Mining overlying strata separation layer dynamic development in-situ monitoring method based on DAS

Abstract

The invention discloses a DAS-based mining overlying strata separation layer dynamic development in-situ monitoring method, which comprises the following steps: collecting data of a stope face, drilling a monitoring hole based on the data, putting a distributed vibration optical fiber into the monitoring hole, and coupling and solidifying the optical fiber and a hole wall by adopting a coupling material; the distributed vibration optical fiber transmits a sensed vibration signal to the DAS demodulator, and the DAS demodulator pre-processes data and then transmits the data to the ground DAS monitoring terminal; the ground DAS monitoring terminal carries out overlying strata separation layer characteristic parameter extraction on the data; and constructing an overlying strata separation layer development stage identification model, dividing the model based on the extracted overlying strata separation layer characteristic parameters, performing threshold comparison, and performing early warning when the overlying strata separation layer characteristic parameters exceed an early warning threshold. The overlying strata separation layer can be monitored in time and comprehensively, and the risk of separation layer water damage is reduced.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of coal mining safety monitoring, and particularly relates to a dynamic development in-situ monitoring method for overburden strata separation based on DAS. BACKGROUND

[0002] In the process of coal mining, with the operation of mining, the movement of overburden strata will cause the separation between strata. If water accumulates in the separation space, separation water is formed. Once the key layer breaks, water inrush accidents are easily caused, which brings great safety hazards to mine operation. The occurrence of separation water disasters not only threatens the safety of the mine, but also may cause great losses to the life safety of personnel and mine facilities. Therefore, it is particularly important to timely monitor the dynamic changes of separation and make accurate prediction.

[0003] At present, the traditional overburden strata separation monitoring means in the coal mining industry mainly includes separation meters or borehole stress meters. These traditional methods belong to "point type" monitoring, that is, data collection is carried out through the setting of multiple sensors or monitoring points. Since these methods are based on local point measurement, there are monitoring blind areas, and it is difficult to fully capture the continuous distribution of separation in space. At the same time, the real-time performance and dynamic tracking ability of the traditional methods are insufficient, and most of them rely on manual reading or instantaneous collection, which makes it difficult to accurately capture the dynamic evolution process of separation. In addition, the traditional instruments have weak anti-interference ability and poor environmental adaptability, and are easily disturbed by the water-bearing property of strata and mining activities, so the stability is poor. Electronic sensors are prone to failure in the humid and corrosive environment of the coal mine, resulting in loss of monitoring data and inability to fully and accurately record the dynamic evolution process of separation from "micro crack" to "large cavity". Therefore, the existing technology has certain limitations in separation monitoring and water disaster prediction. SUMMARY

[0004] The purpose of the present application is to provide a dynamic development in-situ monitoring method for overburden strata separation based on DAS, which can timely and comprehensively monitor the overburden strata separation and reduce the risk of separation water disaster.

[0005] The technical solution adopted by the present application is a dynamic development in-situ monitoring method for overburden strata separation based on DAS, comprising the following steps: Step 1, collecting data of the mining face, and drilling a monitoring hole based on the data, and lowering a distributed vibration optical fiber into the monitoring hole, and coupling and solidifying the optical fiber with the hole wall by using a coupling material; Step 2, the distributed vibration optical fiber transmits the sensed vibration signals to a DAS demodulator, and after the data is preprocessed by the DAS demodulator, it is transmitted to a ground DAS monitoring terminal; Step 3, the ground DAS monitoring terminal extracts overburden strata separation characteristic parameters from the data; Step 4, a model for identifying the development stage of the overburden separation layer is constructed, the extracted overburden separation layer characteristic parameters are divided by the model for identifying the development stage of the overburden separation layer, and the threshold is compared, and when the warning threshold is exceeded, warning is carried out.

[0006] The application is also characterized in that, The data of the mining face collected in step 1 include the parameters of the lithology distribution and thickness of the overburden, the parameters of the key layer of the overburden, the dip angle, the strike, the tendency of the overburden, the strike length, the tendency length, the coal seam thickness, the mining speed and the advancing direction of the working face.

[0007] The monitoring hole in step 1 is arranged on the side or the rear of the advancing direction of the working face, and the aquifer and the existing engineering interference area are avoided; when the rock layer is horizontal or the dip angle is less than 15°, the monitoring hole starts vertically, otherwise, the angle of the monitoring hole is parallel to the dip angle of the rock layer or forms an angle of 5°-10°. The drilling depth of the monitoring hole is to penetrate the uppermost key layer and extend 5m-10m below the mining influence depth.

[0008] The coupling material in step 1 is an epoxy resin-based composite coupling material, the compressive strength is greater than or equal to 30MPa, the linear expansion coefficient is less than or equal to 15x10 -6 / ℃, and the acoustic impedance matching degree with the surrounding rock is greater than or equal to 90%; when the coupling is solidified, the segmented injection method is adopted to gradually fill from the hole bottom to the hole opening, so as to ensure that the coupling material uniformly fills the small gap between the optical fiber and the hole wall, and the solidification time is 24h-48h.

[0009] The sampling frequency of the DAS demodulator in step 2 is 1kHz-10kHz, the data preprocessing of the DAS demodulator includes filtering the downhole high-frequency random noise and periodic mechanical interference contained in the data, the downhole high-frequency random noise is filtered by a low-pass analog filter with a preset cutoff frequency to filter the high-frequency electromagnetic interference generated by the downhole motor and cable; the periodic mechanical interference adopts an adaptive Notch filtering algorithm, which dynamically generates a frequency notch channel by real-time identification of the characteristic frequency of the noise, and filters the periodic interference in the frequency band; Then, the spatial-temporal matrix data is generated based on the filtered data, and the spatial-temporal matrix data includes the vibration signal amplitude, signal propagation time of each monitoring point on the distributed vibration optical fiber and the three-dimensional spatial position information of the corresponding monitoring point.

[0010] The extracted overburden separation layer characteristic parameters in step 3 include the energy value, dominant frequency and frequency change rate of the vibration signal, and time-frequency spectrum characteristics.

[0011] The overburden strata separation development stage identification model in step 4 is constructed based on an SVM model, during model training, historical separation data of the same geological condition mine as the working face in the present work is collected, and after labeling, standardization and abnormal value elimination, a data set is formed; the pretreated data set is divided into a training set and a verification set according to 7:3, a linear kernel is used as a kernel function, an optimal C value is used as a penalty parameter, and a model is trained through a sequential minimal optimization algorithm; The overburden strata separation development stage identification model is divided into a generation period, a development period and a mutation period, When the vibration energy is low and stable, the dominant frequency is in the low frequency band and the frequency change rate is gentle, and the time-frequency spectrum energy is concentrated in the low frequency and presents a single peak, the generation period is divided. When the vibration energy breaks through the initial threshold and continues to rise, the dominant frequency rises to the medium frequency band and the frequency change rate accelerates, and the time-frequency spectrum energy migrates to the medium frequency and appears double-band peak superposition, the development period is divided. When the vibration energy breaks through the early warning threshold and increases suddenly, the dominant frequency jumps to the high frequency band and the frequency change rate increases suddenly, and the time-frequency spectrum energy concentrates on the high frequency and presents a single peak, the mutation period is divided.

[0012] The early warning is in the form of sound and light alarm, short message push or terminal pop-up window.

[0013] The distributed vibration optical fiber is a mine-used flame-retardant single-mode optical fiber with a diameter of 900-1250 mu m, and has acid and alkali corrosion resistance.

[0014] The beneficial effects of the present application are: The DAS-based in-situ monitoring method for mining overburden strata separation of the present application uses a distributed vibration optical fiber as a sensing medium, and combines with a global drilling layout design, so that continuous monitoring of the mining overburden strata in a range of several kilometers can be realized, the spatial distribution characteristics of the separation of different layers can be completely captured, information loss caused by local point monitoring can be avoided, and through the overburden strata separation development stage identification model based on the collected data, the development stage of the separation is divided into a generation period, a development period and a mutation period, and the early warning threshold is set, once the monitoring finds that the separation value of a certain monitoring point exceeds the threshold, early warning can be carried out in time through various forms such as sound and light, short message, etc.; the separation monitoring is directly related to water disaster warning, the risk of water inrush accident caused by overburden strata separation is effectively reduced, and the safety of mine operation is ensured.

[0015] In addition, by using the optical fiber made of mine-used flame-retardant and acid-alkali-resistant material and the underground explosion-proof DAS demodulator meeting the underground safety standard, the present application can work stably for a long time in the complex environment of high dust, strong electromagnetic field and moisture corrosion in the mine, and the monitoring interruption caused by equipment failure is reduced. BRIEF DESCRIPTION OF DRAWINGS

[0016] Figure 1This is a flowchart illustrating the in-situ monitoring method for dynamic development of overburden delamination based on DAS according to the present invention. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] Example 1 This invention relates to an in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS, such as... Figure 1 As shown, it includes the following steps: Step 1: Collect data from the mining face, drill monitoring holes based on the data, insert distributed vibration optical fibers into the monitoring holes, and use coupling materials to couple and solidify the optical fibers with the hole walls. Furthermore, the collected data on the longwall face include the distribution and thickness parameters of the overburden lithology, parameters of key overburden layers, dip angle, strike, and dip direction of the overburden; strike length, dip length, coal seam thickness, mining speed, and advance direction of the longwall face.

[0019] Specifically, the lithological distribution and thickness parameters of the overburden include the lithology of each stratum from the coal seam roof to the surface, the thickness of each individual layer, and the characteristics of interlayer interfaces. This type of data determines areas prone to delamination and is the basis for judging "where delamination may occur".

[0020] Key parameters of the overburden include mechanical parameters such as the burial depth, thickness, elastic modulus, and compressive strength of the key layer.

[0021] The strike length, dip length, coal seam thickness, mining speed, and advance direction of the working face determine the distribution range and transmission path of mining stress, which are used to identify the "significant mining stress zone" and ensure that boreholes are laid in the area most susceptible to mining stress triggering.

[0022] Step 2: The distributed vibration fiber transmits the sensed vibration signal to the DAS demodulator, and after the DAS demodulator preprocesses the data, it is transmitted to the ground DAS monitoring terminal. Step 3: The ground-based DAS monitoring terminal extracts the characteristic parameters of the overburden delamination from the data; Step 4: Construct a model for identifying the development stage of overburden separation layer, based on the extracted overburden separation layer feature parameters. Divide the overburden separation layer according to the model and compare the thresholds. Issue an early warning when the threshold is exceeded.

[0023] Example 2 Based on Example 1 above, in this embodiment, the monitoring holes in step 1 of the DAS-based in-situ monitoring method for dynamic development of mining overburden delamination are arranged to the side or rear of the working face advancing direction to avoid the interference zone between the aquifer and the existing engineering; the purpose is to arrange monitoring holes in areas with significant mining stress and easy delamination development. To adapt to the rock strata orientation and ensure the fiber optic coupling effect, when the rock strata are horizontal or the dip angle is less than 15°, the monitoring hole is started vertically; otherwise, the angle of the monitoring hole is parallel to the rock strata dip angle or at an angle of 5° to 10°. The drilling depth of the monitoring borehole is to penetrate the uppermost key stratum and extend 5m to 10m below the depth affected by mining. Because the failure of the key stratum will cause multiple layers of delamination below it, it is necessary to cover the easily delaminating sections above and below the key stratum, while avoiding the expansion of the mining area and causing monitoring blind spots, so as to ensure that the overburden delamination can be monitored throughout the entire mining cycle.

[0024] Furthermore, if a strong aquifer exists below the key layer, the depth should be controlled 3-5m above the aquifer (or a casing should be used to isolate the aquifer) to prevent the borehole from conducting water.

[0025] Example 3 Based on Example 2 above, this embodiment uses an epoxy resin-based composite coupling material in step 1 of the DAS-based in-situ monitoring method for dynamic development of overburden delamination during mining. This material possesses both high strength and environmental adaptability, with a compressive strength ≥30MPa, capable of withstanding the stress generated by rock strata compression during overburden mining and preventing damage to the coupling layer under pressure; its coefficient of linear expansion is ≤15×10⁻⁶. -6 / ℃, which can adapt to downhole temperature fluctuations and prevent gaps from appearing between the coupling material and the borehole wall and optical fiber due to temperature changes; the acoustic impedance matching degree with the surrounding rock is ≥90%, which can minimize the transmission loss of vibration signals between "surrounding rock-coupling material-optical fiber" and ensure the accurate transmission of small vibration signals in the overburden.

[0026] The coupling curing operation employs a segmented injection method from the bottom to the opening: Coupling material is injected starting from the bottom of the monitoring hole. After this segment initially fills the tiny gaps between the optical fiber and the hole wall and expels air, the next segment is injected upwards, gradually filling the hole opening. This method avoids the presence of residual voids at the bottom due to material gravity when injecting from the opening, ensuring that the coupling material uniformly wraps the optical fiber and fills the pores throughout the entire hole. After injection, the curing time is controlled at 24-48 hours. Once the material is fully cured, the optical fiber and the surrounding rock of the hole wall form a tight and stable "rock-optical fiber co-deformation body," laying the structural foundation for the accurate sensing and transmission of subsequent vibration signals.

[0027] Example 4 Based on Example 3 above, in step 2 of the DAS-based in-situ monitoring method for dynamic development of overburden separation in mining activities, the sampling frequency of the DAS demodulator is 1kHz~10kHz. The DAS demodulator performs data preprocessing, including filtering downhole high-frequency random noise and periodic mechanical interference contained in the data.

[0028] High-frequency random noise in the well is filtered out by a low-pass analog filter with a preset cutoff frequency to remove high-frequency electromagnetic interference generated by downhole motors and cables, thus preventing high-frequency noise from entering the subsequent digital processing stage and causing signal "contamination". At the same time, the filter adopts a wide dynamic range design to ensure that the low-frequency vibration signal of the overburden is completely preserved and is not over-filtered and attenuated.

[0029] The periodic mechanical interference is addressed using an adaptive Notch filtering algorithm. This algorithm identifies the characteristic frequency of the noise in real time and dynamically generates a frequency notch channel to filter out the periodic interference in that frequency band. Compared to complex filtering algorithms, this method has a small computational load and a fast response time, and can complete the processing in milliseconds. This meets the downhole "real-time monitoring" requirements while avoiding the misfiltering of effective vibration signals from the overburden.

[0030] After two stages of preliminary filtering, redundant interference in the electrical signal is significantly reduced, leaving only the effective signal directly related to overburden vibration.

[0031] Then, based on the filtered data, a spatiotemporal matrix data is generated, which includes the vibration signal amplitude, signal propagation time, and three-dimensional spatial location information of each monitoring point on the distributed vibration optical fiber.

[0032] Specifically, the generation of spatiotemporal matrix data is the core step in associating discrete vibration signals with the "time-space" dimension. By integrating multi-dimensional information, it achieves precise recording of "when, where, and how intense the vibration occurs." First, the demodulator discretely samples the filtered electrical signal at a preset sampling frequency. By calculating the voltage / current peak value at each sampling point, it is converted into the "vibration signal amplitude," which characterizes the vibration intensity. Then, based on the transmission speed and return time difference of the optical signal in the optical fiber, the transmission time from the vibration signal's generation location on the sensing optical fiber to the demodulator is calculated to help verify the accuracy of the spatial location. Combining the pre-calibrated data of the monitoring borehole with the spatial resolution of the optical fiber itself, the sensing optical fiber is divided into several continuous monitoring units, each unit corresponding to a unique "three-dimensional spatial coordinate," ensuring that the vibration signal can be accurately located to the specific stratum of the overburden.

[0033] Finally, a two-dimensional spatiotemporal matrix is ​​constructed using the "time dimension" and "spatial location dimension" as the rows and columns of the matrix: each row of the matrix corresponds to a sampling time, and each column corresponds to a monitoring unit on the sensing fiber; each element in the matrix integrates the "vibration signal amplitude" and "propagation time" of that time and that monitoring unit to form a structured data set with a one-to-one correspondence of "time-space-vibration characteristics", that is, complete spatiotemporal matrix data.

[0034] Example 5 Based on Example 4 above, the characteristic parameters of overburden separation extracted in step 3 of the DAS-based in-situ monitoring method for dynamic development of overburden separation in this invention include the energy value of the vibration signal, the dominant frequency and the rate of frequency change, and the time-spectrum characteristics.

[0035] Vibration signal energy value is a core indicator for measuring the intensity of overburden vibration. Vibration is weak during the delamination initiation stage, strengthens during the development stage, and increases sharply during the abrupt change stage. Its extraction revolves around "time domain signal amplitude conversion + time window integration": 1. Retrieve the hourly vibration signal amplitude of each monitoring unit from the spatiotemporal matrix data, and first perform "baseline calibration" on the amplitude - deduct the DC component of the signal to ensure that the energy calculation only reflects the dynamic vibration of the overburden, rather than the static environmental influence.

[0036] 2. Using a "sliding time window," the vibration signal amplitude within each window is processed as follows: The amplitude squared at each sampling point within the window is calculated, the squared amplitudes are summed, and then multiplied by the sampling interval to obtain the "instantaneous vibration energy value" within that time window. The energy values ​​of the continuous window are then smoothed to eliminate energy fluctuations caused by accidental pulse interference, ensuring that the energy value can stably reflect the continuous activity of the overlying strata, thus obtaining the vibration signal energy value.

[0037] The dominant frequency reflects the "main frequency component of the vibration", and the frequency change rate reflects the "dynamic evolution speed of the frequency". Both are extracted through "frequency domain conversion + dynamic tracking".

[0038] Specifically, the extraction of dominant frequencies requires first performing a fast Fourier transform on the preprocessed time-domain signal to convert it into a frequency-domain power spectral density map. In the power spectral density map, the frequency corresponding to the peak of the power spectrum is located. This frequency is the dominant frequency within the time window, representing the main frequency component of the overburden vibration. Subsequently, it is necessary to compare the dominant frequencies of adjacent monitoring units to verify spatial consistency and eliminate outliers caused by local interference.

[0039] To extract the frequency change rate, the dominant frequencies of the same monitoring unit are first sequentially connected in time to form a continuous "time-dominant frequency" sequence. Then, the difference between the dominant frequencies in two consecutive time windows is calculated, and the difference is divided by the time interval between the two windows to obtain the frequency change rate for that period. This reflects the dynamic evolution speed of the dominant frequency and helps to determine the acceleration or deceleration trend of ablation development.

[0040] For non-stationary vibration signals of overburden delamination, time-spectrum feature extraction prioritizes the use of short-time Fourier transform: the preprocessed vibration signal is divided into multiple overlapping short-time windows, and a fast Fourier transform is performed on each window separately to generate a three-dimensional time-frequency-power spectrum. Core features are extracted from the spectrum, including the distribution range of energy-concentrated frequency bands, the migration law of frequency bands over time, and the time nodes and amplitudes of frequency jumps. In complex scenarios such as abrupt delamination, wavelet transform (using the db4 wavelet basis) is added to further improve the time-frequency resolution, accurately capture the details of instantaneous frequency changes, and fully characterize the dynamic features of overburden vibration in the spatiotemporal dimension.

[0041] Furthermore, warnings are issued via audible and visual alarms, SMS notifications, or pop-up windows on the terminal.

[0042] Furthermore, the distributed vibration optical fiber is a mining-grade flame-retardant single-mode optical fiber with a diameter of 900μm~1250μm and possesses acid and alkali corrosion resistance.

[0043] Example 6 Based on the above embodiment 5, the overburden delamination development stage identification model in step 4 of the DAS-based in-situ monitoring method for dynamic development of overburden delamination in mining is constructed based on the SVM model. During model training, historical delamination data of mines with the same geological conditions and on-site calibration data of the current working face are collected. After labeling the stage, the data is standardized and outliers are removed to form a dataset. Outlier removal employs the 3σ principle to filter feature parameters, calculating the mean μ and standard deviation σ for each feature class, and removing samples that exceed the range [μ-3σ, μ+3σ]. After outlier removal, Z-score standardization is used to unify the dimensions of feature parameters, ensuring that the weights of each feature are balanced during model training.

[0044] The preprocessed dataset was divided into a training set and a validation set in a 7:3 ratio. The model was trained using a linear kernel as the kernel function, the optimal C value as the penalty parameter, and a sequence minimization optimization algorithm. The overlying stratification development stage identification model divides the development into three stages: generation stage, development stage, and abrupt change stage. The generation period is defined as when the vibration energy is low and stable, the dominant frequency is in the low frequency range and the frequency change rate is slow, and the time spectrum energy is concentrated in the low frequency and shows a single peak. The development stage is defined as the period when the vibration energy exceeds the initial threshold and continues to rise, the dominant frequency rises to the mid-frequency band and the rate of frequency change accelerates, and the time-spectrum energy migrates to the mid-frequency band and the peaks of the two frequency bands overlap. The period of sudden change is defined as when the vibration energy exceeds the warning threshold and increases sharply, the dominant frequency jumps to the high frequency band and the frequency change rate increases sharply, and the time spectrum energy is instantly concentrated at the high frequency and presents a single peak.

[0045] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0046] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for in-situ monitoring of dynamic development of overburden delamination based on DAS, characterized in that, Includes the following steps: Step 1: Collect data from the mining face, drill monitoring holes based on the data, insert distributed vibration optical fibers into the monitoring holes, and use coupling materials to couple and solidify the optical fibers with the hole walls. Step 2: The distributed vibration fiber transmits the sensed vibration signal to the DAS demodulator, and after the DAS demodulator preprocesses the data, it is transmitted to the ground DAS monitoring terminal. Step 3: The ground-based DAS monitoring terminal extracts the characteristic parameters of the overburden delamination from the data; Step 4: Construct a model for identifying the development stage of overburden separation, with the goal of dividing the separation layer into different stages. Based on the extracted overburden separation feature parameters, the model is used to divide the overburden separation into different stages and compare the thresholds. If the threshold is exceeded, an early warning is issued.

2. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, The data collected in step 1 for the longwall face includes the distribution and thickness parameters of the overburden lithology, parameters of the key overburden layers, dip angle, strike, and dip direction of the overburden; strike length, dip length, coal seam thickness, mining speed, and advance direction of the longwall face.

3. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, In step 1, the monitoring holes are set to the side or rear of the working face in the direction of advancement to avoid the interference zone between the aquifer and the existing project; when the rock strata are horizontal or the dip angle is less than 15°, the monitoring holes are started vertically; otherwise, the angle of the monitoring holes is parallel to the dip angle of the rock strata or at an angle of 5° to 10°. The drilling depth of the monitoring hole is to penetrate the uppermost key layer and extend to 5m~10m below the depth affected by mining.

4. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, The coupling material in step 1 is an epoxy resin-based composite coupling material with a compressive strength ≥30MPa and a coefficient of linear expansion ≤15×10⁻⁶. -6 / ℃, acoustic impedance matching degree with surrounding rock ≥90%; during coupling curing, a segmented injection method is adopted, gradually filling from the bottom of the hole to the opening to ensure that the coupling material uniformly fills the tiny gaps between the optical fiber and the hole wall, and the curing time is 24h~48h.

5. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, In step 2, the sampling frequency of the DAS demodulator is 1kHz~10kHz. The DAS demodulator preprocesses the data, including filtering the downhole high-frequency random noise and periodic mechanical interference contained in the data. The downhole high-frequency random noise is filtered out by the low-pass analog filter with a preset cutoff frequency to remove the high-frequency electromagnetic interference generated by downhole motors and cables. The periodic mechanical interference is addressed using an adaptive Notch filtering algorithm, which dynamically generates a frequency notch channel by identifying the characteristic frequency of the noise in real time, thereby filtering out the periodic interference in that frequency band. Then, based on the filtered data, a spatiotemporal matrix data is generated, which includes the vibration signal amplitude, signal propagation time, and three-dimensional spatial location information of each monitoring point on the distributed vibration optical fiber.

6. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, The overlying delamination characteristic parameters extracted in step 3 include the energy value of the vibration signal, the dominant frequency and the rate of frequency change, and the time-spectral characteristics.

7. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 6, characterized in that, In step 4, the overburden delamination development stage identification model is constructed based on the SVM model. During model training, historical delamination data of mines with the same geological conditions and on-site calibration data of the current working face are collected. After labeling the stage, the data is standardized and outliers are removed to form a dataset. The preprocessed dataset is divided into a training set and a validation set in a 7:3 ratio. The model is trained using a linear kernel as the kernel function, the optimal C value as the penalty parameter, and a sequence minimum optimization algorithm. The overlying stratification development stage identification model divides the development into three stages: generation stage, development stage, and abrupt change stage. The generation period is defined as when the vibration energy is low and stable, the dominant frequency is in the low frequency range and the frequency change rate is slow, and the time spectrum energy is concentrated in the low frequency and shows a single peak. The development stage is defined as the period when the vibration energy exceeds the initial threshold and continues to rise, the dominant frequency rises to the mid-frequency band and the rate of frequency change accelerates, and the time-spectrum energy migrates to the mid-frequency band and the peaks of the two frequency bands overlap. The period of sudden change is defined as when the vibration energy exceeds the warning threshold and increases sharply, the dominant frequency jumps to the high frequency band and the frequency change rate increases sharply, and the time spectrum energy is instantly concentrated at the high frequency and presents a single peak.

8. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, The warning is issued via audible and visual alarms, SMS push notifications, or terminal pop-ups.

9. The in-situ monitoring method for dynamic development of mining-induced overburden delamination based on DAS according to claim 1, characterized in that, The distributed vibration optical fiber is a mining flame-retardant single-mode optical fiber with a diameter of 900μm~1250μm and has acid and alkali corrosion resistance.

Images