An underground space positioning method, system, device and medium

Abstract

This invention provides a method, system, device, and medium for underground spatial positioning, belonging to the field of underground goaf detection. It includes acquiring surface deformation data using time-series InSAR technology; performing two filters on the surface deformation data based on standard deviation and DBSCAN clustering to extract continuous subsidence hotspot areas; iteratively optimizing goaf parameters using a genetic-particle swarm optimization algorithm to obtain the optimal goaf parameters for the area under study; and achieving underground goaf positioning based on the optimal goaf parameters. This improves the accuracy of underground parameter acquisition and the accuracy of actual underground goaf positioning.

Description

Technical Field

[0001] This invention belongs to the field of underground mining area detection, specifically involving an underground space positioning method, system, equipment and medium. Background Technology

[0002] The demand for and activities related to the development of underground resources and space are growing rapidly. This development causes underground disturbances that are transmitted to the surface, leading to abnormal landforms and features, damaging surrounding resources and the environment, and impacting production and the safety of life and property. Therefore, to prevent and control these disturbance-induced resource, environmental, and safety problems, it is essential to clearly define their spatial and temporal locations.

[0003] Spaceborne synthetic radar aperture interferometry (InSAR) has proven to be an effective remote sensing tool for mapping ground deformation caused by various natural or human activities. InSAR captures not only continuous deformation regions induced by underground mining goaf disturbances but also discrete deformation regions caused by many other factors (such as measurement errors and atmospheric errors). Directly using the raw deformation results for underground mining goaf location will reduce the accuracy of location due to these errors, thus affecting decision-making. Most existing methods for locating underground mining goafs use probabilistic integral models to describe the nonlinear relationship between underground mining goafs and surface deformation, primarily applied to the location of underground coal mining goafs. However, this model requires numerous parameters, and its location accuracy is highly correlated with the accuracy of parameter acquisition. In practical underground mining goaf location applications, obtaining these model parameters is challenging. Summary of the Invention

[0004] To address the problem of inaccurate positioning of existing underground mining areas, this invention provides an underground space positioning method, system, equipment, and medium.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A method for locating underground spaces includes the following steps:

[0007] Based on the temporal SAR images of the area under study, long-term series surface deformation data of the area are obtained; the surface deformation data are filtered once using the standard deviation; and the DBSCANS unsupervised machine learning clustering algorithm is used to filter the data based on the first filtering result to obtain a continuous high-density subsidence point set.

[0008] The parameters of the goaf are iteratively optimized based on the genetic-particle swarm optimization algorithm to obtain the optimal goaf parameters for the area under study, and the underground goaf is located based on the optimal goaf parameters.

[0009] The iterative optimization of goaf parameters based on the genetic-particle swarm optimization algorithm specifically involves: remotely interpreting the continuous high-density subsidence point set to obtain the initial threshold range of goaf parameters; randomly generating initial underground goaf parameters within the initial threshold range; constructing a mapping relationship between the underground goaf and surface deformation, and mapping the underground goaf parameters to surface subsidence parameters based on the mapping relationship; constructing a fitness function by subtracting the surface subsidence parameters from the InSAR monitoring results; iteratively optimizing the fitness function until the minimum value is reached or the number of iterations reaches the upper limit, which is the optimal goaf parameter.

[0010] Preferably, the surface deformation data is filtered once using the standard deviation, specifically including the following steps:

[0011] Key parameters related to setting the filtering threshold Specifically:

[0012] ;

[0013] in, This represents the average value of the pixels within the window. This represents the standard deviation of pixels within the window;

[0014] Based on the set window size, the study area is traversed, and a window stability coefficient is generated at each location. ;

[0015] After the traversal is complete, The region with the smallest value is taken as the deformation-stable region of the area to be studied, and the standard deviation of this region is used as the first filtering threshold.

[0016] The surface deformation data is filtered according to the first filtering threshold. Points with pixel values ​​below the threshold are considered to be significant deformation points and are retained, while the rest are discarded, resulting in surface deformation data after one filtering.

[0017] Preferably, the unsupervised machine learning clustering algorithm DBSCANS filtering based on the first filtering result specifically includes the following steps:

[0018] Based on the conditions of the area under study and prior information, the distance threshold, a core input parameter of the DBSCAN algorithm, is determined. and minimum points The value;

[0019] Randomly select one filtered surface deformation data point p as the point. If at point p... If the number of sample points in the neighborhood of point p is greater than or equal to min, then p is taken as the core point, and all unvisited sample points in the neighborhood of p are recursively checked and included in the same cluster until the cluster cannot be expanded; if the number of sample points in the neighborhood of point p is greater than or equal to min, then p is taken as the core point, and all unvisited sample points in the neighborhood of p are recursively checked and included in the same cluster, until the cluster can no longer ... If there are fewer than min points in the neighborhood, then point p is marked as a noise point;

[0020] After traversing all points, the algorithm ends, removing noise points and points not included in the cluster, resulting in a continuous high-density set of settlement points.

[0021] Preferably, the goaf parameters specifically include strike length, dip length, burial depth, dip angle, and azimuth angle.

[0022] Preferably, the mapping relationship between the underground goaf and surface deformation is as follows:

[0023] ;

[0024] Where L, W, and d represent the strike length, dip length, and burial depth of the goaf, respectively. ( ) represents the coordinates of the surface deformation point in the fault plane coordinate system. , , These represent the deformations induced by underground mining voids in the east-west horizontal direction, the north-south horizontal direction, and the vertical direction, respectively. , and The sub-tables represent the functional relationships between underground goaf parameters and surface deformation in the east-west, north-south, and vertical directions.

[0025] The present invention also provides an underground space positioning system, specifically comprising:

[0026] The data filtering module is used to acquire long-term surface deformation data of the study area based on the time-series SAR images of the study area; the surface deformation data is filtered once using the standard deviation; and the unsupervised machine learning clustering algorithm DBSCANS is used to filter the data based on the result of the first filtering to obtain a continuous high-density subsidence point set.

[0027] The spatial positioning module is used to iteratively optimize the parameters of the goaf area based on the genetic-particle swarm optimization algorithm to obtain the optimal goaf area parameters for the area under study, and to locate the underground goaf area based on the optimal goaf area parameters.

[0028] The iterative optimization of goaf parameters based on the genetic-particle swarm optimization algorithm specifically involves: remotely interpreting the continuous high-density subsidence point set to obtain the initial threshold range of goaf parameters; randomly generating initial underground goaf parameters within the initial threshold range; constructing a mapping relationship between the underground goaf and surface deformation, and mapping the underground goaf parameters to surface subsidence parameters based on the mapping relationship; constructing a fitness function by subtracting the surface subsidence parameters from the InSAR monitoring results; iteratively optimizing the fitness function until the minimum value is reached or the number of iterations reaches the upper limit, which is the optimal goaf parameter.

[0029] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps described in the underground space positioning method.

[0030] The present invention also provides a computer-readable storage medium storing a computer program that, when loaded by a processor, can execute the steps described in the underground space positioning method.

[0031] The underground space positioning method provided by this invention has the following beneficial effects:

[0032] This invention first utilizes a two-step filtering algorithm based on standard deviation and DBSCAN clustering to filter the original InSAR results twice, eliminating discrete deformation points caused by errors and retaining deformation points induced by continuous underground mining voids. For the retained deformation points, a genetic-particle swarm optimization algorithm is used to perform a nonlinear solution space search for the geometric parameters of the underground mining voids, combining the relationship between the underground mining voids and surface deformation. The diversity preservation mechanism of the genetic algorithm avoids premature convergence, while the rapid convergence of the particle swarm optimization algorithm quickly approximates the optimal solution, obtaining the optimal mining void parameters. This achieves the location of underground mining voids, significantly improving the positioning accuracy. Attached Figure Description

[0033] To more clearly illustrate the embodiments and design schemes of the present invention, the accompanying drawings required for this embodiment will be briefly described below. The drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0034] Figure 1 This is a flowchart of an underground space positioning method according to an embodiment of the present invention.

[0035] Figure 2 This is an InSAR spatiotemporal baseline map in an embodiment of the present invention.

[0036] Figure 3 This represents the time-series InSAR cumulative settlement of the study area in this embodiment of the invention.

[0037] Figure 4 This is the processing flow of the two-step filtering algorithm in this embodiment of the invention.

[0038] Figure 5 This is an example of extracting the azimuth angle of fire zone 1 in this embodiment of the invention.

[0039] Figure 6 This is an example of fire zone 2 azimuth angle extraction in this embodiment of the invention.

[0040] Figure 7 The above are the measured results of the underground goaf area in the embodiments of the present invention.

[0041] Figure 8 This is a schematic diagram of the location results of the underground goaf in an embodiment of the present invention. Detailed Implementation

[0042] To enable those skilled in the art to better understand and implement the technical solutions of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention and should not be construed as limiting the scope of protection of the present invention.

[0043] Example

[0044] This invention provides a method for locating underground spaces, such as... Figure 1 As shown, the specific steps include:

[0045] Step 1: Obtain surface deformation data using time-series InSAR technology.

[0046] First, time-series SAR images of the study area are acquired using the same type of SAR sensor. Then, time-series InSAR technology is used to obtain long-term surface deformation information of the study area.

[0047] Currently, the main methods used for time-series InSAR processing are permanent scatterer synthetic aperture radar interferometry (PS-InSAR), small baseline set-based synthetic aperture radar interferometry (SBAS-InSAR), and distributed scatterer-based synthetic aperture radar interferometry (DS-InSAR).

[0048] The core concept of PS-InSAR technology lies in utilizing multiple single-look SAR images covering the same study area, selecting one image as the master image, and then registering the other SAR images with this master image. In the time series, based on the stability of amplitude and phase information, persistent scatterers (PS) are identified and selected as targets. Through interferometry and terrain removal processing, differential interferometric phases based on PS targets are obtained, and the differential interferometric phases of adjacent PS targets are further subdivided. Based on the characteristics of different phase components in the interferometric phases after the two subdivisions, a deformation phase model is constructed to estimate deformation and residual terrain information.

[0049] SBAS-InSAR technology is based on interferometric pairs of multiple master images, utilizing high-coherence points to recover temporal deformation information of the study area. The workflow is as follows: First, the temporal and spatial baselines of multiple SAR images covering different time points of a specific region are calculated, and interferometric pairs are selected based on appropriate temporal and spatial baseline thresholds. Then, differential interferometry is performed on the selected interferometric pairs, followed by phase unwrapping. Finally, based on the subset formed by freely combined interferograms, the deformation parameters are estimated for the phase equations composed of all interferograms using either the least squares method or singular value decomposition (SVD).

[0050] The advantage of SBAS-InSAR technology lies in its ability to overcome the problem of poor interferometric coherence that may result from the selection of a single reference image in the PS-InSAR method, while reducing the demand for SAR data and improving computational efficiency. DS-InSAR, on the other hand, addresses the issue of missing permanent scatterers in the target area by identifying and selecting distributed scatterers (PS) as targets to ultimately obtain temporal deformation results of the land surface. This method can acquire a sufficient number of monitoring points even in areas lacking permanent scatterers, overcoming the limitation of PS-InSAR technology's reliance on the selection of permanent scatterers.

[0051] The optimal temporal InSAR results are selected and spatially interpolated using the Kriging interpolation method to obtain a large-scale fitted subsidence result for the target area.

[0052] Step 2: Use a two-stage filtering algorithm based on standard deviation and DBSCAN clustering to extract continuous subsidence hotspots from the surface deformation data.

[0053] S21: First-step filtering based on standard deviation. The inversion of underground mining voids mainly relies on the obvious subsidence characteristics of the surface. However, the original InSAR results contain uplift and small-scale subsidence induced by various factors around the study area, and these data cannot provide effective reference information for locating underground mining voids. Therefore, a standard deviation-based filtering method is used to filter the original deformation results. First, the key parameter related to the filtering threshold, the "window stability coefficient," is defined. ":

[0054] ;

[0055] in, This represents the average value of the pixels within the window. This represents the standard deviation of pixels within the window. The window size can be freely set according to the actual study area size (e.g., 200×200). After the window size is determined, the entire study area is traversed, and a window stability coefficient is generated at each location in each iteration. After the traversal is complete, the view The region with the smallest value is identified as the deformation-stable region of the study area, and the standard deviation of this region is used as the threshold for the first filtering step. Finally, the data of the study area is filtered once based on the determined threshold. Pixels with values ​​below the threshold are considered to be significantly deformed and are retained, while the rest are discarded.

[0056] S22: Second-step filtering based on DBSCAN clustering. DBSCAN is an unsupervised machine learning clustering algorithm that can divide regions with sufficient density of points into the same cluster.

[0057] First, determine the core input parameters of the DBSCAN algorithm based on the actual situation and prior information. The value of ), where, is the distance threshold, and min is the minimum number of points.

[0058] The DBSCAN algorithm is used to perform a second filtering on the result of the first filtering step. Specifically, for each point p, if it is within its... If there are fewer than min points in the neighborhood, point p is marked as a noise point; if p is not a noise point, it is considered a density-connected point, and p and all its density-reachable points are recursively included in the same cluster. This process continues until no more points can be included in the current cluster. The algorithm terminates when all points have been visited and each point has been assigned to a cluster or marked as noise; a set of continuous high-density settlement points for locating underground goaf areas is selected.

[0059] Step 3: Establish the relationship between the underground goaf and surface deformation. Considering the difficulty in obtaining geological parameters in practice, an elastic dislocation model with fewer required parameters is chosen to describe the relationship between the underground goaf and surface deformation. During the disturbance of the underground goaf, the subsidence at the top of the underground goaf can be regarded as the tension dislocation in the model, as shown in the following formula:

[0060] ;

[0061] The above equation demonstrates the transformation of ground points in different coordinate systems, involving three coordinate systems in total: the local Cartesian coordinate system of the ground. Fault coordinate system and the fault plane coordinate system Specifically, the coordinates in the local Cartesian coordinate system on the ground are... a little Based on the azimuth of underground space The expression of (the angle between the strike of the working face and true north) and the origin of the fault coordinate system in the local Cartesian coordinate system on the ground. Convert to fault coordinate system Then, based on the depth of the underground space With the inclination angle (the angle between the working surface and the horizontal plane). Transformed into the coordinate system of the fault plane Furthermore, for ground points in the fault coordinate system The surface deformation can be obtained as follows:

[0062] ;

[0063] In the above formula, ; ; ; ; .in, This represents the amount of subsidence at the top of the underground goaf. Represents the Poisson's ratio of the strata. and Let Lamé constant be . This represents the inclination angle of the underground space. Ultimately, the point in the fault plane coordinate system... The surface deformation generated at the location is transformed into fault coordinate system according to the relationship between the three spatial coordinate systems. , and thus , , , Substitute into the above formula , , The relationship between the goaf parameters and surface deformation data is obtained by solving the problem:

[0064] ;

[0065] Where L, W, and d represent the strike length, dip length, and burial depth of the goaf, respectively. ( ) represents the coordinates of the surface deformation point in the fault plane coordinate system. , , These represent the deformations induced by underground mining voids in the east-west horizontal direction, the north-south horizontal direction, and the vertical direction, respectively. , and The occurrence of these features is related to the rectangular shape of the goaf, and is caused by the integration and accumulation during the surface subsidence process solved using an elastic dislocation model. The sub-tables represent the functional relationships between underground goaf parameters and surface deformation in the east-west, north-south, and vertical directions.

[0066] Step 4: Optimize the relationship between low goaf parameters and surface deformation data based on the Genetic-Particle Swarm Optimization (GA-PSO) algorithm to obtain the optimal goaf parameters, and locate the goaf based on the optimal goaf parameters.

[0067] S41: Based on the continuous subsidence hotspot areas obtained in step two, InSAR remote sensing interpretation (i.e., visual interpretation) is used to preliminarily determine the threshold range of underground goaf parameters (strike length, dip length, dip angle, center coordinates, burial depth, height, etc.). Within this threshold range, an initial underground goaf parameter is randomly generated as the initial value for the first generation of iterations.

[0068] The deformation of the first-generation initial value towards the Line of Sight (LOS) of the land surface is denoted as... Vertical deformation is calculated by the relationship between goaf parameters and surface deformation data. Deformation of East and West and North-South Deformation The deformation in three directions is projected onto the satellite's LOS direction based on the satellite's incident angle and heading angle, specifically as follows:

[0069] ;

[0070] in, The angle of incidence of the satellite used; The heading angle of the satellite used.

[0071] The surface deformation calculated by the nonlinear function model is synthesized to the LOS direction and then subtracted from the InSAR monitoring results, which is defined as the fitness function:

[0072] ;

[0073] Where n is the number of pixels in the image, and j is the calculation and technical parameter. This represents the surface deformation values ​​obtained from satellite observations. This represents the simulated surface deformation value generated by forward simulation of the underground space parameters based on the elastic dislocation model in this iteration.

[0074] A lower fitness indicates a more accurate input size parameter. The first generation of parameters is randomly generated within a threshold range, resulting in a higher fitness value. An iterative genetic-particle swarm optimization algorithm is used, reducing the fitness value with each iteration until convergence.

[0075] S42: The genetic-particle swarm optimization algorithm is used to iterate the fitness function. When the fitness function reaches its minimum value or the number of iterations reaches its upper limit, the resulting goaf parameters are the optimal goaf parameters. The underground goaf in the target area is located based on the obtained optimal goaf parameters.

[0076] Genetic algorithms and particle swarm optimization (PSO) are two commonly used algorithms for finding the extrema of nonlinear functions based on solution space search, each with its own advantages and disadvantages. Genetic algorithms, through crossover and mutation, possess better global search capabilities and can escape abnormal extrema; however, their convergence is slower, and the number of iterations required to reach the optimal solution is usually large. PSO, a biomimetic algorithm based on the gregarious foraging behavior of birds, achieves extremely fast convergence by assigning different search directions to all individuals; however, its search direction is influenced by the optimal individual, making it prone to getting trapped in local optima and difficult to escape. The genetic-particle swarm optimization algorithm, combining these two algorithms for iterative optimization of the fitness function, inherits both the global search capability of genetic algorithms and the fast convergence of particle swarm optimization.

[0077] To verify the true effectiveness of this invention, application examples are used to further illustrate the invention:

[0078] Example 1:

[0079] Step 1: Obtain large-scale temporal surface deformation using temporal InSAR technology. Region A was selected as the study area. Region A has a typical temperate continental climate with diurnal temperature variations exceeding 10℃, summer high temperatures reaching 41.9℃, and annual precipitation less than 100 mm. The hot and dry climate contributes to a high tendency for coal spontaneous combustion. A fire zone in Region A is a typical case of coal fire disaster caused by an abandoned coal mine. Related firefighting projects began in 2017. Existing drilling data indicates that the coal seams in this fire zone are mostly thicker than 10 meters, with a depth of approximately 200 meters and a width between 80 and 260 meters. The dip angle of the coal seams is mostly above 70°, and there are numerous surrounding fissures. The fire is fierce, spreads rapidly, and the firefighting task is extremely challenging.

[0080] Thirty-eight Sentinel-1A scenes from January 2015 to December 2017 were selected, with azimuth and range resolutions of 5m and 20m respectively. The spatiotemporal baselines are as follows: Figure 2 As shown in the figure. This time period includes different phases before and after fire control, facilitating comparative analysis of subsidence phenomena before and after control. Furthermore, a 30m precision SRTM digital elevation model (DEM) was selected as an external reference for geocoding and topographic phase compensation. The temporal cumulative InSAR deformation monitoring results for the study area are shown below. Figure 3 As shown.

[0081] Step Two: A two-step filtering algorithm was used to remove discrete settlement points. First, a standard deviation-based first-step filtering algorithm was used, with a window size of 200×200 pixels to traverse the entire study area. The standard deviation of the window with the minimum global window stability coefficient (9.2 mm) was selected as the threshold for the first-step filtering. This reduced the number of deformation points from 88,330 to 5,774. Then, surface thermal anomalies were identified using thermal infrared imagery, and the parameters of the DBSCAN algorithm in the second-step filtering were determined based on the range of these anomalies. The threshold value is (125m, 408). This threshold is used to perform a second filtering on the results of the first step, further reducing the number of deformation points to 2151. The specific process is as follows: Figure 4 As shown.

[0082] Step 3: Locating the underground goaf.

[0083] Using the RM-DBC algorithm, the azimuth angles of the two fire zones were pre-determined as 61° and 48° based on the InSAR surface deformation monitoring results from step one. Figure 5 and Figure 6 As shown. Furthermore, based on the study area, the inversion ranges for other parameters are determined as follows: strike length. , tendency length center coordinates center coordinates The suspected coal fire area ② has the following characteristics: directional length , tendency length center coordinates center coordinates .

[0084] Subsequently, the GA-PSO algorithm, combined with the aforementioned initial threshold settings, was used to input the surface deformation data filtered by the two-step filtering algorithm. After 1000 iterations, the final inversion result, i.e., the parameters of the underground mining void to be determined, was output and compared with the actual data measured by the natural point method. Figure 7 As shown, the results are presented in Tables 1 and 2. Figure 8 As shown.

[0085] Table 1 Inversion Results of Underground Goaf 1

[0086]

[0087] Table 2 Inversion Results of Underground Mining Goaf Area 2

[0088]

[0089] This invention also provides an InSAR-based positioning system for underground goaf areas, specifically comprising:

[0090] The data filtering module is used to obtain long-term surface deformation data of the study area based on the time-series SAR images of the study area; the surface deformation data is filtered once using the standard deviation; and the unsupervised machine learning clustering algorithm DBSCANS is used to filter the data based on the result of the first filtering to obtain a continuous high-density set of subsidence points.

[0091] The spatial positioning module is used to iteratively optimize the parameters of the goaf based on the genetic-particle swarm optimization algorithm to obtain the optimal goaf parameters for the area under study, and to locate the underground goaf based on the optimal goaf parameters.

[0092] The genetic-particle swarm optimization algorithm is used to iteratively optimize the parameters of the goaf. Specifically, the following steps are taken: remote sensing interpretation of a continuous high-density subsidence point set is used to obtain the initial threshold range of the goaf parameters; initial underground goaf parameters are randomly generated within the initial threshold range; a mapping relationship between the underground goaf and surface deformation is constructed, and the underground goaf parameters are mapped to surface subsidence parameters based on the mapping relationship; the surface subsidence parameters are subtracted from the InSAR monitoring results to construct a fitness function; the fitness function is iteratively optimized, and the goaf parameters obtained when the minimum value is reached or the number of iterations reaches the upper limit are the optimal goaf parameters.

[0093] The modules in the aforementioned InSAR-based underground goaf positioning system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the computer device's memory as software, so that the processor can call and execute the corresponding operations of each module.

[0094] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps in an embodiment of an underground space positioning method. Specific implementation methods can be found in the method embodiments, and will not be repeated here.

[0095] Furthermore, the present invention also provides a non-transitory computer-readable storage medium containing instructions, on which a computer program is stored. For example, a memory containing instructions that can be executed by a processor of a computer device to perform the above-described method. For example, the non-transitory computer-readable storage medium may be a ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc. When the computer program is executed by the processor, it can implement the steps in an embodiment of an underground space positioning method. Specific implementation methods can be found in the method embodiments, which will not be repeated here.

[0096] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0097] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0098] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0099] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.

[0100] It should be noted that the specific embodiments described above enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way. Therefore, although the present invention has been described in detail in this specification and embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the present invention; and all technical solutions and improvements that do not depart from the spirit and scope of the present invention are covered within the protection scope of the present invention patent. No reference numerals in the claims should be construed as limiting the scope of the claims. Any simple variations or equivalent substitutions of technical solutions that can be readily obtained by those skilled in the art within the scope of the technology disclosed in the present invention are within the protection scope of the present invention.

Claims

1. A method for positioning in underground space, characterized in that, Includes the following steps: Based on the temporal SAR images of the area under study, obtain long-term series surface deformation data of the area under study; The surface deformation data were filtered once using the standard deviation, and the DBSCANS unsupervised machine learning clustering algorithm was then used to filter the data based on the first filtering result to obtain a continuous high-density set of settlement points. The parameters of the goaf are iteratively optimized based on the genetic-particle swarm optimization algorithm to obtain the optimal goaf parameters for the area under study, and the underground goaf is located based on the optimal goaf parameters. The iterative optimization of goaf parameters based on the genetic-particle swarm optimization algorithm specifically involves: remotely interpreting the continuous high-density subsidence point set to obtain the initial threshold range of goaf parameters; randomly generating initial underground goaf parameters within the initial threshold range; constructing a mapping relationship between underground goaf and surface deformation; and mapping the underground goaf parameters to surface subsidence parameters based on the mapping relationship. The fitness function is constructed by subtracting the surface subsidence parameters from the InSAR monitoring results. The optimal goaf parameters are obtained by iteratively optimizing the fitness function until the minimum value is reached or the number of iterations reaches the upper limit.

2. The underground space positioning method according to claim 1, characterized in that, The surface deformation data is filtered once using the standard deviation, specifically including the following steps: Key parameters related to setting the filtering threshold Specifically: ； in, This represents the average value of the pixels within the window. This represents the standard deviation of pixels within the window; Based on the set window size, the study area is traversed, and a window stability coefficient is generated at each location. ; After the traversal is complete, The region with the smallest value is taken as the deformation-stable region of the area to be studied, and the standard deviation of this region is used as the first filtering threshold. The surface deformation data is filtered according to the first filtering threshold. Points with pixel values ​​below the threshold are considered to be significant deformation points and are retained, while the rest are discarded, resulting in surface deformation data after one filtering.

3. The underground space positioning method according to claim 2, characterized in that, The DBSCANS filtering algorithm for unsupervised machine learning clustering based on the first filtering result specifically includes the following steps: Based on the conditions of the area under study and prior information, the distance threshold, a core input parameter of the DBSCAN algorithm, is determined. and minimum points The value; If a point is randomly selected from the filtered surface deformation data, and the data is at point p, then... If the number of sample points in the neighborhood of point p is greater than or equal to min, then p is taken as the core point. All unvisited sample points in the neighborhood of p are recursively checked and included in the same cluster until the cluster cannot be expanded further. ... If there are fewer than min points in the neighborhood, then point p is marked as a noise point; After traversing all points, the algorithm ends, removing noise points and points not included in the cluster, resulting in a continuous high-density set of settlement points.

4. The underground space positioning method according to claim 1, characterized in that, The parameters of the goaf specifically include strike length, dip length, burial depth, dip angle, and azimuth angle.

5. The underground space positioning method according to claim 4, characterized in that, The mapping relationship between the underground goaf and surface deformation is as follows: ； Where L, W, and d represent the strike length, dip length, and burial depth of the goaf, respectively. ( ) represents the coordinates of the surface deformation point in the fault plane coordinate system. , , These represent the deformations induced by underground mining voids in the east-west horizontal direction, the north-south horizontal direction, and the vertical direction, respectively. , and The sub-tables represent the functional relationships between underground goaf parameters and surface deformation in the east-west, north-south, and vertical directions.

6. An underground space positioning system, characterized in that, include: The data filtering module is used to obtain long-term surface deformation data of the study area based on the time-series SAR images of the study area. The surface deformation data is filtered once using the standard deviation, and the DBSCANS unsupervised machine learning clustering algorithm is used to filter the first filtering result to obtain a continuous high-density set of settlement points. The spatial positioning module is used to iteratively optimize the parameters of the goaf area based on the genetic-particle swarm optimization algorithm to obtain the optimal goaf area parameters for the area under study, and to locate the underground goaf area based on the optimal goaf area parameters. The iterative optimization of goaf parameters based on the genetic-particle swarm optimization algorithm specifically involves: remotely interpreting the continuous high-density subsidence point set to obtain the initial threshold range of goaf parameters; randomly generating initial underground goaf parameters within the initial threshold range; constructing a mapping relationship between underground goaf and surface deformation; and mapping the underground goaf parameters to surface subsidence parameters based on the mapping relationship. The fitness function is constructed by subtracting the surface subsidence parameters from the InSAR monitoring results. The optimal goaf parameters are obtained by iteratively optimizing the fitness function until the minimum value is reached or the number of iterations reaches the upper limit.

7. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is loaded by the processor, it is able to perform the steps of the method according to any one of claims 1 to 5.

Images