Optical cable early warning method and system based on passive inductance integration

Abstract

The invention relates to the technical field of optical cable monitoring, and discloses an optical cable early warning method based on passive sensing integration, and the method comprises the steps: carrying out the passive sensing signal collection of an optical cable disposed in a target region, so as to obtain a disturbance signal transmitted along the optical cable; performing preprocessing operation on the disturbance signal to obtain a normalized disturbance signal sequence; meanwhile, environment data are collected through a multi-parameter sensor network; preprocessing the disturbance signal and the environment data, constructing a multi-dimensional feature vector containing gas concentration sudden change features, temperature distribution abnormity and vibration frequency-amplitude, performing disturbance event identification based on the multi-dimensional feature vector, and determining corresponding disturbance event data; and combining the disturbance event data with the spatial distribution of the optical fiber sensing signal and the vibration wave propagation time difference, positioning the positioning result data corresponding to the occurrence of the abnormal event, generating an early warning instruction based on the positioning result data, and pushing the early warning instruction to the user terminal through the monitoring platform. According to the invention, the accuracy of optical cable safety monitoring can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of optical cable monitoring, and in particular to an optical cable early warning method and system based on passive telepathy integration. Background Art

[0002] At present, the safety monitoring of underground optical cables mainly relies on traditional point sensors (such as single-point vibration sensors and gas sensors) or video surveillance systems, which have some defects. For example, point sensors can only monitor locally and are prone to missing hidden damage along the optical cable; video surveillance relies on visible light imaging, which cannot penetrate soil or complex environments. Existing systems usually monitor a single parameter (such as vibration or gas concentration) in isolation and lack the fusion analysis of multiple parameters (such as vibration frequency, temperature field, and gas leakage). This leads to a high false alarm rate (>15%) in event identification and makes it difficult to distinguish between normal construction and destructive behavior.

[0003] From the above, we can see that how to improve the accuracy of optical cable safety monitoring still needs to be solved. Summary of the Invention

[0004] In order to improve the accuracy of optical cable safety monitoring, the present application provides an optical cable early warning method and system based on passive telepathy integration.

[0005] In the first aspect, the present application provides an optical cable early warning method based on passive synaesthesia integration, which adopts the following technical solutions:

[0006] An optical cable early warning method based on passive synaesthesia integration, comprising:

[0007] Passive sensing signals are collected from optical cables deployed in the target area to obtain disturbance signals propagating along the cables; preprocessing operations are performed on the disturbance signals to obtain a normalized disturbance signal sequence; and environmental data is collected through a multi-parameter sensor network, wherein the environmental data includes gas concentration, temperature, and vibration acceleration;

[0008] Preprocessing the disturbance signal and environmental data to construct a multidimensional feature vector including gas concentration mutation characteristics, temperature distribution anomalies, and vibration frequency-amplitude, and performing disturbance event identification based on the multidimensional feature vector to determine corresponding disturbance event data, where the disturbance event data includes gas disturbance, temperature disturbance, and vibration disturbance;

[0009] The disturbance event data is combined with the spatial distribution of the optical fiber sensing signal and the time difference of vibration wave propagation to locate the positioning result data corresponding to the abnormal event. An early warning instruction is generated based on the positioning result data and pushed to the user terminal through the monitoring platform.

[0010] Optionally, the multi-parameter sensor network includes:

[0011] Non-dispersive infrared technology is used to monitor CO2, CO, and CH4 concentrations, and measurements are made by integrating a MEMS photoelectric sensor chip and an infrared light source. TO-Can metal packaging and an integrated hydrophobic coating lens are used, and nitrogen / vacuum packaging is used to reduce internal water vapor content, improving temperature measurement accuracy in high temperature and high humidity environments. Piezoelectric or capacitive mass-spring structures are used, with a range covering ±2g to ±16g and a sampling frequency ≥1000Hz.

[0012] Optionally, during the disturbance event identification process, the method further includes:

[0013] Monitor the concentrations of CO2, CO, and CH4 in real time, and calculate the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered; perform fast Fourier transform on the original vibration signal to decompose it into different frequency bands; analyze whether there is a significant energy peak in the low-frequency band to determine whether there is excavation activity; analyze whether there is a significant energy peak in the high-frequency band to determine whether there is mechanical shock; and make a comprehensive judgment based on the acceleration amplitude to identify the specific disturbance type; monitor the surface temperature distribution of the equipment in real time, identify local hot spots on the surface of the monitoring equipment, and calculate the temperature change rate corresponding to the local hot spots. If the temperature difference corresponding to the temperature change rate changes by ≥15℃ / min, a fire risk warning is triggered.

[0014] Optionally, the method further comprises:

[0015] Analyze the disturbance signal propagating along the optical cable to determine the arrival time of the vibration wave at different locations. Calculate the arrival time difference of the individual shadows based on the arrival time at different locations. Combine the arrival time difference with the vibration wave propagation speed and calculate the preliminary location result of the abnormal event through geometric relationship.

[0016] After obtaining the preliminary positioning result, a corresponding weight is assigned to each measurement point according to the noise characteristics of the sensor, and the preliminary positioning result is optimized based on the weight optimized using the weighted least squares method to obtain the corresponding positioning result data.

[0017] Optionally, the method further comprises:

[0018] Build a three-dimensional virtual model of the underground space, import data from various sensors in real time, overlay gas concentration, temperature field, and vibration distribution data to form a dynamic visualization interface, and annotate the dynamic visualization interface on a GIS map to display the location, type, and time of abnormal events in real time;

[0019] When excessive gas concentration or abnormal temperature is detected, the regional fan is automatically started and the valve is closed. The video surveillance system is linked to capture the on-site image and upload it to the monitoring platform.

[0020] Optionally, the intelligent analysis platform further includes:

[0021] Regularly collect status data of each monitoring device and store it in the database according to categories;

[0022] Provide rich chart display functions based on the status data, customize query results by selecting time period, device type and other conditions, and export the query results.

[0023] In a second aspect, the present application provides an optical cable early warning system based on passive synaesthesia integration, which adopts the following technical solutions:

[0024] An optical cable early warning system based on passive synaesthesia integration, comprising:

[0025] The data acquisition module performs passive sensing signal acquisition on the optical cables deployed in the target area to obtain disturbance signals propagating along the optical cables; pre-processes the disturbance signals to obtain a normalized disturbance signal sequence; and simultaneously collects environmental data through a multi-parameter sensor network, wherein the environmental data includes gas concentration, temperature, and vibration acceleration;

[0026] a disturbance event data determination module, which preprocesses the disturbance signal and environmental data and constructs a multidimensional feature vector including gas concentration mutation characteristics, temperature distribution anomalies, and vibration frequency-amplitude, and performs disturbance event identification based on the multidimensional feature vector to determine corresponding disturbance event data, where the disturbance event data includes gas disturbance, temperature disturbance, and vibration disturbance;

[0027] The positioning result data positioning module combines the disturbance event data with the spatial distribution of the optical fiber sensing signal and the time difference of the vibration wave propagation to locate the positioning result data corresponding to the abnormal event, generates an early warning instruction based on the positioning result data, and pushes it to the user terminal through the monitoring platform.

[0028] Optionally, it also includes:

[0029] The concentration monitoring module is used to monitor the concentrations of CO2, CO, and CH4 in real time and calculate the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered;

[0030] The judgment module performs a fast Fourier transform on the original vibration signal to decompose it into different frequency bands. It analyzes whether there are significant energy peaks in the low-frequency band to determine whether there is mining activity; it analyzes whether there are significant energy peaks in the high-frequency band to determine whether there is mechanical impact. It also makes a comprehensive judgment based on the acceleration amplitude to identify the specific disturbance type.

[0031] Temperature change rate calculation module. This module monitors the surface temperature distribution of the equipment in real time, identifies local hot spots on the surface, and calculates the corresponding temperature change rate. If the corresponding temperature difference sudden change is ≥15°C / min, a fire risk warning is triggered.

[0032] In a third aspect, the present application provides an optical cable early warning system based on passive synaesthesia integration, which adopts the following technical solutions:

[0033] An optical cable early warning system based on passive synaesthesia integration includes a processor running a program of any one of the above-mentioned optical cable early warning methods based on passive synaesthesia integration.

[0034] In a fourth aspect, the present application provides a storage medium, which adopts the following technical solution:

[0035] A storage medium stores a program of any one of the above-mentioned optical cable early warning methods based on passive synaesthesia integration.

[0036] In summary, this application includes at least one of the following beneficial technical effects:

[0037] First, multi-dimensional data such as sudden changes in gas concentration, abnormal temperature distribution, and vibration frequency-amplitude are collected in real time using NDIR gas sensors (CO2, CO, CH4), thermopile array temperature sensors, and MEMS vibration sensors to construct a multi-dimensional feature vector. Second, a machine learning algorithm is used to intelligently analyze the feature vector. For example, the FFT decomposition of the low-frequency (0.5-20Hz) and high-frequency (20-100Hz) bands of the vibration signal is used, and the acceleration amplitude (≥0.5g) is combined to distinguish between disturbance types such as excavation and mechanical impact. The sliding window mutation slope (≥5% vol / min) and temperature difference mutation (≥15°C / min) are used to accurately identify gas leaks or fire risks. This multi-parameter collaborative analysis significantly reduces the false alarm rate of traditional methods (such as distinguishing normal construction from sabotage) and improves the accuracy of event type identification through feature fusion, enabling optical cable safety monitoring to shift from "passive response" to "active and precise early warning."

[0038] By calculating the arrival time difference of the vibration wave at different positions of the optical cable and combining it with the wave speed (340m / s) to infer the event position, and then correcting the noise interference through the weighted least squares method, the positioning result is further optimized. At the same time, a three-dimensional model of the underground space is constructed based on digital twin technology, gas, temperature, and vibration data are superimposed in real time, and the spatiotemporal distribution of abnormal events is visualized through GIS maps. In addition, the system supports linkage control (such as automatically starting fans and closing valves when gas concentration exceeds the standard) and video capture, forming a "perception-analysis-positioning-response" closed loop. This precise positioning and real-time linkage mechanism solves the problems of large positioning error (>±50 meters) and response delay (several hours) in traditional methods, enabling optical cable safety monitoring to have "point-to-point" early warning capabilities and second-level emergency response efficiency, significantly improving the intelligence level and safety of optical cable operation and maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 The present invention is a flowchart showing an optical cable early warning method based on passive synaesthesia integration according to an exemplary embodiment.

[0040] Figure 2 The present invention is a structural block diagram of an optical cable early warning method device based on passive synaesthesia integration according to an exemplary embodiment. DETAILED DESCRIPTION

[0041] Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings.

[0042] Throughout this specification, reference to the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0043] The present application embodiment discloses an optical cable early warning method based on passive synaesthesia integration, referring to Figure 1 ,include:

[0044] S100, passive sensing signal collection is performed on the optical cables laid in the target area to obtain the disturbance signal propagating along the optical cables; the disturbance signal is preprocessed to obtain a normalized disturbance signal sequence; and environmental data is collected through a multi-parameter sensor network. The environmental data includes gas concentration, temperature and vibration acceleration.

[0045] First, passive optical cable sensing signal collection is carried out. The specific implementation process includes: laying the optical cable in the target area (such as underground pipelines, cable trenches, etc.) so that it is in direct contact with the surrounding environment, and sensing external disturbances (such as excavation, settlement, vibration) through the physical deformation of the optical cable itself (such as strain or micro-displacement).

[0046] It should be noted here that optical cables, as distributed sensors, use distributed fiber optic sensing technology (such as optical time-domain reflectometry, or OTDR) to monitor disturbance signals propagating along the cables in real time. The disturbance signal originates from external events (such as mechanical excavation or subsidence) that cause deformation of the optical cable. The optical signal (such as phase or intensity) in the cable changes due to Rayleigh or Brillouin scattering, and these changes are converted into electrical signals. Key parameters include a sampling frequency of ≥1000Hz to ensure the capture of high-frequency vibration details, and signal coverage that can reach the entire length of the optical cable (e.g., 10-50 kilometers), enabling full-area sensing.

[0047] Based on the above steps, there is no need to deploy additional sensors. The optical cable itself can cover a large area, eliminating the blind spots of traditional point sensors. It only relies on the physical changes of the optical signal, reducing energy consumption and maintenance costs.

[0048] Then the data collection and processing of the multi-parameter sensor network is carried out. The specific execution process includes:

[0049] 1. NDIR gas sensor (CO2, CO, CH4 concentration monitoring), the hardware components include:

[0050] The infrared light source uses a MEMS photoelectric sensor chip and an infrared light source (such as a tungsten lamp or LED) to emit broadband infrared light. The gas chamber and filter selectively transmit infrared light of a specific wavelength (such as 3.4μm for CH4 and 4.26μm for CO2) after the gas to be measured enters the gas chamber. The pyroelectric detector receives the remaining light intensity after gas absorption and converts it into an electrical signal.

[0051] Workflow: When infrared light passes through the gas chamber, the target gas (such as CO2) absorbs infrared light of a specific wavelength, causing the light intensity to attenuate. The detector measures this change in light intensity and calculates the concentration using the Lambert-Beer law (formula: A = ε·c·l, where A is absorbance, ε is the molar absorptivity, c is the concentration, and l is the optical path length). Dual-wavelength calibration uses two filters (target wavelength and reference wavelength) to eliminate background interference by comparing the transmittance difference, improving accuracy (for example, CO concentration detection accuracy is ±30ppm). Using specific wavelength filters prevents cross-interference (for example, distinguishing the absorption peaks of CO and CO2).

[0052] 2. Thermopile temperature sensor (temperature monitoring), the hardware components include:

[0053] The TO-Can package structure uses a metal base and nickel-plated anti-oxidation material, with an integrated thermopile chip and memory chip. The lens is coated with a nano-hydrophobic coating (contact angle ≥ 150°) to suppress measurement errors caused by condensation. Nitrogen / vacuum packaging is evacuated or filled with nitrogen before packaging to reduce internal water vapor content (humidity ≤ 0.1%).

[0054] Working process: The thermopile chip absorbs infrared radiation from the target area and converts temperature changes into a voltage signal. The voltage value is converted into a temperature value with an accuracy of ±2°C using calibration data (stored on the chip). It should be noted here that the metal package and hydrophobic coating ensure stable operation in a high-humidity environment of 95% RH.

[0055] TO-Can packaging reduces environmental interference, and the hydrophobic coating avoids the effects of condensation, making it suitable for high-temperature and high-humidity underground spaces. By monitoring local hot spots (such as ≥100°C) or sudden temperature changes (≥15°C / min), it provides temperature-dimensional evidence for abnormal events.

[0056] 3.MEMS vibration sensor (vibration acceleration monitoring), the hardware components include: Hardware components:

[0057] Piezoelectric or capacitive structure. In the piezoelectric type, when the mass block is vibrated, the piezoelectric ceramic generates an electric charge (such as quartz or lead zirconate titanate), which is suitable for high-frequency vibration (20-100Hz); in the capacitive type, the displacement of the mass block changes the capacitance pole distance, and high sensitivity (such as ±0.5g range) is achieved through MEMS micromachining technology.

[0058] Measuring range and sampling frequency: The measuring range covers ±2g to ±16g, supporting low-frequency (0.5-20Hz, excavation activities) and high-frequency (mechanical shock) vibration analysis; the sampling frequency is ≥1000Hz to ensure the capture of vibration details (such as the low-frequency vibration characteristics of mechanical excavation).

[0059] Workflow: Vibration causes the mass to displace, and the piezoelectric sensor outputs a charge or the capacitive sensor outputs a capacitance change; the signal is amplified, filtered, and converted into an acceleration value (such as g); the data is transmitted in real time to the ACU (area control unit) via the RS-485 interface.

[0060] Through the above steps, low-frequency (excavation) and high-frequency (mechanical impact) vibrations can be distinguished, and the disturbance type can be accurately classified in combination with the acceleration amplitude (≥0.5g). The piezoelectric type is suitable for high frequency and high acceleration, and the capacitive type is suitable for low-frequency micro-vibration, covering complex disturbance scenarios.

[0061] Furthermore, data preprocessing and normalization are performed. The specific implementation process includes:

[0062] For disturbance signal preprocessing, corresponding noise reduction is required, using wavelet denoising or Kalman filtering to eliminate environmental noise (such as electromagnetic interference and temperature drift); in standardization processing, the optical cable signal is normalized to a unified dimension (such as voltage or dB unit) to eliminate sensor differences.

[0063] For environmental data synchronization and calibration: it is necessary to ensure timestamp alignment and time consistency of gas, temperature, and vibration data (such as through GPS clock synchronization); and perform corresponding drift compensation, perform zero-point calibration on the temperature sensor to eliminate long-term drift; and perform cross-sensitivity correction on the gas sensor.

[0064] Based on the above steps, data quality can be guaranteed, noise and drift can be eliminated, and a reliable foundation can be provided for subsequent feature extraction; and the data format is unified to support the construction of multi-dimensional feature vectors (such as vibration frequency, gas mutation slope, and temperature distribution).

[0065] S200, preprocessing the disturbance signal and environmental data and constructing a multidimensional feature vector including gas concentration mutation characteristics, temperature distribution anomalies, and vibration frequency-amplitude, and performing disturbance event identification based on the multidimensional feature vector to determine the corresponding disturbance event data, which includes gas disturbance, temperature disturbance, and vibration disturbance.

[0066] Among them, in the process of disturbance event identification and multi-dimensional feature vector construction, the overall execution process is:

[0067] First, data preprocessing is performed to process the disturbance signal, reduce noise (such as wavelet denoising) and perform baseline correction on the vibration signal collected by the optical cable to eliminate environmental noise interference (such as electromagnetic interference and temperature drift). Then, environmental data synchronization is performed to align the timestamps of gas concentration, temperature, and vibration data to ensure the spatiotemporal consistency of multi-parameter data; and standardization is performed to convert data of different dimensions (such as gas concentration %vol, temperature ℃, acceleration g) into a unified standardized value to facilitate subsequent analysis.

[0068] Then comes the construction of a multi-dimensional feature vector, which extracts parameters such as the concentration change rate and cumulative increment within the sliding window to form gas dimensional characteristics. This requires analyzing the spatial distribution of the equipment surface temperature and identifying local hot spots and sudden changes in temperature differences. At the same time, the low-frequency (0.5-20Hz) and high-frequency (20-100Hz) frequency band energies of the vibration signal are decomposed through FFT, and combined with the acceleration amplitude to form vibration dimensional characteristics.

[0069] Furthermore, disturbance events can be identified: a classification model can be applied, multi-dimensional features can be input into a trained machine learning model (such as SVM or random forest), and the event type (such as gas leakage, mechanical excavation, fire) can be output. Here, it is necessary to determine the threshold judgment rules and set the threshold based on domain knowledge (such as a CO concentration mutation slope ≥ 5% vol / min triggers an early warning) to assist the model classification results.

[0070] By integrating gas, temperature, and vibration data, false alarms from a single sensor can be avoided (for example, distinguishing normal construction from destructive behavior); through feature engineering and machine learning, complex disturbance types can be identified (such as the vibration differences between excavation and settlement).

[0071] During the disturbance event identification process, the method further includes:

[0072] S210, real-time monitoring of CO2, CO, and CH4 concentrations, and calculation of the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered.

[0073] Among them, real-time monitoring and sliding window calculation are first performed: the NDIR sensor outputs CO2, CO, and CH4 concentration data in real time, with a sampling frequency of ≥1Hz; the data is divided into continuous sliding windows (such as a window of 1 minute, moving forward 1 second per second); the concentration change rate within each window is calculated (such as the current concentration minus the initial concentration, divided by the time difference).

[0074] Then, threshold judgment and warning triggering are performed. It can be a mutation slope judgment. If the concentration change rate in the window is ≥5% vol / min (such as the CO concentration rises by 5% within 1 minute), it is judged as an abnormal leak. It can also be a cumulative amount judgment. The cumulative increase of the concentration in the window is calculated. If it exceeds 120% of the set threshold (such as the set threshold is 100ppm, the cumulative amount exceeds 120ppm), an early warning is triggered.

[0075] Then, linkage control is carried out: a gas leak alarm is sent to the monitoring platform, the type of leaked gas is marked, the relevant valves are closed, the fan is started (wind speed ≥ 5m / s) and the video surveillance is linked to capture the on-site image.

[0076] Through the mutation slope and cumulative amount threshold, leakage risks (such as pipeline rupture or valve failure) can be quickly identified; by combining sliding windows and multi-threshold judgments, normal fluctuations and abnormal leakage (such as natural CO2 concentration changes and man-made leakage) can be distinguished.

[0077] S220 performs a fast Fourier transform on the original vibration signal to decompose it into different frequency bands; analyzes whether there are significant energy peaks in the low-frequency band to determine whether there is mining activity; analyzes whether there are significant energy peaks in the high-frequency band to determine whether there is mechanical impact; and makes a comprehensive judgment based on the acceleration amplitude to identify the specific disturbance type.

[0078] Among them, vibration signal preprocessing is first performed: wavelet denoising is performed on the original vibration acceleration signal (sampling rate ≥ 1000Hz) to eliminate high-frequency noise interference; then the signal is divided into frames of fixed length (such as 1 second / frame) to ensure the time domain locality of spectrum analysis.

[0079] Then, fast Fourier transform (FFT) decomposition is performed to divide the corresponding frequency bands: the low-frequency band (0.5-20Hz) corresponds to low-frequency vibrations such as mechanical excavation and settlement; the high-frequency band (20-100Hz) corresponds to high-frequency vibrations such as mechanical impact and equipment failure; energy distribution analysis is required, and FFT is performed on each frame signal to calculate the energy distribution of each frequency band.

[0080] Further energy peak determination and comprehensive determination:

[0081] For low-frequency band judgment, if the low-frequency band energy peak is significantly higher than the background noise (such as a sudden increase in low-frequency vibration energy in excavation activities), it is judged to be an excavation activity; for high-frequency band judgment, if the high-frequency band energy peak is significantly higher than the background noise (such as a sudden increase in high-frequency vibration energy in mechanical impact), it is judged to be a mechanical impact; for acceleration amplitude judgment, the acceleration amplitude (such as ≥0.5g) is combined to further verify the event type (such as low-frequency vibration in excavation is usually accompanied by a smaller acceleration, while the high-frequency vibration acceleration of mechanical impact is larger).

[0082] Finally, the event type is output: based on the above analysis results, the specific disturbance type (such as "mechanical excavation" or "mechanical impact") is output.

[0083] Through frequency band energy analysis and acceleration amplitude, different disturbance types can be accurately identified (such as distinguishing normal construction vibration from destructive behavior). Combining frequency domain and time domain characteristics can avoid misjudgment of a single indicator (such as distinguishing high-frequency vibration caused by strong winds from mechanical shock).

[0084] S230 monitors the surface temperature distribution of the equipment in real time, identifies local hot spots on the surface of the monitoring equipment, and calculates the temperature change rate corresponding to the local hot spots. If the temperature difference corresponding to the temperature change rate changes by ≥15°C / min, a fire risk warning is triggered.

[0085] S300: Combine the disturbance event data with the spatial distribution of the optical fiber sensor signal and the propagation time difference of the vibration wave to locate the positioning result data corresponding to the abnormal event. Generate an early warning instruction based on the positioning result data and push it to the user terminal through the monitoring platform.

[0086] Among them, vibration wave time difference analysis and positioning calculation are performed first. The specific execution process includes:

[0087] Vibration wave arrival time recording: Sensor network deployment, fiber optic sensor nodes are placed at regular intervals (e.g., 10 meters) in the target area (e.g., along the optical cable) to form a spatial monitoring network; timestamp recording, each node records the arrival time of the disturbance signal (e.g., vibration wave) in real time, accurate to the millisecond level. For example, node A detects vibration at 10:00:00, and node B detects the same vibration at 10:00:01.

[0088] Time difference calculation and propagation speed analysis: By comparing the arrival times recorded by adjacent nodes, the time difference of vibration wave propagation is calculated. For example, node B detects the vibration 1 second later than node A, and the time difference is 1 second. Based on the known propagation speed of vibration waves in the medium (such as 340 m / s in optical cables), the distance of vibration wave propagation is calculated in combination with the time difference. For example, if the time difference is 1 second, the propagation distance is 340 m / s × 1 second = 340 meters.

[0089] Positioning model construction: Hyperbolic positioning principle, assuming that an event occurs at a certain point between two nodes, the distance difference between the point and the two nodes is determined based on the time difference and propagation speed, forming a hyperbolic trajectory; multi-node joint positioning, through the time difference data of at least three nodes, constructs the intersection of multiple hyperbolic trajectories, and ultimately determines the specific coordinates of the event (such as longitude and latitude or optical cable mileage).

[0090] Through time difference analysis and hyperbola model, the area where the event occurred can be quickly determined, with the error controlled within ±10 meters. By utilizing the distribution characteristics of fiber optic sensing nodes, global positioning can be achieved, avoiding the blind spot problem of traditional point sensors.

[0091] Then, the positioning results are optimized and the accuracy is improved. The specific implementation process includes:

[0092] 1. Sensor noise modeling:

[0093] Noise source analysis identifies signal interference sources at each node (such as signal attenuation at optical cable connections and electromagnetic interference). By analyzing historical data to calculate noise levels, for example, a node may experience large time difference measurement errors due to environmental interference. By quantifying noise levels, we ensure that high-precision nodes contribute more to positioning results and reduce interference from noisy nodes. Once noise sources are identified, we can optimize sensor deployment or environmental shielding (for example, adding a shielding layer to reduce electromagnetic interference).

[0094] 2. Preliminary Positioning Calculation: Record the arrival time of the vibration wave. Deploy multiple sensor nodes along the optical cable (e.g., one node every 10 meters) and record the arrival time of the disturbance signal in real time. For example, if node 1 detects a vibration wave at 10:00:00, and node 2 detects the same vibration wave at 10:00:01, time difference and propagation speed analysis: Calculate the time difference by comparing the time differences between adjacent nodes. For example, if node 2 detects the vibration 1 second later than node 1, the time difference is 1 second. Determine the propagation speed. Based on the known propagation speed of the vibration wave in the optical cable (e.g., 340 meters / second), calculate the propagation distance. For example, a time difference of 1 second corresponds to a propagation distance of 340 meters.

[0095] Geometric positioning: Hyperbolic positioning assumes an event occurs at a point between two nodes. Based on the time difference and propagation speed, the distance difference between that point and the two nodes is determined, forming a hyperbolic trajectory. Multi-node joint positioning uses time difference data from at least three nodes to construct the intersection of multiple hyperbolic trajectories to preliminarily determine the event coordinates (for example, at the 120th meter of the optical cable). The initial positioning error range is ±10 meters. Time difference and geometric relationships are used to quickly narrow the event range, providing initial coordinates for subsequent optimization. Furthermore, the sensor network along the optical cable is utilized to avoid the positioning blind spots of traditional point sensors.

[0096] 3. Weighted optimization algorithm improves accuracy:

[0097] Weight matrix construction: Weight allocation: assign weights to each node based on the noise modeling results (e.g., node A has a weight of 0.8, node B has a weight of 0.2); weight coefficients are added to construct a weight matrix, where high-weight nodes have a greater impact on the final result.

[0098] Optimization model establishment: The objective function is to minimize the weighted residual and build an optimization model. For example, the contribution of each node data is adjusted in a weighted manner to minimize the error.

[0099] Iterative correction and error reduction: Lagrangian algorithm application: Combined with constraints (such as distance difference), weights and positioning coordinates are iteratively adjusted. Newton method iterations gradually correct parameters and reduce errors. For example: initial error ±10 meters → ±8 meters after the first iteration → ±5 meters after the second iteration.

[0100] Final positioning result: After multiple iterations, the positioning error can be reduced to within ±5 meters, and the specific location of the event is determined (for example, within ±5 meters of the 120th meter of the optical cable).

[0101] Through weight distribution and iterative optimization, the influence of noise nodes on positioning results is significantly reduced.

[0102] 4. Application of positioning results and early warning push:

[0103] Data integration and verification: Multi-source data fusion, combining the spatial distribution data of fiber optic sensors, the time difference of vibration wave propagation and weighted optimization results, generates the final positioning coordinates; cross-validation, using data from other sensors (such as temperature and gas concentration) to assist in verifying the positioning results. For example, if the temperature near the positioning point rises abnormally, it is further confirmed that the incident is a fire.

[0104] Linked control and early warning: Early warning push, the positioning coordinates (such as "120 meters of the optical cable") and event type (such as "gas leakage") are pushed to the user terminal through the monitoring platform; automatic control, triggering emergency operations such as valve closing and fan starting to reduce the impact of accidents.

[0105] Through high-precision positioning, the fault point can be quickly located and personnel can be guided to handle it accurately; and by combining noise modeling and optimization algorithms, false alarms and missed alarms can be reduced, thereby improving the overall performance of the system.

[0106] In the embodiment of the present application, S100 to S300 are the core steps of the entire process of the optical cable safety monitoring system, among which S100 is responsible for collecting and preprocessing multi-source sensor data (such as optical cable vibration, gas concentration, temperature, etc.), and laying the foundation for subsequent analysis through noise reduction, synchronization and standardization; S200 accurately identifies the type of disturbance event (such as leakage, excavation, fire) through multi-dimensional feature extraction (such as gas mutation slope, vibration frequency band energy, temperature anomaly distribution) combined with machine learning models and threshold rules; S300 uses the time difference of vibration wave propagation and the spatial distribution of optical fiber sensing to achieve sub-meter positioning through noise modeling and weighted optimization algorithm, and finally generates early warning instructions and pushes them to the user terminal in real time through the monitoring platform, linking the control equipment (such as closing valves, starting sprinklers), forming a closed loop of "perception-analysis-positioning-response", significantly improving the early warning, precise disposal and emergency response efficiency of safety hazards.

[0107] In an embodiment of the present application, the method further includes:

[0108] The first step is to build a three-dimensional virtual model of the underground space, import data from various sensors in real time, overlay gas concentration, temperature field and vibration distribution data to form a dynamic visualization interface, and mark it on the GIS map of the dynamic visualization interface to display the location, type and time of abnormal events in real time.

[0109] Among them, the three-dimensional model is constructed first:

[0110] Static data such as the geological structure of the underground space, optical cable laying paths, valve positions, and fan distribution are imported into the system, and a three-dimensional virtual model is generated using BIM (Building Information Modeling) or GIS (Geographic Information System) technology. For example, optical cable paths are marked with different colors, and valve and fan positions are marked with icons. The location information of each sensor (such as the coordinates of optical cable nodes, temperature sensors, and gas sensors) is associated in the model to ensure that the data corresponds one-to-one with the physical equipment.

[0111] Then perform real-time data overlay and visualization:

[0112] Receive sensor data (such as gas concentration, temperature, and vibration intensity) in real time and map the data to the corresponding position in the 3D model through the API interface. For example, if a node detects a CH4 concentration of 3% vol, a red warning will be displayed at that location in the model.

[0113] Gas concentration thermodynamic map, using color depth to indicate concentration (e.g., red for high concentration, green for normal); temperature field distribution map, superimposing temperature data to highlight local hot spots (e.g., where the temperature in a certain area reaches 120°C); vibration intensity distribution, showing the propagation path of vibration waves through different colors or ripple effects.

[0114] GIS map annotation: On the bird's-eye view GIS map of the 3D model, icons are used to mark the location of abnormal events in real time (such as "gas leakage at the 200th meter of the optical cable"), and the event type (such as "CO concentration exceeds the standard") and occurrence time (such as "2025-04-25 11:00:09") are displayed.

[0115] User interaction and alarm prompts: The system refreshes data every second to ensure that the visual interface is synchronized with the real-time status; when an anomaly is detected, the corresponding position in the model flashes red and is accompanied by an alarm sound (such as a beep). Users can click the anomaly icon to view detailed data (such as concentration curve, vibration waveform).

[0116] In the second step, when the gas concentration exceeds the standard or the temperature is detected to be abnormal, the regional fan is automatically started and the valve is closed, and the video surveillance system is linked to capture the on-site image and upload it to the monitoring platform.

[0117] Among them, abnormal event triggering condition judgment: the system monitors sensor data in real time, and automatically triggers the linkage mechanism when the gas concentration exceeds the threshold (such as CO ≥ 2% vol) or the temperature is abnormal (such as local temperature ≥ 100°C and temperature difference mutation ≥ 10°C / min).

[0118] Device control command issuance: The system generates a control command (such as "start regional fan A, close valve B") and sends the command to the on-site device controller through protocols such as Modbus and MQTT. For example, after receiving the command, the fan runs at a wind speed of 5m / s and the valve closes. After execution, the device returns status information (such as "fan started, valve closed") to ensure a closed-loop command.

[0119] Video surveillance linkage: The system calls the preset camera position (such as the camera closest to the anomaly point) and automatically switches to the camera screen; triggers the camera to capture three on-site images (such as a panoramic view of the leak point and a local close-up) and record a 1-minute video clip; the captured images and videos are uploaded to the monitoring platform via the network, marked with time, location and event type for subsequent analysis.

[0120] Monitoring platform display and notification: An alarm window pops up on the main interface of the monitoring platform, displaying the event type, location, time, and associated images. For example, the title is: "Gas leak alarm (200 meters of optical cable)", and the details are: "CO concentration 3% vol, fan A has been started, and valve B has been closed." Multi-terminal push: Alarm summaries are sent to administrators via SMS, APP push, or email to ensure timely response.

[0121] By constructing a three-dimensional virtual model of the underground space and overlaying real-time data on the GIS map, the precise positioning and dynamic visualization of abnormal events are achieved, allowing operation and maintenance personnel to intuitively view the location, type and parameters of the event; the linkage control function automatically triggers fan startup, valve closure and video capture, shortening the manual intervention time to within 10 seconds, effectively curbing the expansion of accidents; at the same time, captured images and videos provide visual evidence for post-event analysis; multiple systems work together to form a "perception-analysis-response-recording" closed loop, combined with knowledge base technology to optimize alarm strategies and anomaly detection, significantly improving the ability to prevent and control safety hazards, and realizing three-dimensional perception, intelligent decision-making and automated disposal.

[0122] In the embodiment of the present application, the intelligent analysis platform further includes:

[0123] The first step is to collect the status data of each monitoring device regularly and store it in the database according to categories.

[0124] Among them, through Modbus, MQTT or HTTP protocols, status data is collected from each monitoring device (such as gas sensors, temperature sensors, and optical cable vibration monitoring nodes) at regular intervals (such as every 5 minutes), including equipment operating status (such as online / offline), parameter values (such as temperature, vibration intensity), fault codes, etc.; data collection is triggered using scheduled task tools (such as Linux's crontab or Kubernetes CronJob) or message queues (such as RabbitMQ) to ensure the periodicity and reliability of data collection.

[0125] Then, outliers (such as temperature data that suddenly changes beyond the physical range) or missing values are removed, and metadata such as timestamps and device IDs are added. Data is stored in different database tables or partitions by sensor type (such as gas sensor and temperature sensor). For example, gas concentration data is stored in the "Gas_Sensor" table, and vibration data is stored in the "Vibration_Node" table. Data is stored in different database tables or partitions by sensor type (such as gas sensor and temperature sensor). For example, gas concentration data is stored in the "Gas_Sensor" table, and vibration data is stored in the "Vibration_Node" table.

[0126] The collection process is monitored in real time. If the device is found to be offline or the data is interrupted (for example, no data is collected for three consecutive times), an alarm is triggered and recorded in the log system. After storage is completed, CRC verification or data volume statistics are used to ensure data integrity.

[0127] The second step is to provide rich chart display functions based on status data, customize query results by selecting time period, device type and other conditions, and export query results.

[0128] First, configure data visualization and select chart types: a variety of chart templates are provided for users to choose from, such as:

[0129] Line charts show the changing trends of equipment parameters over time (e.g., temperature fluctuations within 24 hours); bar charts / pie charts show the distribution of equipment types or the percentage of fault types; and heat maps show the density of equipment status in a specific area of the underground space (e.g., the distribution of sensors in high-risk areas). Users can drag the timeline to select a time period (e.g., "Last Hour" or "2023-01-01 to 2023-01-07") or filter equipment types (e.g., "Fiber Optic Cable Vibration Node") using the drop-down menu.

[0130] Then, query and result generation: condition combination query, users can set multiple condition combination query, for example:

[0131] Time condition: select "Last 7 days"; Device condition: filter "Temperature sensor" and "Status is offline"; Parameter condition: set "Temperature ≥ 80℃".

[0132] Query execution: The system quickly retrieves data through SQL or NoSQL query statements (such as MongoDB's aggregation pipeline) and returns the results to the front end.

[0133] Finally, export and share results: Multiple export formats are available, such as Excel spreadsheets, PDF reports, CSV files, or JSON data packages. Query results are converted to a specified format (e.g., Excel spreadsheets must include headers, data columns, and units). Once the file is generated, users can download it via a link or share it directly to their email or cloud drive.

[0134] The intelligent analysis platform regularly collects and categorizes the status data of each monitoring device to ensure data integrity and traceability. It also provides rich charting capabilities, supports custom queries, and exports results. This allows operations personnel to intuitively monitor equipment operating trends, promptly identify potential faults, and make informed decisions and optimize systems based on detailed data analysis. Furthermore, data quality monitoring and redundant backups enhance system reliability, while flexible data sharing and export capabilities meet the needs of audits and cross-departmental collaboration, ultimately improving the overall operational efficiency and management level of underground space safety monitoring.

[0135] The present application embodiment discloses an optical cable early warning system based on passive synaesthesia integration, referring to Figure 2 ,include:

[0136] Data acquisition module 001 performs passive sensing signal acquisition on the optical cables deployed in the target area to obtain disturbance signals propagating along the optical cables; performs preprocessing operations on the disturbance signals to obtain a normalized disturbance signal sequence; and simultaneously collects environmental data through a multi-parameter sensor network, where the environmental data includes gas concentration, temperature, and vibration acceleration;

[0137] The disturbance event data determination module 002 pre-processes the disturbance signal and environmental data and constructs a multi-dimensional feature vector containing gas concentration mutation characteristics, temperature distribution anomalies and vibration frequency-amplitude, and performs disturbance event identification based on the multi-dimensional feature vector to determine the corresponding disturbance event data. The disturbance event data includes gas disturbance, temperature disturbance and vibration disturbance.

[0138] The positioning result data positioning module 003 combines the disturbance event data with the spatial distribution of the optical fiber sensing signal and the time difference of the vibration wave propagation to locate the positioning result data corresponding to the abnormal event, generates an early warning instruction based on the positioning result data, and pushes it to the user terminal through the monitoring platform.

[0139] The optical cable early warning system based on passive synaesthesia integration in this application also includes:

[0140] The concentration monitoring module is used to monitor the concentrations of CO2, CO, and CH4 in real time and calculate the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered;

[0141] The judgment module performs a fast Fourier transform on the original vibration signal to decompose it into different frequency bands. It analyzes whether there are significant energy peaks in the low-frequency band to determine whether there is mining activity; it analyzes whether there are significant energy peaks in the high-frequency band to determine whether there is mechanical impact. It also makes a comprehensive judgment based on the acceleration amplitude to identify the specific disturbance type.

[0142] Temperature change rate calculation module. This module monitors the surface temperature distribution of the equipment in real time, identifies local hot spots on the surface, and calculates the corresponding temperature change rate. If the corresponding temperature difference sudden change is ≥15°C / min, a fire risk warning is triggered.

[0143] An embodiment of the present application further discloses an optical cable early warning system based on passive synaesthesia integration, comprising a processor running a program of any one of the above-mentioned optical cable early warning methods based on passive synaesthesia integration.

[0144] The embodiment of the present application further discloses a storage medium storing a program of any one of the above-mentioned optical cable early warning methods based on passive synaesthesia integration.

[0145] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. An optical cable early warning method based on passive synaesthesia integration, characterized in that: include: Passive sensing signals are collected from optical cables deployed in the target area to obtain disturbance signals propagating along the cables; preprocessing operations are performed on the disturbance signals to obtain a normalized disturbance signal sequence; and environmental data is collected through a multi-parameter sensor network, wherein the environmental data includes gas concentration, temperature, and vibration acceleration; Preprocessing the disturbance signal and environmental data to construct a multidimensional feature vector including gas concentration mutation characteristics, temperature distribution anomalies, and vibration frequency-amplitude, and performing disturbance event identification based on the multidimensional feature vector to determine corresponding disturbance event data, where the disturbance event data includes gas disturbance, temperature disturbance, and vibration disturbance; The disturbance event data is combined with the spatial distribution of the optical fiber sensing signal and the time difference of vibration wave propagation to locate the positioning result data corresponding to the abnormal event. An early warning instruction is generated based on the positioning result data and pushed to the user terminal through the monitoring platform.

2. The optical cable early warning method based on passive synaesthesia integration according to claim 1 is characterized in that: The multi-parameter sensor network comprises: Use non-dispersive infrared technology to monitor CO2, CO, and CH4 concentrations by integrating MEMS photoelectric sensor chips with infrared light sources; The TO-Can metal package is integrated with a hydrophobic coating lens, and nitrogen / vacuum packaging is used to reduce internal water vapor content, thereby improving temperature measurement accuracy in high temperature and high humidity environments. It adopts piezoelectric or capacitive mass-spring structure, with a measuring range of ±2g to ±16g and a sampling frequency of ≥1000Hz.

3. The optical cable early warning method based on passive synaesthesia integration according to claim 1 is characterized in that: During the disturbance event identification process, the method further includes: Real-time monitoring of CO2, CO, and CH4 concentrations, and calculation of the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered; Perform a fast Fourier transform on the original vibration signal to decompose it into different frequency bands. Analyze whether there are significant energy peaks in the low-frequency band to determine whether there is mining activity; analyze whether there are significant energy peaks in the high-frequency band to determine whether there is mechanical impact. Combined with the acceleration amplitude, a comprehensive judgment is made to identify the specific disturbance type. Monitor the surface temperature distribution of the equipment in real time, identify local hot spots on the surface of the monitoring equipment, and calculate the temperature change rate corresponding to the local hot spots. If the temperature difference corresponding to the temperature change rate changes by ≥15℃ / min, a fire risk warning will be triggered.

4. The optical cable early warning method based on passive synaesthesia integration according to claim 1 is characterized in that: The method also includes: Analyze the disturbance signal propagating along the optical cable to determine the arrival time of the vibration wave at different locations. Calculate the arrival time difference of the individual shadows based on the arrival time at different locations. Combine the arrival time difference with the vibration wave propagation speed and calculate the preliminary location result of the abnormal event through geometric relationship. After obtaining the preliminary positioning result, a corresponding weight is assigned to each measurement point according to the noise characteristics of the sensor, and the preliminary positioning result is optimized based on the weight optimized using the weighted least squares method to obtain the corresponding positioning result data.

5. The optical cable early warning method based on passive synaesthesia integration according to claim 1 is characterized in that: The method also includes: Build a three-dimensional virtual model of the underground space, import data from various sensors in real time, overlay gas concentration, temperature field, and vibration distribution data to form a dynamic visualization interface, and annotate the dynamic visualization interface on a GIS map to display the location, type, and time of abnormal events in real time; When it is detected that the gas concentration exceeds the standard or the temperature is abnormal, the regional fan is automatically started and the valve is closed, and the video surveillance system is linked to capture the on-site image and upload it to the monitoring platform.

6. The optical cable early warning method based on passive synaesthesia integration according to claim 1 is characterized in that: The intelligent analysis platform also includes: Regularly collect status data of each monitoring device and store it in the database according to categories; Provide rich chart display functions based on the status data, customize query results by selecting time period, device type and other conditions, and export the query results.

7. An optical cable early warning system based on passive synaesthesia integration, characterized in that: include: The data acquisition module performs passive sensing signal acquisition on the optical cables deployed in the target area to obtain disturbance signals propagating along the optical cables; pre-processes the disturbance signals to obtain a normalized disturbance signal sequence; and simultaneously collects environmental data through a multi-parameter sensor network, wherein the environmental data includes gas concentration, temperature, and vibration acceleration; a disturbance event data determination module, which preprocesses the disturbance signal and environmental data and constructs a multidimensional feature vector including gas concentration mutation characteristics, temperature distribution anomalies, and vibration frequency-amplitude, and performs disturbance event identification based on the multidimensional feature vector to determine corresponding disturbance event data, where the disturbance event data includes gas disturbance, temperature disturbance, and vibration disturbance; The positioning result data positioning module combines the disturbance event data with the spatial distribution of the optical fiber sensing signal and the time difference of the vibration wave propagation to locate the positioning result data corresponding to the abnormal event, generates an early warning instruction based on the positioning result data, and pushes it to the user terminal through the monitoring platform.

8. The optical cable early warning system based on passive synaesthesia integration according to claim 7 is characterized in that: Also includes: The concentration monitoring module is used to monitor the concentrations of CO2, CO, and CH4 in real time and calculate the mutation slope within the sliding window. If the mutation slope is ≥5% vol / min or the cumulative amount exceeds the threshold of 120%, a gas leak warning is triggered; The judgment module performs a fast Fourier transform on the original vibration signal to decompose it into different frequency bands. It analyzes whether there are significant energy peaks in the low-frequency band to determine whether there is mining activity; it analyzes whether there are significant energy peaks in the high-frequency band to determine whether there is mechanical impact. It also makes a comprehensive judgment based on the acceleration amplitude to identify the specific disturbance type. The temperature change rate calculation module monitors the surface temperature distribution of the equipment in real time, identifies local hot spots on the surface of the monitoring equipment, and is used to calculate the temperature change rate corresponding to the local hot spots. If the temperature difference corresponding to the temperature change rate suddenly changes by ≥15℃ / min, a fire risk warning is triggered.

9. An optical cable early warning system based on passive synaesthesia integration, characterized in that: The method comprises a processor running a program of an optical cable early warning method based on passive synaesthesia integration as claimed in any one of claims 1 to 6.

10. A storage medium, characterized in that: A program is stored for the optical cable early warning method based on passive synaesthesia integration as claimed in any one of claims 1 to 6.

Images