0139] Trigger differentiated alarm pushes based on alarm level through multiple channels;

[0140] Example: Level 1 alarm: audible and visual alarm from the operations and maintenance center + SMS or telephone notification from on-duty personnel + pop-up in the mobile app; Level 2 alarm: pop-up in the app + SMS notification; Level 3 alarm: platform message reminder; the push content includes "event type, precise location (grid + relative coordinates), location reliability and handling requirements";

[0141] Based on the alarm level, joint operation and maintenance personnel will carry out differentiated operation and maintenance coordination, including automatic linkage triggering and manual handling coordination.

[0142] Among them, automatic linkage triggering: For level one or level two alarms, automatic linkage actions are triggered, including: monitoring linkage: real-time images from three cameras around the event location are retrieved and pushed to the operation and maintenance platform; sensor linkage: supplementary monitoring sensors (such as pipeline pressure sensors) in the event area are activated and data is transmitted back in real time; dispatch linkage: based on the geographical grid matching of operation and maintenance personnel, electronic work orders (including location navigation and handling specifications) are automatically generated and dispatched to the nearest operation and maintenance team.

[0143] Manual handling and collaboration: Level 3 alarms or special scenarios (such as extreme weather preventing automatic dispatch) can be handled manually by operations and maintenance personnel. A detailed "Event Details Page" is provided, supporting the viewing of the entire chain of data. Operations and maintenance personnel can manually enter dispatch information and designate the person in charge of handling the situation. The entire progress of "order acceptance - departure - arrival - handling completion" can be tracked synchronously. If there is no response within the time limit, the alert will be upgraded.

[0144] Maintenance personnel can upload real-time progress updates (such as text, on-site photos, and videos) via mobile devices. Event IDs are automatically associated and stored, and the maintenance progress is tracked synchronously. Progress can be traced back (e.g., "10:05 Order received → 10:12 Arrival at the site → 10:30 Completion of the process").

[0145] After maintenance is completed, submit a handling result form and perform verification according to the event type: For fault-related events (leakage, cutting), upload a comparison of data before and after handling (such as pressure recovery curve, vibration signal disappearance record); for interference-related events (equipment false alarm), mark the interference source (such as "pump start-up and shutdown") and upload evidence; for personnel-related events (intrusion, construction), upload "on-site personnel confirmation record" (such as construction permit document).

[0146] Once the verification is successful, the event is marked as closed, and a confirmation notification of completion is sent to the operations and maintenance personnel; if the verification fails (if there is no comparative data), the event is returned for supplementation.

[0147] The entire event data is archived to the operation and maintenance database using the event ID as the index. The archived content includes: source data: scene parameter package, candidate signal fragments; process data: three-level decision records, location calibration logs, alarm push records; handling data: work order information, handling result sheet, on-site evidence; the archived data adopts cold and hot separation storage, the hot data of the past 30 days supports second-level retrieval, and the historical cold data is compressed and retained for 1 year;

[0148] Monthly event analysis reports are generated regularly, and key indicators are statistically analyzed, including: accuracy (the matching rate between high-confidence events and actual handling results (target ≥95%)), handling timeliness (average response time for alarms at each level), and false alarm rate (the proportion of non-high-confidence events that are ultimately determined to be interference). Based on the event analysis reports, key parameters (such as decision thresholds and benchmark weights) are adjusted, and scenario parameter packages and benchmark libraries are updated synchronously to achieve a complete closed loop of event handling and technology iteration.

[0149] Example 2

[0150] Please see Figure 3 As shown, parts not described in detail in this embodiment are described in Embodiment 1. A distributed optical fiber synchronous vibration monitoring method is provided, including:

[0151] S1: Obtain the parameter requirements of the target monitoring scene, call the preset scene parameter mapping table to perform scene adaptation configuration, and generate scene parameter package and synchronous trigger signal;

[0152] S2: Based on the scene parameter package and the synchronous trigger signal, the corresponding frequency vibration signal is screened. By combining the energy threshold with the coordinated abrupt change in temperature and polarization state, candidate target signal segments are extracted. Event fingerprints and scene markers are added to the candidate target signal segments, and they are encapsulated into a synchronous spatiotemporal data package.

[0153] S3: Based on synchronous spatiotemporal data packets, access the synchronous data pool to bind 3D data, generate a set of suspicious events through three-level judgment and edge scene complementation rules; then obtain real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generate a list of high-confidence events;

[0154] S4: Based on a list of high-confidence events, the first reporting node broadcasts an inquiry signal containing event fingerprints through dual communication channels. Neighboring nodes respond and lock the smallest geographic grid segment. Then, the static benchmark library and dynamic anchor point library are called for calibration to generate grid-level relative position and location confidence.

[0155] S5: Based on grid-level relative location and fixed location reliability, classify alarm levels and execute differentiated operation and maintenance linkage; archive full-link data, and regularly optimize full-link parameters to form a closed loop.

[0156] Example 3

[0157] This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the operation mode of the distributed optical fiber synchronous vibration monitoring system described above.

[0158] Since the electronic device described in this embodiment is the electronic device used to implement the distributed optical fiber synchronous vibration monitoring method described in this application embodiment, those skilled in the art can understand the specific implementation method and various variations of the electronic device in this embodiment based on the distributed optical fiber synchronous vibration monitoring method described in this application embodiment. Therefore, how the electronic device implements the method in this application embodiment will not be described in detail here. Any electronic device used by those skilled in the art to implement the distributed optical fiber synchronous vibration monitoring method described in this application embodiment falls within the scope of protection of this application.

[0159] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0160] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for users of ordinary technical skills, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A distributed optical fiber synchronous vibration monitoring system, characterized in that, include: Scene adaptation unit: Obtains the parameter requirements of the target monitoring scene, calls the preset scene parameter mapping table to perform scene adaptation configuration, and generates scene parameter package and synchronous trigger signal; Optical domain acquisition unit: Based on scene parameter package and synchronous trigger signal, it filters vibration signals of corresponding frequency, and by combining energy reaching the standard with the coordinated abrupt change of temperature and polarization state, it intercepts candidate target signal segments; it marks the candidate target signal segments with event fingerprints and scene tags, and encapsulates them into synchronous spatiotemporal data packets; Event decoupling unit: Based on synchronous spatiotemporal data packets, it accesses the synchronous data pool and binds 3D data. It generates a set of suspicious events through three-level judgment and edge scene complementation rules. Then, it obtains real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generates a list of high-confidence events. Location calibration unit: Based on a list of high-confidence events, the first reporting node broadcasts an inquiry signal containing event fingerprints through dual communication channels, and neighboring nodes respond to lock the smallest geographic grid segment; Then, the static benchmark library and dynamic anchor point library are called for calibration to generate grid-level relative position and positional reliability. Incident handling unit: Based on grid-level relative location and fixed location reliability, alarm levels are classified and differentiated operation and maintenance linkages are executed; Archive end-to-end data and regularly optimize end-to-end parameters to form a closed loop.

2. The distributed optical fiber synchronous vibration monitoring system according to claim 1, characterized in that, The generation methods of the scene parameter package and the synchronization trigger signal include: The target monitoring scene information is acquired and parsed into structured parameter requirements. The preset scene parameter mapping table is retrieved to match the corresponding benchmark parameters, forming a scene parameter package. The initial configuration is performed based on the scene parameter package, and a synchronization trigger signal is generated and distributed to the entire link.

3. The distributed optical fiber synchronous vibration monitoring system according to claim 2, characterized in that, The methods for intercepting candidate target signal segments include: Based on the scene parameter package and the synchronous trigger signal, laser pulses are emitted and backscattered signals from optical fibers are received. The vibration signal corresponding to the target frequency is obtained by directional frequency filtering. The energy value of the vibration signal is collected in real time and compared with the reference parameters. The energy qualified signal is selected by screening. At the same time, the temperature and polarization state data of the corresponding optical fiber position are collected. The temperature and polarization coordinated abrupt change qualified signal is determined by baseline comparison. For vibration signals that meet both criteria, segments are extracted according to preset rules, and after quality verification, they are marked as candidate target signal segments.

4. The distributed optical fiber synchronous vibration monitoring system according to claim 3, characterized in that, The method of encapsulating the event fingerprint and scene marker of the candidate target signal segment into a synchronous spatiotemporal data packet includes: For candidate target signal segments, an event fingerprint is generated by matching the scene parameter package with the actual frequency characteristics of the signal, and at the same time, the scene type identifier is bound as a scene marker. The tagged candidate target signal segments are embedded with synchronization timestamps, fiber optic geoidentifiers, and signal quality information, and then encapsulated into synchronization spatiotemporal data packets.

5. A distributed optical fiber synchronous vibration monitoring system according to claim 4, characterized in that, The methods for generating the suspicious event set include: Based on synchronous spatiotemporal data packets, candidate target signal segments are accessed into a pre-set synchronous data pool, and temperature and polarization state data with the same geographic identifier and time sequence are retrieved to form a three-dimensional data group. Then, high-confidence candidates are marked using a three-level decision rule: First, a dynamic baseline is constructed based on vibration data during event-free periods to verify the vibration amplitude. If the amplitude meets the standard, the event is judged as a primary triggering pass. Then, the coordinated abrupt change in temperature or polarization state is verified. If the amplitude meets the standard, the event is judged as a suspicious event. Finally, the consistency of vibration, temperature and polarization state three modes is verified for suspicious events. If the consistency meets the standard, the event is marked as a high-confidence candidate. Based on high-confidence candidates and combined with edge scene complement rules, a set of suspicious events with quality labels is generated.

6. A distributed optical fiber synchronous vibration monitoring system according to claim 5, characterized in that, The method of obtaining real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generating a list of high-confidence events, includes: The suspicious event set is grouped by geographic grid, and multi-dimensional completion verification is performed on suspicious events at the edge of the group. If they pass the verification, they are upgraded to candidates and included in the group. Duplicate events are eliminated by clustering events within a group based on their spatiotemporal correlation; then, modal differentiation is performed by combining event fingerprints and physical quantity characteristics to filter out real events and determine their types. From the real events categorized into groups, events with high confidence levels and clear types are selected and labeled as standardized event types to form a list of high-confidence events.

7. A distributed optical fiber synchronous vibration monitoring system according to claim 6, characterized in that, The first reporting node broadcasts an inquiry signal containing the event fingerprint through dual communication channels, and neighboring nodes respond by locking the smallest geographic grid segment in the following ways: Select the node closest to the event and with the best signal strength from the distributed monitoring nodes as the first reporting node; The first reporting node generates an inquiry signal containing event fingerprints based on a list of high-confidence events, and broadcasts it to upstream and downstream neighboring nodes in parallel through primary and backup dual communication channels. Each neighboring node retrieves local data for matching and returns valid response information. The first reporting node summarizes the valid response information and filters valid response nodes according to matching degree and response time; then, by combining the geographical coordinates of the first reporting node and valid response nodes, the smallest geographical grid segment where the event is located is located.

8. A distributed optical fiber synchronous vibration monitoring system according to claim 7, characterized in that, The generation methods for the grid-level relative position and location confidence include: Using the locked smallest geographic grid segment as the search range, the built-in static benchmark library and dynamic anchor point library are called to match all surrounding static benchmark points and historical event anchor points in the same scene and environment as two types of benchmark points, thereby obtaining the grid-level relative position of the event relative to the benchmark point. The location reliability of the relative position is obtained by evaluating the confidence level through a confidence level assessment mechanism.

9. A distributed optical fiber synchronous vibration monitoring system according to claim 8, characterized in that, The method of classifying alarm levels and executing differentiated operation and maintenance linkage based on grid-level relative position and fixed position confidence includes: Based on grid-level relative location, different alarm levels are divided according to event type and location reliability, and alarms are pushed through multiple channels. Based on the alarm level, joint maintenance personnel will carry out differentiated maintenance coordination and track the maintenance progress in real time. After the operation and maintenance is completed, submit and verify the handling result form. If the verification is successful, mark the event as closed and notify the operation and maintenance personnel. Archive end-to-end data using event IDs as indexes, and regularly generate event analysis reports to optimize end-to-end parameters, forming a closed loop.

10. A distributed optical fiber synchronous vibration monitoring method, implemented based on the distributed optical fiber synchronous vibration monitoring system described in claims 1-9, characterized in that, include: S1: Obtain the parameter requirements of the target monitoring scene, call the preset scene parameter mapping table to perform scene adaptation configuration, and generate scene parameter package and synchronous trigger signal; S2: Based on the scene parameter package and the synchronous trigger signal, the corresponding frequency vibration signal is screened. By combining the energy threshold with the coordinated abrupt change in temperature and polarization state, candidate target signal segments are extracted. Event fingerprints and scene markers are added to the candidate target signal segments, and they are encapsulated into a synchronous spatiotemporal data package. S3: Based on synchronous spatiotemporal data packets, access the synchronous data pool to bind 3D data, generate a set of suspicious events through three-level judgment and edge scene complementation rules; then obtain real events by decoupling concurrent events through event fingerprint clustering and modality differentiation, and generate a list of high-confidence events; S4: Based on a high-confidence event list, the first reporting node broadcasts an inquiry signal containing the event fingerprint through dual communication channels, and neighboring nodes respond to lock the smallest geographic grid segment; Then, the static benchmark library and dynamic anchor point library are called for calibration to generate grid-level relative position and positional reliability. S5: Based on grid-level relative location and fixed location reliability, classify alarm levels and execute differentiated operation and maintenance linkage; archive full-link data, and regularly optimize full-link parameters to form a closed loop.

Images