A coal seam thickness real-time inversion method based on seismic dispersion characteristics with excavation

Abstract

The coal seam thickness real-time inversion method based on the seismic dispersion characteristics of the excavation in the application uses the excavator source, adopts the advantage frequency band adaptive processing, combines the dynamic inversion model, systematically solves the problems of stopping excavation construction, static lag, high frequency dependence and offline processing in the traditional channel wave technology, and improves the economy, real-time, adaptability and accuracy. The method reduces the loss of downtime, reduces the equipment cost, realizes the update of the minute-level geological data, solves the problem of low-frequency random signal processing, and can control the dynamic inversion error within the meter level. The effective signal energy of the underground seismic data in the coal mine is low, the signal-to-noise ratio is poor, and it is difficult to accurately pick up the first arrival. The data used for the conventional channel wave inversion coal thickness is the static data of one-time acquisition, and the data of the excavation is the dynamic data of continuous reception, which can continuously correct the group velocity value and improve the accuracy of the group velocity, so that the coal seam thickness obtained is more accurate.

Description

Technical Field

[0001] This invention belongs to the field of geophysical exploration technology, specifically relating to a real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation. Background Technology

[0002] Channel waves are guided waves formed when seismic waves propagate through low-velocity interlayers in coal seams and their roof and floor. Their dispersion characteristics (the dependence of phase velocity on frequency) are closely related to the coal seam thickness. Therefore, channel waves are a relatively reliable exploration method for detecting coal thickness. Traditional channel wave exploration uses artificial seismic sources (such as explosives or controlled seismic sources) to generate seismic waves, and uses a sensor array deployed in the roadway to receive the signals, combining the dispersion curves to invert the coal thickness. This method has mature applications in coal mine geological exploration, but it requires pausing tunneling operations during construction, affecting mining efficiency. Furthermore, channel wave detection technology can only obtain static coal thickness information and cannot dynamically monitor real-time changes in the coal seam ahead during tunneling; the dedicated seismic source equipment and complex construction process increase economic costs. Current patents and literature on channel wave coal thickness inversion mostly focus on static scenarios, with the core being the generation of high-frequency channel waves through active seismic sources and the establishment of a dispersion curve database. Earthquake detection technology during tunneling uses the vibration generated by the cutting teeth of the tunneling machine cutting the coal seam as the seismic source. It achieves advanced geological exploration by collecting vibration signals in real time. It has the natural advantage of "no need to stop tunneling" and is also a means of dynamic monitoring.

[0003] However, existing coal seam thickness inversion models cannot be directly applied to underground tunneling scenarios because the active seismic source spectrum is controllable, while the seismic source frequency band is random and the energy is concentrated in the low-frequency band. Traditional inversion models rely on static calibration, which makes it difficult to adapt to the continuous changes in coal seam thickness during tunneling. At the same time, existing systems rely on ground workstations for processing and cannot be integrated with underground tunneling machines. Summary of the Invention

[0004] The purpose of this invention is to provide a real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation, so as to solve the problem that existing methods are unable to dynamically reflect the continuous changes in coal seam thickness.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0006] A real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation includes the following steps: Step 1: Deploy the observation system at the working face of the target area, preset multiple drilling cycles at fixed intervals, and synchronously collect seismic data during each drilling cycle; Command the tunneling machine to perform drilling operations until the first drilling cycle is completed; The observation system includes multiple acquisition stations, and each acquisition station includes multiple detectors. The seismic wave amplitudes collected by any detector at all sampling times within a drilling cycle are recorded as a seismic trace; The seismic gathers detected by all geophones form the seismic data acquired during this drilling cycle. Step 2: Preprocess the seismic data acquired during this drilling cycle, remove distorted and invalid seismic traces, and use the seismic trace set of all preprocessed seismic traces as valid seismic records. Step 3: For the reflected waves in the effective seismic records, the reverse time migration imaging algorithm is used to perform reverse time imaging, and the location with the highest energy in the reverse time imaging result is determined as the location of the tunnel boring machine. Step 4: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is filtered multiple times, and the filtered frequencies are subjected to inverse Fourier transform to obtain the corresponding time signal, i.e., the time spectrum of the effective seismic record. Step 5: Based on the relative position of each detector to the tunneling machine, convert the time spectrum into a dispersion spectrum; Step 6: Perform frequency division processing on the dispersion spectrum to obtain the dominant frequency band; Bandpass filtering of the excavation seismic data was performed using the dominant frequency band, and the group velocity was obtained by tomography of the filtered results, which was denoted as the reference group velocity. Step 7: Based on the coal seam thickness of the area exposed in this drilling cycle, use a multinomial fitting method to establish the relationship between coal seam thickness and group velocity. The thickness of the coal seam in the area not exposed during this drilling cycle was calculated based on the relationship between coal seam thickness and group velocity. Step 8: Continue drilling with the tunneling machine until the next drilling cycle is completed. Repeat steps 1-7, and use the group velocity obtained in this drilling cycle to correct the baseline group velocity, as shown in the following formula:

[0007] in, This represents the corrected reference group velocity, in m / s; Represents the reference group velocity, in m / s; This represents the group velocity obtained in this drilling cycle, in m / s; The corrected group velocity is redefined as the reference group velocity. The thickness of the coal seam in the area not exposed during this drilling cycle was calculated using the benchmark group velocity. Step 9: Repeat steps 1-8 until the tunneling machine completes all drilling cycles.

[0008] The present invention also has the following features: Furthermore, in step 2, the valid seismic records are filtered using a notch filter to remove 50Hz power frequency interference.

[0009] Furthermore, step 3 includes the following steps: Multiple possible tunneling machine locations are defined within the target area. Reverse-time migration imaging is used to perform reverse-time imaging on the reflected waves from the valid seismic records to obtain the energy value at each possible tunneling machine location, as shown in the following formula:

[0010] in, Indicates any possible location of the tunneling machine in the target area. The energy value at that point is dimensionless. Indicates the half-length of the hour window, in seconds; This indicates the epicenter obtained through ray tracing. To the The travel time to each receiving point is s; Indicates the total number of detectors; Indicates the first The reflected wave deflection field of each detector; This represents a local time variable within an hourly window. ; To prevent the minimum value where the denominator is 0; The point corresponding to the maximum energy value is selected as the location of the tunneling machine, i.e., the location of the seismic source.

[0011] Furthermore, step 4 includes the following sub-steps: Step 41: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is a two-dimensional matrix, with rows representing seismic traces, columns representing frequencies, and elements representing amplitude values. Step 42: For the frequency domain spectrum, perform multiple filtering operations using narrowband filters with different center frequencies; Step 43: Perform an inverse Fourier transform on the frequency domain spectrum after multiple filterings to obtain the corresponding time spectrum; The time spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent time, and the elements in the matrix represent the energy amplitude.

[0012] Further, step 5 uses the following formula to convert the time spectrum into a dispersion spectrum;

[0013] in, The dispersion spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent the group velocity, and each element in the matrix represents the energy amplitude. Represents group velocity, in m / s; Represents the time spectrum; Indicates the location of the tunneling machine; Indicates the detector position; The time, in seconds, is the time it takes for seismic waves to travel from the location of the tunnel boring machine to the detector. Furthermore, in step 6, the dispersion spectrum is sequentially divided into 14 center frequencies ranging from 10 to 75 Hz with an interval of 5 Hz to obtain multiple single-frequency data volumes in different frequency bands; the single-frequency data volume with the highest resolution is selected from the multiple single-frequency data volumes as the dominant frequency band. Bandpass filtering of the excavation seismic data was performed using the dominant frequency band, and the group velocity was obtained by tomographic imaging of the filtered results.

[0014] Compared with the prior art, the present invention has the following technical effects: This invention presents a real-time coal seam thickness inversion method based on the dispersion characteristics of seismic data obtained during tunneling. Utilizing the tunneling machine's seismic source and employing adaptive processing with advantageous frequency bands, combined with a dynamic inversion model, it systematically solves the problems of construction stoppage, static lag, high-frequency dependence, and offline processing inherent in traditional channel wave technology. This method achieves improvements in economy, real-time performance, adaptability, and accuracy. It reduces downtime losses, lowers equipment costs, achieves minute-level geological data updates, overcomes the challenge of low-frequency random signal processing, and can control dynamic inversion errors within the meter range. Underground coal mine seismic data has low effective signal energy and poor signal-to-noise ratio, making it difficult to accurately pick up the first arrival. This technology avoids the conventional first arrival picking step by obtaining the dispersion spectrum from the original seismic data obtained during tunneling. Conventional channel wave inversion of coal thickness uses static data acquired only once, while tunneling data is continuously received dynamic data, which can continuously correct group velocity values ​​and improve the accuracy of group velocity, thus obtaining a more accurate coal seam thickness. Attached Figure Description

[0015] Figure 1This is a diagram of a fiber optic observation system in a seismic borehole during excavation, according to one embodiment of the present invention. Figure 2 This is a coal thickness distribution diagram of the working face in one embodiment of the invention. Detailed Implementation

[0016] It should be noted that, unless otherwise specified, all components in this invention are components known in the prior art.

[0017] The following are specific embodiments of the present invention. It should be noted that the present invention is not limited to the following specific embodiments. All equivalent modifications made based on the technical solutions of this application fall within the protection scope of the present invention.

[0018] A real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation includes the following steps: Step 1, as follows Figure 1 As shown, an observation system is deployed at the working face in the target area to collect seismic data during one drilling cycle. The observation system includes multiple acquisition stations, and each acquisition station includes multiple detectors; The seismic wave amplitudes collected by any detector at all sampling times within a drilling cycle are recorded as a seismic trace; The seismic gathers detected by all geophones form the seismic data acquired during this drilling cycle. Step 2: Preprocess the seismic data during excavation, remove distorted and invalid seismic traces, and use the seismic trace set of all preprocessed seismic traces as valid seismic records. It should be noted that distorted and invalid seismic traces are adaptively determined by those skilled in the art based on the actual working scenario. This embodiment does not impose limitations, but several typical scenarios are given: Invalid seismic traces are those that cannot provide any valid signal at all, including: Seismic traces with zero recorded values ​​or extremely low noise levels and no waveform changes due to component damage, such as sensor failure or fiber optic cable breakage, do not contain stratigraphic information.

[0019] The signal-to-noise ratio is extremely low, the effective signal is overwhelmed by noise, the waveform is chaotic and irregular, and the seismic trace is invalid.

[0020] Distorted seismic traces record signals, but the signals are distorted and cannot reflect the true wavefield. In this field, they typically include seismic traces with flattened tops or bottoms, exhibiting a flat-top or flat-bottom shape; seismic traces with waveforms similar to other adjacent seismic traces but with completely opposite polarities; and seismic traces with waveforms similar to normal traces but with abnormally small amplitudes or abnormal low-frequency oscillations. Such seismic traces also need to be discarded.

[0021] Those skilled in the art can manually identify and eliminate such cases, or they can automatically eliminate them using conventional methods in the field.

[0022] In this step, a notch filter is used to filter and remove 50Hz power frequency interference. The filtered seismic traces were then compressed to 20dB.

[0023] Step 3: For the reflected waves in the effective seismic records, the reverse time migration imaging algorithm is used to perform reverse time imaging, and the location with the highest energy in the reverse time imaging result is determined as the location of the tunnel boring machine. This step specifically includes the following steps: Multiple possible tunneling machine locations are set within the target area; Possible locations for the tunneling machine shall be determined by those skilled in the art based on the actual working conditions; For the reflected waves in the valid seismic records, a reverse time migration imaging algorithm is used to perform reverse time imaging to obtain the energy value at each possible tunneling machine location, as shown in the following formula:

[0024] in, Indicates any possible location of the tunneling machine in the target area. The energy value at that point is dimensionless. Indicates the half-length of the hour window, in seconds; This indicates the epicenter obtained through ray tracing. To the The travel time to each receiving point is s; Indicates the total number of detectors; Indicates the first The reflected wave deflection field of each detector; This represents a local time variable within an hourly window. ; To prevent the minimum value where the denominator is 0; The point corresponding to the maximum energy value is selected as the location of the tunneling machine, i.e., the location of the seismic source.

[0025] Step 4: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is filtered multiple times, and the filtered frequencies are subjected to inverse Fourier transform to obtain the corresponding time signal, i.e., the time spectrum of the effective seismic record. Step 4 includes the following sub-steps: Step 41: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is a two-dimensional matrix, with rows representing seismic traces, columns representing frequencies, and elements representing amplitude values. Step 42, let Distance from the epicenter x Earthquake records observed at the location x(t) The Fourier transform results, using different center frequencies ( i Narrowband filter (=1,2,3,…) right Perform multiple filtering operations; The narrowband filter is shown in the following equation:

[0026] in, Here is the mathematical representation of a narrowband filter, where As the frequency independent variable, The center frequency; where The value is in the range [ f min , f max ] Above △ f The first of the equally spaced divisions i There are N discrete values ​​and a total of Nf discrete frequencies.

[0027] Indicates the first The center frequency of a narrowband filter, in Hz; This represents any narrowband filter; Represents the relative width of a narrowband filter; it is dimensionless. The sharpness parameter representing the peak value of the Gaussian function is dimensionless. Represents the Gaussian function; After multiple filtering operations by a narrowband filter, a series of output spectra are obtained:

[0028] Step 43: Perform an inverse Fourier transform on the frequency domain spectrum after multiple filterings to obtain the corresponding time spectrum; The time spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent time, and the elements in the matrix represent the energy amplitude.

[0029] Step 5: Based on the relative position of each detector to the tunneling machine, convert the time spectrum into a dispersion spectrum; Specifically, the time spectrum is converted into a dispersion spectrum using the following formula;

[0030] in, The dispersion spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent the group velocity, and each element in the matrix represents the energy amplitude. Represents group velocity, in m / s; Represents the time spectrum; Indicates the location of the tunneling machine; Indicates the detector position; This indicates the time it takes for seismic waves to travel from the location of the tunnel boring machine to the geophone. Step 6: Perform frequency division processing on the dispersion spectrum to obtain the dominant frequency band; Bandpass filtering of the excavation seismic data was performed using the dominant frequency band, and group velocity tomography was performed on the filtered results. The dispersion spectrum was sequentially divided into 14 center frequencies ranging from 10 to 75 Hz with a 5 Hz interval to obtain a series of single-frequency data volumes in different frequency bands. The single-frequency data volume with the strongest energy and highest resolution was selected from the multiple single-frequency data volumes as the dominant frequency band. It should be noted that the above is only one implementation method, and the actual implementation in the field is not limited to this scheme. Those skilled in the art can select the advantageous frequency bands that reflect changes in coal seam thickness better, have relatively strong energy, and have high resolution.

[0031] The dominant frequency band was selected to perform bandpass filtering on the seismic data during excavation, and the filtered results were used for group velocity tomography.

[0032] Among them, selecting the dominant frequency band to perform bandpass filtering on the seismic data during excavation and performing group velocity tomography on the filtered results are routine operations in this field. Specifically, the dominant frequency band selected after frequency division of the dispersion spectrum is used as the passband range of the bandpass filter for bandpass filtering, so as to retain the frequency band components most sensitive to changes in coal seam thickness and suppress noise and interference in other frequency bands. Then, based on the filtered seismic data, the group velocity value at each location within the target area is retrieved using a tomographic imaging algorithm, ultimately obtaining the group velocity. In this field, group velocity refers to the spatial distribution of group velocity values.

[0033] The group velocity can be expressed as follows:

[0034] in, This represents the group velocity, which is the group velocity value at each location within the target area. Rows of a matrix Ng This indicates the number of detectors, and the locations of multiple detectors represent multiple locations within the target area. Columns of a matrix N This represents the number of sampling points for the velocity of each detector group; The elements in the matrix represent the group velocity value at any sampling point of any detector.

[0035] Step 7: Based on the coal seam thickness revealed in this drilling cycle, establish the relationship between coal seam thickness and group velocity using a multinomial fitting method. Polynomial fitting is a commonly used regression method for establishing the relationship between coal seam thickness h and group velocity v. Its basic idea is to approximate the functional relationship between the two using an nth-degree polynomial, which can better describe the possible tortuous relationship between coal thickness and group velocity. The thickness of the coal seam in the area not exposed during this drilling cycle was calculated based on the relationship between coal seam thickness and group velocity. Step 8: When the next drilling cycle is completed, repeat steps 1-7 and use the group velocity obtained this time to correct the group velocity of the previous drilling cycle. In this embodiment, group velocity is the speed at which the overall energy of a group of mixed-frequency waves propagates forward. In a dispersive medium, waves of different frequencies propagate at different speeds, and group velocity is a function of frequency. The coal seam thickness is obtained by inverting the dispersion curve.

[0036] For the same location, if the coal seam is undisturbed, the theoretical true value of the group velocity is fixed. However, since the seismic source is random and the noise environment is changing, the group velocity value calculated independently each time will have random errors. By correcting the group velocity, we can both preserve historical information (stable areas) and respond quickly to new changes.

[0037] Repeat step 8 until the drilling operation is complete.

Claims

1. A method for real-time inversion of coal seam thickness based on seismic dispersion characteristics during excavation, characterized in that, Includes the following steps: Step 1: Deploy the observation system at the working face of the target area, preset multiple drilling cycles at fixed intervals, and synchronously collect seismic data during each drilling cycle; Command the tunneling machine to perform drilling operations until the first drilling cycle is completed; The observation system includes multiple acquisition stations, and each acquisition station includes multiple detectors. The seismic wave amplitudes collected by any detector at all sampling times within a drilling cycle are recorded as a seismic trace; The seismic gathers detected by all geophones form the seismic data acquired during this drilling cycle. Step 2: Preprocess the seismic data acquired during this drilling cycle, remove distorted and invalid seismic traces, and use the seismic trace set of all preprocessed seismic traces as valid seismic records. Step 3: For the reflected waves in the effective seismic records, the reverse time migration imaging algorithm is used to perform reverse time imaging, and the location with the highest energy in the reverse time imaging result is determined as the location of the tunnel boring machine. Step 4: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is filtered multiple times, and the filtered frequencies are subjected to inverse Fourier transform to obtain the corresponding time signal, i.e., the time spectrum of the effective seismic record. Step 5: Based on the relative position of each detector to the tunneling machine, convert the time spectrum into a dispersion spectrum; Step 6: Perform frequency division processing on the dispersion spectrum to obtain the dominant frequency band; Bandpass filtering of the excavation seismic data was performed using the dominant frequency band, and the group velocity was obtained by tomography of the filtered results, which was denoted as the reference group velocity. Step 7: Based on the coal seam thickness of the area exposed in this drilling cycle, use a multinomial fitting method to establish the relationship between coal seam thickness and group velocity. The thickness of the coal seam in the area not exposed during this drilling cycle was calculated based on the relationship between coal seam thickness and group velocity. Step 8: Continue drilling with the tunneling machine until the next drilling cycle is completed. Repeat steps 1-7, and use the group velocity obtained in this drilling cycle to correct the baseline group velocity, as shown in the following formula: in, This represents the corrected reference group velocity, in m / s; Represents the reference group velocity, in m / s; This represents the group velocity obtained in this drilling cycle, in m / s; The corrected group velocity is redefined as the reference group velocity. The thickness of the coal seam in the area not exposed during this drilling cycle was calculated using the benchmark group velocity. Step 9: Repeat steps 1-8 until the tunneling machine completes all drilling cycles.

2. The real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation as described in claim 1, characterized in that, In step 2, the valid seismic records are filtered using a notch filter to remove 50Hz power frequency interference.

3. The real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation as described in claim 2, characterized in that, Step 3 includes the following steps: Multiple possible tunneling machine locations are defined within the target area. Reverse-time migration imaging is used to perform reverse-time imaging on the reflected waves from the valid seismic records to obtain the energy value at each possible tunneling machine location, as shown in the following formula: in, Indicates any possible location of the tunneling machine in the target area. The energy value at that point is dimensionless. Indicates the half-length of the hour window, in seconds; This indicates the epicenter obtained through ray tracing. To the The travel time to each receiving point is s; Indicates the total number of detectors; Indicates the first The reflected wave deflection field of each detector; This represents a local time variable within an hourly window. ; To prevent the minimum value where the denominator is 0; The point corresponding to the maximum energy value is selected as the location of the tunneling machine, i.e., the location of the seismic source.

4. The real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation as described in claim 3, characterized in that, Step 4 includes the following sub-steps: Step 41: Perform a Fourier transform on each seismic trace in the valid seismic record to transform the time-domain signal to the frequency domain and obtain the corresponding frequency domain spectrum. The frequency domain spectrum is a two-dimensional matrix, with rows representing seismic traces, columns representing frequencies, and elements representing amplitude values. Step 42: For the frequency domain spectrum, perform multiple filtering operations using narrowband filters with different center frequencies; Step 43: Perform an inverse Fourier transform on the frequency domain spectrum after multiple filtering to obtain the corresponding time spectrum; The time spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent time, and the elements in the matrix represent the energy amplitude.

5. The real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation as described in claim 4, characterized in that, Step 5 uses the following formula to convert the time spectrum into a dispersion spectrum; in, The dispersion spectrum is a two-dimensional matrix, where the rows represent the center frequency, the columns represent the group velocity, and each element in the matrix represents the energy amplitude. Represents group velocity, in m / s; Represents the time spectrum; Indicates the location of the tunneling machine; Indicates the detector position; The time, in seconds, represents the time it takes for a seismic wave to travel from the location of the tunnel boring machine to the detector.

6. The real-time inversion method for coal seam thickness based on seismic dispersion characteristics during excavation as described in claim 5, characterized in that, In step 6, the dispersion spectrum is divided into 14 center frequencies ranging from 10 to 75 Hz with an interval of 5 Hz to obtain multiple single-frequency data volumes in different frequency bands; the single-frequency data volume with the highest resolution is selected as the dominant frequency band from the multiple single-frequency data volumes. Bandpass filtering of the excavation seismic data was performed using the dominant frequency band, and the group velocity was obtained by tomographic imaging of the filtered results.

Images