A mine ecological safety identification method based on a convolutional neural network model

Abstract

This invention discloses a mine ecological safety identification method based on a convolutional neural network model, belonging to the field of mine ecological safety identification technology. This invention introduces a convolutional neural network (CNN) to automatically optimize resistance values. Combining time dimensions and dynamic factors, it adjusts the resistance surface in real time through monitoring, enabling it to dynamically respond to the impacts of terrain changes, vegetation density fluctuations, and meteorological changes. It can capture the real-time impacts of sudden rainfall, mine expansion, or vegetation restoration, resulting in a highly timely resistance surface that accurately reflects the dynamic characteristics of the ecosystem. Simultaneously, it combines circuit theory and deep learning to generate multiple potential migration paths, avoiding reliance on a single minimum resistance path and providing more diverse migration options. Furthermore, it dynamically optimizes path weights using a time series model, prioritizing the protection of channels with high ecological value and high migration demand, ensuring sufficient resilience and adaptability of the migration network.

Description

Technical Field

[0001] This invention relates to the field of mine ecological safety identification technology, and in particular to a mine ecological safety identification method based on a convolutional neural network model. Background Technology

[0002] In the identification of ecological safety in mines, traditional methods mostly use static ecological resistance surfaces and single-path minimum cumulative resistance models to identify ecological corridors and key ecological nodes. They rely on remote sensing images or geographic information system (GIS) data to construct resistance surfaces, mapping different land use types, topography, vegetation cover and other factors to fixed resistance values, thereby describing ecological corridors in the ecosystem.

[0003] However, the mining environment is greatly affected by human mining, and ecological elements such as landform and vegetation cover change significantly over time. This makes it difficult for traditional static resistance surfaces to accurately reflect the dynamic characteristics of the mining ecology, resulting in low timeliness and applicability in ecological identification.

[0004] To address this issue, some traditional solutions involve increasing monitoring frequency and zoning to identify and establish ecological corridors in advance to cope with changes in the ecological environment. Specific methods include setting up ecological corridors in major ecologically sensitive areas, areas with good vegetation cover, or areas where key species are active, and adjusting the resistance surface according to the current situation. In addition, some solutions improve the model to increase the number of pathways. However, due to the complexity and dynamism of mining ecosystems, such static solutions are difficult to respond in real time, and corridor adjustments usually require manual intervention, making it impossible to achieve real-time updates and true adaptability. Therefore, there is an urgent need for a mine ecological safety identification method based on a convolutional neural network model to solve this problem. Summary of the Invention

[0005] In view of the aforementioned existing problems, the present invention is proposed.

[0006] This invention provides a mine ecological safety identification method based on a convolutional neural network model. It addresses the problems that static resistance surfaces and fixed single-path models cannot adapt to rapid environmental changes, resulting in insufficient connectivity identification or large biases, and failing to accurately reflect the diversity of species migration paths. Traditional methods also struggle to achieve dynamic updates and multi-path identification of ecological corridors, and lack real-time and high-resolution response mechanisms.

[0007] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0008] This invention provides a method for identifying mine ecological safety based on a convolutional neural network model, comprising step S1: mine data acquisition.

[0009] Collect mine data, including mine topography data, vegetation density, meteorological data, and mining records.

[0010] Step S2, Dynamic resistance surface modeling,

[0011] Based on the mine data collected in step S1, a dynamic ecological resistance surface is constructed to reflect the spatial distribution of the degree of obstruction to species migration in each region of the ecosystem. The resistance value of each region depends on ecological factors.

[0012] Step S3: Multi-path identification of ecological corridors.

[0013] By combining deep learning with circuit theory, multiple ecological corridors, i.e. possible migration routes, are identified and generated on the dynamic ecological resistance surface. This is not limited to the path of least resistance, but takes into account ecosystem factors, including species migration needs and resource distribution.

[0014] In step S3, based on deep learning models and time series analysis, the priority of each migration route is adjusted to prioritize the protection of ecological corridors for species with high ecological value and high demand.

[0015] Step S4: Dynamic identification of ecological pinch points and obstacle points.

[0016] Based on the dynamic ecological resistance surface constructed in step S2, pinch points and obstacle points in the ecosystem are identified, namely bottlenecks and blocking areas in the migration path.

[0017] Step S5, Ecological management based on dynamic results,

[0018] Based on the ecological corridors, splints, and obstacles identified in steps S3 and S4, targeted ecological protection and restoration measures are implemented. The identification results are transmitted to the ecological management system in real time, allowing the mine management to prioritize the restoration or protection of key corridors and splints, thereby reducing the impact of human activities on ecological connectivity.

[0019] Furthermore, in step S2, a convolutional neural network (CNN) is used to dynamically adjust the resistance value based on the topographic data, vegetation density, and water distribution in the meteorological data of the mining area, thereby simulating the obstruction of species migration by the mining area ecosystem.

[0020] Furthermore, in step S2, based on the data collected in step S1, a convolutional neural network (CNN) is used to dynamically adjust the resistance value according to the topographic data, vegetation density, and water distribution in the meteorological data of the mining area. Specific steps include:

[0021] Terrain data (slope) ), vegetation density and meteorological data (water body distribution) Perform spatial rasterization:

[0022] ,in, Represents pixels The initial resistance value, This represents a function that normalizes various ecological factors, keeping the range within a certain range. between, Represents pixels The slope value indicates the terrain gradient; the steeper the slope, the higher the resistance. Represents pixels The higher the vegetation cover density, the lower the resistance. Represents pixels The water body distribution characteristics show that the resistance is lower the closer to the water body.

[0023] Based on the initial resistance value A convolutional neural network (CNN) is used to dynamically adjust the resistance, capturing local spatial characteristics and reflecting the interaction of ecological factors.

[0024] ,

[0025] in, Represents pixels The dynamically adjusted resistance value, This represents the ReLU activation function, used to preserve the non-linear characteristics of the output. The weight matrix represents the convolution kernel, indicating the weights at different positions. Indicates the size of the convolution kernel. This represents the bias term, used to adjust the baseline level of the overall resistance value. The local neighborhood pixel value represents the initial resistance value.

[0026] Furthermore, in step S2, an environmental change response mechanism is introduced:

[0027] Introduction time and dynamic environmental factors Based on time series data to capture the impact of environmental changes, the dynamic resistance value over time is calculated:

[0028] ,in, Represents pixels In time The final dynamic resistance value, These represent the weighting coefficients of dynamic environmental factors, used to adjust the degree of influence of environmental changes on the resistance value. Indicates time Next pixel Dynamic environmental factors (such as sudden rainfall or vegetation restoration) are obtained from measured data and range from [specific range missing]. ,

[0029] A dynamic ecological resistance surface is generated based on the final dynamic resistance value:

[0030] ,

[0031] in, Indicates time The dynamic ecological resistance surface It indicates the total regional extent of the ecosystem.

[0032] Furthermore, in step S3, multi-path identification of ecological corridors is performed, specifically including:

[0033] In dynamic ecological resistance Above, the current intensity during species migration is calculated based on circuit theory.

[0034] ,

[0035] in, Represents pixels The current intensity represents the flow intensity in the ecological migration network. This represents the dynamic ecological resistance surface generated in step S2. and Let them represent the set of origin points and the set of destination points for migration in the ecosystem, respectively. Indicates the starting point and target point The voltage value is set to and ;

[0036] Using current intensity Identify potential migration routes:

[0037] ,

[0038] Representing a path In pixels The indicated value, For path indexing, This indicates that the path passes through this pixel. This indicates that the application was not approved. Represents the unit step function, when the input is greater than... Time output Otherwise output , Representing a path The current threshold is used to control path flow. The number of paths is adjusted to generate multiple evenly distributed migration paths;

[0039] Introducing a deep learning model to dynamically optimize path weights based on ecological factors:

[0040] ,

[0041] in, Let represent the objective function, whose value is the weighted sum of the resistance costs and negative impacts of all paths. This represents the total number of candidate migration routes. These represent the first, second, and third weighting parameters, respectively, which are used to measure the impact of drag, terrain gradient, and dynamic environmental factors. Represents pixels The terrain gradient corresponds to the slope cost of migration. Indicates time Next pixel Dynamic environmental factors, such as rainfall, can be assessed using a deep learning framework to optimize path distribution, making it more concentrated in low-resistance areas while reducing the negative impact of dynamic environments on paths.

[0042] Furthermore, in step S3, adjusting path priority based on time series data specifically includes:

[0043] Combine time series models to predict the importance of each path and dynamically adjust the priority:

[0044] ,

[0045] in, Representing a path In time priority, These represent the adjustment weights for the first priority and the second priority, respectively. Representing a path Migration demand at time t, based on time series prediction of species migration demand;

[0046] Using time series analysis to predict future migration demand:

[0047] ,

[0048] in, Representing a path In time The migration needs, The time decay factor representing migration demand. The response coefficient represents the response to environmental changes. Indicates time Real-time changes in environmental factors

[0049] Based on path priority, select paths with high ecological value for protection.

[0050] Furthermore, in step S4, based on the circuit theory model, the current density in different regions is calculated to identify high-importance nodes and obstacles hindering species migration in the ecological network, wherein the high-importance nodes are pinch points.

[0051] Furthermore, in step S4, the location and priority of pinch points and obstacle points are adjusted according to changes in mining operations, including expansion of the mining area and destruction of vegetation.

[0052] In step S4, combining the vegetation density changes and meteorological changes collected in step S1, the agent-based modeling method is used to simulate the frequency and pressure of different species passing through pinch points and obstacle points in the ecological corridor, and dynamically adjust the positions of pinch points and obstacle points.

[0053] Furthermore, in step S4, the current density is calculated based on the resistance surface:

[0054] Using circuit theory, the current distribution on the ecological resistance surface is simulated, and the current density in each region is calculated:

[0055] ,

[0056] in, Represents pixels The current density, i.e., the degree of stress at which a point serves as a migration route within an ecological network. , Represents the starting set and the final set They define the origin and destination areas of the migration, respectively. Represents pixels Conduction rate, Pixels representing ecological resistance surfaces The resistance value, the current density here Reflecting pixels in the ecosystem The importance of connectivity in migration networks; high current density indicates bottleneck locations in the path that are crucial to migration.

[0057] Based on the current density map, regions with high current concentrations are used as pinch points: ,in, Represents pixels Is it a pinch point? Represented as grip point, Indicates a non-grip point. This indicates an indicator function that outputs when the condition is true. Otherwise output , Indicates the current density threshold;

[0058] Obstacles, i.e., blocked areas, are detected using conductivity and rate of change of resistance.

[0059] ,in, Represents pixels Is it an obstacle point? Represented as obstacle points, Indicates non-obstacle points. Represents pixels The rate of change of the resistance value represents the dynamic trend of the degree of obstruction. This represents the threshold for the rate of change of resistance.

[0060] Furthermore, in step S4, the grip points and obstacle points are adjusted dynamically according to the mining process:

[0061] Based on the expanded mining area, the current density threshold is redefined: ,in, Indicates time The pinch current density threshold after dynamic adjustment This represents the influence coefficient of current density in the expanded mining area. Indicates time The proportion of mining expansion areas in the lower mine should be increased, the current density threshold should be raised, and misjudgment of non-critical nodes as pinch points should be avoided.

[0062] The criteria for determining obstacle points are adjusted based on the dynamic changes in vegetation destruction. ,in, Indicates time The threshold for the rate of change of resistance after dynamic adjustment. This represents the coefficient indicating the influence of vegetation destruction on the rate of change of resistance. Indicates time The lower the vegetation cover loss rate, the higher the vegetation destruction rate, the lower the judgment threshold, and the more sensitively it can identify possible obstacle points;

[0063] In step S4, the grip points and obstacle points are dynamically adjusted based on the proxy simulation, specifically including:

[0064] Agent-based modeling was used to simulate the frequency of species passing through grips in ecological corridors:

[0065] ,in, Indicates time Sub-agents through pinch points frequency, This represents the total number of agent individuals participating in the migration in the simulation. Indicates time Sub-agent Location, Indicates if the agent After the pinch point Output Otherwise output , Reflecting the pinch point In the migration network, high-frequency pinch points should be prioritized for protection based on their importance.

[0066] Simulate the stress level of the agent passing through the obstacle point: ,in, Indicates time Lower obstacle point The level of agent pressure, Indicates agent The dynamic ecological resistance surface at the location, the pressure level of the obstacle point reflects its negative impact on the migration path, and obstacle points with high pressure need to be repaired or mitigated first.

[0067] The beneficial effects of this invention are:

[0068] This invention introduces a convolutional neural network (CNN) to automatically optimize resistance values. By combining the time dimension and dynamic environmental factors, it adjusts the resistance surface in real time, enabling it to dynamically respond to the effects of terrain changes, vegetation density fluctuations, and meteorological changes. It can capture the real-time impact of changes such as sudden rainfall, mining area expansion, or vegetation restoration. The generated resistance surface has high timeliness and accurately reflects the dynamic characteristics of the ecosystem.

[0069] This invention combines circuit theory and deep learning to generate multiple potential migration paths, avoiding reliance on a single path of least resistance and providing more diverse migration options. It also dynamically optimizes path weights using time series models, prioritizing the protection of channels with high ecological value and high migration demand, thus ensuring that the migration network has sufficient resilience and adaptability.

[0070] This invention, through the combination of deep learning and circuit theory, enables the automatic identification and dynamic adjustment of ecological resistance surfaces, migration paths, pinch points, and obstacle points. It can adapt to the complex changes in the mine's ecological environment without human intervention. By using surrogate simulation methods to quantify the dynamic pressure levels of pinch points and obstacle points, protection and restoration decisions become more accurate and adaptable.

[0071] This invention, by combining real-time monitoring data, enables the model to automatically and dynamically adjust resistance values ​​and corridor layout. At the same time, it adaptively optimizes the priority of pinch points and obstacle points, greatly improving adjustment efficiency. Furthermore, by transmitting the identification results to the ecological management system in real time, mine staff can quickly obtain dynamic adjustment suggestions, thereby improving management efficiency.

[0072] This invention employs a proxy simulation method that can simulate the behavioral patterns of different species in ecological corridors. Through quantitative analysis of pinch point frequency and obstacle point pressure, it ensures that ecological corridors can meet the ecological needs of multiple species. Based on the simulation results, it prioritizes the protection of corridors for species with high demand, thus ensuring the diversity and sustainability of the ecological network. Attached Figure Description

[0073] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] Figure 1 This is a flowchart illustrating the mine ecological safety identification method based on a convolutional neural network model according to the present invention. Detailed Implementation

[0075] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0076] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0077] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0078] Example 1, referring to Figure 1 This embodiment provides a method for identifying mine ecological safety based on a convolutional neural network model, including the following steps:

[0079] Step S1, Mine data acquisition,

[0080] Collect mine data, including mine topography data, vegetation density, meteorological data, and mining records.

[0081] Step S2, Dynamic resistance surface modeling,

[0082] Based on the mine data collected in step S1, a dynamic ecological resistance surface is constructed to reflect the spatial distribution of the degree of obstruction to species migration in each region of the ecosystem. The resistance value of each region depends on ecological factors.

[0083] In step S2, a convolutional neural network (CNN) is used to dynamically adjust the resistance value based on the topographic data, vegetation density, and water distribution in the meteorological data of the mining area, thereby simulating the obstruction of species migration by the mining area ecosystem.

[0084] In step S2, based on the data collected in step S1, a convolutional neural network (CNN) is used to dynamically adjust the resistance value according to the topographic data, vegetation density, and water distribution in the meteorological data of the mining area. Specific steps include:

[0085] Terrain data (slope) ), vegetation density and meteorological data (water body distribution) Perform spatial rasterization:

[0086] ,in, Represents pixels The initial resistance value, This represents a function that normalizes various ecological factors, keeping the range within a certain range. between, Represents pixels The slope value indicates the terrain gradient; the steeper the slope, the higher the resistance. Represents pixels The higher the vegetation cover density, the lower the resistance. Represents pixels The water body distribution characteristics show that the resistance is lower the closer to the water body.

[0087] Based on the initial resistance value A convolutional neural network (CNN) is used to dynamically adjust the resistance, capturing local spatial characteristics and reflecting the interaction of ecological factors.

[0088] ,

[0089] in, Represents pixels The dynamically adjusted resistance value, This represents the ReLU activation function, used to preserve the non-linear characteristics of the output. The weight matrix represents the convolution kernel, indicating the weights at different positions. Indicates the size of the convolution kernel. This represents the bias term, used to adjust the baseline level of the overall resistance value. This represents the local neighborhood pixel value of the initial resistance value;

[0090] In step S2, an environmental change response mechanism is introduced:

[0091] Introduction time and dynamic environmental factors Based on time series data to capture the impact of environmental changes, the dynamic resistance value over time is calculated:

[0092] ,in, Represents pixels In time The final dynamic resistance value, These represent the weighting coefficients of dynamic environmental factors, used to adjust the degree of influence of environmental changes on the resistance value. Indicates time Next pixel Dynamic environmental factors (such as sudden rainfall or vegetation restoration) are obtained from measured data and range from [specific range missing]. ,

[0093] A dynamic ecological resistance surface is generated based on the final dynamic resistance value:

[0094] ,

[0095] in, Indicates time The dynamic ecological resistance surface Indicates the total regional extent of the ecosystem;

[0096] Specifically, this involves combining space and time, incorporating the time dimension into the resistance surface model, and employing dynamic environmental factors. Reflecting sudden environmental changes, such as mining expansion or vegetation restoration, enhancing the timeliness of resistance surfaces, and simultaneously employing... Automatic learning captures the interaction between topography, vegetation, and water distribution, enabling the resistance surface to more accurately reflect ecosystem characteristics. A dynamic response mechanism is introduced, combined with real-time monitoring data. It automatically and dynamically adjusts the resistance value without human intervention, generating a dynamic ecological resistance surface. Dynamically reflects migration resistance in the ecological environment;

[0097] Step S3: Multi-path identification of ecological corridors.

[0098] By combining deep learning with circuit theory, multiple ecological corridors, i.e. possible migration routes, are identified and generated on the dynamic ecological resistance surface. This is not limited to the path of least resistance, but takes into account ecosystem factors, including species migration needs and resource distribution.

[0099] In step S3, based on deep learning models and time series analysis, the priority of each migration route is adjusted to prioritize the protection of ecological corridors for species with high ecological value and high demand.

[0100] Step S3 involves multi-path identification of ecological corridors, specifically including:

[0101] In dynamic ecological resistance Above, the current intensity during species migration is calculated based on circuit theory.

[0102] ,

[0103] in, Represents pixels The current intensity represents the flow intensity in the ecological migration network. This represents the dynamic ecological resistance surface generated in step S2. and Let them represent the set of origin points and the set of destination points for migration in the ecosystem, respectively. Indicates the starting point and target point The voltage value is set to and ;

[0104] Using current intensity Identify potential migration routes:

[0105] ,

[0106] Representing a path In pixels The indicated value, Path Index Represents the unit step function, when the input is greater than... Time output Otherwise output , Representing a path The current threshold is used to control path flow. The number of paths is adjusted to generate multiple evenly distributed migration paths;

[0107] Introducing a deep learning model to dynamically optimize path weights based on ecological factors:

[0108] ,

[0109] in, Let represent the objective function, whose value is the weighted sum of the resistance costs and negative impacts of all paths. This represents the total number of candidate migration routes. These represent the first, second, and third weighting parameters, respectively, which are used to measure the impact of drag, terrain gradient, and dynamic environmental factors. Represents pixels The terrain gradient corresponds to the slope cost of migration. Indicates time Next pixel Dynamic environmental factors, such as rainfall, are considered. Through a deep learning framework, the path distribution is optimized to concentrate it in low-resistance areas, while reducing the negative impact of the dynamic environment on the path.

[0110] In step S3, path priority is adjusted based on time series data:

[0111] Combine time series models to predict the importance of each path and dynamically adjust the priority:

[0112] ,

[0113] in, Representing a path In time priority, These represent the adjustment weights for the first priority and the second priority, respectively. Representing a path Migration demand at time t, based on time series prediction of species migration demand;

[0114] Using time series analysis to predict future migration demand:

[0115] ,

[0116] in, Representing a path In time The migration needs, The time decay factor representing migration demand. The response coefficient represents the response to environmental changes. Indicates time Real-time changes in environmental factors

[0117] Based on path priority, select high ecological value paths for protection;

[0118] Specifically, by identifying migration paths through circuit theory, optimizing path characteristics through deep learning, and adjusting path priorities through time series analysis, high-value corridors are ultimately prioritized for protection. This approach combines circuit flow with multi-path optimization and introduces dynamic environment and time series analysis to achieve intelligent spatiotemporal adjustment. At the same time, deep learning is used to optimize the objective function and balance resistance, gradient, and dynamic environmental factors.

[0119] Step S4: Dynamic identification of ecological pinch points and obstacle points.

[0120] Based on the dynamic ecological resistance surface constructed in step S2, pinch points and obstacle points in the ecosystem are identified, namely bottlenecks and blocking areas in the migration path.

[0121] In step S4, based on the circuit theory model, the current density in different regions is calculated to identify high-importance nodes and obstacles hindering species migration in the ecological network. The high-importance nodes are called pinch points.

[0122] In step S4, the location and priority of pinch points and obstacle points are adjusted according to changes in mining operations, including expansion of the mining area and destruction of vegetation.

[0123] In step S4, combining the vegetation density changes and meteorological changes collected in step S1, the agent-based modeling method is used to simulate the frequency and pressure of different species passing through pinch points and obstacle points in the ecological corridor, and dynamically adjust the positions of pinch points and obstacle points.

[0124] In step S4, the current density is calculated based on the resistance surface:

[0125] Using circuit theory, the current distribution on the ecological resistance surface is simulated, and the current density in each region is calculated:

[0126] ,in, Represents pixels The current density, i.e., the degree of stress at which a point serves as a migration route within an ecological network. , Represents the starting set and the final set They define the origin and destination areas of the migration, respectively. Represents pixels Conduction rate, Pixels representing ecological resistance surfaces The resistance value, the current density here Reflecting pixels in the ecosystem The importance of connectivity in migration networks; high current density indicates bottleneck locations in the path that are crucial to migration.

[0127] Based on the current density map, regions with high current concentrations are used as pinch points: ,in, Represents pixels Is it a pinch point? Represented as grip point, Indicates a non-grip point. This indicates an indicator function that outputs when the condition is true. Otherwise output , Indicates the current density threshold;

[0128] Obstacles, i.e., blocked areas, are detected using conductivity and rate of change of resistance.

[0129] ,in, Represents pixels Is it an obstacle point? Represented as obstacle points, Indicates non-obstacle points. Represents pixels The rate of change of the resistance value represents the dynamic trend of the degree of obstruction. This represents the threshold for the rate of change of resistance, used to define the criteria for determining obstacle points. Areas where the rate of change of resistance is higher than the threshold are marked as obstacle points, representing dynamic areas that may hinder species migration.

[0130] In step S4, the grip points and obstacle points are adjusted according to the dynamics of mining operations:

[0131] Based on the expanded mining area, the current density threshold is redefined: ,in, Indicates time The pinch current density threshold after dynamic adjustment This represents the influence coefficient of current density in the expanded mining area. Indicates time The proportion of mining expansion areas in the lower mine should be increased, the current density threshold should be raised, and misjudgment of non-critical nodes as pinch points should be avoided.

[0132] The criteria for determining obstacle points are adjusted based on the dynamic changes in vegetation destruction. ,in, Indicates time The threshold for the rate of change of resistance after dynamic adjustment. This represents the coefficient indicating the influence of vegetation destruction on the rate of change of resistance. Indicates time The lower the vegetation cover loss rate, the higher the vegetation destruction rate, the lower the judgment threshold, and the more sensitively it can identify possible obstacle points;

[0133] In step S4, the grip points and obstacle points are dynamically adjusted based on the proxy simulation, specifically including:

[0134] Agent-based modeling was used to simulate the frequency of species passing through grips in ecological corridors:

[0135] ,in, Indicates time Sub-agents through pinch points frequency, This represents the total number of agent individuals participating in the migration in the simulation. Indicates time Sub-agent Location, Indicates if the agent After the pinch point Output Otherwise output , Reflecting the pinch point In the migration network, high-frequency pinch points should be prioritized for protection based on their importance.

[0136] Simulate the stress level of the agent passing through the obstacle point: ,in, Indicates time Lower obstacle point The level of agent pressure, Indicates agent The dynamic ecological resistance surface at the location, the pressure level of the obstacle point reflects its negative impact on the migration route, and obstacle points with high pressure need to be repaired or mitigated first.

[0137] Specifically, in step S4, pinch points and obstacle points are identified using current density and resistance change rate, and the identification threshold is dynamically adjusted to adapt to changes in mine expansion and vegetation destruction. At the same time, the importance of pinch points and the pressure level of obstacle points are quantified by proxy simulation, and key locations and repair strategies are dynamically adjusted in a priority manner.

[0138] In this article, an ecological corridor refers to a passageway in an ecological network that connects two or more habitats, providing a means for species migration, population exchange, and the functional flow of the ecosystem.

[0139] Step S5, Ecological management based on dynamic results,

[0140] Based on the ecological corridors, splints, and obstacles identified in steps S3 and S4, targeted ecological protection and restoration measures are implemented. The identification results are transmitted to the ecological management system in real time, allowing the mine management to prioritize the restoration or protection of key corridors and splints, thereby reducing the impact of human activities on ecological connectivity.

[0141] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for identifying mine ecological safety based on a convolutional neural network model, characterized in that: include, Step S1, Mine data acquisition, Collect mine data, including mine topography data, vegetation density, meteorological data, and mining records. Step S2, Dynamic resistance surface modeling, Based on the mine data collected in step S1, a dynamic ecological resistance surface is constructed to reflect the spatial distribution of the degree of obstruction to species migration in each region of the ecosystem. The resistance value of each region depends on ecological factors. Step S3: Multi-path identification of ecological corridors. By combining deep learning with circuit theory, multiple ecological corridors, i.e. possible migration routes, are identified and generated on a dynamic ecological resistance surface. In step S3, based on deep learning models and time series analysis, the priority of each migration route is adjusted to prioritize the protection of ecological corridors for species with high ecological value and high demand. Step S4: Dynamic identification of ecological pinch points and obstacle points. Based on the dynamic ecological resistance surface constructed in step S2, pinch points and obstacle points in the ecosystem are identified, namely bottlenecks and blocking areas in the migration path. Step S5, Ecological management based on dynamic results, Based on the ecological corridors, pinch points, and obstacle points identified in steps S3 and S4, targeted ecological protection and restoration measures are implemented. In step S2, a convolutional neural network (CNN) is used to dynamically adjust the resistance value based on the topographic data, vegetation density, and water distribution in the meteorological data of the mining area. In step S2, an environmental change response mechanism is introduced: Introduction time and dynamic environmental factors Based on time series data to capture the impact of environmental changes, the dynamic resistance value over time is calculated: ， in, Represents pixels In time The final dynamic resistance value, Represents pixels The dynamically adjusted resistance value, These represent the weighting coefficients of dynamic environmental factors, used to adjust the degree of influence of environmental changes on the resistance value. Indicates time Next pixel The dynamic environmental factors are obtained from measured data and range from [missing information]. , A dynamic ecological resistance surface is generated based on the final dynamic resistance value: ， in, Indicates time The dynamic ecological resistance surface Indicates the total regional extent of the ecosystem; In step S4, agent-based modeling is used to simulate the frequency of species passing through grips in the ecological corridor: ,in, Indicates time Sub-agents through pinch points frequency, This represents the total number of agent individuals participating in the migration in the simulation. Indicates time Sub-agents Location, Indicates if the agent is an individual After the pinch point Output Otherwise output .

2. The mine ecological safety identification method based on a convolutional neural network model according to claim 1, characterized in that, In step S2, based on the data collected in step S1, a convolutional neural network (CNN) is used to dynamically adjust the resistance value according to the topographic data, vegetation density, and water distribution in the meteorological data of the mining area. Specific steps include: Spatial rasterization of terrain data: ， in, Represents pixels The initial resistance value, This represents a function that normalizes various ecological factors, keeping the range within a certain range. between, Represents pixels The slope value indicates the terrain gradient; the steeper the slope, the higher the resistance. Represents pixels The higher the vegetation cover density, the lower the resistance. Represents pixels The water body distribution characteristics show that the resistance is lower the closer to the water body. Based on the initial resistance value A convolutional neural network (CNN) is used to dynamically adjust the resistance, capturing local spatial characteristics and reflecting the interaction of ecological factors. ， in, Represents pixels The dynamically adjusted resistance value, This represents the ReLU activation function, used to preserve the non-linear characteristics of the output. The weight matrix represents the convolution kernel, indicating the weights at different positions. Indicates the size of the convolution kernel. This represents the bias term, used to adjust the baseline level of the overall resistance value. The local neighborhood pixel value represents the initial resistance value.

3. The mine ecological safety identification method based on a convolutional neural network model according to claim 2, characterized in that, Step S3 involves multi-path identification of ecological corridors, specifically including: In dynamic ecological resistance Above, the current intensity during species migration is calculated based on circuit theory: ， in, Represents pixels The current intensity represents the flow intensity in the ecological migration network. This represents the dynamic ecological resistance surface generated in step S2. and Let them represent the set of origin points and the set of destination points for migration in the ecosystem, respectively. Representing the starting point and target point The voltage value is set to and ; Using current intensity Identify potential migration routes: ， Representing a path In pixels The indicated value, For path index, Represents the unit step function, when the input is greater than... Time output Otherwise output , Representing a path The current threshold is used to control path flow. The number of paths is adjusted to generate multiple evenly distributed migration paths; Introducing a deep learning model to dynamically optimize path weights based on ecological factors: ， in, Let represent the objective function, whose value is the weighted sum of the resistance costs and negative impacts of all paths. This represents the total number of candidate migration routes. These represent the first, second, and third weighting parameters, respectively, which are used to measure the impact of drag, terrain gradient, and dynamic environmental factors. Represents pixels The terrain gradient corresponds to the slope cost of migration. It represents the unit of integration in two-dimensional space.

4. The mine ecological safety identification method based on a convolutional neural network model according to claim 3, characterized in that, In step S3, adjusting path priority based on time series data specifically includes: Combine time series models to predict the importance of each path and dynamically adjust the priority: ， in, Representing a path In time priority, These represent the adjustment weights for the first priority and the second priority, respectively. Representing a path Migration needs at time t; Using time series analysis to predict future migration demand: ， in, Representing a path In time The migration needs, The time decay factor representing migration demand. The response coefficient represents the response to environmental changes. Indicates time Real-time changes in environmental factors Based on path priority, select paths with high ecological value for protection.

5. The mine ecological safety identification method based on a convolutional neural network model according to claim 4, characterized in that, In step S4, based on the circuit theory model, the current density in different regions is calculated to identify high-importance nodes and obstacles that hinder species migration in the ecological network. The high-importance nodes are called pinch points.

6. The mine ecological safety identification method based on a convolutional neural network model according to claim 5, characterized in that, In step S4, the positions and priorities of the grip points and obstacle points are adjusted according to changes in mining operations; In step S4, combining the vegetation density changes and meteorological changes collected in step S1, the agent-based modeling method is used to simulate the frequency and pressure of different species passing through pinch points and obstacle points in the ecological corridor, and dynamically adjust the positions of pinch points and obstacle points.

7. The mine ecological safety identification method based on a convolutional neural network model according to claim 6, characterized in that, In step S4, the current density is calculated based on the resistance surface: Using circuit theory, the current distribution on the ecological resistance surface is simulated, and the current density in each region is calculated: ,in, Represents pixels The current density, i.e., the degree of stress at which a point serves as a migration route within an ecological network. Represents pixels Conduction rate, This represents the dynamic ecological resistance surface generated in step S2; Based on the current density map, regions with high current concentrations are used as pinch points: ,in, Represents pixels Is it a pinch point? Represented as grip point, Indicates a non-grip point. This indicates an indicator function that outputs when the condition is true. Otherwise output , Indicates the current density threshold; Obstacles, i.e., blocked areas, are detected using conductivity and rate of change of resistance. ,in, Represents pixels Is it an obstacle point? Represented as obstacle points, Indicates non-obstacle points. Represents pixels The rate of change of resistance value, This represents the threshold for the rate of change of resistance.

8. The mine ecological safety identification method based on a convolutional neural network model according to claim 7, characterized in that, In step S4, the grip points and obstacle points are adjusted according to the dynamics of mining operations: Based on the expanded mining area, the current density threshold is redefined: ,in, Indicates time The pinch current density threshold after dynamic adjustment This represents the influence coefficient of current density in the expanded mining area. Indicates time The proportion of underground mining expansion areas; The criteria for determining obstacle points are adjusted based on the dynamic changes in vegetation destruction. ,in, Indicates time The threshold for the rate of change of resistance after dynamic adjustment. This represents the coefficient indicating the influence of vegetation destruction on the rate of change of resistance. Indicates time Lower vegetation cover loss rate; Step S4 also includes: Simulate the stress level of the agent passing through the obstacle point: ,in, Indicates time Lower obstacle point The level of agent pressure, Represents agent individual The dynamic ecological resistance surface at the location, the pressure level of the obstacle point reflects its negative impact on the migration path, and obstacle points with high pressure need to be repaired or mitigated first.

Images