A Wavelet Domain-Based Slot Wave Attenuation Imaging Method

Abstract

This invention discloses a wavelet domain-based channel wave attenuation imaging method. The specific process is as follows: S1, Transmission channel wave detection: A shot is fired at one end of the roadway on one side of the working face, and a detector is deployed in the roadway on the other side of the working face to receive the channel wave; S2, Wavelet domain attenuation imaging inversion: The attenuation coefficients of the model are transformed to the wavelet domain to obtain the attenuation coefficients of different grids within the working face; S3, Based on the attenuation coefficients of different grids within the working face obtained in step S2, an attenuation map of the working face is output. Using this invention, the model parameters of traditional uniform grid model inversion systems can be reduced, improving the reliability of the inversion results. Simultaneously, it enhances the imaging and identification capabilities of geological anomalies with abrupt attribute changes, thereby improving the accuracy of structural detection within the working face and providing direct technical support for working face mining.

Description

Technical Field

[0001] This invention relates to the field of imaging technology for detecting the internal structure of coal mine working faces, and specifically to a groove wave attenuation imaging method based on the wavelet domain. Background Technology

[0002] Seismic waves excited within coal seams interfere with each other to form channel waves. These channel wave signals are affected by discontinuous geological bodies along their propagation path, such as river scour zones, coal seam thinning zones, faults, collapse columns, variations in coal seam thickness, gangue distribution, and erosion zones, causing changes in their velocity, amplitude, phase, and frequency. The high energy, ease of identification, and significant dispersion effect of channel waves allow for the detection of geological structures and internal anomalies within coal seams (Liu Gang, 2022). Channel wave seismic exploration is an effective method for mine structural detection (Dang Baoquan et al., 2022), enabling the detection of coal seam discontinuities by utilizing channel waves excited and propagating within the coal seam.

[0003] The main imaging methods for amplitude energy attenuation of coal mine channel waves include four frequency domain methods: spectral ratio method, centroid frequency shift method, rise time method, and amplitude attenuation method. Among them, the rise time method and amplitude attenuation method belong to the time domain; the centroid frequency shift method and spectral ratio method belong to the frequency domain. However, current imaging methods for different attenuation types all use direct inversion methods, which have problems such as large number of inversion model parameters, large condition number of inversion matrix, and underdetermined inversion system (Yang Yanjun, 2022). In addition, when encountering abrupt changes in the internal structure of the working face (such as faults and collapse columns), there is a problem that it is impossible to characterize the boundary of abrupt changes in abnormal structures.

[0004] The wavelet analysis multi-scale decomposition method for inverting slot wave attenuation coefficients transforms the direct solution of attenuation coefficients into an inversion problem of wavelet coefficients in the wavelet domain. Utilizing the inherent connections between wavelet scales and the sparsity of wavelet coefficients, it effectively avoids the objective function falling into local extrema and the problem of high computational cost. Multi-scale inversion starts at a large scale of wavelet decomposition. At larger scales, the objective function exhibits stronger convexity and fewer local extrema under the model grid, which is beneficial for convergence to the global optimum. Multi-scale decomposition theory based on different wavelet transforms is applied to traditional model parameters. During the inversion process, the model is automatically divided into multiple scales according to the resolution limit of the data. This method serves to partition the grid at different scales. Iterative inversion is first performed on a large-scale grid to obtain a stable, coarse model. This initial inversion result is then used as the initial model for inversion on a smaller-scale grid until the global optimum of the objective function is found. The advantages of this inversion strategy are reduced dependence on the initial model selection, greater suitability for imaging large-disturbance anomalies, and effective improvement in imaging resolution and reliability (Gao Jian et al., 2016). Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention aims to provide a wavelet domain-based groove attenuation imaging method. By applying wavelet multi-scale decomposition to groove attenuation imaging in coal mine working faces, this method is of great significance in reducing the underdeterminacy problem of the inversion system and improving the imaging of abrupt anomalies.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A wavelet domain-based groove attenuation imaging method, the specific process of which is as follows:

[0008] S1. Transmission slot wave detection: The blasting point is activated in the roadway on one side of the working face, and the slot wave is received by the detectors deployed in the roadway on the other side of the working face.

[0009] S2. Wavelet Domain Attenuation Imaging Inversion: Transform the attenuation coefficient of the model to the wavelet domain to obtain the attenuation coefficients of different grids inside the working surface.

[0010] S3. Based on the different mesh attenuation coefficients inside the working surface obtained in step S2, output the working surface attenuation map.

[0011] Furthermore, in step S1, after acquiring slot wave data using a detector, the slot wave data is preprocessed, including bad channel removal, polarity correction, and slot wave extraction.

[0012] Furthermore, the specific process of step S2 is as follows:

[0013] When the slot wave energy generated at the shot point is A0, and the slot wave energy received by the detector is A, then:

[0014] A=A0e -xd (1)

[0015] In the formula, the transmission slot wave attenuation coefficient d is the ray length, Q is the quality factor, f is the slot wave frequency, and v is the phase velocity corresponding to the slot wave frequency f; taking the natural logarithm of both sides of equation (1) yields:

[0016] xd=lnA0-lnA (2)

[0017] When there is a ray y i Passing through attenuation coefficients x1, x2, ..., x n The mesh within the working surface, and the ray lengths of different meshes are d. i1 ,d i2 ,…,d in When the total attenuation along the i-th ray path is:

[0018]

[0019] In the formula, Given the current total length of the ray, the total attenuation of the ray is lnA0 - lnA. j A j Indicates the j-th grid trough wave Energy; by combining the attenuation observation data of all rays, the target inversion matrix can be established:

[0020] Gm=d (4)

[0021] In the formula, d represents the observed data, i.e., the received slot wave energy; m represents the attenuation coefficient of different grids within the working surface; and G represents the sensitivity matrix.

[0022] Using the properties of orthogonal wavelet transform, we have W -1 =W T W T W = I, where I represents the identity matrix, and equation (4) can be written as:

[0023] Δd=GW T WΔm (5)

[0024] Δm is the update amount of the attenuation coefficient, and W is the wavelet transform matrix; take This represents the update amount of the attenuation coefficient in the wavelet domain; The two-dimensional wavelet transform of the sensitivity matrix G, Each row represents the two-dimensional wavelet transform coefficients corresponding to the row in G.

[0025]

[0026] Equation (6) represents the wavelet coefficients of equation (5) at different scales, which means that the original solution of the attenuation coefficient equation is transformed into the solution of the wavelet coefficients in the wavelet domain.

[0027] After obtaining the wavelet coefficients, according to It can be obtained That is, the solution can be obtained. Perform inverse wavelet transform to obtain the update amount Δm of the attenuation coefficient, and then obtain the attenuation coefficient of different grids inside the working surface.

[0028] The beneficial effects of this invention are as follows: by using this invention, the model parameters of the traditional uniform grid model inversion system can be reduced, the reliability of the inversion results can be improved, and the imaging and identification capabilities of geological anomalies with abrupt changes in properties can be improved, thereby improving the accuracy of internal structural detection of the working face and providing direct technical support for working face mining. Attached Figure Description

[0029] Figure 1 This is a flowchart illustrating the method of Embodiment 1 of the present invention;

[0030] Figure 2 This is a schematic diagram of the arrangement of the channel wave observation system in Embodiment 1 of the present invention;

[0031] Figure 3 This is a schematic diagram of the attenuation coefficient model and wavelet coefficient model in Embodiment 2 of the present invention. Detailed Implementation

[0032] The present invention will be further described below with reference to the accompanying drawings. It should be noted that this embodiment is based on the present technical solution and provides detailed implementation methods and specific operation processes, but the protection scope of the present invention is not limited to this embodiment.

[0033] Example 1

[0034] This embodiment provides a wavelet domain-based slot wave attenuation imaging method, such as... Figure 1 As shown, the specific process is as follows:

[0035] S1. Transmission slot wave detection: A blasting point is activated in the roadway on one side of the working face, and the slot wave is received by detectors deployed in the roadway on the other side of the working face. For example... Figure 2 As shown, the firing point in tunnel 2 is the firing point, and the receiving point yi in tunnel 1 is the detector position.

[0036] S2, Wavelet domain attenuation imaging inversion:

[0037] When the slot wave energy generated at the shot point is A0, and the slot wave energy received by the detector is A, then:

[0038] A=A0e -xd (1)

[0039] In the formula, the transmission slot wave attenuation coefficient d is the ray length, Q is the quality factor, f is the slot wave frequency, and v is the phase velocity corresponding to the slot wave frequency f; taking the natural logarithm of both sides of equation (1) yields:

[0040] xd=lnA0-lnA(2)

[0041] Figure 2 The diagram shows a gridded working surface. For example, there is a ray y in the region. i Passing through attenuation coefficients x1, x2, ..., x n The grid is defined, and the ray lengths of different grids are d. i1 d i2 , ..., d in Then the total attenuation along the i-th ray path is:

[0042]

[0043] In the formula, Given the current total length of the ray, the total attenuation of the ray is lnA0 - lnA. j A j Indicates the j-th grid trough wave Energy, combined with attenuation observation data of all rays, allows us to construct the target inversion matrix:

[0044] Gm=d (4)

[0045] In the formula, d represents the observed data, i.e., the received slot wave energy; m represents the attenuation coefficient of different grids within the working surface; and G represents the sensitivity matrix.

[0046] The key to wavelet domain attenuation imaging inversion technology is transforming the model's attenuation coefficients to the wavelet domain. Utilizing the properties of orthogonal wavelet transform, we have W... -1 =W T W T W = I, where I represents the identity matrix, and equation (4) can be written as:

[0047] Δd=GW T WΔm (5)

[0048] Δm is the update amount of the attenuation coefficient, and W is the wavelet transform matrix; take This represents the update amount of the attenuation coefficient in the wavelet domain; The two-dimensional wavelet transform of the sensitivity matrix G, Each row represents the two-dimensional wavelet transform coefficients corresponding to the row in G.

[0049]

[0050] Equation (6) represents the wavelet coefficients of equation (5) at different scales, which means that the original solution of the attenuation coefficient equation is transformed into the solution of the wavelet coefficients in the wavelet domain.

[0051] After obtaining the wavelet coefficients, according to It can be obtained That is, the solution can be obtained. Perform inverse wavelet transform to obtain the update amount Δm of the attenuation coefficient, and then obtain the attenuation coefficient of different grids inside the working surface.

[0052] S3. Based on the different mesh attenuation coefficients inside the working surface obtained in step S2, output the working surface attenuation map.

[0053] Furthermore, in this embodiment, in step S1, after acquiring the slot wave data, the slot wave data is preprocessed, and the preprocessing includes bad channel removal, polarity correction and slot wave extraction.

[0054] Example 2

[0055] Figure 3 Figures (a) and (b) show the mesh diagrams of the coal mine working face attenuation coefficient model and the wavelet coefficient model in the wavelet domain, respectively. The model is divided into grids with a spacing of 10m, representing a working face length of 1000m and a width of 320m. The model includes 32 vertical grids and 100 horizontal grids, for a total of 3200 grids. In conventional inversion, 3200 channel wave attenuation coefficients need to be inverted. The wavelet coefficient model has 125 non-zero elements, meaning that only 125 non-zero wavelet coefficients need to be inverted during the inversion. Therefore, it can be seen that the parameters of the wavelet domain inverted model are reduced to 96.09% compared to the original 3200 model parameters, while the observed data remains unchanged. This greatly improves the stability of the inversion equation and the reliability of the inversion.

[0056] For those skilled in the art, various corresponding changes and modifications can be made based on the above technical solutions and concepts, and all such changes and modifications should be included within the protection scope of the claims of this invention.

Claims

1. A wavelet domain-based slot wave attenuation imaging method, characterized in that, The specific process is as follows: S1. Transmission slot wave detection: The blasting point is activated in the roadway on one side of the working face, and the slot wave is received by the detectors deployed in the roadway on the other side of the working face. S2. Wavelet Domain Attenuation Imaging Inversion: Transform the attenuation coefficient of the model to the wavelet domain to obtain the attenuation coefficients of different grids inside the working surface. The specific process is as follows: When the slot wave energy generated at the shot point is A0, and the slot wave energy received by the detector is A, then: A=A0e -xd (1) In the formula, the transmission slot wave attenuation coefficient d is the ray length, Q is the quality factor, f is the slot wave frequency, and v is the phase velocity corresponding to the slot wave frequency f; taking the natural logarithm of both sides of equation (1) yields: xd=lnA0-lnA (2) When there is a ray y i Passing through attenuation coefficients x1, x2, ..., x n The mesh within the working surface, and the ray lengths of different meshes are d. i1 ,d i2 ,…,d in When the total attenuation along the i-th ray path is: In the formula, Given the current total length of the ray, the total attenuation of the ray is lnA0 - lnA. j A j This represents the slot wave energy in the j-th grid; by combining the attenuation observation data of all rays, the target inversion matrix can be established: Gm=d (4) In the formula, d represents the observed data, i.e., the received slot wave energy; m represents the attenuation coefficient of different grids within the working surface; and G represents the sensitivity matrix. Using the properties of orthogonal wavelet transform, we have W -1 =W T W T W = I, where I represents the identity matrix, and equation (4) can be written as: Δd=GW T WΔm (5) Δm is the update amount of the attenuation coefficient, and W is the wavelet transform matrix; take This represents the update amount of the attenuation coefficient in the wavelet domain; The two-dimensional wavelet transform of the sensitivity matrix G, Each row represents the two-dimensional wavelet transform coefficients corresponding to the row in G. Equation (6) represents the wavelet coefficients of equation (5) at different scales, which means that the original solution of the attenuation coefficient equation is transformed into the solution of the wavelet coefficients in the wavelet domain. After obtaining the wavelet coefficients, according to It can be obtained That is, the solution can be obtained. Perform inverse wavelet transform to obtain the update amount Δm of the attenuation coefficient, and then obtain the attenuation coefficient of different grids inside the working surface; S3. Based on the different mesh attenuation coefficients inside the working surface obtained in step S2, output the working surface attenuation map.

2. The method according to claim 1, characterized in that, In step S1, after acquiring slot wave data using a detector, the slot wave data is preprocessed, including bad channel removal, polarity correction, and slot wave extraction.

Images