Monitoring and evaluation method and module for reconstruction of stratum heterogeneity characteristics of dump

Abstract

The application discloses a kind of monitoring and evaluation method and module of stratum heterogeneity feature of dump reconstruction, belong to the field of mine environmental protection and ecological restoration. Image data of dump is obtained, the spectral feature, texture feature and color feature of rock-soil body of dump are extracted to construct the classification feature set of rock-soil type of target mine dump area, the spatial distribution result of rock-soil type of dump is obtained by classification using classification feature set, stratum similarity index and stratum homogeneity index are calculated, and the heterogeneity feature of dump can be evaluated using stratum similarity index and stratum homogeneity index. It can comprehensively depict stratum heterogeneity feature of dump reconstruction from two dimensions of vertical and horizontal, two scales of whole and local, provide theoretical support and decision basis for dump process optimization and ecological restoration strategy formulation.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of mine environmental protection and ecological restoration, and specifically relates to a method and module for monitoring and evaluating the heterogeneity characteristics of reconstructed strata of a dump. BACKGROUND

[0002] Open-pit mining forms a large dump, and the stripped material is loosely accumulated in the dump, showing high spatial heterogeneity, which easily leads to water and soil loss, soil degradation and environmental pollution problems, becoming the core area of ecological degradation in the open-pit mine area, and bringing great challenges to ecological restoration and sustainable development of the mine area.

[0003] The prior art disclosed in Patent Publication No. CN114912680A describes a large internal dump instability hazard degree evaluation method, which relates to the field of geotechnical engineering disaster prevention and mitigation. The method combines qualitative and quantitative analysis based on triangular fuzzy numbers and generalized induced ordered weighted average (GIOWA) operators, and mainly includes: determining the scheme set, attribute set, decision set and evaluation set according to the actual engineering; constructing a multi-person fuzzy judgment matrix to determine the weight of each index; determining the evaluation matrix according to the expert language evaluation data; performing data aggregation to obtain the evaluation value; obtaining the evaluation result based on the evaluation value, and performing sorting and optimization. This method has large calculation amount, low evaluation efficiency and high use cost.

[0004] Existing research on the dump mainly focuses on the surface environment of the dump, and the spatial heterogeneity of the internal stratum structure is still insufficient. The spatial heterogeneity of the surface environment of the dump is largely controlled by the internal stratum structure. Existing research shows that the spatial heterogeneity characteristics formed by the disordered accumulation of rock-soil bodies in the dump often lead to non-uniform settlement, resulting in the development of surface crack networks, further changing the pore structure, hydraulic properties and nutrient availability of the reconstructed soil, and limiting the vertical expansion and ecological niche occupation ability of plant roots. SUMMARY

[0005] In view of the problems existing in the prior art, the present application provides a method and module for monitoring and evaluating the heterogeneity characteristics of reconstructed strata of a dump, which can comprehensively characterize the heterogeneity characteristics of the reconstructed strata of the dump from the vertical and horizontal dimensions, and the overall and local scales, providing theoretical support and decision basis for dump process optimization and ecological restoration strategy formulation.

[0006] To achieve the above object, the application provides the following technical scheme: a method for monitoring and evaluating the heterogeneous characteristics of the reconstructed stratum of a dump, image data of the dump are acquired, spectral features, texture features and color features of the rock-soil body of the dump are extracted to construct a classification feature set of the rock-soil type of the dump area of the target mining area, the classification feature set is used for classification to obtain a spatial distribution result of the rock-soil type of the dump, a stratum similarity index and a stratum homogeneity index are calculated using the spatial distribution result of the rock-soil type of the dump, and the heterogeneous characteristics of the dump can be evaluated using the stratum similarity index and the stratum homogeneity index.

[0007] The specific steps are as follows:

[0008] S1: high-resolution image data of the dump area of the target mining area are acquired using a UAV, orthophoto image DOM and digital elevation model DEM of the dump area are generated, and an original stratum sequence of the mining area is generated according to geological data and field survey data;

[0009] S2: spectral features, texture features and color features of the rock-soil body of the dump are extracted based on the orthophoto image DOM obtained in S1 to construct a classification feature set of the rock-soil type of the dump area of the target mining area, the rock-soil type is classified using a trained neural network, the spatial distribution of the rock-soil body in the dump area is divided, and a spatial distribution result of the rock-soil type of the dump is obtained;

[0010] S3: a reconstructed stratum sequence of the dump is constructed according to the spatial distribution result of the rock-soil type of the dump obtained in S2, and a stratum similarity index SSI is calculated;

[0011] S4: the composition proportion of the rock-soil type in each layer is counted according to the reconstructed stratum sequence of the dump, and a stratum homogeneity index GHI is calculated accordingly;

[0012] S5: the overall retention degree of the stratum of the dump is judged using the stratum similarity index SSI and the stratum homogeneity index GHI, the higher the SSI and GHI, the more ordered the stratum of the dump, the more conducive to the stability of the slope of the dump, and the easier the subsequent ecological restoration.

[0013] Further, in S1, the reconstruction method of the original stratum sequence of the mining area is as follows:

[0014] S201: the original stratum sequence of the target mining area is reconstructed based on the coal mine drilling column chart and field survey data , wherein represents the i-th layer in the original stratum, which is composed of a single rock-soil type without considering coal seams in an ideal state.

[0015] ​Furthermore, in S2, the neural networks used to classify soil and rock types are Maximum Likelihood Classification (MLC), Random Forest (RF), and Support Vector Machine (SVM).

[0016] Furthermore, in S3, the method for calculating the stratigraphic similarity index is as follows:

[0017] S301: Divide the DOM and DEM images of the spoil heap area, which record height information, into multiple evaluation units at equal intervals along the coal mining direction.

[0018] S302: Within each evaluation unit, based on the spatial distribution results of soil and rock types in the spoil heap obtained in step S2, the vertical direction is divided into multiple stratigraphic units according to preset vertical intervals to form a reconstructed stratigraphic sequence. Since the strata of the spoil heap are composed of a disordered mixture of various soil and rock types, the lithological composition of the spoil heap in the reconstructed stratigraphic sequence is represented in vector form as follows: ,in This represents the proportion of type k soil and rock in layer j;

[0019] S303: Based on the original stratigraphic sequence of the mining area Reconstructing the stratigraphic sequence with spoil heaps Inter-layer matching is performed using the order constraint principle, that is, when Layers and Interlayer matching pairs Layers can only be with The strata and subsequent strata are matched sequentially to maintain the logical order between the original strata; meanwhile, the similarity between each pair of matched strata is determined by... Layer in The proportion of soil and rock types in the layer is measured, and then a dynamic programming method is used to optimize the overall matching path to maximize the global similarity.

[0020] S304: Calculate the Vertical Stratigraphic Similarity Index (SSI) for this evaluation unit by combining the similarity of all matching pairs, using the following formula:

[0021] ,

[0022] In the formula, This refers to the number of stratigraphic layers in the spoil heap. Indicates the original strata In the spoil heap layer The proportion of soil and rock types in the data.

[0023] Furthermore, in S4, the method for calculating the formation homogeneity index is as follows:

[0024] S401: In each evaluation unit, the stratum sequence is reconstructed according to the divided dump, the composition proportion of the rock-soil type in each layer is counted, and the stratum homogeneity index GHI is calculated according to the composition proportion, and the calculation formula is as follows:

[0025] ,

[0026] ,

[0027] In the formula, is the proportion of the rock-soil type in the stratum; is the total number of rock-soil types; is the layer homogeneity index GHI; the higher the GHI value, the more uniform the rock-soil composition in the stratum, the lower the mixing degree, and the smaller the disturbance in the stripping and disposal process, and the better the structure preservation degree.

[0028] A monitoring and evaluation method for the non-homogeneity characteristics of a reconstructed stratum of a dump, characterized by comprising an image acquisition unit, an image processing unit, a spectral information extraction unit, a rock-soil type classification unit, a rock-soil type spatial distribution calculation unit of the dump, a stratum similarity index calculation unit, a stratum homogeneity index calculation unit, and a dump preservation degree judgment unit.

[0029] The image acquisition unit is used to acquire high-resolution image data of the target mining area dump region;

[0030] The image processing unit is used to generate orthophoto images DOM and digital elevation models DEM of the dump region from the high-resolution image data of the dump region;

[0031] The spectral information extraction unit is used to extract the spectral characteristics, texture characteristics and color characteristics of the rock-soil body of the dump to construct the classification feature set of the rock-soil type of the target mining area dump region;

[0032] The rock-soil type spatial distribution calculation unit is used to classify the rock-soil type, divide the spatial distribution of the rock-soil body in the dump region, and obtain the rock-soil type spatial distribution of the dump;

[0033] The stratum similarity index calculation unit calculates the stratum homogeneity index from the rock-soil type spatial distribution result of the dump;

[0034] The stratum homogeneity index calculation unit obtains the stratum homogeneity index of the rock-soil distribution in the reconstructed stratum at the same horizontal height in the reconstructed stratum sequence of the dump;

[0035] The non-homogeneity degree judgment unit judges the non-homogeneity degree of the stratum of the dump by using the stratum similarity index and the stratum homogeneity index.​

[0036] A computer device includes a processor and a memory, the processor being electrically connected to the memory, the memory being used to store instructions and data, and the processor being used to execute a method for monitoring and evaluating the heterogeneous characteristics of reconstructed strata in a spoil heap.

[0037] A computer-readable storage medium storing a computer program adapted to be loaded and executed by a processor for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps.

[0038] Beneficial Effects: Addressing the heterogeneous characteristics of spoil heaps, this study utilizes the Stratigraphic Similarity Index (SSI) and the Geometric Homogeneity Index (GHI) to assess the reconstructed strata. SSI quantifies the overall preservation of the original stratigraphic structure within the reconstructed strata; GHI assesses the homogeneity of soil and rock distribution at the same horizontal level, revealing differences in internal stratigraphic structure even with similar SSI values. These two indicators comprehensively characterize the heterogeneous features of reconstructed strata in spoil heaps from both vertical and horizontal dimensions, and from both overall and local scales, providing theoretical support and decision-making basis for optimizing spoil heap management processes and formulating ecological restoration strategies. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps according to the present invention.

[0040] Figure 2 This is a schematic diagram of the evaluation unit for the spoil heap in an embodiment of the present invention.

[0041] Figure 3 The following is a classification diagram of the distribution of soil and rock types in the spoil heap in this embodiment of the invention: Area I in the diagram represents the upper strata composed of various soil and rock materials; Area II represents the concentrated dumping area of ​​gangue in the spoil heap; Area III represents the lower strata mainly composed of gray sandstone.

[0042] Figure 4 This is a diagram showing the evaluation results of the similarity index of the spoil heap layer in an embodiment of the present invention.

[0043] Figure 5 This is a spatial distribution map of the homogeneity index of the waste dumping site in an embodiment of the present invention; in the map, area I represents the waste gangue dumping area; area II corresponds to the upper strata; area III represents the bottom strata; and area IV represents the middle strata zone. Detailed Implementation

[0044] The invention will now be further described with reference to the accompanying drawings.

[0045] like Figure 1 As shown, this invention discloses a method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps. The specific steps are as follows:

[0046] S1: Obtain high-resolution image data of the target mine dump area by using a drone, generate orthophoto image DOM and digital elevation model DEM of the dump area; generate the original stratum sequence of the mine area according to geological data and field survey data; the reconstruction method of the original stratum sequence of the mine area is:

[0047] S101: Reconstruct the original stratum sequence of the target mine area based on the coal mine drilling column chart and the field survey data , wherein represents the i-th layer in the original stratum, which is composed of a single rock-soil type without considering coal seams in an ideal state.

[0048] S2: Based on the orthophoto image DOM obtained in S1, extract the spectral features, texture features and color features of the rock-soil body of the dump to construct the classification feature set of the rock-soil type of the target mine dump area, classify the rock-soil type by using a neural network, divide the spatial distribution of the rock-soil body in the dump area, and obtain the spatial distribution result of the rock-soil type of the dump; three network models of maximum likelihood classification MLC, random forest RF and support vector machine SVM are used to classify the rock-soil type of the image after adding the characteristic band, and the SVM classification result is used to replace the MLC classification result for the mudstone type, and a training set and a verification set are artificially generated to complete the model training.

[0049] S3: According to the spatial distribution result of the rock-soil type of the dump obtained in S2, construct the reconstructed stratum sequence of the dump, calculate the stratum similarity index, use the stratum similarity index to evaluate the overall similarity between the reconstructed stratum and the original stratum established in step S1 in the vertical stratum sequence structure, reflect the overall retention degree of the original stratum structure characteristics in the reconstructed stratum, and use the stratum homogeneity index to evaluate the homogeneity of the rock-soil distribution in the same horizontal height; the calculation method of the stratum similarity index is:

[0050] S301: Divide the DOM and DEM images of the dump area into multiple evaluation units along the mining direction of the main coal at equal intervals;

[0051] S302: In each evaluation unit, according to the spatial distribution result of the rock-soil type of the dump obtained in S2, divide the vertical direction into multiple stratum units according to the preset vertical interval to form the reconstructed stratum sequence ; since the stratum of the dump is composed of multiple rock-soil types in disorder, the lithology composition in the reconstructed stratum sequence is represented in vector form as , wherein represents the proportion of the k-th rock-soil type in the j-th layer;

[0052] S303: According to the original stratum sequence of the mine area​ and the dump reconfigure the stratum sequence , the principle of sequential constraint is adopted for interlayer matching, that is, when layer and layer constitute a matching pair, layer can only be matched with layer and the strata after it in sequence to maintain the logical order between the original strata; at the same time, the similarity between each matching stratum pair is measured by the proportion of rock-soil types of layer in layer, and then the dynamic programming method is used to optimize the overall matching path to maximize the global similarity;

[0053] S304: Calculate the vertical stratum similarity index SSI of the evaluation unit by integrating the similarity of all matching pairs, as follows:

[0054] ,

[0055] In the formula, is the number of strata divided in the dump, represents the original stratum in the dump stratum .

[0056] S4: Calculate the stratum homogeneity index SSI from the spatial distribution of rock-soil types in the dump obtained in S2, and use the stratum homogeneity index to evaluate the degree of chaos of rock-soil distribution within the same horizontal height of the reconstructed stratum. The higher the stratum similarity index SSI and the stratum homogeneity index GHI, the more ordered the dump stratum is, the more conducive to the stability of the dump slope, and the easier the subsequent ecological restoration is. The calculation method of the stratum homogeneity index is:

[0057] S401: In each evaluation unit, according to the dump reconfigured stratum sequence divided in S302, the proportion of rock-soil types within each layer is counted, and the stratum homogeneity index GHI is calculated accordingly, as follows:

[0058] ,

[0059] ,

[0060] In the formula, is the proportion of the th rock-soil type in the stratum; is the total number of rock-soil types; is the layer homogeneity index GHI; the higher the GHI value, the more uniform the rock-soil composition within the stratum, the lower the degree of mixing, and the better the structure preservation during the stripping and disposal process.

[0061] Example: Monitoring and evaluation of the heterogeneity characteristics of the reconstructed stratum in the dump of Haerwus Coal Mine:

[0062] Step (1): Data collection and preprocessing: In July 2024, DJI Mavic 3 multispectral version unmanned aerial vehicle was used to collect remote sensing data of the dump of Haerwus Coal Mine; the original image spatial resolution was 5 cm x 5 cm, and after radiation correction and image splicing preprocessing, the orthographic image DOM and digital elevation model DEM of the dump were generated. To improve the calculation efficiency, the image resolution was resampled to 20 cm x 20 cm. In addition, along the tangent direction of the coal mining direction, every 100 m was divided into an evaluation unit, a total of 24 spatial units were constructed, as the basic spatial unit for subsequent reconstruction of stratum structure heterogeneity disturbance feature identification and quantitative analysis, as shown in Figure 2 .

[0063] Step (2): Classification sample set construction: Based on the preprocessed high-resolution remote sensing image, visual interpretation method was used to construct the classification sample set of rock-soil types in the dump, a total of 706 sample points, and according to the proportion of 7:3, it was divided into training set and validation set.

[0064] Step (3): Extract classification features: According to the original spectral information of unmanned aerial vehicle remote sensing image, including blue (Blue), green (Green), red (Red) three spectral bands, the texture spectral features and color spectral features of different rock-soil bodies were obtained, and the classification feature set of different rock-soil body types in the dump of open-pit coal mine was constructed; among them, the texture spectral features include mean (Mean), variance (Variance), homogeneity (Homogeneity), contrast (Contrast), dissimilarity (Dissimilarity), entropy (Entropy), second moment (Second Moment) and correlation (Correlation) 8 parameters; color spectral features include excess red index (Excess Red Index, ExR), excess green index (Excess Green Index, ExG), excess blue index (Excess Blue Index, ExB), green-red difference index (Green-Red Difference Index, GRDI), green-blue difference index (Green-Blue Difference Index, GBDI), and also include hue (H), saturation (S) and brightness (V) in HSV color space, the above 19 classification features, as shown in Table 1:

[0065] Table 1: Classification feature set and description of rock-soil types in the dump

[0066] ;

[0067] where r, g, b represent the red, green, and blue bands of the image, respectively; M, N are the dimensions of the gray level co-occurrence matrix; i, j represent the gray level indices corresponding to the gray values of the reference pixel and the adjacent pixel in the matrix; d is the pixel distance, is the direction angle (e.g., 0°, 45°, 90°, 135°), which together determines the construction method of the gray level co-occurrence matrix; represents the probability of the simultaneous occurrence of gray levels i and j at a given distance and direction; μ is the mean, is the variance, which represents the concentration tendency and dispersion degree of the gray distribution, respectively.

[0068] Step (4): Classification feature screening: the separability of the texture spectral features and color spectral features between different rock-soil types is calculated using the J-M distance method. From the 19 classification features, the feature bands with a J-M distance greater than 1.8 are selected to ensure that the selected features have good inter-class discrimination ability. A total of 6 color spectral features, including H, S, GRDI, GBDI, ExR, and ExB, are selected.

[0069] Step (5): Feature band dimension reduction: principal component analysis (PCA) is performed on the selected 6 color spectral features for dimension reduction, and the top two principal component bands with the highest contribution are selected as the final classification feature bands. The cumulative contribution rate of the top two principal component bands reaches 98.84%, effectively retaining the main information of the original data and improving the classification accuracy.

[0070] Step (6): Rock-soil type classification and optimization: three types of neural networks, including maximum likelihood classification (MLC), random forest (RF), and support vector machine (SVM), are used to classify the rock-soil types in the image after adding the feature bands. By comparing the classification accuracies of different algorithms and conducting detailed analysis based on the confusion matrix, the final optimization scheme is adopted, which replaces the SVM classification results for the mudstone type based on the MLC classification results. The training set and validation set generated in step (2) are used for model training. The overall classification accuracy of this scheme reaches 76.92%, and the Kappa coefficient is 0.72. The classification results are shown in Figure 3 , Figure 3 where region I represents the upper stratum composed of multiple rock-soil materials; region II represents the concentrated discharge area of the waste rock in the dump; and region III represents the lower stratum mainly composed of gray sandstone.

[0071] Step (7): Stratum similarity index calculation: first, based on the coal mine drilling column chart and field survey data, the original stratum sequence of the study area is reconstructed where represents the i-th stratum in the original stratum, layer, which is composed of a single rock-soil type in an ideal state. Subsequently, within each evaluation unit, the vertical direction is divided into multiple stratigraphic units according to the aforementioned rock-soil classification results and at a set vertical interval (5 m), forming a reconstructed stratigraphic sequence ; unlike the original stratum, the dump stratum formed through the processes of stripping, mining, and transportation composed of a disordered mixture of multiple rock-soil types, the lithological composition of which can be represented in vector form as where represents the proportion of the kth rock-soil type in the jth layer. The original stratigraphic sequence and the dump stratigraphic sequence are constructed, and then layer-to-layer matching is performed using the sequential constraint matching principle, i.e., when and are matched, can only be matched with and the strata after it to maintain the logical order between the original strata. At the same time, the similarity between each pair of matched strata is measured by the proportion of rock-soil types in in , and then the dynamic programming method is used to optimize the overall matching path to maximize the global similarity. Finally, the vertical stratigraphic similarity index of the evaluation unit is calculated by synthesizing the similarity of all matching pairs, and the formula is as follows:

[0072] ,

[0073] where is the number of stratigraphic layers divided in the dump, and represents the proportion of rock-soil types in the original stratum in the dump stratum ;

[0074] The calculation results of stratigraphic similarity are shown in Figure 4As shown, based on spatial location characteristics, the waste dump of the Haerwusu open-pit coal mine was gridded into 24 evaluation units (numbered Unit0-Unit23). These 24 evaluation units were then divided into three structural zones. Evaluation units numbered Unit0-Unit6 are located in the low-elevation area of ​​the waste dump (<1150m), mainly consisting of the lower components of the original strata, such as gangue, mudstone, and grayish-white sandstone. The stratigraphic similarity index (SSI) values ​​of most units are stable at around 0.6, indicating a relatively complete stratigraphic sequence structure. Evaluation units numbered Unit7-Unit12 have a higher overall elevation and contain more upper stripped material. The SSI curves in this area show a significant decrease, reflecting a higher intensity of disturbance to the upper strata caused by mining. Evaluation units numbered Unit13-Unit23 mainly cover the waste dump area and the endwall area of ​​the mining area. The terrain is complex, and the SSI curves in each evaluation unit show dramatic fluctuations, making it difficult to identify clear vertical structural features. The overall degree of disturbance is high, reflecting stronger spatial heterogeneity.

[0075] Step (8): Calculation of stratigraphic homogeneity index: Within each evaluation unit, the stratigraphic sequence is reconstructed based on the spoil heaps defined in the above steps. The composition ratio of soil and rock types within each layer is statistically analyzed, and the stratigraphic homogeneity index is calculated accordingly. The calculation formula is as follows:

[0076] ,

[0077] ,

[0078] In the formula, For the first The proportion of each type of soil and rock in this stratum; This represents the total number of soil and rock types. This is the information entropy value.

[0079] like Figure 5 As shown, the distribution characteristics of the stratigraphic homogeneity index (GHI) within the three-dimensional space of the spoil heap are illustrated. Overall, the GHI value varies between 0.04 and 0.98, reflecting the significant spatial heterogeneity of the stratigraphic structure within the spoil heap. The bottom strata and waste disposal areas (such as regions I and III) exhibit relatively high overall GHI values, displaying a continuous, high-value distribution, indicating that the material composition in these areas is relatively homogeneous, the depositional structure is relatively stable, and the internal consistency of the structure is high. In contrast, with increasing depositional height, especially in region II in the north, the GHI value decreases significantly, and the spatial distribution becomes more complex. Combined with the soil and rock type classification results of the spoil heap, it can be seen that the material sources in this area are more diverse, and the depositional process is more complex, resulting in poor overall stratigraphic homogeneity and reflecting a greater degree of disturbance from mining.

[0080] The analysis results of SSI and GHI show that the lower layer of the dump is of strong integrity, high consistency and good vertical continuity, while the upper layer is more complex, with loose combination, high degree of mixed materials and typical "stable lower and chaotic upper" spatial heterogeneous structure.

[0081] The above-described embodiments are merely preferred modes of the present application, and are not intended to limit the scope of the present application. Various modifications and improvements to the technical solutions of the present application made by those skilled in the art without departing from the design spirit of the present application shall fall within the protection scope of the present application as defined by the claims.

Claims

1. A method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in a spoil heap, characterized in that, The specific steps are as follows: S1: Use drones to acquire high-resolution image data of the target mining area's spoil heap, and generate orthophoto DOM and digital elevation model (DEM) of the spoil heap area; generate the original stratigraphic sequence of the mining area based on geological data and field survey data; S2: Based on the orthophoto DOM obtained in S1, extract the spectral features, texture features and color features of the soil and rock bodies in the spoil heap to construct a classification feature set of soil and rock types in the spoil heap area of ​​the target mining area. Use the trained neural network to classify the soil and rock types, divide the spatial distribution of soil and rock bodies in the spoil heap area, and obtain the spatial distribution results of soil and rock types in the spoil heap. S3: Based on the spatial distribution results of soil and rock types of the spoil heap obtained in S2, construct the reconstructed stratigraphic sequence of the spoil heap and calculate the stratigraphic similarity index (SSI). The method for calculating the stratigraphic similarity index is as follows: S301: Divide the DOM and DEM images of the spoil heap area, which record height information, into multiple evaluation units at equal intervals along the coal mining direction. S302: Within each evaluation unit, based on the spatial distribution results of soil and rock types in the spoil heap obtained in step S2, the vertical direction is divided into multiple stratigraphic units according to preset vertical intervals to form a reconstructed stratigraphic sequence. Since the strata of the spoil heap are composed of a disordered mixture of various soil and rock types, the lithological composition of the spoil heap in the reconstructed stratigraphic sequence is represented in vector form as follows: ,in This represents the proportion of type k soil and rock in layer j; S303: Based on the original stratigraphic sequence of the mining area Reconstructing the stratigraphic sequence with spoil heaps Inter-layer matching is performed using the order constraint principle, that is, when Layers and Interlayer matching pairs Layers can only be with The strata and subsequent strata are matched sequentially to maintain the logical order between the original strata; meanwhile, the similarity between each pair of matched strata is determined by... Layer in The proportion of soil and rock types in the strata is measured, and then a dynamic programming method is used to optimize the overall matching path to maximize global similarity; original stratigraphic sequence ,in Representing the first in the original strata layer; S304: Calculate the Vertical Stratigraphic Similarity Index (SSI) for this evaluation unit by combining the similarity of all matching pairs, using the following formula: ， In the formula, This refers to the number of stratigraphic layers in the spoil heap. Indicates the original strata In the spoil heap layer The proportion of soil and rock types in the content; S4: Based on the reconstructed stratigraphic sequence of the spoil heap, the composition ratio of soil and rock types in each layer is statistically analyzed, and the stratigraphic homogeneity index GHI is calculated accordingly. S5. Use the stratigraphic similarity index (SSI) and the stratigraphic homogeneity index (GHI) to determine the overall retention degree of the strata at the spoil heap site.

2. The method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps according to claim 1, characterized in that: In S1, the method for reconstructing the original stratigraphic sequence of the mining area is as follows: S201: Reconstructing the original stratigraphic sequence of the target mining area based on coal mine borehole columnar sections and field survey data. ,in Representing the first in the original strata In an ideal scenario, coal seams are not considered; the layer consists of a single type of rock and soil.

3. The method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps according to claim 2, characterized in that: In S2, the neural networks used to classify soil and rock types are Maximum Likelihood Classification (MLC), Random Forest (RF), and Support Vector Machine (SVM).

4. The method for monitoring and evaluating the heterogeneity characteristics of reconstructed strata in spoil heaps according to claim 1, characterized in that: In S4, the formation homogeneity index is calculated as follows: S401: Within each evaluation unit, reconstruct the stratigraphic sequence based on the spoil heaps defined in step S302, statistically analyze the composition ratio of soil and rock types within each layer, and calculate the stratigraphic homogeneity index GHI accordingly. The calculation formula is as follows: ， ， In the formula, For the first The proportion of each type of soil and rock in this stratum; This represents the total number of soil and rock types.

5. A module used in the monitoring and evaluation method for the heterogeneity characteristics of reconstructed strata in a spoil heap as described in any one of claims 1-4, characterized in that: It includes an image acquisition unit, an image processing unit, a spectral information extraction unit, a soil and rock type classification unit, a soil and rock type spatial distribution calculation unit for spoil heaps, a stratigraphic similarity index calculation unit, a stratigraphic homogeneity index calculation unit, and a spoil heap retention degree judgment unit. Image acquisition unit is used to acquire high-resolution image data of the target mining area's spoil heap area; The image processing unit is used to generate orthophotos (DOM) and digital elevation models (DEM) of the spoil heap area from high-resolution image data of the spoil heap area. The spectral information extraction unit is used to extract the spectral, texture, and color features of the soil and rock bodies in the spoil heap to construct a classification feature set of soil and rock types in the spoil heap area of ​​the target mining area; The spatial distribution calculation unit for soil and rock types in the spoil heap is used to classify soil and rock types, divide the spatial distribution of soil and rock bodies within the spoil heap area, and obtain the spatial distribution of soil and rock types in the spoil heap. The stratigraphic similarity index calculation unit calculates the stratigraphic homogeneity index based on the spatial distribution results of soil and rock types in the spoil heap; The stratigraphic homogeneity index calculation unit obtains the stratigraphic homogeneity index of the distribution of soil and rock within the same horizontal height of the reconstructed stratigraphic sequence in the constructed spoil heap through the stratigraphic homogeneity index; The heterogeneity assessment unit for spoil heaps uses stratigraphic similarity index and stratigraphic homogeneity index to determine the heterogeneity of the strata at the spoil heap site.

6. A computer device, characterized in that, It includes a processor and a memory, the processor being electrically connected to the memory, the memory being used to store instructions and data, and the processor being used to execute the monitoring and evaluation method for reconstructing the heterogeneity characteristics of the strata in the spoil heap as described in any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program adapted to be loaded by a processor and executed by the monitoring and evaluation method for reconstructing the heterogeneity characteristics of the strata in the spoil heap as described in any one of claims 1-4.

Images