Method and device for predicting vegetation coverage rate of mining area
By constructing a hybrid additive land impact model and a land likelihood regression model, combined with nonlinear modeling techniques, the problem of accuracy in vegetation cover prediction in mining areas was solved, achieving more accurate vegetation cover prediction and supporting the sustainable development of mining areas.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENHUA SHENDONG COAL GRP
- Filing Date
- 2026-01-08
- Publication Date
- 2026-05-15
AI Technical Summary
Existing technologies struggle to accurately predict vegetation cover within mining areas, especially in complex mining environments. Linear models cannot account for the nonlinear characteristics and complex feedback mechanisms of vegetation cover, leading to inaccurate predictions.
A hybrid additive land impact model and a land likelihood regression model are constructed. By combining nonlinear modeling methods, vegetation impact factors are obtained, the weights of impact points are determined, and a land likelihood weighted nonlinear model is constructed to comprehensively consider the environmental conditions and land use status of different areas in the mining area.
It improves the accuracy and robustness of vegetation coverage prediction, enabling a more comprehensive understanding of the changing patterns of vegetation coverage in mining areas and providing technical support for the sustainable development of mining areas.
Smart Images
Figure CN122048112A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of artificial intelligence technology, and in particular to a method and apparatus for predicting vegetation coverage in mining areas. Background Technology
[0002] With the increasing demand for coal and the accelerated pace of mining, large areas of land have been destroyed, vegetation in mining areas has degraded, and the ecological environment has been damaged, hindering the sustainable development of mining areas. Since vegetation cover is a crucial indicator for evaluating the health of the ecological environment in mining areas, predicting vegetation cover can play a vital role in understanding the healthy development status of mining areas and promoting their sustainable development. Therefore, accurately predicting vegetation cover in mining areas is of paramount importance. Summary of the Invention
[0003] This disclosure is made in view of the above-mentioned problems. This disclosure provides a method, apparatus, equipment, medium, and product for predicting vegetation cover in mining areas.
[0004] According to one aspect of this disclosure, a method for predicting vegetation cover in a mining area is provided, comprising: Obtain vegetation impact factors for the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage rate of the mining area; Based on the vegetation impact factors, a mixed additive land impact model is constructed for the target mining area; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors in the target mining area; The weights of each influencing point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and the vegetation influence factors; wherein, the land likelihood regression model is used to determine the vegetation coverage of the target mining area; The land likelihood regression model is modified based on the mixed additive land impact model until the prediction accuracy reaches a preset accuracy threshold, resulting in a land likelihood weighted nonlinear model. The vegetation coverage of the target mining area is then predicted using the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction result.
[0005] Furthermore, according to one aspect of this disclosure, constructing a mixed additive land impact model corresponding to the target mining area based on the vegetation impact factors includes: Determine the correlation between the vegetation influencing factors and the vegetation coverage of the target mining area; Vegetation impact factors whose correlation is greater than a preset correlation threshold are identified as target vegetation impact factors. Determine the basis functions, and obtain the smooth function by fitting based on the basis functions; Based on the target vegetation impact factor and the smoothing function, the mixed additive land impact model is constructed.
[0006] Furthermore, the mixed additive land impact model described in one aspect of this disclosure meets the following conditions: ; Wherein, Y is used to indicate the vegetation cover rate calculated by the mixed additive land impact model. β 0 represents the intercept term. S j ( ) indicates the first ( ). j A smoothing function for each target vegetation influencing factor. j The value ranges from 1 to m, where m is the number of target vegetation influencing factors. X j For the first j One target vegetation influencing factor, For land type processing functions, ε This is the error term of the mixed additive land impact model.
[0007] Furthermore, the processing function for the land type according to one aspect of this disclosure meets the following conditions: ; β k This is the coefficient corresponding to the k-th land use type, where k ranges from 1 to h, and h is the number of land use types. k (LandType) is the indicator function corresponding to the kth land use type; in, k1 is used to indicate that the land use type is a coal mining area.
[0008] Furthermore, according to one aspect of this disclosure, the weights of each influencing point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and the vegetation influence factors, including: In the target mining area, a target point and a target area are determined; wherein, the land use type of the target point is a coal mining area, and the target area is an area generated with the target point as the center; Identify influence points in the target area that share the same land use type as the target point; Based on the long radius and short radius corresponding to the target area, the weight of each influence point in the target mining area is determined; A land likelihood regression model is constructed based on the weights and the vegetation influence factors.
[0009] Furthermore, according to one aspect of this disclosure, determining the weight of each influence point in the target mining area based on the long and short radii corresponding to the target area includes: Based on formula Determine the weight of each of the aforementioned influencing points in the target mining area; Among them, W i，p Let d be the weight of the p-th influencing point relative to the target point i, where p ranges from 1 to q, q is the number of influencing points in the target mining area, and d is the weight of the p-th influencing point relative to the target point i. ip Let R1 be the distance from the p-th influencing point to the target point i, R2 be the minor radius, R1 be the major radius, and y be the distance from the p-th influencing point to the target point i. p Let P be the p-th influencing point, and P be the set of influencing points.
[0010] Furthermore, the land likelihood regression model described in one aspect of this disclosure meets the following conditions: ; Among them, Y i β0(u) represents the vegetation cover predicted by the land likelihood regression model. i ,v i ) is the intercept of the target point, β j (u i ,v i ) represents the parameter of the j-th target vegetation impact factor, where j ranges from 1 to m, and m is the number of target vegetation impact factors. W i，p Let X be the weight of the p-th influencing point relative to the target point i. ij Let ε be the value of the j-th target vegetation impact factor. i This is the error term corresponding to the target point.
[0011] Furthermore, according to one aspect of this disclosure, a land likelihood weighted nonlinear model is obtained by combining the mixed additive land impact model and the land likelihood regression model for modification, comprising: Determine the nonlinear relationships corresponding to the target vegetation impact factors in the mixed additive land impact model; The nonlinear relationship is replaced with the linear relationship corresponding to the target vegetation influence factor in the land likelihood regression model to obtain the land likelihood weighted nonlinear model.
[0012] Furthermore, the land likelihood weighted nonlinear model according to one aspect of this disclosure meets the following conditions: ; Among them, Y ’ i W represents the vegetation cover predicted by the land likelihood weighted nonlinear model. i，p Let be the weight of the p-th influencing point relative to the target point i.β 0 represents the intercept term, f j ( ) is the smoothing function of the j-th target vegetation influence factor, X ij Let be the value of the j-th target vegetation influence factor. For land type processing functions, ε This is the error term.
[0013] According to another aspect of this disclosure, a device for predicting vegetation cover in a mining area is provided, comprising: The acquisition module is used to acquire vegetation impact factors of the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage of the mining area. The first model construction module is used to construct a mixed additive land impact model corresponding to the target mining area based on the vegetation impact factors; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors of the target mining area; The second model construction module is used to determine the weight of each influencing point in the target mining area, and to construct a land likelihood regression model based on the weight and the vegetation influence factor; wherein, the land likelihood regression model is used to determine the vegetation coverage of each influencing point; The correction module is used to correct the prediction based on the mixed additive land influence model and the land likelihood regression model until the prediction accuracy reaches a preset accuracy threshold, thereby obtaining a land likelihood weighted nonlinear model. The land likelihood weighted nonlinear model is then used to predict the vegetation coverage of the target mining area, and the vegetation coverage prediction result is obtained.
[0014] According to another aspect of this disclosure, an electronic device is provided, comprising: a processor, a memory, and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is in operation, the processor communicates with the memory via the bus, and the machine-readable instructions, when executed by the processor, perform the steps of the first aspect above, or any possible implementation of the first aspect.
[0015] According to another aspect of this disclosure, a computer-readable storage medium is provided that stores a computer program, which, when executed by a processor, performs the steps of the first aspect or any possible implementation thereof.
[0016] According to another aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps of the method described in the above aspect.
[0017] As will be described in detail below, a method and apparatus for predicting vegetation cover in mining areas according to embodiments of this disclosure are provided. A mixed additive land impact model is constructed by acquiring vegetation influencing factors. This model integrates vegetation influencing factors and uses nonlinear modeling to overcome the shortcomings of traditional methods that do not consider the complex negative feedback mechanisms and nonlinear characteristics of these factors, thereby providing a more comprehensive and accurate understanding of the changing patterns and trends of vegetation cover in mining areas. Furthermore, a land likelihood regression model is constructed by determining the weights of influencing points, and then modified using the mixed additive land impact model to obtain a land likelihood weighted nonlinear model. This process incorporates the spatial heterogeneity reflected in the differences in environmental conditions and land use in different areas of the mining area into the model. The coupling of these two models improves the robustness and applicability of the model, making its prediction of vegetation cover in mining areas more accurate and providing strong technical support for the sustainable development of mining areas.
[0018] It should be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further illustration of the claimed technology. Attached Figure Description
[0019] The above and other objects, features, and advantages of this disclosure will become more apparent from the more detailed description of the embodiments thereof in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this disclosure and form part of the specification. They are used together with the embodiments of this disclosure to explain the disclosure and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.
[0020] Figure 1 A flowchart illustrating a method for predicting vegetation coverage in mining areas, as provided in this embodiment of the disclosure.
[0021] Figure 2 A schematic diagram of basis function fitting for the method of predicting vegetation coverage in mining areas provided in this embodiment of the disclosure.
[0022] Figure 3 This disclosure provides a schematic diagram illustrating the determination of the influence radius of a method for predicting vegetation coverage in mining areas, based on an embodiment.
[0023] Figure 4 A schematic diagram of the influence points of the method for predicting vegetation coverage in mining areas provided in this embodiment of the disclosure.
[0024] Figure 5 A schematic diagram of the weight curve of the method for predicting vegetation coverage in mining areas provided in the embodiments of this disclosure.
[0025] Figure 6 A schematic diagram illustrating the accuracy verification of the method for predicting vegetation coverage in mining areas provided in this embodiment of the disclosure.
[0026] Figure 7 This is a schematic diagram of a device for predicting vegetation coverage in a mining area, provided as an embodiment of this disclosure.
[0027] Figure 8 This is a schematic diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation
[0028] To make the objectives, technical solutions, and advantages of this disclosure more apparent, exemplary embodiments according to this disclosure will now be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this disclosure, and not all embodiments of this disclosure. It should be understood that this disclosure is not limited to the exemplary embodiments described herein.
[0029] The components of the embodiments of this disclosure, typically described and illustrated in the accompanying drawings, can be arranged and designed in a variety of different configurations. Therefore, the following detailed description of the embodiments of this disclosure provided in the drawings is not intended to limit the scope of the claimed disclosure, but merely to illustrate selected embodiments of the disclosure. All other embodiments obtained by those skilled in the art based on the embodiments of this disclosure without inventive effort are within the scope of this disclosure.
[0030] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.
[0031] In this document, the term "and / or" merely describes a relationship, indicating that three relationships can exist. For example, A and / or B can represent three cases: A alone, A and B simultaneously, and B alone. Furthermore, the term "at least one" in this document means any combination of at least two of any one or more elements. For example, including at least one of A, B, and C can mean including any one or more elements selected from the set consisting of A, B, and C.
[0032] Research has revealed that with the increasing demand for coal and the accelerated pace of mining, large areas of land have been destroyed, vegetation in mining areas has degraded, and the ecological environment has been damaged, hindering the sustainable development of mining areas. Since vegetation cover is a crucial indicator for evaluating the health of the ecological environment in mining areas, predicting vegetation cover can play a vital role in understanding the healthy development status of mining areas and promoting their sustainable development. Therefore, accurately predicting vegetation cover in mining areas is of paramount importance.
[0033] Among related technologies, it is difficult to achieve dynamic measurement of vegetation cover over a large mining area. Currently, remote sensing technology, with its advantages of large-scale coverage, multi-temporal applicability, applicability to different scales, and continuous observation capabilities, has become an important tool for vegetation cover inversion modeling in mining areas. Among these methods, statistical models, primarily based on linear regression, are widely used and can establish empirical statistical relationships between vegetation cover and other natural or anthropogenic factors.
[0034] However, due to the complexity of actual conditions in mining areas, the analysis of vegetation coverage involves multiple interacting factors such as climate change, soil properties, and human activities. The impact of these factors on vegetation coverage may not be linear and involves complex feedback mechanisms. The nonlinear characteristics of vegetation coverage make it impossible for linear models to accurately model and analyze the vegetation coverage in mining areas.
[0035] Based on the above research, this disclosure provides a method for predicting vegetation cover in mining areas. By acquiring vegetation influencing factors, a mixed additive land impact model is constructed. This model integrates vegetation influencing factors and uses nonlinear modeling to overcome the shortcomings of traditional methods that do not consider the complex negative feedback mechanisms and nonlinear characteristics of these factors. This allows for a more comprehensive and accurate understanding of the changing patterns and trends of vegetation cover in mining areas. Furthermore, a land likelihood regression model is constructed by determining the weights of influencing points, and then modified using the mixed additive land impact model to obtain a land likelihood weighted nonlinear model. This process incorporates the spatial heterogeneity reflected in the differences in environmental conditions and land use in different areas of the mining area. The coupling of these two models improves the robustness and applicability of the model, making its prediction of vegetation cover in mining areas more accurate and providing strong technical support for the sustainable development of mining areas.
[0036] To facilitate understanding of this embodiment, a method for predicting vegetation coverage in mining areas disclosed in this disclosure will first be described in detail. The execution entity of this method is generally an electronic device with a certain computing capability. In some possible implementations, the method for predicting vegetation coverage in mining areas can be implemented by a processor calling computer-readable instructions stored in memory.
[0037] See Figure 1 The diagram shows a flowchart of a method for predicting vegetation coverage in mining areas according to an embodiment of this disclosure. The method includes steps S101 to S104, wherein: S101. Obtain vegetation impact factors for the target mining area; wherein, vegetation impact factors are used to indicate factors affecting the vegetation coverage rate of the mining area.
[0038] In the embodiments of this disclosure, vegetation influencing factors include, but are not limited to: land use type of the target mining area, natural factors of the mining area, and human activity factors of the mining area.
[0039] Among them, the natural factors of the mining area include: temperature, precipitation, soil properties, elevation, slope and aspect of the target mining area.
[0040] Among them, human activity factors in the mining area include: the population of the target mining area and the amount of mining.
[0041] First, data can be acquired through domestic and international remote sensing satellites, drone aerial photography, and other data acquisition channels to obtain long-term remote sensing images covering the target mining area. Second, remote sensing images of appropriate resolution can be preprocessed, including cloud removal, noise reduction, and geometric correction. Finally, land use types in the target mining area can be identified using methods such as supervised classification.
[0042] Here, meteorological data (i.e., temperature, precipitation, etc.) from natural factors in the mining area can be collected through methods such as weather station observations. Soil properties in the mining area can be obtained through methods such as on-site sampling and analysis. Topographic data (i.e., elevation, slope, and aspect, etc.) of the mining area can be extracted from DEM (Digital Elevation Model) data.
[0043] S102. Based on vegetation impact factors, construct a mixed additive land impact model corresponding to the target mining area; wherein, the mixed additive land impact model is used to determine the relationship between vegetation coverage and vegetation impact factors in the target mining area.
[0044] In embodiments of this disclosure, a nonlinear relationship between vegetation influencing factors can be constructed. Then, a smoothing function for each vegetation influencing factor can be determined based on the data distribution and model requirements.
[0045] Subsequently, based on the determined smoothing function, a mixed additive land impact model is constructed. This mixed additive land impact model is a nonlinear model.
[0046] S103. Determine the weight of each influencing point in the target mining area, and construct a land likelihood regression model based on the weights and vegetation influence factors; wherein, the land likelihood regression model is used to determine the vegetation coverage of the target mining area.
[0047] In the embodiments of this disclosure, a target point in a target mining area can be determined, and an influence point within a preset range of the target point (hereinafter referred to as the target area) can be determined.
[0048] After identifying the points of influence, their weights can be determined based on the influence radius corresponding to the preset range of the target point.
[0049] After determining the weights of each influencing point, a land likelihood regression model can be constructed based on these weights and vegetation influence factors.
[0050] S104. Based on the mixed additive land impact model and the land likelihood regression model, the prediction accuracy is corrected until the preset accuracy threshold is reached to obtain the land likelihood weighted nonlinear model. The vegetation coverage of the target mining area is predicted by the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction result.
[0051] In embodiments of this disclosure, the linear relationship in the land likelihood regression model can be replaced by the nonlinear relationship in the additive land influence model to obtain a land likelihood weighted nonlinear model.
[0052] After determining the land likelihood weighted nonlinear model, the vegetation influence factors of the target mining area can be input into the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction results of the target mining area output by the land likelihood weighted nonlinear model.
[0053] In the embodiments of this disclosure, firstly, vegetation impact factors of the target mining area are obtained; wherein, vegetation impact factors are used to indicate factors affecting the vegetation coverage of the mining area; secondly, a mixed additive land impact model is constructed based on the vegetation impact factors; wherein, the mixed additive land impact model is used to determine the vegetation coverage of the target mining area; nextly, the weights of each impact point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and vegetation impact factors; wherein, the land likelihood regression model is used to determine the vegetation coverage of each impact point; finally, the land likelihood regression model is modified based on the mixed additive land impact model to obtain a land likelihood weighted nonlinear model, and the vegetation coverage of the target mining area is predicted using the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction result.
[0054] In the above embodiments, a mixed additive land impact model is constructed by acquiring vegetation influencing factors. This mixed additive land impact model integrates vegetation influencing factors and uses nonlinear modeling techniques to fill the gaps in traditional methods that do not consider the complex negative feedback mechanisms and nonlinear characteristics of these factors, thereby more comprehensively and accurately grasping the changing patterns and trends of vegetation coverage in mining areas. Furthermore, a land likelihood regression model is constructed by determining the weights of influencing points, and then modified using the mixed additive land impact model to obtain a land likelihood weighted nonlinear model. This process incorporates the spatial heterogeneity reflected in the differences in environmental conditions and land use in different areas of the mining area into the model. The coupling of these two models improves the robustness and applicability of the model, making its prediction of vegetation coverage in mining areas more accurate, and providing strong technical support for the sustainable development of mining areas.
[0055] In an optional implementation, a mixed additive land impact model is constructed based on vegetation impact factors, specifically including the following steps: First, determine the correlation between vegetation influencing factors and vegetation coverage in the target mining area; Secondly, vegetation impact factors with a correlation greater than a preset correlation threshold are identified as target vegetation impact factors. Secondly, the basis functions are determined, and a smoothing function is obtained by fitting based on the basis functions; Finally, a mixed additive land impact model is constructed based on the target vegetation impact factor and the smoothing function.
[0056] In the embodiments of this disclosure, the correlation between vegetation influencing factors and the vegetation coverage of the target mining area can be determined by using correlation coefficient determination methods. For example, the correlation between vegetation influencing factors and the vegetation coverage of the target mining area can be determined by methods such as Pearson correlation coefficient and Spearman rank correlation coefficient.
[0057] After determining the correlation between each vegetation influencing factor and the target mining area, a target correlation with a correlation greater than a preset correlation threshold can be identified, and the vegetation influencing factor corresponding to the target correlation can be identified as the target vegetation influencing factor.
[0058] Here, the target vegetation impact factor can be identified as the independent variable in the mixed additive land impact model.
[0059] Here, the smoothing function of each target vegetation impact factor can be determined based on the data distribution of each target vegetation impact factor and the requirements of the mixed additive land impact model.
[0060] The smoothing function is composed of multiple basis functions, and each basis function is multiplied by its corresponding coefficient to form the smoothing function.
[0061] Reference Figure 2 The diagram shown is a schematic diagram of basis function fitting for the prediction method of vegetation coverage in mining areas provided in this embodiment of the present disclosure. It shows that when there are too many basis functions, overfitting is likely to occur, and when there are too few basis functions, the fitting effect is poor.
[0062] In this embodiment, there are 7 basis functions, including polynomial basis functions, smooth spline functions, and Gaussian radial basis functions.
[0063] Here, the mixed additive land impact model meets the following conditions: ; Where Y is used to indicate the vegetation cover calculated by the mixed additive land impact model. β 0 represents the intercept term. S j( ) indicates the first ( ). j A smoothing function for each target vegetation influencing factor. j The value ranges from 1 to m, where m is the number of target vegetation influencing factors. X j For the first j One target vegetation influencing factor, For land type processing functions, ε This is the error term for the mixed additive land impact model.
[0064] Here, the land type processing function meets the following conditions: ; β k This is the coefficient corresponding to the k-th land use type, where k ranges from 1 to h, and h is the number of land use types. k (LandType) is the indicator function corresponding to the kth land use type; in, k1 is used to indicate that the land use type is a coal mining area.
[0065] Here, land use types include: open-pit mines, coal mining areas, spoil heaps, and reclamation areas. Other areas can be understood as including five types of land use characteristic of mining areas, i.e., h=5.
[0066] In an optional implementation, the weights of each influencing point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and vegetation influence factors, specifically including the following steps: First, the target point and target area are determined in the target mining area; the land use type of the target point is a coal mining area, and the target area is the area generated with the target point as the center. Secondly, identify the influence points in the target area that share the same land use type as the target point; Secondly, based on the long and short radii corresponding to the target area, the weights of each influencing point in the target mining area are determined; Finally, a land likelihood regression model was constructed based on weights and vegetation influence factors.
[0067] In the embodiments of this disclosure, the target area can be determined based on a preset radius of influence (i.e., the aforementioned long radius and short radius) and the target point.
[0068] Reference Figure 3 The diagram shown illustrates the determination of the influence radius of the prediction method for vegetation coverage in mining areas provided in this embodiment of the present disclosure, wherein: First, several short radii r and long radii E are preset.
[0069] Here, based on the actual conditions of the Zhundong mining area and the resolution of the images used in this embodiment, the possible values for the long and short radii are determined as shown in Table 1 below: Table 1
[0070] Secondly, the short and long radii are processed to obtain a combination of long and short radii.
[0071] One can combine the short radius and the long radius using the Cartesian product to obtain the combination of the short and long radii.
[0072] For example, given the short radius r = (r1, r2, r3, ...) and the long radius E = (E1, E2, E3, ...), after performing a Cartesian product, we get the combination of the short and long radii {(E1, E1), (E1, E2), (E1, E3), (E2, E1), (E2, E2), (E2, E3), (E3, E1), (E3, E2), (E3, E3) ...}.
[0073] Secondly, local pseudo-equations can be constructed based on combinations of long and short radii.
[0074] Among these, local fitting equations (i.e., local pseudo-equations) can be established for the combination of long and short radii respectively.
[0075] Finally, the optimal combination of long and short radii can be determined by judging the Akaike information content criterion based on the local pseudo-equation.
[0076] The Akaike Information Content Criterion (AIC) meets the following conditions: ; Where g is the number of combinations of long and short radii, and L is the likelihood function.
[0077] Here, the combination of long and short radii with the smallest AIC can be selected as the optimal combination of long and short radii.
[0078] Here, refer to Figure 4 The diagram shown illustrates the influence points of the method for predicting vegetation coverage in mining areas provided in this embodiment of the present disclosure, wherein: The target point K is used as the center of the circle, and the short radius R1 and the long radius R2 are used as the radii of the target point, forming the pixels covered by the short radius R1 and the pixels covered by the long radius R2, and the above pixels are defined as the target area.
[0079] Then, points in the target area with the same land use type as the target point can be identified as influence points.
[0080] For example, for a pixel between the long radius R2 and the short radius R1, if the land use type of the pixel corresponding to the pixel is the same as that of the target point, it is included in the set of affected points p; otherwise, it is not included in the set of affected points P.
[0081] Here, based on the long and short radii corresponding to the target area, the weights of each influencing point in the target mining area are determined, including: Based on formula Determine the weight of each influencing point in the target mining area; Among them, W i，p Let d be the weight of the p-th influencing point relative to the target point i, where p ranges from 1 to q, q is the number of influencing points in the target mining area, and d is the weight of the p-th influencing point relative to the target point i. ip Let R1 be the distance from the p-th influencing point to the target point i, R2 be the minor radius, R1 be the major radius, and y be the distance from the p-th influencing point to the target point i. p Let P be the p-th influencing point, and P be the set of influencing points.
[0082] Here, refer to Figure 5 The diagram shown is a schematic representation of the weight curve of the method for predicting vegetation coverage in mining areas provided in this embodiment of the present disclosure, wherein: The vertical axis represents the weight value, with a maximum value of 1; the horizontal axis represents the distance between the affected point and the target point.
[0083] Figure 5 In this diagram, X represents the target point, y1 represents the influence points within the short radius R1, y2 represents the influence points within the long radius R2, and d represents the distance from each influence point to the target point. ij Let be the distance from the j-th influence point within the short radius to the target point i.
[0084] Here, the weights of each influencing point in the target mining area are used to indicate the distance between the influencing point and the target point. In this application, a near-Gaussian weighting function is used as the decay function to calculate the weights of different influencing points.
[0085] Here, the land likelihood regression model meets the following conditions: ; Among them, Y i β0(u) represents the vegetation cover predicted by the land likelihood regression model. i ,v i ) is the intercept of the target point, β j (u i ,v i ) represents the parameter of the j-th target vegetation impact factor, where j ranges from 1 to m, and m is the number of target vegetation impact factors. W i，p Let X be the weight of the p-th influencing point relative to the target point i. ij Let ε be the value of the j-th target vegetation impact factor. iThis is the error term corresponding to the target point.
[0086] In an optional implementation, the land likelihood regression model is modified based on the hybrid additive land impact model to obtain a land likelihood weighted nonlinear model, specifically including the following steps: First, determine the nonlinear relationships corresponding to the target vegetation impact factors in the mixed additive land impact model; Then, the nonlinear relationship is replaced with the linear relationship corresponding to the target vegetation influence factor in the land likelihood regression model to obtain the land likelihood weighted nonlinear model.
[0087] In the embodiments of this disclosure, the smoothing function part of the target vegetation influence factor in the mixed additive land influence model can be determined, and the smoothing function part can be used as the nonlinear relationship corresponding to the target vegetation influence factor in the mixed additive land influence model.
[0088] Subsequently, the linear relationship corresponding to the target vegetation impact factor in the land likelihood regression model can be determined, and the nonlinear relationship corresponding to the target vegetation impact factor in the mixed additive land impact model can be used to replace the linear relationship corresponding to the target vegetation impact factor in the land likelihood regression model.
[0089] Here, the nonlinear relationship corresponding to the target vegetation influence factor in the mixed additive land impact model is: The linear relationship between the target vegetation influence factors in the land likelihood regression model is as follows: .
[0090] Here, β0(u) can be used as an example. i ,v i ) and ε i The terms are combined to obtain the error term ε.
[0091] Here, the land likelihood weighted nonlinear model meets the following conditions: ; Among them, Y ’ i W represents the vegetation cover predicted by the land likelihood weighted nonlinear model. i，p Let be the weight of the p-th influencing point relative to the target point i. β 0 represents the intercept term, f j ( ) is the smoothing function of the j-th target vegetation influence factor, X ij Let be the value of the j-th target vegetation influence factor. For land type processing functions, ε This is the error term.
[0092] Among them, in the land likelihood weighted nonlinear model f in j( ) is equivalent to S in j .
[0093] In the embodiments of this disclosure, the accuracy of the land likelihood weighted nonlinear model can be verified by hierarchical k-fold cross-validation. The model is trained using a training dataset, and the model parameters and structure are continuously adjusted based on the prediction results to improve the accuracy and performance of the model.
[0094] Here, refer to Figure 6 The diagram shown is a schematic representation of the accuracy verification of the prediction method for vegetation coverage in mining areas provided in this embodiment of the present disclosure, wherein: First, the original dataset is divided into five layers based on the characteristic land use type of the mining area (open-pit mine, coal area, spoil heap, reclamation area, and other areas). Each layer is further divided into K sets of similar size that do not overlap.
[0095] The sets of each layer are combined to form k subsets. For each subset, the following steps are performed: the current fold is used as the validation set, and the remaining K-1 folds are merged as the training set. The training set is used to train the model, and the validation set is used to evaluate the model's performance and record the accuracy. The above steps are repeated until each subset has served as a validation set at least once. The average of all accuracies obtained from the K validations is taken as the final performance evaluation result of the model.
[0096] In the above embodiments, the accuracy verification method stratifies the original dataset according to the characteristic land use types of the mining area, which can ensure that the proportion of each land use type in each compromise is consistent, reduce the randomness caused by uneven sample selection during the accuracy verification process, improve the applicability of the model in complex areas such as mining areas, improve the accuracy of the model in predicting vegetation cover, and continuously adjust the model parameters and structure according to the prediction results, optimize the model configuration, and improve the model accuracy and performance.
[0097] Based on the same inventive concept, this disclosure also provides a device for predicting vegetation coverage in mining areas, which corresponds to the method for predicting vegetation coverage in mining areas. Since the principle of the device in this disclosure is similar to the method for predicting vegetation coverage in mining areas described above, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.
[0098] Reference Figure 7 The diagram shown is a schematic of a device for predicting vegetation coverage in a mining area according to an embodiment of this disclosure. The device includes: an acquisition module 71, a first model construction module 72, a second model construction module 73, and a correction module 74; wherein, The acquisition module is used to acquire vegetation impact factors of the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage of the mining area. The first model construction module is used to construct a mixed additive land impact model corresponding to the target mining area based on the vegetation impact factors; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors of the target mining area; The second model construction module is used to determine the weight of each influencing point in the target mining area, and to construct a land likelihood regression model based on the weight and the vegetation influence factor; wherein, the land likelihood regression model is used to determine the vegetation coverage of each influencing point; The correction module is used to correct the prediction based on the mixed additive land influence model and the land likelihood regression model until the prediction accuracy reaches a preset accuracy threshold, thereby obtaining a land likelihood weighted nonlinear model. The land likelihood weighted nonlinear model is then used to predict the vegetation coverage of the target mining area, and the vegetation coverage prediction result is obtained.
[0099] This disclosure discloses a mixed additive land impact model by acquiring vegetation influencing factors. This model integrates vegetation influencing factors and uses nonlinear modeling to overcome the shortcomings of traditional methods that do not consider the complex negative feedback mechanisms and nonlinear characteristics of these factors. This allows for a more comprehensive and accurate understanding of the changing patterns and trends of vegetation cover in mining areas. Furthermore, a land likelihood regression model is constructed by determining the weights of influencing points, and then modified using the mixed additive land impact model to obtain a land likelihood weighted nonlinear model. This process incorporates the spatial heterogeneity reflected in the differences in environmental conditions and land use in different areas of the mining area. The coupling of these two models improves the robustness and applicability of the model, making its prediction of vegetation cover in mining areas more accurate and providing strong technical support for the sustainable development of mining areas.
[0100] The processing flow of each module in the device and the interaction flow between each module can be referred to the relevant descriptions in the above method embodiments, and will not be detailed here.
[0101] Corresponding to Figure 1 In addition to the image detection method described in this disclosure, an electronic device 800 is also provided, such as... Figure 8 The diagram shown is a structural schematic of an electronic device 800 provided in an embodiment of this disclosure, including: The system includes a processor 81, a memory 82, and a bus 83. The memory 82 stores execution instructions and includes main memory 821 and external memory 822. The main memory 821, also called internal memory, temporarily stores the computational data in the processor 81, as well as data exchanged with external memory such as a hard disk. The processor 81 exchanges data with the external memory 822 through the main memory 821. When the electronic device 800 is running, the processor 81 communicates with the memory 82 through the bus 83, causing the processor 81 to execute the following instructions: Obtain vegetation impact factors for the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage rate of the mining area; Based on the vegetation impact factors, a mixed additive land impact model is constructed for the target mining area; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors in the target mining area; The weights of each influencing point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and the vegetation influence factors; wherein, the land likelihood regression model is used to determine the vegetation coverage of the target mining area; The land likelihood regression model is modified based on the mixed additive land impact model until the prediction accuracy reaches a preset accuracy threshold, resulting in a land likelihood weighted nonlinear model. The vegetation coverage of the target mining area is then predicted using the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction result.
[0102] The basic principles of this disclosure have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this disclosure are merely examples and not limitations, and should not be considered as essential features of each embodiment of this disclosure. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the scope of this disclosure to the necessity of employing the aforementioned specific details for implementation.
[0103] The block diagrams of devices, apparatuses, devices, and systems disclosed herein are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.
[0104] Additionally, as used herein, the "or" used in a list of items beginning with "at least one" indicates a separate list, such that a list of, for example, "at least one of A, B, or C" means A or B or C, or AB or AC or BC, or ABC (i.e., A and B and C). Furthermore, the word "exemplary" does not imply that the described example is preferred or better than other examples.
[0105] It should also be noted that in the systems and methods of this disclosure, the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered as equivalent solutions to this disclosure.
[0106] Various changes, substitutions, and modifications can be made to the technology described herein without departing from the teachings defined by the appended claims. Furthermore, the scope of the claims of this disclosure is not limited to the specific aspects of the processes, machines, manufactures, events, means, methods, and actions described above. Currently existing or later-developed processes, machines, manufactures, events, means, methods, or actions that perform substantially the same function or achieve substantially the same result as the corresponding aspects described herein can be utilized. Therefore, the appended claims include such processes, machines, manufactures, events, means, methods, or actions within their scope.
[0107] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this disclosure. Therefore, this disclosure is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features disclosed herein.
[0108] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this disclosure to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations therein.
Claims
1. A method for predicting vegetation coverage in mining areas, characterized in that, include: Obtain vegetation impact factors for the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage rate of the mining area; Based on the vegetation impact factors, a mixed additive land impact model is constructed for the target mining area; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors in the target mining area; The weights of each influencing point in the target mining area are determined, and a land likelihood regression model is constructed based on the weights and the vegetation influence factors; wherein, the land likelihood regression model is used to determine the vegetation coverage of the target mining area; The land likelihood regression model is modified based on the mixed additive land impact model until the prediction accuracy reaches a preset accuracy threshold, resulting in a land likelihood weighted nonlinear model. The vegetation coverage of the target mining area is then predicted using the land likelihood weighted nonlinear model to obtain the vegetation coverage prediction result.
2. The method as described in claim 1, characterized in that, The construction of the mixed additive land impact model corresponding to the target mining area based on the vegetation impact factors includes: Determine the correlation between the vegetation influencing factors and the vegetation coverage of the target mining area; Vegetation impact factors whose correlation is greater than a preset correlation threshold are identified as target vegetation impact factors. Determine the basis functions, and obtain the smooth function by fitting based on the basis functions; Based on the target vegetation impact factor and the smoothing function, the mixed additive land impact model is constructed.
3. The method as described in claim 2, characterized in that, The mixed additive land impact model meets the following conditions: ； Wherein, Y is used to indicate the vegetation cover rate calculated by the mixed additive land impact model. β 0 represents the intercept term. S j ( ) indicates the first ( ). j A smoothing function for each target vegetation influencing factor. j The value ranges from 1 to m, where m is the number of target vegetation influencing factors. X j For the first j One target vegetation influencing factor, For land type processing functions, ε This is the error term of the mixed additive land impact model.
4. The method as described in claim 3, characterized in that, The processing function for the land type meets the following conditions: ； β k This is the coefficient corresponding to the k-th land use type, where k ranges from 1 to h, and h is the number of land use types. k (LandType) is the indicator function corresponding to the kth land use type; in, k1 is used to indicate that the land use type is a coal mining area.
5. The method as described in claim 1, characterized in that, The step of determining the weights of each influencing point in the target mining area and constructing a land likelihood regression model based on the weights and the vegetation influence factors includes: In the target mining area, a target point and a target area are determined; wherein, the land use type of the target point is a coal mining area, and the target area is an area generated with the target point as the center; Identify influence points in the target area that share the same land use type as the target point; Based on the long radius and short radius corresponding to the target area, the weight of each influence point in the target mining area is determined; A land likelihood regression model is constructed based on the weights and the vegetation influence factors.
6. The method as described in claim 5, characterized in that, The determination of the weights of each influencing point in the target mining area based on the long and short radii corresponding to the target area includes: Based on formula Determine the weight of each of the aforementioned influencing points in the target mining area; Among them, W i，p Let d be the weight of the p-th influencing point relative to the target point i, where p ranges from 1 to q, q is the number of influencing points in the target mining area, and d is the weight of the p-th influencing point relative to the target point i. ip Let R1 be the distance from the p-th influencing point to the target point i, R2 be the minor radius, R1 be the major radius, and y be the distance from the p-th influencing point to the target point i. p Let P be the p-th influencing point, and P be the set of influencing points.
7. The method as described in claim 6, characterized in that, The land likelihood regression model meets the following conditions: ； Among them, Y i β0(u) represents the vegetation cover predicted by the land likelihood regression model. i ,v i ) is the intercept of the target point, β j (u i ,v i ) represents the parameter of the j-th target vegetation impact factor, where j ranges from 1 to m, and m is the number of target vegetation impact factors. W i，p Let X be the weight of the p-th influencing point relative to the target point i. ij Let ε be the value of the j-th target vegetation impact factor. i This is the error term corresponding to the target point.
8. The method as described in claim 1, characterized in that, The modified land likelihood weighted nonlinear model, based on the combination of the mixed additive land impact model and the land likelihood regression model, includes: Determine the nonlinear relationships corresponding to the target vegetation impact factors in the hybrid additive land impact model; The nonlinear relationship is replaced with the linear relationship corresponding to the target vegetation influence factor in the land likelihood regression model to obtain the land likelihood weighted nonlinear model.
9. The method as described in claim 8, characterized in that, The land likelihood weighted nonlinear model meets the following conditions: ； Among them, Y ’ i W represents the vegetation cover predicted by the land likelihood weighted nonlinear model. i，p Let be the weight of the p-th influencing point relative to the target point i. β 0 represents the intercept term, f j ( ) is the smoothing function of the j-th target vegetation influence factor, X ij Let be the value of the j-th target vegetation influence factor. For land type processing functions, ε This is the error term.
10. A device for predicting vegetation coverage in a mining area, characterized in that, include: The acquisition module is used to acquire vegetation impact factors of the target mining area; wherein, the vegetation impact factors are used to indicate factors affecting the vegetation coverage of the mining area. The first model construction module is used to construct a mixed additive land impact model corresponding to the target mining area based on the vegetation impact factors; wherein, the mixed additive land impact model is used to determine the relationship between the vegetation coverage rate and the vegetation impact factors of the target mining area; The second model construction module is used to determine the weight of each influencing point in the target mining area, and to construct a land likelihood regression model based on the weight and the vegetation influence factor; wherein, the land likelihood regression model is used to determine the vegetation coverage of each influencing point; The correction module is used to correct the prediction based on the mixed additive land influence model and the land likelihood regression model until the prediction accuracy reaches a preset accuracy threshold, thereby obtaining a land likelihood weighted nonlinear model. The land likelihood weighted nonlinear model is then used to predict the vegetation coverage of the target mining area, and the vegetation coverage prediction result is obtained.