Optical cable burial depth detection method based on distributed optical fiber sensing technology

Abstract

The invention proposes an optical cable burial depth detection method based on a distributed optical fiber sensing technology, and relates to the technical field of optical fiber sensing, and the method comprises the steps: collecting the burial state change data of an optical cable, and the burial state change data comprise temperature data and vibration data; constructing a spatio-temporal data matrix corresponding to the temperature data, carrying out data processing on the spatio-temporal data matrix according to a sliding window algorithm, obtaining spatial temperature distribution data, selecting a plurality of reference features of the spatial temperature distribution data in any time period to construct a feature matrix, and sending the feature matrix into an optical cable temperature-burial depth relation model, obtaining an optical cable prediction depth corresponding to the temperature data; constructing a ground noise environment model based on the vibration data, obtaining a noise field distribution characteristic function corresponding to the ground noise environment model based on the ground noise environment model, performing Fourier transform on the noise field distribution characteristic function to obtain a noise correlation function, and obtaining a ground environment parameter according to the ground environment parameter and the noise correlation function. Calculating an optical cable prediction depth corresponding to the vibration data; and calculating the burying depth of the optical cable based on the optical cable prediction depth corresponding to the temperature data, the optical cable prediction depth corresponding to the vibration data and a depth prediction algorithm. According to the invention, the accuracy of predicting the embedding depth of the optical cable can be improved.

Description

Technical Field

[0001] The present invention relates to the field of optical fiber sensing technology, and in particular to a method for detecting the buried depth of an optical cable based on distributed optical fiber sensing technology. Background Art

[0002] With the rapid development of communication technology, the safety and stability of optical cables, as the main carrier of information transmission, are of vital importance. However, the buried depth of optical cables directly affects their safety and stability. Traditional methods for measuring the buried depth of optical cables, such as excavation and sonar, have problems such as complex operation, low efficiency, and easy damage to optical cables.

[0003] The Chinese patent publication with the publication number CN116662877A discloses a sample evaluation method for pattern recognition of distributed optical fiber sensing technology, which includes: determining the evaluation value of the correlation of sample features and the accuracy evaluation value of the sample class by calculating the covariance between the sample features, and assigning different weights to different samples according to the two results. However, the above method will still introduce background noise caused by external natural environment interference, and using a single covariance calculation to evaluate the correlation and accuracy of sample features will also lead to increased measurement errors. Therefore, it is very necessary to provide a method for detecting the buried depth of optical cables based on distributed optical fiber sensing technology to improve the accuracy of predicting the buried depth of optical cables. Summary of the invention

[0004] In view of this, the present invention proposes a method for detecting the buried depth of optical cables based on distributed optical fiber sensing technology. By constructing a spatiotemporal data matrix and applying a sliding window algorithm for processing, the key features of the spatial temperature distribution are effectively extracted, and the optical cable temperature-buried depth relationship model is enabled to capture short-term and long-term temperature change patterns, and combined with the ground noise environment model and noise correlation function, the prediction accuracy and model adaptability are improved.

[0005] The present invention provides a method for detecting the buried depth of an optical cable based on distributed optical fiber sensing technology, the method comprising:

[0006] Collecting buried state change data of the optical cable, wherein the buried state change data includes temperature data and vibration data;

[0007] Constructing a spatiotemporal data matrix corresponding to the temperature data, and performing data processing on the spatiotemporal data matrix according to a sliding window algorithm to obtain spatial temperature distribution data, selecting multiple reference features of the spatial temperature distribution data in any time period to construct a feature matrix, and sending the feature matrix to an optical cable temperature-burial depth relationship model to obtain an optical cable predicted depth corresponding to the temperature data;

[0008] Constructing a ground noise environment model based on the vibration data, and obtaining a noise field distribution characteristic function corresponding to the ground noise environment model based on the ground noise environment model, performing Fourier transform on the noise field distribution characteristic function to obtain a noise correlation function, and calculating a predicted depth of the optical cable corresponding to the vibration data according to ground environment parameters and the noise correlation function;

[0009] The buried depth of the optical cable is calculated based on the predicted depth of the optical cable corresponding to the temperature data, the predicted depth of the optical cable corresponding to the vibration data, and a depth prediction algorithm.

[0010] On the basis of the above technical solution, preferably, the data processing of the spatiotemporal data matrix according to the sliding window algorithm to obtain the spatial temperature distribution data specifically includes:

[0011] Setting the length of the sliding window, and performing sliding window division on the space-time data matrix based on the sliding window length, and obtaining window space data after the space-time data matrix is ​​sorted within the sliding window;

[0012] Based on the window space data and the window temperature relationship function, the window temperature corresponding to each window space data is calculated, and the window temperatures corresponding to all the window space data are integrated to form the spatial temperature distribution data.

[0013] On the basis of the above technical solution, preferably, the method further comprises:

[0014] Acquire burial information of the optical cable, wherein the burial information includes a sample point position, a sample point depth, and a first detection temperature set corresponding to the sample point position and the sample point depth;

[0015] Changing the sample point position and sample point depth of the optical cable to obtain a second detection temperature set;

[0016] Based on the first detection temperature set, the second detection temperature set and the window sliding algorithm, a sample point temperature distribution data set is obtained, and the sample point temperature distribution data set is standardized to obtain a standard temperature distribution data set;

[0017] Based on the standard temperature distribution data set and grey correlation function, an optical cable temperature-burial depth relationship model is constructed.

[0018] More preferably, the expression of the temperature feature vector of the i-th sample point in the feature matrix is:

[0019]

[0020]

[0021] ΔT(i, t) = T t (i, t)-T t (i, t-1)

[0022]

[0023]

[0024] Where M(i) represents the mean value function of the i-th sample point, T t (i, t) represents the temperature of the i-th sample point at the t-th time, n represents the number of sample points, S(i) represents the standard deviation function of the i-th sample point, ΔT(i, t) represents the temperature difference function of the i-th sample point at the t-th time, v(i, t) represents the rate of change function of the i-th sample point at the t-th time, Δt represents the time difference between two adjacent sample points, ω(i) represents the temperature feature vector of the i-th sample point, max() represents the maximum value function, Represents the average rate of change function of i sample points.

[0025] More preferably, the grey relational function can be expressed as:

[0026]

[0027] Among them, γ i represents the grey relational function of the i-th sample point, j represents the j-th factor index, x j (i) represents the i-th sample point data in the j-th factor index, y(i) represents the output data of the i-th sample point, represents the resolution coefficient, n represents the number of sample points, max() represents the maximum function, and min() represents the minimum function.

[0028] More preferably, the building of the ground noise environment model based on the vibration data specifically includes:

[0029] Preprocessing the vibration data to obtain standard vibration data, and extracting frequency features corresponding to the standard vibration data, wherein the frequency features include a noise source wave number and a noise source angular frequency;

[0030] A ground noise environment model is constructed based on the Helmholtz equation, the noise source wave number and the noise source angular frequency.

[0031] More preferably, the ground noise environment model is expressed as:

[0032]

[0033] in, represents the Laplace operator, k represents the wave number of the noise source, P(r, z) represents the sound pressure at the position where the horizontal coordinate is r and the vertical coordinate is z, ω represents the angular frequency of the noise source, z s Represents the height difference between the noise source and the ground, N s (ω,r s ) indicates an angular frequency of ω and a position of r s The noise source is , and δ() represents the Dirac delta function.

[0034] More preferably, the noise correlation function is expressed as:

[0035] Γ ω (r1,r2;z1,z2)= <P(r1,z1)P * (r2,z2)>

[0036]

[0037]

[0038] Among them, Γ ω represents the cross spectral density, r1 represents the horizontal coordinate of the first detection point, r2 represents the horizontal coordinate of the second detection point, z1 represents the height difference between the first detection point and the ground, z2 represents the height difference between the second detection point and the ground, P(r1, z1) represents the sound pressure at the first detection point, P*(r2, z2) represents the complex conjugate of the sound pressure at the second detection point, R represents the difference between the horizontal coordinates of the first detection point and the horizontal coordinates of the second detection point, Φ τ () represents the modal depth function, J0() represents the zero-order Bessel function, κ τ represents the real part of the horizontal wave number, α τ represents the imaginary part of the horizontal wave number, Γ t (R,z1,z2) represents the noise correlation function, q represents the intensity of the noise source, ρ represents the density of the waveguide medium, and τ represents the mode number.

[0039] More preferably, the method further comprises:

[0040] Acquire real-time depth data according to the depth prediction algorithm, and obtain a burial depth alarm threshold from a database to determine whether the real-time depth data of the sample point is less than the burial depth alarm threshold;

[0041] If the real-time depth data of the sample point is less than the buried depth alarm threshold, the geographical location of the sample point is obtained, and an alarm signal and the geographical location of the sample point are sent to the mobile terminal of the staff.

[0042] The optical cable buried depth detection method based on distributed optical fiber sensing technology provided by the present invention has the following beneficial effects compared with the prior art:

[0043] (1) By constructing a spatiotemporal data matrix and applying a sliding window algorithm to process it, the key features of the spatial temperature distribution are effectively extracted and converted into spatial temperature distribution data, which can significantly improve the model's ability to capture complex temperature-depth relationships. At the same time, data from any time period is selected for analysis, so that the optical cable temperature-burial depth relationship model can adapt to different time scales, so that the optical cable temperature-burial depth relationship model can capture short-term and long-term temperature change patterns, and can handle nonlinear relationships and adapt to changes in different time scales, thereby improving the model's adaptability. By constructing a ground noise environment model, the complex ground noise environment can be accurately described and characterized. Fourier transform can convert time domain signals into frequency domain, which is conducive to analyzing the frequency components of the signal. At the same time, the noise correlation function is used to reflect the correlation between noise signals at different locations. The optical cable temperature-burial depth relationship model is combined with the ground noise environment model and the noise correlation function, thereby improving the accuracy of the prediction of the optical cable burial depth.

[0044] (2) By obtaining the sample point location, depth and corresponding temperature set, the integrity and representativeness of the data are ensured. At the same time, the sample point location and depth are changed to obtain the second detection temperature set, which improves the model's adaptability to different environments and conditions. The sample point temperature distribution data set is standardized to eliminate the influence of different scales and units, improve the data comparability and model stability, and combine the grey correlation function to construct the optical cable temperature-burial depth relationship model, so that the temperature-burial depth relationship model can handle small sample data and incomplete information, thereby improving the accuracy and reliability of the optical cable temperature-burial depth relationship model. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A flow chart of a method for detecting buried depth of optical cables based on distributed optical fiber sensing technology provided by the present invention;

[0047] Figure 2 A schematic diagram of vibration detection for underground buried optical cables provided by the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0049] The present application embodiment discloses a method for detecting the buried depth of an optical cable based on distributed optical fiber sensing technology. Figure 1 As shown, the method comprises steps S1 to S4.

[0050] Step S1, collecting buried state change data of the optical cable, wherein the buried state change data includes temperature data and vibration data.

[0051] In this step, a distributed optical fiber temperature sensing system is used to collect temperature data along the optical cable path, and an appropriate spatial sampling interval (e.g., 1 meter) and time sampling interval (e.g., 10 minutes) are set to record temperature values ​​at different depths and times to form temperature-space-time three-dimensional data. That is, the distributed optical fiber temperature sensing system is used to continuously collect temperature data of the optical fiber under test, and a spatiotemporal data matrix containing dual dimensions of time t and space l is constructed as the original signal T t,l .

[0052] The distributed fiber optic temperature sensing system is mainly composed of a light source, a fiber optic sensor, a spectrum analyzer and a data processing unit. Its spatial resolution is usually 0.5-2 meters, and the temperature resolution can generally reach 0.1°C or higher. The sampling interval can be set according to demand and can range from 5 minutes to 1 hour. Long-term temperature collection covers different seasons to capture the temperature change pattern, while recording day and night temperature changes, especially in shallow buried areas, and recording the impact of special weather such as rain and snow on temperature.

[0053] Use a distributed acoustic wave sensing system to collect vibration signals along the optical cable, set appropriate sampling frequency (e.g. 1000Hz) and spatial resolution (e.g. 1 meter), and record vibration waveform data at different locations and times. Among them, the distributed acoustic wave sensing system consists of a light source, an optical fiber sensor, a photodetector, and a signal processing unit. The sampling frequency is usually set to 500Hz-2000Hz, which is determined according to the highest frequency to be monitored. The spatial resolution is generally 1-10 meters, depending on the performance of the optical fiber sensor and actual needs. The acquisition time can be set according to actual needs, which can be continuous acquisition or timed acquisition. For the collected vibration waveform data, observe the vibration amplitude, duration and other characteristics, perform Fourier transform, analyze the frequency characteristics of the vibration signal, and use methods such as short-time Fourier transform or wavelet transform to analyze the time-frequency characteristics of the signal.

[0054] Step S2, construct a spatiotemporal data matrix corresponding to the temperature data, and process the spatiotemporal data matrix according to the sliding window algorithm to obtain spatial temperature distribution data, select multiple reference features of the spatial temperature distribution data in any time period to construct a feature matrix, and send the feature matrix into the optical cable temperature-burial depth relationship model to obtain the optical cable predicted depth corresponding to the temperature data.

[0055] By constructing a spatiotemporal data matrix and applying a sliding window algorithm for processing, the key features of the spatial temperature distribution are effectively extracted and converted into spatial temperature distribution data. This process not only optimizes the data structure, but also significantly improves the model's ability to capture complex temperature-depth relationships. At the same time, data from any time period is selected for analysis, so that the optical cable temperature-burial depth relationship model can adapt to different time scales, and the optical cable temperature-burial depth relationship model can capture short-term and long-term temperature change patterns, and can handle nonlinear relationships and adapt to changes at different time scales, thereby improving the prediction accuracy and model adaptability.

[0056] Step S2 also includes steps S21 to S22.

[0057] Step S21, setting the length of the sliding window, and dividing the space-time data matrix into sliding windows based on the sliding window length, and obtaining window space data after the space-time data matrix is ​​sorted in the sliding window.

[0058] In this step, a sliding window size is set to m+1, and the original signal T in the window is t,l Sort in order, that is:

[0059]

[0060] The spatial signal y of the sorted sliding window m+1 , perform corresponding processing and calculate the temperature t of the sliding window:

[0061]

[0062] in, Indicates The temperature of the sample point is t(i), and t(i) represents the temperature of the i-th sample point after processing.

[0063] Step S22, based on the window space data and the window temperature relationship function, calculate the window temperature corresponding to each window space data, and integrate the window temperatures corresponding to all the window space data to form spatial temperature distribution data.

[0064] In this step, the moving step size of the sliding window is set to k, and the temperature of each window is calculated by sliding, and integrated into the processed spatial temperature distribution data.

[0065] In this embodiment, the spatial temperature distribution data of any period of time is selected, and five statistical data are extracted: the average value, the standard deviation, the maximum temperature difference, the average rate of change, and the maximum rate of change. After normalization, they are used as feature vectors to construct the full-line feature matrix. The expression of the temperature feature vector of the i-th sample point in the feature matrix is:

[0066]

[0067]

[0068] ΔT(i,t)=T t (i,t)-T t (i,t-1)

[0069]

[0070]

[0071] Where M(i) represents the mean value function of the i-th sample point, T t (i,t) represents the temperature of the i-th sample point at the t-th time, n represents the number of sample points, S(o) represents the standard deviation function of the i-th sample point, ΔT(i,t) represents the temperature difference function of the i-th sample point at the t-th time, v(i,t) represents the rate of change function of the i-th sample point at the t-th time, Δt represents the time difference between two adjacent sample points, ω(i) represents the temperature feature vector of the i-th sample point, max() represents the maximum value function, Represents the average rate of change function of i sample points.

[0072] In one example, the process of constructing an optical cable temperature-burial depth relationship model includes: obtaining the burial information of the optical cable, wherein the burial information includes the sample point position, the sample point depth, and a first detection temperature set corresponding to the sample point position and the sample point depth; changing the sample point position and the sample point depth of the optical cable to obtain a second detection temperature set; based on the first detection temperature set, the second detection temperature set, and the window sliding algorithm, obtaining a sample point temperature distribution data set, and standardizing the sample point temperature distribution data set to obtain a standard temperature distribution data set; based on the standard temperature distribution data set and the grey correlation function, constructing an optical cable temperature-burial depth relationship model.

[0073] The process of acquiring the buried information of the optical cable includes: planning the sampling points, recording the locations of the sampling points, measuring the depth of the sampling points, recording the environmental parameters, collecting the temperature data, recording the meteorological data and establishing the sample point information database.

[0074] Planning sampling points includes selecting multiple representative sample points along the optical cable path and ensuring that the sample points cover different geographical environments and burial conditions (such as different soil types, surface cover, etc.). A sufficient number of sampling points are set to ensure statistical significance. Usually, there are no less than 30 sampling points. The spacing between sampling points can be determined based on the total length of the optical cable and terrain changes, for example, one point is selected every 100 meters or 500 meters.

[0075] Recording the location of sample points includes using high-precision GPS equipment to record the geographic coordinates (longitude, latitude) of each sample point, while recording the distance relative to the starting point of the optical cable (such as 1500 meters of the optical cable), and drawing a distribution map of the sampling points to ensure that the sampling points are evenly distributed and representative.

[0076] Measuring the depth of the sample point includes using a laser rangefinder or an ultrasonic depth meter to measure the depth of the sample point when it is buried. For shallower burials, a ruler can be used to directly measure and record the measured depth with an accuracy of centimeters.

[0077] The recorded environmental parameters include soil type (such as sand, clay, rock, etc.); surface cover (such as bare soil, grass, asphalt pavement, etc.); groundwater level; surrounding environmental characteristics (such as whether it is close to buildings, water bodies, etc.).

[0078] Collecting temperature data (first detected temperature set) includes continuous temperature monitoring using a distributed fiber optic temperature sensing system, setting an appropriate spatial resolution, usually 0.5 to 1 meter, and setting an appropriate time resolution, such as collecting data every 5 or 10 minutes, and continuously collecting data for at least 24 hours, preferably a week or longer, to capture daily changes and short-term weather effects.

[0079] Recording meteorological data includes recording meteorological parameters such as temperature, humidity, and precipitation during the sampling period. A portable weather station can be used or data can be obtained from the nearest meteorological observation station to record the weather conditions during the sampling period (sunny, cloudy, rainy, etc.).

[0080] Establishing the sample point information database includes assigning a unique identifier to each sample point, creating a data table containing the following geographic coordinates (longitude, latitude), relative distance, measurement depth, soil type, surface cover type, and environmental characteristic description fields, and associating temperature data with sample point information. At the same time, any special circumstances that may affect the temperature of the optical cable (such as underground pipelines, geothermal activity, etc.) are recorded, and areas that may require special treatment (such as river crossings, under roads, etc.) are marked.

[0081] For changing the location and depth of the sample points of the optical cable, the change strategy can be planned, such as determining the purpose of the change (such as verifying the performance of the model at different depths, the impact of the test location on temperature, etc.), determining the scope of the change (such as changing 20% ​​of the sample points, or adding 1-2 new points near each original sample point), etc. A new location can be selected near the original sample point, usually within a range of 5-10 meters, to ensure that the new location does not interfere with the original optical cable path or other infrastructure. For some of the original sample points, change their burial depth, the new depth can be deeper or shallower than the original, usually ranging from 0.5-2 meters. For the newly added location, small-scale excavation and burial work is required. For the point with changed depth, it may be necessary to re-excavate and adjust the depth of the optical cable.

[0082] The steps of obtaining the burial information of the optical cable and changing the sample point position and the sample point depth of the optical cable are repeated to obtain more diverse data samples and enhance the generalization ability of the optical cable temperature-burial depth relationship model.

[0083] The temperature data obtained in the above steps are cleaned to remove outliers and noise, and to handle missing data. Interpolation methods can be used or incomplete records can be deleted, and the temperature data are aligned with timestamps and location information. According to the temporal resolution of the data and the research objectives, an appropriate sliding window length (such as 24 hours) is selected, and the sliding step size is set, which is usually less than the window length (such as 1 hour) to ensure overlap between windows. The temperature data in each sliding window is sorted, and the basic statistics in the window (average temperature, maximum and minimum temperature, temperature standard deviation, temperature median and quartile, and temperature change feature extraction) are calculated; the rate of change of temperature in the window (such as hourly temperature change) is calculated, and the temperature change trend is extracted. Simple linear regression can be used to obtain the slope and identify temperature fluctuation patterns, such as daily cycles or other periodic changes. All extracted features are combined into a comprehensive data set, with each sample point corresponding to a row of data, including the following: sample point ID, location information (latitude and longitude or relative coordinates), timestamp, statistical features (average temperature, standard deviation, etc.), change features (rate of change, trend, etc.), spatial features (spatial gradient, etc.), time series features (autocorrelation coefficient, etc.), frequency domain features, and environmental related features. Calculate the correlation between features, remove highly correlated redundant features, and use principal component analysis (PCA) or other dimensionality reduction techniques to reduce the dimensionality of the feature space. Standardize all numerical features, usually using Z-score or Min-Max methods, to ensure that all features are in the same numerical range for subsequent model processing.

[0084] For building a model based on a standard temperature distribution data set and a grey correlation function, this step includes selecting a suitable reference sequence, which is usually sample point data of known depth, calculating the grey correlation between the comparison sequence and the reference sequence, and constructing a correlation matrix to reflect the strength of the relationship between different features and burial depth.

[0085] The grey relational function can be expressed as:

[0086]

[0087] Among them, γ i represents the grey relational function of the i-th sample point, j represents the j-th factor index, x j (i) represents the i-th sample point data in the j-th factor index, y(i) represents the output data of the i-th sample point, represents the resolution coefficient, n represents the number of sample points, max() represents the maximum function, and min() represents the minimum function.

[0088] The sorted data set is randomly shuffled and divided into training set, validation set, and test set according to preset proportions, such as 70% for training set, 15% for validation set, and 15% for test set, to ensure the independence and validity of the data set. A support vector machine (SVM) model is built based on the training set, and the validation set is used for cross-validation to optimize the model's parameters and performance. The model is then tested, evaluated, optimized, and adjusted using the test set to ultimately obtain a trained model.

[0089] Step S3, constructing a ground noise environment model based on the vibration data, and obtaining a noise field distribution characteristic function corresponding to the ground noise environment model based on the ground noise environment model, performing Fourier transform on the noise field distribution characteristic function to obtain a noise correlation function, and calculating the predicted depth of the optical cable corresponding to the vibration data based on the ground environment parameters and the noise correlation function.

[0090] By constructing a ground noise environment model, the complex ground noise environment can be accurately described and characterized. Fourier transform can convert time domain signals into frequency domain, which is conducive to analyzing the frequency components of the signal and thus obtaining more useful information. At the same time, the noise correlation function is used to reflect the correlation between noise signals at different locations. By combining the ground environment parameters and the noise correlation function, the depth corresponding to the vibration data can be estimated.

[0091] Step S3 also includes steps S31 to S32.

[0092] Step S31 , preprocessing the vibration data to obtain standard vibration data, and extracting frequency features corresponding to the standard vibration data, wherein the frequency features include the wave number of the noise source and the angular frequency of the noise source.

[0093] In this step, the vibration data is cleaned to remove outliers and fill in missing data. Low-pass, high-pass or band-pass filters are used to remove unnecessary frequency components in the vibration data. The vibration data is amplitude standardized. At the same time, the long time series data is divided into appropriate time windows. Spectral analysis is performed using methods such as short-time Fourier transform or wavelet transform to extract key features such as noise source wave number and noise source angular frequency from the spectrum.

[0094] Step S32: constructing a ground noise environment model based on the Helmholtz equation, the noise source wave number and the noise source angular frequency.

[0095] In this step, the noise source model is established based on the extracted wave number and angular frequency by using the Helmholtz equation to describe the propagation of sound waves in the ground medium and setting appropriate boundary conditions according to the actual environment.

[0096] The expression of the ground noise environment model is:

[0097]

[0098] in, represents the Laplace operator, k represents the wave number of the noise source, P(r, z) represents the sound pressure at the position where the horizontal coordinate is r and the vertical coordinate is z, ω represents the angular frequency of the noise source, z s Represents the height difference between the noise source and the ground, i.e. r s The vertical component, N s (ω,r s ) indicates an angular frequency of ω and a position of r s The noise source is , and δ() represents the Dirac delta function.

[0099] For the ground noise environment model, the following boundary conditions are usually considered:

[0100] Free surface condition (ground surface): At z = 0 (assuming the ground surface is the z = 0 plane), it should satisfy This means that the vertical gradient of sound pressure at the surface is zero, that is, the sound waves are completely reflected at the surface.

[0101] Radiation conditions (deep): When z→∞, it should satisfy: This is the Sommerfeld radiation condition, which ensures that the wave energy decays at a distance and does not reflect back from infinity.

[0102] Periodic or decay conditions in the horizontal direction: Depending on the problem, it may be necessary to set periodic boundary conditions or decay conditions in the horizontal direction.

[0103] If the model contains multiple layers, the continuity of pressure and particle velocity must be satisfied at the interface between the layers:

[0104] P1=P2

[0105]

[0106] Where ρ is the density of the waveguide medium, ρ1 represents the medium on one side of the interface, and ρ2 represents the medium on the other side of the interface.

[0107] See also Figure 2 , Figure 2 This is a schematic diagram of vibration detection, where the expression of the noise correlation function is:

[0108] Γ ω (r1,r2;z1,z2)= <P(r1,z1)P * (r2,z2)>

[0109]

[0110]

[0111] Among them, Γ ω represents the cross spectral density, r1 represents the horizontal coordinate of the first detection point, r2 represents the horizontal coordinate of the second detection point, z1 represents the height difference between the first detection point and the ground, z2 represents the height difference between the second detection point and the ground, P(r1, z1) represents the sound pressure at the first detection point, P*(r2, z2) represents the complex conjugate of the sound pressure at the second detection point, R represents the difference between the horizontal coordinates of the first detection point and the horizontal coordinates of the second detection point, Φ τ () represents the modal depth function, J0() represents the zero-order Bessel function, κ τ represents the real part of the horizontal wave number, α τ represents the imaginary part of the horizontal wave number, Γ t (R,z1,z2) represents the noise correlation function, q represents the intensity of the noise source, ρ represents the density of the waveguide medium, and τ represents the mode number.

[0112] By deriving and integrating the above formula, we can get the time derivative of the noise correlation function:

[0113]

[0114] When the ground noise conditions are determined, q, ρ, κ τ and α τ Assuming it to be a fixed value, the amplitude A of the time derivative of the noise correlation function C Depends on the depth relationship between the two receiving points, its expression is:

[0115]

[0116] The above formula reflects a vertical coherence characteristic. For an ideal solid space, such as bedrock and sedimentary layers, this characteristic is related to the relative distance between the two receiving points, that is, it has nothing to do with the absolute position. Therefore, the relationship can be written as:

[0117] A C ∝Φ′|z2-z1

[0118] Therefore, the depth change between two receiving points can be estimated by extracting the amplitude of the time derivative of the noise correlation function.

[0119] In one example, since the integration in the time derivative of the noise correlation function is performed in the frequency domain, The contributions of different modes are represented in the time derivative of the noise correlation function, and Φ τ (z1)Φ τ (z2) directly reflects the influence of the depth of the receiving point.

[0120] The depth-amplitude relationship is established. If only the amplitude of a single dominant mode is considered, the amplitude can be expressed as:

[0121]

[0122] When z1 is fixed, the amplitude of only considering the change of z2 can be expressed as:

[0123] A(z2,ω)∝Φ τ (z2)gC(ω)

[0124] Among them, C(ω) contains all terms that are independent of z2.

[0126] For modal function approximation, in general, simple functions can be used to approximate the modal function, such as: or Where H is the dielectric thickness and β is the attenuation coefficient associated with z2.

[0127] When using the sin approximation:

[0128] When using exp approximation: A(z2,ω)∝exp(-βz2)C(ω)

[0129] By performing a logarithmic transformation on the amplitude corresponding to z2 when using the exp approximation, we can obtain: Log(A)=-βz2+log(C).

[0130] Then repeat the above process for different frequencies, and the frequency-related model parameters can be expressed as:

[0131] A(z2,ω)=C(ω)exp(-β(ω)z2)

[0132] Similarly, when z1 is fixed, only the amplitude of z2 change is considered, and the frequency-dependent model parameters can be expressed as:

[0133] A(z1,ω)=D(ω)exp(-μ(ω)z1)

[0134] Among them, D(ω) contains all terms that are independent of z1, and μ represents the attenuation coefficient related to z1.

[0135] For a receiving point of unknown depth, the depth corresponding to z2 can be calculated using the following steps:

[0136] For each frequency ω, calculate z2: z2(ω)=-log(A(ω) / C(ω)) / β(ω).

[0137] For the integration of multi-frequency results, z2=Σ(w(ω)*z2(ω)) / Σw(ω).

[0138] Wherein, w(ω) represents a weight coefficient, which can be determined based on the signal-to-noise ratio or fitting quality of each frequency.

[0139] Step S4, calculating the buried depth of the optical cable based on the predicted depth of the optical cable corresponding to the temperature data, the predicted depth of the optical cable corresponding to the vibration data, and a depth prediction algorithm.

[0140] In this step, the depth prediction algorithm can be expressed as: Z = λz t +ξz sa

[0141] Where λ represents the weight coefficient corresponding to the predicted depth of the optical cable, z t represents the predicted depth of the optical cable, ξ represents the weight coefficient corresponding to the predicted depth of the optical cable, and z sa Indicates the predicted depth of the optical cable.

[0142] Step S4 also includes steps S41 to S42.

[0143] Step S41, obtaining real-time depth data according to the depth prediction algorithm, and obtaining the burial depth alarm threshold from the database, and determining whether the real-time depth data of the sample point is less than the burial depth alarm threshold.

[0144] In this step, the real-time depth data obtained from the depth prediction algorithm is compared with a preset threshold value obtained from a system configuration file or a database.

[0145] Step S42: If the real-time depth data of the sample point is less than the buried depth alarm threshold, the geographical location of the sample point is obtained, and an alarm signal and the geographical location of the sample point are sent to the mobile terminal of the staff.

[0146] In this step, the burial depth alarm threshold can be set to a reasonable burial depth alarm threshold according to the actual burial requirements of the optical cable. For example, if the standard burial depth of the optical cable is 1 meter, the burial depth alarm threshold can be set to 0.8 meters. The burial depth alarm threshold can be a fixed value or dynamically adjusted according to different environmental conditions and soil characteristics. Combined with the mileage pile number or GPS coordinates of the optical cable line, the geographical location of the abnormal point can be accurately located. The GIS system or electronic map can be used to intuitively display the specific location of the abnormal point on the optical cable line. The location information of the abnormal point can include longitude and latitude coordinates, kilometer marks, nearby landmarks, etc., which is convenient for maintenance personnel to quickly find the problem. Alarm information can be sent to relevant maintenance personnel through mobile APP, SMS, email and other methods. The alarm information may include the location coordinates of the abnormal point, burial depth data, temperature change trend, etc., to provide a basis for subsequent maintenance. Different levels of alarms can be set, such as slight deviations, serious anomalies, etc., to trigger different levels of response measures. Establish a complete alarm management system, record alarm history, analyze fault modes, and optimize alarm strategies. After receiving the alarm, maintenance personnel can rush to the scene and take corresponding maintenance measures, such as replenishing backfill soil, adjusting the burial depth, etc. It can also combine historical data analysis to predict possible problem areas and take preventive maintenance in advance.

[0147] By constructing a spatiotemporal data matrix and applying a sliding window algorithm to process it, the key features of the spatial temperature distribution are effectively extracted and converted into spatial temperature distribution data, which can significantly improve the model's ability to capture complex temperature-depth relationships. At the same time, data from any time period is selected for analysis, so that the optical cable temperature-burial depth relationship model can adapt to different time scales, and the optical cable temperature-burial depth relationship model can capture short-term and long-term temperature change patterns, and can handle nonlinear relationships and adapt to changes in different time scales, thereby improving the model's adaptability. By constructing a ground noise environment model, the complex ground noise environment can be accurately described and characterized. Fourier transform can convert time domain signals to frequency domain, which is conducive to analyzing the frequency components of the signal. At the same time, the noise correlation function is used to reflect the correlation between noise signals at different locations. The optical cable temperature-burial depth relationship model is combined with the ground noise environment model and the noise correlation function, thereby improving the accuracy of the prediction of the optical cable burial depth.

[0148] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting buried depth of optical cable based on distributed optical fiber sensing technology, characterized in that: The method comprises: Collecting buried state change data of the optical cable, wherein the buried state change data includes temperature data and vibration data; Constructing a spatiotemporal data matrix corresponding to the temperature data, and performing data processing on the spatiotemporal data matrix according to a sliding window algorithm to obtain spatial temperature distribution data, selecting multiple reference features of the spatial temperature distribution data in any time period to construct a feature matrix, and sending the feature matrix to an optical cable temperature-burial depth relationship model to obtain an optical cable predicted depth corresponding to the temperature data; Constructing a ground noise environment model based on the vibration data, and obtaining a noise field distribution characteristic function corresponding to the ground noise environment model based on the ground noise environment model, performing Fourier transform on the noise field distribution characteristic function to obtain a noise correlation function, and calculating a predicted depth of the optical cable corresponding to the vibration data according to ground environment parameters and the noise correlation function; The buried depth of the optical cable is calculated based on the predicted depth of the optical cable corresponding to the temperature data, the predicted depth of the optical cable corresponding to the vibration data, and a depth prediction algorithm.

2. The method according to claim 1, characterized in that The processing of the spatiotemporal data matrix according to the sliding window algorithm to obtain spatial temperature distribution data specifically includes: Setting the length of the sliding window, and performing sliding window division on the space-time data matrix based on the sliding window length, and obtaining window space data after the space-time data matrix is ​​sorted within the sliding window; Based on the window space data and the window temperature relationship function, the window temperature corresponding to each window space data is calculated, and the window temperatures corresponding to all the window space data are integrated to form the spatial temperature distribution data.

3. The method according to claim 1, characterized in that The method further comprises: Acquire burial information of the optical cable, wherein the burial information includes a sample point position, a sample point depth, and a first detection temperature set corresponding to the sample point position and the sample point depth; Changing the sample point position and sample point depth of the optical cable to obtain a second detection temperature set; Based on the first detection temperature set, the second detection temperature set and the window sliding algorithm, a sample point temperature distribution data set is obtained, and the sample point temperature distribution data set is standardized to obtain a standard temperature distribution data set; Based on the standard temperature distribution data set and grey correlation function, an optical cable temperature-burial depth relationship model is constructed.

4. The method according to claim 1, characterized in that The expression of the temperature feature vector of the i-th sample point in the feature matrix is: ΔT(i,t)=T t (i,t)-T t (i,t-1) Where M(i) represents the mean value function of the i-th sample point, T t (i, t) represents the temperature of the i-th sample point at the t-th time, n represents the number of sample points, S(i) represents the standard deviation function of the i-th sample point, ΔT(i, t) represents the temperature difference function of the i-th sample point at the t-th time, ν(i, t) represents the rate of change function of the i-th sample point at the t-th time, Δt represents the time difference between two adjacent sample points, ω(i) represents the temperature feature vector of the i-th sample point, max() represents the maximum value function, Represents the average rate of change function of i sample points.

5. The method according to claim 3, characterized in that The grey relational function can be expressed as: Among them, γ i represents the grey relational function of the i-th sample point, j represents the j-th factor index, x j (i) represents the i-th sample point data in the j-th factor index, y(i) represents the output data of the i-th sample point, represents the resolution coefficient, n represents the number of sample points, max() represents the maximum function, and min() represents the minimum function.

6. The method according to claim 1, characterized in that The method of constructing a ground noise environment model based on vibration data specifically includes: Preprocessing the vibration data to obtain standard vibration data, and extracting frequency features corresponding to the standard vibration data, wherein the frequency features include a noise source wave number and a noise source angular frequency; A ground noise environment model is constructed based on the Helmholtz equation, the noise source wave number and the noise source angular frequency.

7. The method according to claim 6, characterized in that The expression of the ground noise environment model is: in, represents the Laplace operator, k represents the wave number of the noise source, P(r, z) represents the sound pressure at the position where the horizontal axis is r and the vertical axis is z, ω represents the angular frequency of the noise source, z s Represents the height difference between the noise source and the ground, N s (ω,r s ) indicates an angular frequency of ω and a position of r s The noise source is δ(), and it represents the Dirac delta function.

8. The method according to claim 7, characterized in that The expression of the noise correlation function is: C ω (r1,r2;z1,z2)= <P(r1,z1)P * (r2,z2)> Among them, Γ ω represents the cross spectral density, r1 represents the horizontal coordinate of the first detection point, r2 represents the horizontal coordinate of the second detection point, z1 represents the height difference between the first detection point and the ground, z2 represents the height difference between the second detection point and the ground, P(r1, z1) represents the sound pressure at the first detection point, P*(r2, z2) represents the complex conjugate of the sound pressure at the second detection point, R represents the difference between the horizontal coordinates of the first detection point and the horizontal coordinates of the second detection point, Φ τ () represents the modal depth function, J0() represents the zero-order Bessel function, κ τ represents the real part of the horizontal wave number, α τ represents the imaginary part of the horizontal wave number, Γ t (R,z1,z2) represents the noise correlation function, q represents the intensity of the noise source, ρ represents the density of the waveguide medium, and τ represents the mode number.

9. The method according to claim 3, characterized in that The method further comprises: Acquire real-time depth data according to the depth prediction algorithm, and obtain a burial depth alarm threshold from a database to determine whether the real-time depth data of the sample point is less than the burial depth alarm threshold; If the real-time depth data of the sample point is less than the buried depth alarm threshold, the geographical location of the sample point is acquired, and an alarm signal and the geographical location of the sample point are sent to the mobile terminal of the staff.

Images