Seismic and electric collaborative detection and identification method and system for coal mine goaf

Abstract

This invention provides a method and system for seismic-electrical co-detection and identification of coal mine goaf areas, relating to the field of coal mine detection technology. The method includes: acquiring seismic exploration data and electrical resistivity tomography data, and generating a seismic multi-attribute response set; generating an electrical multi-attribute response set; performing gridded mapping to form a seismic-electrical dual-response unit matrix; constructing a discrete unit system for the mining area and establishing a geological constraint parameter library; performing seismic-electrical response consistency constraint fusion to generate a seismic-electrical coupled response tensor field; performing goaf anomaly identification, structural zoning, and risk level determination, and outputting the spatial risk assessment results of the goaf area. This invention solves the technical problem that existing technologies typically directly perform fixed-weight fusion or empirical comprehensive interpretation of different physical field data, which easily leads to response conflicts when local anomaly responses are affected, reducing the stability and reliability of goaf identification results.

Description

Technical Field

[0001] This invention relates to the field of coal mine detection technology, specifically to a method and system for seismic-electrical combined detection and identification of coal mine goaf areas. Background Technology

[0002] During coal mining, as the coal face advances, underground rock strata are easily affected by mining activities, leading to the formation of goafs, fracture zones, and localized subsidence areas. When residual cavities, fractured rock masses, or accumulated water exist within the goaf, it not only affects the stability of the surrounding rock but may also induce safety accidents such as roof collapse, water inrush, and surface subsidence. Therefore, accurate detection and spatial identification of goafs are crucial technical means to ensure safe mine production during coal mining.

[0003] Existing technologies typically perform fixed-weight fusion or empirical comprehensive interpretation of different physical field data, which makes it impossible to dynamically adjust the consistency between seismic and electrical responses in different regions according to actual geological conditions. This leads to response conflicts when local abnormal responses are affected by factors such as lithological differences, water content changes, and pore structure, reducing the stability and reliability of goaf identification results. Summary of the Invention

[0004] This application provides a method and system for seismic-electric collaborative detection and identification of coal mine goaf areas, aiming to solve the technical problem that existing technologies usually directly perform fixed-weight fusion or empirical comprehensive interpretation of different physical field data, which easily leads to response conflicts when local abnormal responses are affected, thus reducing the stability and reliability of goaf identification results.

[0005] The first aspect disclosed in this application provides a method for seismic-electrical co-detection and identification of coal mine goaf areas. The method includes: acquiring seismic exploration data and electrical resistivity exploration data of the target mining area; extracting seismic wave reflection attributes, energy attenuation attributes, and wave impedance change attributes from the seismic exploration data to generate a seismic multi-attribute response set; extracting apparent resistivity distribution attributes, electrical gradient attributes, and electrical anomaly enhancement attributes from the electrical resistivity exploration data to generate an electrical multi-attribute response set; mapping the seismic multi-attribute response set and the electrical multi-attribute response set to a grid based on a spatial three-dimensional grid model to form a seismic-electrical dual-response unit matrix; constructing a discrete unit system of the mining area based on the spatial three-dimensional grid model and establishing a geological constraint parameter library, which includes at least lithological background parameters, water content range parameters, and pore structure parameters; performing seismic-electrical response consistency constraint fusion on the seismic-electrical dual-response unit matrix based on the geological constraint parameter library to generate a seismic-electrical coupled response tensor field; and performing goaf anomaly identification, structural zoning, and risk level determination based on the seismic-electrical coupled response tensor field, outputting a spatial risk assessment result for the goaf area.

[0006] The second aspect of this application discloses a seismic-electrical co-detection and identification system for coal mine goaf areas. The system is used in conjunction with the aforementioned seismic-electrical co-detection and identification method for coal mine goaf areas. The system includes: a data acquisition module for acquiring seismic exploration data and electrical resistivity exploration data of the target mining area, and extracting seismic wave reflection attributes, energy attenuation attributes, and wave impedance change attributes from the seismic exploration data to generate a multi-attribute seismic response set; an attribute extraction module for extracting apparent resistivity distribution attributes, electrical gradient attributes, and electrical anomaly enhancement attributes from the electrical resistivity exploration data to generate a multi-attribute electrical response set; and a gridding mapping module for mapping the multi-attribute seismic data based on a spatial three-dimensional grid model. The seismic-electric dual-response unit matrix is ​​formed by meshing the set of physical response and the set of electrical multi-attribute responses. A parameter library construction module is used to construct a discrete unit system for the mining area based on the spatial three-dimensional mesh model and establish a geological constraint parameter library, which includes at least lithological background parameters, water content range parameters, and pore structure parameters. A constraint fusion module is used to perform seismic-electric response consistency constraint fusion on the seismic-electric dual-response unit matrix based on the geological constraint parameter library, generating a seismic-electric coupled response tensor field. A level determination module is used to identify anomalies, structural zoning, and risk level determination in the goaf based on the seismic-electric coupled response tensor field, outputting the spatial risk assessment results of the goaf.

[0007] One or more technical solutions provided in this application have at least the following beneficial effects: By extracting various seismic attributes such as seismic wave reflection, energy attenuation, and wave impedance changes, a multi-attribute seismic response set reflecting changes in the subsurface medium structure is formed, improving the expressive power of seismic anomaly information. By extracting electrical attributes such as apparent resistivity, electrical gradient, and anomaly enhancement, a multi-attribute electrical response set characterizing subsurface conductivity is formed, improving the identification accuracy of electrical anomaly information. By establishing a unified three-dimensional spatial grid, spatial correspondence between seismic and electrical data is achieved, forming a unified data representation structure and providing a foundation for multi-source information fusion. By introducing geological constraint information such as lithology, water content, and pore structure, a geological constraint parameter library is established, improving the consistency between multi-source detection results and actual geological conditions. Through seismoelectric response consistency analysis and adaptive weighted fusion under geological constraints, the reliability of multi-source data fusion is improved, enhancing the authenticity of anomaly responses in goaf areas. Through comprehensive analysis of the seismoelectric coupled response tensor field, anomaly identification, spatial zoning, and risk level determination in goaf areas are achieved, improving the accuracy of goaf risk assessment.

[0008] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0009] Figure 1 This is a schematic diagram of the seismic-electric collaborative detection and identification method for coal mine goaf provided in the embodiments of this application.

[0010] Figure 2 This is a schematic diagram of the structure of a seismic-electric collaborative detection and identification system for coal mine goaf provided in an embodiment of this application.

[0011] Figure labeling: Data acquisition module 10, attribute extraction module 20, grid mapping module 30, parameter library construction module 40, constraint fusion module 50, level determination module 60. Detailed Implementation

[0012] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0013] Example 1, as Figure 1 As shown in the embodiments of this application, a method for seismic-electric collaborative detection and identification of coal mine goaf areas is provided, the method comprising: Seismic exploration data and electrical resistivity tomography data of the target mining area are acquired, and seismic wave reflection attributes, energy attenuation attributes and wave impedance change attributes are extracted from the seismic exploration data to generate a seismic multi-attribute response set.

[0014] Seismic survey lines or 3D seismic observation systems are deployed within the target mining area to collect data on reflected waves, diffracted waves, and changes in propagation energy at subsurface interfaces. After denoising, static correction, amplitude compensation, and anomaly correction of the raw seismic records, parameters such as reflection amplitude, reflection phase axis continuity, spectral attenuation, and abrupt changes in wave impedance are extracted. Among these, seismic wave reflection attributes reflect changes in subsurface structural interfaces, energy attenuation attributes characterize energy loss in fractured, cavitary, or loose media, and wave impedance variation attributes characterize differences in medium density and wave velocity. These parameters are uniformly encoded according to spatial location and attribute type to form a multi-attribute seismic response set.

[0015] The apparent resistivity distribution attribute, electrical gradient attribute, and electrical anomaly enhancement attribute are extracted from the electrical exploration data to generate a multi-attribute electrical response set.

[0016] Resistivity, high-density electrical resistivity, or transient electromagnetic (TEM) methods are used to explore the target mining area, obtaining potential differences, current intensities, and inverted resistivity data at different measuring points or survey lines. After processing the acquired data, including electrode contact anomaly removal, measurement error correction, background field normalization, and interference signal attenuation, apparent resistivity distribution, electrical gradient changes, and anomaly enhancement responses are extracted. Apparent resistivity distribution describes differences in the conductivity of the subsurface medium; electrical gradient reflects the strength of electrical changes at adjacent spatial locations; and electrical anomaly enhancement attributes highlight anomalous bodies such as low-resistivity aquifers, high-resistivity cavities, or fractured zones. All electrical parameters are organized uniformly according to coordinate location to generate a multi-attribute electrical response set.

[0017] Based on a spatial three-dimensional grid model, the seismic multi-attribute response set and the electrical multi-attribute response set are mapped to a grid to form a seismoelectric dual-response unit matrix.

[0018] A three-dimensional spatial grid model is established based on the mining area's planar boundary, detection depth, and preset grid scale, dividing the target mining area into several volume elements with unique spatial coordinates. Seismic multi-attribute response sets and electrical multi-attribute response sets are projected onto their respective grid elements. For locations where measurement points do not overlap, attribute values ​​can be supplemented using interpolation, neighborhood assignment, or resampling of inversion results. Since seismic and electrical attributes have different dimensions and value ranges, coordinate unification, scale normalization, and feature dimension matching are required. Ultimately, each grid element corresponds to a set of seismic response parameters and a set of electrical response parameters, thus forming a seismoelectric dual-response element matrix.

[0019] Based on the aforementioned three-dimensional spatial grid model, a discrete unit system for the mining area is constructed, and a geological constraint parameter library is established. The geological constraint parameter library includes at least lithological background parameters, water content range parameters, and pore structure parameters.

[0020] Based on the established spatial three-dimensional grid model, geological data from the mining area, borehole exposure data, measured data from roadways, and existing stratigraphic interpretation results are incorporated to construct a discrete unit system for the mining area. Each grid unit is assigned geological constraint parameters such as lithological background, water cut range, pore structure, and formation stability. Specifically, the lithological background parameter defines the normal physical property range of different rock strata, the water cut range parameter corrects for the influence of water on electrical response, and the pore structure parameter describes the degree of fracture, breakage, or cavity development. After parameter encoding, normalization, and structured storage, a geological constraint parameter library corresponding to the spatial grid is formed.

[0021] Based on the geological constraint parameter library, the seismoelectric dual-response unit matrix is ​​subjected to seismoelectric response consistency constraint fusion to generate a seismoelectric coupled response tensor field.

[0022] The geological constraint parameter library is associated with the seismoelectric dual-response element matrix according to grid number, so that each spatial element simultaneously contains seismic response, electrical response, and geological background information. Within the same or adjacent grids, the spatial overlap, direction of change, and intensity differences between seismic and electrical anomalies are compared to calculate the seismoelectric response consistency index. When seismic attenuation enhancement, resistivity anomalies, and geological fracturing conditions are consistent, the weight of multi-source anomalies in that region is increased; when the two exhibit inconsistent behavior due to factors such as water content and lithological differences, the weight allocation is adjusted according to the geological constraint parameters. After weighted combination and spatial continuity correction, a seismoelectric coupled response tensor field is generated.

[0023] Based on the seismoelectric coupling response tensor field, anomaly identification, structural zoning, and risk level determination of the goaf are performed, and the spatial risk assessment results of the goaf are output.

[0024] Based on the seismoelectric coupled response tensor field, the comprehensive anomaly intensity, spatial continuity, and geometric distribution characteristics of each grid unit are determined. Regions exhibiting strong reflection discontinuities, significant energy attenuation, abrupt changes in wave impedance, and accompanying electrical anomalies are identified as suspected goaf areas or goaf-affected areas. Further, based on the connectivity, burial depth, boundary clarity, and correspondence with roadway excavation data, these areas are divided into cavity areas, fractured and collapsed areas, water-bearing anomaly areas, and transitional influence areas. Finally, combining anomaly intensity, water-bearing probability, and surrounding rock stability, each zone is classified into different risk levels, and the spatial risk assessment results for the goaf areas are output.

[0025] Furthermore, generating a multi-attribute seismic response set includes: The seismic exploration data is subjected to denoising filtering, gather static correction, and anomalous signal correction to obtain denoised and corrected seismic signal data. Based on the denoised and corrected seismic signal data, features are extracted from the first arrival time parameter, reflection amplitude parameter, spectral attenuation parameter, and wave impedance change parameter during the seismic wave propagation process to generate a basic seismic attribute vector set. The basic seismic attribute vector set is subjected to spatial coordinate association processing and multi-scale interpolation reconstruction processing to obtain continuous seismic attribute distribution data corresponding to each spatial location. Based on the continuous seismic attribute distribution data, attribute normalization combination processing is performed to establish a multi-attribute seismic response set containing multiple seismic response characteristic parameters.

[0026] The acquired raw seismic exploration data undergoes quality checks to identify missing traces, anomalous traces, and random interference signals. Bandpass filtering, median filtering, or frequency domain filtering methods are used to remove environmental noise and high-frequency interference while retaining effective reflection information. Then, trace collection static correction is implemented based on topographic relief, low-velocity zones, and geophone elevation differences to ensure that reflection events corresponding to the same subsurface reflection interface remain consistent on the time axis. Finally, anomalous signals are corrected by combining the continuity of adjacent traces and amplitude variation patterns to eliminate the influence of anomalous amplitudes and distorted waveforms, resulting in denoised and corrected seismic signal data with a unified time reference and controlled noise, providing a stable data foundation for subsequent attribute extraction.

[0027] Using denoised and corrected seismic signal data as input, waveform analysis and attribute calculations are performed on each record. First arrival time parameters are extracted based on the first wave arrival location to reflect changes in seismic wave propagation velocity; reflection amplitude parameters are extracted based on changes in peak reflected wave value, envelope energy, and amplitude to characterize the reflection intensity at subsurface interfaces; spectral attenuation parameters are calculated by combining energy changes in different frequency bands to reflect the degree of absorption of high-frequency energy by fracture zones or cavities; and wave impedance change parameters are calculated based on velocity models and density parameters to describe abrupt changes in medium properties. These parameters are then combined and encoded according to the spatial location of the measuring points to form a set of basic seismic attribute vectors.

[0028] Based on the coordinates of the survey lines, shot points, and receiver points, the basic seismic attribute vector set is spatially located, mapping each attribute value to its corresponding three-dimensional spatial location. Since the original sampling points are typically discretely distributed, multi-scale interpolation reconstruction is performed for regions with different sampling densities using methods such as inverse distance weighted interpolation, kriging interpolation, or spline interpolation. This ensures that the discrete attributes are continuously extended to the entire study area. Multi-scale interpolation reconstruction involves performing local detail compensation and overall trend fitting at different spatial scales to balance local anomaly resolution with overall continuity, ultimately obtaining a continuously distributed seismic attribute field.

[0029] The various attributes in the continuous seismic attribute distribution data are processed to unify their dimensions, making different physical attributes comparable. Based on spatial location correspondences, attributes such as first arrival time, reflection amplitude, spectral attenuation, and wave impedance variation are combined and encoded, and a unified attribute index is established, so that each spatial location corresponds to a complete set of seismic response characteristics. The seismic multi-attribute response set is a data set indexed by spatial location and containing multiple seismic attributes. It not only retains information about each individual attribute but also maintains the correspondence between attributes, providing a unified data representation for subsequent seismoelectric joint mapping and multi-source fusion.

[0030] Furthermore, generating a set of electrical multi-attribute responses includes: The electrical exploration data is processed by electrode anomaly removal, measurement error correction, and background electric field normalization to obtain standardized electrical measurement data. Based on the standardized electrical measurement data, the apparent resistivity distribution characteristics, electrical gradient variation characteristics, and anomalous response enhancement characteristics of the subsurface medium are extracted to generate a basic electrical response vector set. The basic electrical response vector set is then subjected to spatial inversion calculation and regional continuous reconstruction processing to obtain continuous electrical response distribution data for the corresponding spatial location. Based on the continuous electrical response distribution data, multi-parameter fusion encoding processing is performed to establish a multi-attribute electrical response set containing multiple electrical response feature parameters.

[0031] The acquired electrical exploration data undergoes a completeness check to identify and remove or correct abnormal measurement points caused by excessive contact resistance, electrode detachment, measurement drift, and environmental electromagnetic interference. Measurement errors are corrected based on instrument calibration parameters and repeated observation results to reduce the impact of instrument drift and human error. Background electric field normalization is performed on the potential and apparent resistivity data of each measurement point, taking into account the background electric field distribution of the target area, to eliminate overall offsets caused by variations in power supply conditions and natural electric fields in different survey areas. Specifically, background electric field normalization uses the stable background value of the region as a reference to uniformly correct the response of each measurement point, ensuring that standardized electrical measurement data have a consistent data benchmark.

[0032] Using standardized electrical measurement data as input, the apparent resistivity distribution characteristics of the subsurface medium are calculated based on the current, potential difference, and inversion results at each measuring point, reflecting the differences in conductivity among different media. The electrical gradient variation characteristics are calculated based on the apparent resistivity variation amplitude between adjacent measuring points or adjacent grids to describe the rate of change in subsurface electrical properties. Combining local background values ​​and neighborhood statistical characteristics, enhanced calculations are performed on anomalous areas deviating from the normal distribution range, resulting in enhanced electrical anomaly characteristics. Anomaly enhancement is achieved by increasing the contrast between the anomalous response and the background response, making low-resistivity aquifer areas, high-resistivity cavitation areas, and fracture-developed areas more prominent. Finally, a basic electrical response vector set is formed according to spatial location.

[0033] A subsurface medium inversion model is established based on the coordinates of the measuring points corresponding to the electrical fundamental response vector set. Two-dimensional or three-dimensional resistivity inversion methods are used to calculate the true electrical distribution at various spatial locations underground, and the discrete measuring point data is converted into a continuous spatial response. Since electrical resistivity measuring points are typically laid out along the measuring lines, to obtain a complete spatial distribution, the inversion results undergo further regional continuous reconstruction. Methods such as inverse distance weighted interpolation, kriging interpolation, or finite element grid interpolation are used to supplement data in unsampled areas, ensuring continuous variation between different measuring lines. Specifically, regional continuous reconstruction restores the continuous electrical field based on the correlation of neighborhood spaces, ultimately obtaining continuous electrical response distribution data covering the entire target mining area.

[0034] Multiple attributes in the continuous electrical response distribution data, such as apparent resistivity, electrical gradient, and anomaly enhancement, are processed with unified dimensions, and a correspondence is established according to spatial grid positions. Based on unified data encoding rules, different types of electrical parameters are sequentially combined so that each spatial location corresponds to a set of electrical characteristic data for a fixed structure. Multi-parameter fusion encoding refers to organizing multiple electrical response parameters into the same data unit according to a unified index while maintaining the independence of each attribute. This does not change the physical meaning of each parameter, but only establishes a unified data representation structure, ultimately forming a multi-attribute electrical response set containing multiple electrical response characteristic parameters. This provides standardized input data for subsequent seismoelectric joint mapping and consistency fusion.

[0035] Furthermore, based on a spatial three-dimensional mesh model, the seismic multi-attribute response set and the electrical multi-attribute response set are meshed and mapped to form a seismoelectric dual-response element matrix, including: A three-dimensional spatial grid model is established based on the target mining area's spatial range, detection depth parameters, and grid resolution parameters. The target mining area is then spatially discretized to generate multiple grid cells with spatial coordinate information. Based on the spatial coordinate information of these grid cells, the seismic multi-attribute response set and the electrical multi-attribute response set are spatially matched to obtain seismic response feature data and electrical response feature data for the corresponding grid locations. The seismic response feature data and electrical response feature data within the same grid cell are then subjected to coordinate unification, scale normalization, and feature dimension matching to generate grid-level seismic-electric dual-source response data. Based on this grid-level seismic-electric dual-source response data, a correspondence between grid cells and multi-source response parameters is established, and the seismic-electric dual-response cell matrix is ​​constructed.

[0036] A unified three-dimensional spatial grid model is established based on the target mining area boundary coordinates, designed exploration depth, and preset grid resolution. The spatial extent is determined by the mining area's planar boundary and elevation range. The exploration depth parameter limits the grid's vertical coverage, and the grid resolution determines the length, width, and height of each grid cell. The target mining area is spatially discretized according to the preset grid size, dividing the continuous underground space into multiple regular grid cells with unique numbers and spatial coordinate information. Each grid cell corresponds to a fixed spatial location, providing a unified spatial positioning reference for different physical field data.

[0037] Using the center coordinates or spatial coverage of each grid cell as the matching basis, the seismic multi-attribute response set and the electrical multi-attribute response set are mapped to the corresponding grid cells. When the sampling point coincides with the grid center location, the corresponding attribute is directly assigned; for cases where the sampling location and grid location do not coincide, the grid attribute value is determined by nearest neighbor matching, inverse distance weighting, or interpolation mapping based on spatial neighborhood relationships. Spatial location matching establishes a correspondence between the detection data and grid cells according to a unified coordinate system, enabling each grid cell to simultaneously associate seismic and electrical attributes, forming seismic response characteristic data and electrical response characteristic data for the corresponding grid location.

[0038] The seismic response and electrical response characteristic data within the same grid cell are processed uniformly. First, positional deviations between different exploration methods are corrected using a unified spatial coordinate system to ensure that all attributes correspond to the same subsurface spatial location. Scale normalization is then performed for different attribute dimensions and value ranges to eliminate the influence of differences in parameters such as amplitude, wave impedance, and apparent resistivity. Feature dimension matching is then completed according to a preset attribute order, ensuring a unified data structure for seismic and electrical attribute dimensions. Specifically, feature dimension matching establishes a unified data format based on a fixed attribute arrangement, ultimately generating the seismic-electric dual-source response data for each grid cell.

[0039] Using the grid cell numbers in the spatial three-dimensional grid model as indexes, the seismic and electrical dual-source response data within the corresponding grid cells are organized and stored according to a unified data structure, establishing a one-to-one correspondence between grid cells and multi-source response parameters. The multi-source response parameters include seismic response parameter groups and electrical response parameter groups. Each parameter retains its original attribute values ​​and category information, without fusion calculations, only forming a unified data representation. The data from all grid cells are arranged and combined according to the grid number order to construct a seismic and electrical dual-response cell matrix, where each row corresponds to a spatial grid cell and each column corresponds to a response attribute, providing a standardized data foundation for subsequent geological constraint fusion and seismic-electric consistency analysis.

[0040] Furthermore, a discrete unit system for the mining area is constructed based on the aforementioned three-dimensional spatial grid model, and a geological constraint parameter library is established, including: Correlation analysis is performed on geological data, borehole information, and lithological distribution data of the target mining area to extract lithological type parameters, water-bearing state parameters, pore structure parameters, and formation stability parameters, generating a set of basic geological parameters. Based on this set of basic geological parameters and a spatial three-dimensional grid model, spatial location matching is performed to map different geological parameters to corresponding grid cells, generating grid-level geological attribute data containing spatial location information. Parameter encoding is performed on each spatial grid cell according to the grid-level geological attribute data to establish the correspondence between grid cells and geological characteristic parameters, generating a geological constraint feature matrix. Based on the geological constraint feature matrix, parameter classification, normalization, and structured storage are performed to construct the geological constraint parameter library.

[0041] Existing geological data, borehole columnar sections, well logging results, and lithological interpretation data for the target mining area were collected and correlated using a unified coordinate system. Based on borehole exposure and stratigraphic distribution information, the properties of strata at each location were analyzed, and lithological type parameters, aquifer state parameters, pore structure parameters, and formation stability parameters were extracted. Specifically, lithological type parameters represent the composition and basic physical properties of different strata; aquifer state parameters represent the water content and occurrence state of the underground medium; pore structure parameters represent the degree of fracture, pore, and pore development; and formation stability parameters reflect the integrity and structural stability of the surrounding rock. Finally, the above parameters were correlated according to spatial location to form a set of basic geological parameters.

[0042] Using a three-dimensional spatial grid model as a unified spatial reference, various parameters in the geological baseline parameter set are mapped to corresponding grid cells according to spatial coordinates. For locations directly exposed by boreholes, parameters are assigned to the corresponding grid cells based on actual coordinates. For areas not directly exposed, spatial interpolation and regional extrapolation are performed by combining data from adjacent boreholes, stratum extension patterns, and stratigraphic interpretation results, ensuring that each grid cell possesses corresponding geological attribute information. Spatial location matching establishes a one-to-one correspondence between geological parameters and grid cells based on a unified coordinate system, converting discrete geological data into continuous spatially distributed data, ultimately generating grid-level geological attribute data containing spatial location information.

[0043] Based on grid-level geological attribute data, parameters such as lithology, water-bearing state, pore structure, and formation stability in each grid unit are uniformly coded. Categorical parameters are represented using a classification numbering method, while continuous parameters are numerically coded according to a preset precision, and a unified data arrangement structure is established based on a fixed attribute order. Parameter coding converts geological information from different sources and of different types into a unified data representation format, forming a standardized data structure while preserving the original physical meaning of each parameter. A correspondence is established between grid unit numbers and corresponding geological characteristic parameters, generating a geological constraint feature matrix where each row corresponds to a spatial grid unit, and each column corresponds to a geological constraint attribute.

[0044] The parameters in the geological constraint feature matrix are categorized and organized according to their attribute types. Lithological parameters, water-bearing parameters, porosity parameters, and stability parameters are assigned to their corresponding data categories. Continuous parameters are then normalized to ensure a unified data scale, while categorical parameters retain their original coding format. After this normalization process, a structured storage format is established based on grid number, parameter category, and attribute number to centrally organize the geological constraint information corresponding to each grid unit. The structured storage uses standardized fields and data indexes to establish a standard data organization form, ultimately constructing a geological constraint parameter library that corresponds one-to-one with the spatial three-dimensional grid. This provides stable geological background data for subsequent seismic-electric response consistency analysis and multi-source constraint fusion.

[0045] Furthermore, based on the geological constraint parameter library, the seismoelectric dual-response element matrix is ​​subjected to seismoelectric response consistency constraint fusion to generate a seismoelectric coupled response tensor field, including: The geological constraint parameter library and the seismoelectric dual-response unit matrix are spatially correlated and matched to generate a set of seismoelectric response units containing geological constraint information. Based on the set of seismoelectric response units, the seismic response parameters, electrical response parameters, and geological constraint parameters corresponding to each spatial grid unit are extracted. Spatial correlation analysis is performed on the seismic response parameters and electrical response parameters to calculate the correlation between responses of different physical fields and generate a seismoelectric response consistency evaluation index. A consistency coefficient matrix corresponding to the spatial grid is constructed based on the seismoelectric response consistency evaluation index. The seismic response weights and electrical response weights in the seismoelectric dual-response unit matrix are dynamically adjusted according to the consistency coefficient matrix to generate an adaptive fusion weight field. The seismoelectric multi-source response data are fused based on the adaptive fusion weight field to generate the seismoelectric coupled response tensor field.

[0046] Using the grid cell numbers in the spatial 3D grid model as a unified index, the geological constraint parameter library and the seismoelectric dual-response element matrix are associated and matched according to the same spatial coordinates. For each grid cell, the corresponding seismic response parameter group, electrical response parameter group, and geological constraint parameter group are read and integrated into the same data cell. Spatial association matching establishes a one-to-one correspondence between data from different sources based on a unified coordinate system, ensuring that the same grid cell simultaneously contains seismic, electrical, and geological information without altering the original physical meaning of each parameter. After completing the association of all grids, a set of seismoelectric response elements containing geological constraint information is generated, providing a unified data foundation for subsequent consistency analysis.

[0047] Using a set of seismoelectric response units as the analysis object, the corresponding seismic response parameters, electrical response parameters, and geological constraint parameters are extracted one by one according to the grid unit number, and then classified and organized according to a unified data structure. Seismic response parameters include attributes such as reflection amplitude, spectral attenuation, and wave impedance changes; electrical response parameters include attributes such as apparent resistivity, electrical gradient, and anomaly enhancement; and geological constraint parameters include attributes such as lithology, water-bearing state, pore structure, and formation stability. The parameters are reorganized according to a fixed attribute order to establish a comprehensive attribute vector for the corresponding grid unit, ensuring that each spatial location has a complete and consistent data representation, providing a unified input for the correlation calculation between different physical fields.

[0048] Spatial correlation analysis was performed on seismic and electrical response parameters within the same grid cell, and the correlation results were corrected by incorporating corresponding geological constraint parameters. First, the consistency of the two types of responses in terms of spatial distribution, variation trends, and anomaly intensity was compared. Then, a correlation was determined based on geological conditions such as lithology, aquifer status, and pore structure. The consistency was increased when the two types of responses exhibited synchronous anomalies in the same area; when the response differences were mainly influenced by lithological variations or aquifer conditions, corrections were made based on geological constraints. Finally, a corresponding seismoelectric response consistency evaluation index was generated for each grid cell to quantitatively characterize the degree of matching between responses from different physical fields.

[0049] Based on the seismoelectric response consistency evaluation index corresponding to each spatial grid unit, a consistency coefficient matrix is ​​established according to the grid number in the spatial three-dimensional grid model. Each grid unit corresponds to a consistency coefficient, which represents the overall consistency between the seismic response and the electrical response at that location. A higher consistency coefficient indicates a higher spatial correspondence between the two types of physical field responses, while a lower consistency coefficient indicates a greater difference between the two types of responses. The consistency coefficients of each grid unit are organized and stored according to the spatial arrangement of the grids, forming a consistency coefficient matrix corresponding to the spatial grid model, providing a unified data basis for subsequent multi-source response weight adjustment.

[0050] Based on the consistency coefficients of each grid cell in the consistency coefficient matrix, the seismic response weights and electrical response weights in the seismoelectric dual-response cell matrix are dynamically adjusted. When the consistency coefficient is high, the weights of the two types of responses are kept relatively balanced; when the consistency coefficient is low, the weights of responses subject to greater disturbance are appropriately reduced in conjunction with geological constraint parameters, while the relative contribution of the other type of response is increased, so that different physical fields have adaptive weight allocation at various spatial locations. The adaptive fusion weight field is a dataset that records the distribution of seismoelectric response weights in each grid cell using a spatial grid as a carrier, reflecting the relative contribution of multi-source responses from different regions to the fusion calculation.

[0051] Based on the adaptive fusion weight field, the seismic response parameters, electrical response parameters and corresponding weights in each grid cell are weighted and combined while keeping the spatial positional relationship unchanged. According to the grid arrangement in the spatial three-dimensional grid model, the fusion results of all grid cells are uniformly organized, and different spatial positions and different response attributes are constructed into a data structure with fixed dimensions, providing a unified data foundation for subsequent anomaly identification, structural zoning and risk level determination in goaf areas.

[0052] Furthermore, the seismoelectric multi-source response data is fused based on the adaptive fusion weight field to generate the seismoelectric coupled response tensor field, including: Based on the adaptive fusion weight field, the seismic response parameters, electrical response parameters, and geological constraint parameters in the seismoelectric dual-response unit matrix are weighted and combined to generate grid-level fused response feature data. The grid-level fused response feature data is then multidimensionally arranged according to the spatial grid positional relationship to construct multidimensional response data containing spatial position, seismic response, electrical response, and geological constraint dimensions. Tensor-encoding is performed on the multidimensional response data to generate a seismoelectric coupled response tensor containing multi-physics response information. Based on spatial continuity constraints, the seismoelectric coupled response tensor undergoes spatial reconstruction and smoothing optimization to obtain a continuously distributed seismoelectric coupled response tensor field.

[0053] Within the same grid cell, various parameters are numerically combined according to their corresponding weights, while retaining the attribute category and spatial location number of each parameter. The weighted combination process calculates each response parameter uniformly based on the weight contribution of different physical fields within the current grid, ensuring that data from different sources participate in feature representation rather than simply being superimposed. Ultimately, each grid cell generates grid-level fused response feature data containing seismic and electrical responses and geological constraint information, providing a foundation for subsequent multidimensional data organization.

[0054] The spatial location dimension is established according to the grid number, and then the seismic response dimension, electrical response dimension and geological constraint dimension are organized in sequence according to the fixed attribute order to ensure that the data structure corresponding to each grid unit is consistent. After all grid data is arranged, multidimensional response data with a unified data structure is constructed to establish a standard input format for tensor quantization coding.

[0055] Using multidimensional response data as input, the spatial location dimension, seismic response dimension, electrical response dimension, and geological constraint dimension are organized into a unified data structure according to a preset dimensional order, and tensor encoding is completed according to fixed indexing rules. Tensor encoding organizes the original two-dimensional or one-dimensional data according to multiple independent dimensions, ensuring that various response parameters retain both their physical properties and the spatial location and correspondences between different dimensions. After encoding, each spatial grid cell corresponds to a data cell in the tensor, and all grids together constitute a seismoelectric coupled response tensor containing multiphysics response information, providing a unified data carrier for subsequent continuous spatial analysis.

[0056] Using the seismoelectric coupled response tensor as the processing object, spatial continuity constraints are established based on the spatial adjacency relationship between adjacent grids. Spatial reconstruction and smoothing optimization are then performed on local discrete anomalies and abrupt boundary regions. Specifically, the continuity correction of local response values ​​is applied based on the response change trends of adjacent grids, ensuring a reasonable transition between adjacent spatial locations while preserving the spatial boundary characteristics of the true anomaly region. The spatial continuity constraint utilizes the generally continuous variation of subsurface medium properties to constrain and correct isolated anomalies and local fluctuations, while spatial reconstruction and smoothing optimization improve data continuity and stability while maintaining the overall spatial structure. Ultimately, a continuously distributed seismoelectric coupled response tensor field is obtained, providing a continuous spatial data foundation for anomaly identification and risk assessment in goaf areas.

[0057] Example 2, based on the same inventive concept as the seismic-electric co-detection and identification method for coal mine goaf in the foregoing examples, such as... Figure 2 As shown in the embodiment of this application, a seismic-electric collaborative detection and identification system for coal mine goaf areas is provided, the system comprising: The data acquisition module 10 is used to acquire seismic exploration data and electrical resistivity exploration data of the target mining area, and extract seismic wave reflection attributes, energy attenuation attributes, and wave impedance change attributes from the seismic exploration data to generate a seismic multi-attribute response set; the attribute extraction module 20 is used to extract apparent resistivity distribution attributes, electrical gradient attributes, and electrical anomaly enhancement attributes from the electrical resistivity exploration data to generate an electrical multi-attribute response set; the gridding mapping module 30 is used to perform gridding mapping between the seismic multi-attribute response set and the electrical multi-attribute response set based on a spatial three-dimensional grid model to form a seismoelectric dual-response unit matrix. The parameter library construction module 40 is used to construct a discrete unit system of the mining area based on the spatial three-dimensional grid model and establish a geological constraint parameter library, which includes at least lithological background parameters, water content range parameters, and pore structure parameters; the constraint fusion module 50 is used to perform seismic-electric dual-response unit matrix seismic-electric response consistency constraint fusion based on the geological constraint parameter library to generate a seismic-electric coupled response tensor field; the level determination module 60 is used to identify anomalies, structural zoning, and risk level determination of the goaf based on the seismic-electric coupled response tensor field, and output the spatial risk assessment results of the goaf.

[0058] Furthermore, the data acquisition module 10 is used to perform the following operation steps: The seismic exploration data is subjected to denoising filtering, gather static correction, and anomalous signal correction to obtain denoised and corrected seismic signal data. Based on the denoised and corrected seismic signal data, features are extracted from the first arrival time parameter, reflection amplitude parameter, spectral attenuation parameter, and wave impedance change parameter during the seismic wave propagation process to generate a basic seismic attribute vector set. The basic seismic attribute vector set is subjected to spatial coordinate association processing and multi-scale interpolation reconstruction processing to obtain continuous seismic attribute distribution data corresponding to each spatial location. Based on the continuous seismic attribute distribution data, attribute normalization combination processing is performed to establish a multi-attribute seismic response set containing multiple seismic response characteristic parameters.

[0059] Furthermore, the attribute extraction module 20 is used to perform the following operation steps: The electrical exploration data is processed by electrode anomaly removal, measurement error correction, and background electric field normalization to obtain standardized electrical measurement data. Based on the standardized electrical measurement data, the apparent resistivity distribution characteristics, electrical gradient variation characteristics, and anomalous response enhancement characteristics of the subsurface medium are extracted to generate a basic electrical response vector set. The basic electrical response vector set is then subjected to spatial inversion calculation and regional continuous reconstruction processing to obtain continuous electrical response distribution data for the corresponding spatial location. Based on the continuous electrical response distribution data, multi-parameter fusion encoding processing is performed to establish a multi-attribute electrical response set containing multiple electrical response feature parameters.

[0060] Furthermore, the mesh mapping module 30 is used to perform the following operation steps: A three-dimensional spatial grid model is established based on the target mining area's spatial range, detection depth parameters, and grid resolution parameters. The target mining area is then spatially discretized to generate multiple grid cells with spatial coordinate information. Based on the spatial coordinate information of these grid cells, the seismic multi-attribute response set and the electrical multi-attribute response set are spatially matched to obtain seismic response feature data and electrical response feature data for the corresponding grid locations. The seismic response feature data and electrical response feature data within the same grid cell are then subjected to coordinate unification, scale normalization, and feature dimension matching to generate grid-level seismic-electric dual-source response data. Based on this grid-level seismic-electric dual-source response data, a correspondence between grid cells and multi-source response parameters is established, and the seismic-electric dual-response cell matrix is ​​constructed.

[0061] Furthermore, the parameter library construction module 40 is used to perform the following operation steps: Correlation analysis is performed on geological data, borehole information, and lithological distribution data of the target mining area to extract lithological type parameters, water-bearing state parameters, pore structure parameters, and formation stability parameters, generating a set of basic geological parameters. Based on this set of basic geological parameters and a spatial three-dimensional grid model, spatial location matching is performed to map different geological parameters to corresponding grid cells, generating grid-level geological attribute data containing spatial location information. Parameter encoding is performed on each spatial grid cell according to the grid-level geological attribute data to establish the correspondence between grid cells and geological characteristic parameters, generating a geological constraint feature matrix. Based on the geological constraint feature matrix, parameter classification, normalization, and structured storage are performed to construct the geological constraint parameter library.

[0062] Furthermore, the constraint fusion module 50 is used to perform the following operation steps: The geological constraint parameter library and the seismoelectric dual-response unit matrix are spatially correlated and matched to generate a set of seismoelectric response units containing geological constraint information. Based on the set of seismoelectric response units, the seismic response parameters, electrical response parameters, and geological constraint parameters corresponding to each spatial grid unit are extracted. Spatial correlation analysis is performed on the seismic response parameters and electrical response parameters to calculate the correlation between responses of different physical fields and generate a seismoelectric response consistency evaluation index. A consistency coefficient matrix corresponding to the spatial grid is constructed based on the seismoelectric response consistency evaluation index. The seismic response weights and electrical response weights in the seismoelectric dual-response unit matrix are dynamically adjusted according to the consistency coefficient matrix to generate an adaptive fusion weight field. The seismoelectric multi-source response data are fused based on the adaptive fusion weight field to generate the seismoelectric coupled response tensor field.

[0063] Furthermore, the constraint fusion module 50 is used to perform the following operation steps: Based on the adaptive fusion weight field, the seismic response parameters, electrical response parameters, and geological constraint parameters in the seismoelectric dual-response unit matrix are weighted and combined to generate grid-level fused response feature data. The grid-level fused response feature data is then multidimensionally arranged according to the spatial grid positional relationship to construct multidimensional response data containing spatial position, seismic response, electrical response, and geological constraint dimensions. Tensor-encoding is performed on the multidimensional response data to generate a seismoelectric coupled response tensor containing multi-physics response information. Based on spatial continuity constraints, the seismoelectric coupled response tensor undergoes spatial reconstruction and smoothing optimization to obtain a continuously distributed seismoelectric coupled response tensor field.

[0064] Through the foregoing detailed description of the seismic-electric co-detection and identification method for coal mine goaf, those skilled in the art can clearly understand the seismic-electric co-detection and identification system for coal mine goaf in this embodiment. Since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and relevant parts can be referred to the method section.

[0065] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for combined seismic and electrical detection and identification of coal mine goaf areas, characterized in that, The method includes: Seismic exploration data and electrical resistivity exploration data of the target mining area are acquired, and seismic wave reflection attributes, energy attenuation attributes and wave impedance change attributes are extracted from the seismic exploration data to generate a seismic multi-attribute response set. The apparent resistivity distribution attribute, electrical gradient attribute, and electrical anomaly enhancement attribute are extracted from the electrical exploration data to generate a multi-attribute electrical response set. Based on a spatial three-dimensional mesh model, the seismic multi-attribute response set and the electrical multi-attribute response set are meshed and mapped to form a seismoelectric dual-response unit matrix. Based on the aforementioned three-dimensional spatial grid model, a discrete unit system for the mining area is constructed, and a geological constraint parameter library is established. The geological constraint parameter library includes at least lithological background parameters, water content range parameters, and pore structure parameters. Based on the geological constraint parameter library, the seismoelectric dual-response unit matrix is ​​fused with seismoelectric response consistency constraints to generate a seismoelectric coupled response tensor field. Based on the seismoelectric coupling response tensor field, anomalies in the goaf are identified, structural zoning and risk level are determined, and the spatial risk assessment results of the goaf are output.

2. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 1, characterized in that, Generate a multi-attribute seismic response set, including: The seismic exploration data is subjected to denoising filtering, gather static correction, and anomaly signal correction to obtain denoised and corrected seismic signal data. Based on the denoised and corrected seismic signal data, features are extracted from the first arrival time parameter, reflection amplitude parameter, spectral attenuation parameter, and wave impedance change parameter during the seismic wave propagation process to generate a set of basic seismic attribute vectors. Spatial coordinate association processing and multi-scale interpolation reconstruction processing are performed on the earthquake basic attribute vector set to obtain continuous earthquake attribute distribution data corresponding to each spatial location; Based on the continuous seismic attribute distribution data, attribute normalization and combination processing is performed to establish a seismic multi-attribute response set containing multiple seismic response characteristic parameters.

3. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 1, characterized in that, Generate a set of electrical multi-attribute responses, including: The electrical exploration data is processed by electrode anomaly removal, measurement error correction, and background electric field normalization to obtain standardized electrical measurement data. Based on the standardized electrical measurement data, the apparent resistivity distribution characteristics, electrical gradient change characteristics, and abnormal response enhancement characteristics of the underground medium are extracted to generate a basic electrical response vector set. Spatial inversion calculation and regional continuum reconstruction processing are performed on the electrical basic response vector set to obtain continuous electrical response distribution data at the corresponding spatial location; Based on the continuous electrical response distribution data, multi-parameter fusion encoding processing is performed to establish a set of electrical multi-attribute responses containing multiple electrical response feature parameters.

4. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 1, characterized in that, Based on a spatial three-dimensional mesh model, the seismic multi-attribute response set and the electrical multi-attribute response set are meshed and mapped to form a seismoelectric dual-response element matrix, including: Based on the spatial range of the target mining area, the detection depth parameters, and the grid resolution parameters, a three-dimensional spatial grid model is established, and the target mining area is spatially discretized to generate multiple grid cells with spatial coordinate information. Based on the spatial coordinate information of the grid cells, the seismic multi-attribute response set and the electrical multi-attribute response set are respectively subjected to spatial location matching processing to obtain the seismic response feature data and electrical response feature data of the corresponding grid locations. The seismic response characteristic data and electrical response characteristic data within the same grid cell are processed by coordinate unification, scale normalization and feature dimension matching to generate grid-level seismic and electrical dual-source response data; Based on the grid-level seismoelectric dual-source response data, establish the correspondence between grid cells and multi-source response parameters, and construct the seismoelectric dual-response cell matrix.

5. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 1, characterized in that, Based on the aforementioned three-dimensional spatial grid model, a discrete unit system for the mining area is constructed, and a geological constraint parameter library is established, including: Correlation analysis was performed on geological data, borehole information and lithological distribution data of the target mining area to extract lithological type parameters, water-bearing state parameters, pore structure parameters and formation stability parameters, and a set of basic geological parameters was generated. Based on the geological parameter set and the spatial three-dimensional grid model, spatial location matching processing is performed to map different geological parameters to corresponding grid cells, generating grid-level geological attribute data containing spatial location information. Based on the grid-level geological attribute data, parameter encoding is performed on each spatial grid unit to establish the correspondence between grid units and geological feature parameters, and a geological constraint feature matrix is ​​generated. Based on the geological constraint feature matrix, parameter classification, normalization, and structured storage are performed to construct the geological constraint parameter library.

6. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 1, characterized in that, Based on the geological constraint parameter library, the seismoelectric dual-response element matrix is ​​subjected to seismoelectric response consistency constraint fusion to generate a seismoelectric coupled response tensor field, including: The geological constraint parameter library and the seismoelectric dual response unit matrix are spatially correlated and matched to generate a set of seismoelectric response units containing geological constraint information. Based on the set of seismoelectric response units, the seismic response parameters, electrical response parameters, and geological constraint parameters corresponding to each spatial grid unit are extracted; Spatial correlation analysis is performed on the seismic response parameters and electrical response parameters to calculate the correlation between responses of different physical fields and generate a seismoelectric response consistency evaluation index. Construct a consistency coefficient matrix corresponding to the spatial grid based on the aforementioned seismic response consistency evaluation index; Based on the consistency coefficient matrix, the seismic response weights and electrical response weights in the seismoelectric dual-response element matrix are dynamically adjusted to generate an adaptive fusion weight field. The seismoelectric multi-source response data are fused based on the adaptive fusion weight field to generate the seismoelectric coupled response tensor field.

7. The method for combined seismic and electrical detection and identification of coal mine goaf areas as described in claim 6, characterized in that, Based on the adaptive fusion weight field, the seismoelectric multi-source response data is fused to generate the seismoelectric coupled response tensor field, including: Based on the adaptive fusion weight field, the seismic response parameters, electrical response parameters, and geological constraint parameters in the seismoelectric dual response unit matrix are weighted and combined to generate grid-level fusion response feature data. The grid-level fused response feature data is processed by multidimensional arrangement based on the spatial grid position relationship to construct multidimensional response data including spatial position dimension, seismic response dimension, electrical response dimension and geological constraint dimension; The multidimensional response data is subjected to tensor encoding processing to generate a seismoelectric coupling response tensor containing multi-physics response information; Based on the spatial continuity constraint, the seismoelectric coupling response tensor is spatially reconstructed and smoothed to obtain a continuously distributed seismoelectric coupling response tensor field.

8. A seismic-electric co-detection and identification system for coal mine goaf areas, characterized in that, The system is used to implement the seismic-electric collaborative detection and identification method for coal mine goaf areas according to any one of claims 1 to 7, the system comprising: The data acquisition module is used to acquire seismic exploration data and electrical resistivity exploration data of the target mining area, and extract seismic wave reflection attributes, energy attenuation attributes and wave impedance change attributes from the seismic exploration data to generate a set of seismic multi-attribute responses. The attribute extraction module is used to extract apparent resistivity distribution attributes, electrical gradient attributes, and electrical anomaly enhancement attributes from the electrical exploration data, and generate a set of electrical multi-attribute responses. The grid mapping module is used to perform grid mapping between the seismic multi-attribute response set and the electrical multi-attribute response set based on the spatial three-dimensional grid model, forming a seismoelectric dual-response unit matrix; The parameter library construction module is used to construct a discrete unit system of the mining area based on the spatial three-dimensional grid model and establish a geological constraint parameter library. The geological constraint parameter library includes at least lithological background parameters, water content range parameters, and pore structure parameters. The constraint fusion module is used to perform seismic-electric response consistency constraint fusion on the seismoelectric dual-response unit matrix based on the geological constraint parameter library, and generate a seismoelectric coupled response tensor field. The risk assessment module is used to identify anomalies, structural zoning, and risk level determination of goaf areas based on the seismoelectric coupling response tensor field, and output the spatial risk assessment results of goaf areas.

Images