A mine environment treatment method and system based on AI virtual simulation technology

Abstract

The application relates to the technical field of mine treatment, in particular to a mine environment treatment method and system based on AI virtual simulation technology. The method comprises the following steps: obtaining a dynamic geological parameter set of a target mine; analyzing the dynamic geological parameter set to determine injection data and soil capillary pressure detection data; determining a composite diffusion mode of a treatment agent according to the injection data and the soil capillary pressure detection data; analyzing the composite diffusion mode to determine a space-time evolution law; constructing an AI dynamic simulation model containing a treatment agent penetration feedback mechanism according to the space-time evolution law; based on the AI dynamic simulation model, performing multi-node data growth simulation on a plurality of treatment schemes to generate a treatment effect evolution atlas containing stratum response time-varying characteristics; analyzing the treatment effect evolution atlas to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme; and determining an optimal treatment scheme implementation path according to the matching degree.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of mine governance, and in particular to a mine environment governance method and system based on AI virtual simulation technology. BACKGROUND

[0002] Mine environment governance, as an important topic in the field of ecological restoration, has the core challenge of the complexity of pollution diffusion mechanism and the uncertainty of governance effect prediction. Especially in the application of in-situ leaching method, after the acid liquor is injected into the ore body through the injection hole, it causes multi-dimensional environmental problems.

[0003] In actual governance, the leaching liquid not recovered during ore extraction will diffuse with surface runoff, and the residual liquor will seep into the groundwater system, forming long-term hidden pollution, causing excessive ammonium nitrogen and sulfate in the soil, and causing changes in soil properties. SUMMARY

[0004] The application provides a mine environment governance method and system based on AI virtual simulation technology to solve the above problems.

[0005] In a first aspect, the application provides a mine environment governance method based on AI virtual simulation technology, which comprises:

[0006] Obtaining a set of dynamic geological parameters of a target mine; analyzing the set of dynamic geological parameters to determine injection data and soil capillary pressure detection data;

[0007] According to the injection data and the soil capillary pressure detection data, a composite diffusion mode of the treatment agent is determined;

[0008] Analyzing the composite diffusion mode to determine the spatio-temporal evolution law; according to the spatio-temporal evolution law, an AI dynamic simulation model containing a treatment agent penetration feedback mechanism is constructed;

[0009] Based on the AI dynamic simulation model, a multi-node data growth simulation is performed on a plurality of governance schemes to generate a governance effect evolution map containing stratum response time-varying characteristics;

[0010] Analyzing the governance effect evolution map to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each governance scheme; according to the matching degree, an optimal governance scheme implementation path is determined.

[0011] By the scheme, the dynamic geological parameter set of the target mine is obtained, which helps to eliminate the migration path prediction deviation caused by single parameter sampling. Analyzing the dynamic geological parameter set, the injection data and the soil capillary pressure detection data are determined, which helps to avoid the lag error of the static parameter library. According to the injection data and the soil capillary pressure detection data, the composite diffusion mode of the treatment agent is determined, which helps to eliminate the misjudgment problem of the traditional linear model to the composite migration path and avoid the over-prediction of the capillary effect on the deep diffusion. Analyzing the composite diffusion mode, the spatio-temporal evolution law is determined, which helps to eliminate the prediction deviation of the residual amount of the agent caused by the parameter solidification. According to the spatio-temporal evolution law, the AI dynamic simulation model containing the penetration feedback mechanism of the treatment agent is constructed, which helps to early warn the potential risk of the decrease of the vegetation survival rate. Based on the AI dynamic simulation model, a plurality of treatment schemes are simulated by multi-node data growth, and the treatment effect evolution atlas containing the time-varying characteristics of the stratum response is generated, which helps to clearly show the standard time of the pollution removal rate and assist the engineering progress control. Analyzing the treatment effect evolution atlas, the matching degree of the action depth of the treatment agent and the vegetation planting condition in each treatment scheme is determined, which helps to reduce the resource waste rate and improve the standard rate of the comprehensive treatment mine. According to the matching degree, the optimal treatment scheme implementation path is determined, which helps to improve the pollution removal rate while reducing the corresponding agent cost increase.

[0012] Optionally, the dynamic geological parameter set further includes a geological structure of the target mine; and the determining the composite diffusion mode of the treatment agent according to the injection data and the soil capillary pressure detection data comprises:

[0013] analyzing the injection data to determine injection hole distribution;

[0014] obtaining an injection hole image based on the injection hole distribution;

[0015] analyzing the injection hole image to determine surface residual liquid characteristics;

[0016] determining residual liquid concentration according to the surface residual liquid characteristics;

[0017] determining a migration rate of the treatment agent according to the residual liquid concentration and the soil capillary pressure detection data;

[0018] analyzing the geological structure to determine stratum fracture strike;

[0019] constructing a three-dimensional pollution diffusion vector field according to the migration rate and the stratum fracture strike;

[0020] determining the composite diffusion mode of the treatment agent according to the three-dimensional pollution diffusion vector field.

[0021] Optionally, the liquid injection data comprises an initial liquid injection concentration; and the determining of the migration rate of the treatment agent based on the residual liquid concentration and the soil capillary pressure detection data comprises:

[0022] analyzing the soil capillary pressure detection data based on the geological structure to determine capillary pressure gradients at different levels;

[0023] calculating the migration rate of the treatment agent according to Darcy's law and the capillary pressure gradients at different levels based on the initial liquid injection concentration and the residual liquid concentration.

[0024] Optionally, the constructing of the AI dynamic simulation model comprising a treatment agent penetration feedback mechanism based on the spatiotemporal evolution law comprises:

[0025] determining spatial concentration distribution characteristics of the treatment agent according to the spatiotemporal evolution law;

[0026] inputting the spatial concentration distribution characteristics into a preset neutralization reaction efficiency prediction model to output time series data of neutralization efficiency of the treatment agent;

[0027] dynamically generating a self-optimization instruction set of reagent injection concentration and liquid injection hole adjustment according to the time series data of neutralization efficiency and a preset mine treatment standard;

[0028] fusing the self-optimization instruction set and the stratum fracture strike to construct the AI dynamic simulation model comprising the treatment agent penetration feedback mechanism.

[0029] Optionally, the constructing of the preset neutralization reaction efficiency prediction model comprises:

[0030] determining chemical properties of the treatment agent according to the liquid injection data;

[0031] determining soil buffer capacity according to the geological structure and the soil capillary pressure detection data;

[0032] establishing a neutralization reaction efficiency prediction model according to the chemical properties and the soil buffer capacity.

[0033] Optionally, the dynamic geological parameter set comprises soil sampling data; and the establishing of the neutralization reaction efficiency prediction model according to the chemical properties and the soil buffer capacity comprises:

[0034] analyzing the soil sampling data to determine calcium carbonate content and cation exchange capacity;

[0035] determining a change relationship between consumption of the treatment agent and PH value according to the chemical properties;

[0036] According to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity and the soil buffer capacity, a neutralization reaction efficiency prediction model is established.

[0037] Optionally, based on the AI dynamic simulation model, a plurality of management schemes are simulated for multi-node data growth to generate a management effect evolution atlas containing time-varying characteristics of stratum response, including:

[0038] According to the soil capillary pressure detection data, the soil capillary ability is determined;

[0039] The migration rate and the soil capillary ability are parameterized to obtain basic parameters and input to the AI dynamic simulation model so that the AI dynamic simulation model performs multi-node data growth simulation for a plurality of management schemes based on the basic parameters;

[0040] The simulation results of the AI dynamic simulation model are obtained, and the simulation results are analyzed to determine the penetration data of each management scheme;

[0041] The penetration data is analyzed to determine the change in penetration depth and the change timestamp;

[0042] According to the change in penetration depth and the change timestamp, a management effect evolution atlas containing time-varying characteristics of stratum response is generated.

[0043] Optionally, the management effect evolution atlas is analyzed to determine the matching degree of the action depth of the management agent and the vegetation planting conditions in each management scheme, including:

[0044] Obtain a set of vegetation planting condition parameters corresponding to each management scheme;

[0045] Based on the management effect evolution atlas, the actual action depth of each management scheme at different time nodes and the corresponding soil parameters are extracted;

[0046] The soil parameters within the actual action depth are matched with the set of vegetation planting condition parameters to determine the matching degree of the action depth of the management agent and the vegetation planting conditions in each management scheme.

[0047] Optionally, the optimal management scheme implementation path is determined according to the matching degree, including:

[0048] A management cost evaluation model is established to calculate the implementation cost parameters of each management scheme;

[0049] A multi-objective optimization function is constructed to perform weight calculation on the matching degree and the implementation cost parameters;

[0050] Select the governance scheme with the highest score and the lowest cost as the optimal implementation path according to the calculation result.

[0051] In a second aspect, the application provides a mine environment governance system based on AI virtual simulation technology, which comprises:

[0052] A parameter set analysis module is configured to obtain a dynamic geological parameter set of a target mine, analyze the dynamic geological parameter set, and determine injection data and soil capillary pressure detection data.

[0053] A mode determination module is configured to determine a composite diffusion mode of a governance agent based on the injection data and the soil capillary pressure detection data.

[0054] A model establishment module is configured to analyze the composite diffusion mode, determine a spatiotemporal evolution rule, and construct an AI dynamic simulation model containing a governance agent penetration feedback mechanism based on the spatiotemporal evolution rule.

[0055] A graph generation module is configured to perform multi-node data growth simulation on a plurality of governance schemes based on the AI dynamic simulation model, and generate a governance effect evolution graph containing stratum response time-varying characteristics.

[0056] A path determination module is configured to analyze the governance effect evolution graph, determine the matching degree of the action depth of the governance agent and the vegetation planting conditions in each governance scheme, and determine an optimal governance scheme implementation path based on the matching degree.

[0057] Optionally, the dynamic geological parameter set further comprises a geological structure of the target mine, and when the mode determination module determines the composite diffusion mode of the governance agent based on the injection data and the soil capillary pressure detection data, it is configured to:

[0058] analyze the injection data to determine injection hole distribution;

[0059] obtain an injection hole image based on the injection hole distribution;

[0060] analyze the injection hole image to determine surface residual liquid characteristics;

[0061] determine residual liquid concentration based on the surface residual liquid characteristics;

[0062] determine the migration rate of the governance agent based on the residual liquid concentration and the soil capillary pressure detection data;

[0063] analyze the geological structure to determine stratum fracture strike;

[0064] construct a three-dimensional pollution diffusion vector field based on the migration rate and the stratum fracture strike;

[0065] According to the three-dimensional pollution diffusion vector field, a composite diffusion mode of the treatment agent is determined.

[0066] Optionally, the liquid injection data includes an initial liquid injection concentration; when the mode determination module determines the migration rate of the treatment agent according to the residual liquid concentration and the soil capillary pressure detection data, it is used for:

[0067] Based on the geological structure, the soil capillary pressure detection data is analyzed to determine the capillary pressure gradient of different levels.

[0068] Based on the initial liquid injection concentration and the residual liquid concentration, the migration rate of the treatment agent is calculated according to Darcy's law and the capillary pressure gradient of different levels.

[0069] Optionally, when the model establishment module constructs an AI dynamic simulation model containing a treatment agent penetration feedback mechanism according to the spatio-temporal evolution law, it is used for:

[0070] According to the spatio-temporal evolution law, the spatial concentration distribution characteristics of the treatment agent are determined.

[0071] The spatial concentration distribution characteristics are input into a preset neutralization reaction efficiency prediction model, and the neutralization efficiency time series data of the treatment agent is output.

[0072] According to the neutralization efficiency time series data and a preset mine treatment standard, a self-optimization instruction set of reagent injection concentration and liquid injection hole position adjustment is dynamically generated.

[0073] The self-optimization instruction set and the stratum fracture trend are fused to construct an AI dynamic simulation model containing a treatment agent penetration feedback mechanism.

[0074] Optionally, the mine environment treatment system based on AI virtual simulation technology further includes a prediction model establishment module, which is used for:

[0075] According to the liquid injection data, the chemical properties of the treatment agent are determined.

[0076] According to the geological structure and the soil capillary pressure detection data, the soil buffer capacity is determined.

[0077] According to the chemical properties and the soil buffer capacity, a neutralization reaction efficiency prediction model is established.

[0078] Optionally, the dynamic geological parameter set includes soil sampling data; when the prediction model establishment module establishes a neutralization reaction efficiency prediction model according to the chemical properties and the soil buffer capacity, it is used for:

[0079] The soil sampling data is analyzed to determine the calcium carbonate content and the cation exchange capacity.

[0080] determine a change relationship between the consumption of the treatment agent and the PH value according to the chemical properties;

[0081] establish a neutralization reaction efficiency prediction model according to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity and the soil buffer capacity.

[0082] Optionally, the atlas generation module generates a treatment effect evolution atlas containing stratum response time-varying characteristics by performing multi-node data growth simulation on a plurality of treatment schemes based on the AI dynamic simulation model, for:

[0083] determine soil capillary capacity according to the soil capillary pressure detection data;

[0084] quantify the migration rate and the soil capillary capacity as parameters to obtain basic parameters and input the basic parameters into the AI dynamic simulation model to enable the AI dynamic simulation model to perform multi-node data growth simulation on a plurality of treatment schemes based on the basic parameters;

[0085] obtain simulation results of the AI dynamic simulation model, analyze the simulation results, and determine penetration data of each treatment scheme;

[0086] analyze the penetration data to determine penetration depth changes and change time stamps;

[0087] generate a treatment effect evolution atlas containing stratum response time-varying characteristics according to the penetration depth changes and the change time stamps.

[0088] Optionally, when the path determination module analyzes the treatment effect evolution atlas to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme, it is used for:

[0089] obtain a set of vegetation planting condition parameters corresponding to each treatment scheme;

[0090] extract the actual action depth of each treatment scheme at different time nodes and the corresponding soil parameters based on the treatment effect evolution atlas;

[0091] perform matching degree calculation on the soil parameters within the actual action depth and the set of vegetation planting condition parameters to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme.

[0092] Optionally, when the path determination module determines an optimal treatment scheme implementation path according to the matching degree, it is used for:

[0093] determine to establish a treatment cost evaluation model to calculate implementation cost parameters of each treatment scheme;

[0094] A multi-objective optimization function is constructed to comprehensively calculate the matching degree and the implementation cost parameter.

[0095] According to the calculation result, the governance scheme with the highest score and the lowest cost is selected as the optimal implementation path. BRIEF DESCRIPTION OF DRAWINGS

[0096] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0097] Figure 1 An application scenario schematic diagram provided by an embodiment of the present application;

[0098] Figure 2 A flowchart of a mine environment governance method based on AI virtual simulation technology provided by an embodiment of the present application;

[0099] Figure 3 A mine environment governance system structure schematic diagram based on AI virtual simulation technology provided by an embodiment of the present application. DETAILED DESCRIPTION

[0100] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions in the embodiments of the present application will be described clearly and completely in the following with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0101] In addition, the term "and / or" in this paper is only to describe the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " in this paper generally represents an "or" relationship between the associated objects unless otherwise specified.

[0102] The embodiments of the present application will be further described in detail below with reference to the drawings of the specification.

[0103] In actual governance, the leaching liquid not recovered during ore extraction will diffuse with surface runoff, and the residual liquid will seep into the groundwater system, forming long-term hidden pollution, leading to excessive ammonium nitrogen and sulfate in the soil, and causing changes in soil properties.

[0104] Based on this, the application provides a mine environment governance method and system based on AI virtual simulation technology. The dynamic geological parameter set of the target mine is obtained, which helps to eliminate the migration path prediction deviation caused by single parameter sampling. The injection data and soil capillary pressure detection data are determined by analyzing the dynamic geological parameter set, which helps to avoid the lag error of the static parameter library. According to the injection data and soil capillary pressure detection data, the composite diffusion mode of the treatment agent is determined, which helps to eliminate the misjudgment problem of the traditional linear model to the composite migration path and avoid the over-prediction of the capillary effect on deep diffusion. The spatio-temporal evolution law is determined by analyzing the composite diffusion mode, which helps to eliminate the prediction deviation of the residual amount of the agent caused by parameter solidification. According to the spatio-temporal evolution law, the AI dynamic simulation model containing the penetration feedback mechanism of the treatment agent is constructed, which helps to early warning the potential risk of the decrease of the vegetation survival rate. Based on the AI dynamic simulation model, the multi-node data growth simulation is performed on several treatment schemes to generate a treatment effect evolution atlas containing the time-varying characteristics of the stratum response, which helps to clearly show the standard time of the pollution removal rate and assist the engineering progress control. The matching degree of the action depth of the treatment agent and the vegetation planting condition in each treatment scheme is determined by analyzing the treatment effect evolution atlas, which helps to reduce the resource waste rate and improve the standard rate of the comprehensive treatment mine. According to the matching degree, the optimal treatment scheme implementation path is determined, which helps to improve the pollution removal rate while reducing the corresponding agent cost increase.

[0105] Figure 1 An application scenario diagram is provided for the application. When the virtual simulation technology is used for mine environment governance, the method provided by the application is applied.

[0106] Specifically, the method provided by the application is applied in any server, and the server interacts with a geological monitoring device. The geological monitoring device can include a three-dimensional geological radar, a centrifuge, a distributed optical fiber sensor, and the like. The dynamic geological parameter set of the target mine is obtained and analyzed by the geological monitoring device to determine the injection data and the soil capillary pressure detection data. According to the injection data and the soil capillary pressure detection data, the composite diffusion mode of the treatment agent is determined and analyzed to obtain the spatio-temporal evolution law. According to the spatio-temporal evolution law, the AI dynamic simulation model containing the penetration feedback mechanism of the treatment agent is constructed. Based on the AI dynamic simulation model, the multi-node data growth simulation is performed on several treatment schemes to generate a treatment effect evolution atlas containing the time-varying characteristics of the stratum response, thereby determining the matching degree of the action depth of the treatment agent and the vegetation planting condition in each treatment scheme, which helps to reduce the resource waste rate and improve the standard rate of the comprehensive treatment mine. According to the matching degree, the optimal treatment scheme implementation path is determined, which helps to improve the pollution removal rate while reducing the corresponding agent cost increase. The specific implementation mode can refer to the following embodiments.

[0107] Figure 2A flowchart of a mine environment treatment method based on an AI virtual simulation technology is provided for an embodiment of the present application. The method of the embodiment can be applied to a server in the above scenarios. As shown in Figure 2 the method includes:

[0108] S201, acquiring a set of dynamic geological parameters of a target mine; analyzing the set of dynamic geological parameters to determine injection data and soil capillary pressure detection data;

[0109] The target mine can be a mining area that needs to be treated.

[0110] The set of dynamic geological parameters can be a set of geological parameters of spatio-temporal evolution characteristics.

[0111] The injection data can be time series data generated during in-situ leaching operations.

[0112] The soil capillary pressure detection data can be a soil capillary force parameter.

[0113] Specifically, the fissure distribution topology of the target mine is obtained by three-dimensional geological radar scanning, the main fissure dip angle, branch fissure density, and connectivity rate parameters are extracted, a distributed optical fiber sensor array is arranged around the injection hole, and the injection pressure, flow rate, and temperature are collected in real time. The soil samples at different depths are obtained by stratified sampling, the capillary pressure-saturation curve of each layer is determined by a centrifuge, and the critical capillary pressure, porosity, and wetting angle are extracted. These data together form the set of dynamic geological parameters.

[0114] Among them, the injection pressure, flow rate, and temperature constitute the injection data; the critical capillary pressure, porosity, and wetting angle constitute the soil capillary pressure monitoring data.

[0115] S202, determining a composite diffusion mode of the treatment agent according to the injection data and the soil capillary pressure detection data;

[0116] The treatment agent can be a chemical agent used to neutralize acidic pollutants.

[0117] The composite diffusion mode can be a double-channel migration model that considers both fissure channel flow rate and capillary vertical seepage rate.

[0118] Specifically, the injection data is analyzed by DTS thermodynamic inversion to determine the spatial distribution of residual liquid concentration; then, based on the Darcy's law correction equation, the soil capillary pressure detection data is introduced, and the porosity-wetting angle joint correction coefficient determined through multiple soil column permeation experiments is introduced. When the main fissure dip angle is large, the fissure channel flow rate and capillary vertical seepage rate determined by the simplified Navier-Stokes equation are vector superimposed to determine the composite diffusion mode of the treatment agent.

[0119] S203, analyze the composite diffusion mode, determine the spatio-temporal evolution law; according to the spatio-temporal evolution law, build an AI dynamic simulation model containing a treatment agent penetration feedback mechanism;

[0120] The spatio-temporal evolution law can be the concentration distribution change law of the treatment agent in three-dimensional space and time dimension.

[0121] The treatment agent penetration feedback mechanism can be a closed-loop control device that dynamically adjusts the injection parameters.

[0122] The AI dynamic simulation model can be a hybrid architecture for simulating the multi-process coupling effect of agent migration-reaction-vegetation restoration.

[0123] Specifically, based on the composite diffusion mode, access the meteorological data interface, real-time obtain rainfall intensity and duration data, calculate the dilution factor of rainfall intensity and duration data on shallow liquid concentration; then, according to the soil cation exchange capacity and calcium carbonate content, dynamically correct the treatment agent consumption rate equation; further, every interval time step, update the agent concentration distribution in the penetration vector field, and calculate the mass flux balance of the capillary zone and the fracture zone; finally, output the three-dimensional concentration field sequence containing the time stamp, form the spatio-temporal evolution law. Subsequently, based on the spatio-temporal evolution law, 3D convolution layers are used to extract the spatio-temporal characteristics of the diffusion mode; at the same time, the full connection layer is used to process the stratum response monitoring data; combined with the spatio-temporal characteristics of the diffusion mode and the stratum response monitoring data, an AI dynamic simulation model containing a treatment agent penetration feedback mechanism is built.

[0124] S204, based on the AI dynamic simulation model, multi-node data growth simulation is performed on several treatment schemes to generate a treatment effect evolution atlas containing stratum response time-varying characteristics;

[0125] The treatment scheme can be an operation combination containing injection parameters, treatment agent types, and vegetation planting time sequence.

[0126] The stratum response time-varying characteristics can be the dynamic change characteristics of the stratum parameters during the treatment process.

[0127] The treatment effect evolution atlas can be a three-dimensional heat map generated by multi-node data growth simulation.

[0128] Specifically, according to the AI dynamic simulation model, time nodes are divided according to treatment stages, and space nodes are divided according to depth layers; further, in the injection period, the agent penetration depth and fracture filling degree are updated; then, in the reaction period, the amount of precipitate of the first neutralization product is calculated, and the porosity parameter is corrected; subsequently, in the vegetation restoration period, the root growth model is introduced, and the soil pH value and porosity are associated, thereby generating a treatment effect evolution atlas containing stratum response time-varying characteristics.

[0129] S205, analyze the governance effect evolution map, determine the matching degree of the action depth of the governance agent and the vegetation planting condition in each governance scheme; according to the matching degree, determine the optimal governance scheme implementation path.

[0130] The action depth can be the vertical coverage range of the governance agent effectively neutralizing the pollutants.

[0131] The vegetation planting condition can be a threshold parameter of the surface soil adapting to the survival of vegetation.

[0132] The matching degree can be a quantitative index of the adaptation of the governance scheme to the mine type.

[0133] The optimal governance scheme implementation path can be a dynamic execution scheme screened out through multi-objective optimization.

[0134] Specifically, based on the governance effect evolution map, the matching degree level is determined by calculating the Pearson correlation coefficient of the action depth and the vegetation planting condition; then, a plurality of geological response curves are generated by Monte Carlo simulation, and the success probability distribution of each governance scheme is calculated; subsequently, based on the resource-stage matrix model in multi-criteria decision analysis, by integrating the cost-benefit ratio evaluation framework of classical operations research and the spatial mapping method of risk matrix, an implementation path decision matrix is constructed, with column dimension being governance stage division, row dimension being resource input level, and cell marking expected effect achievement rate and economic cost ratio; finally, by using the hybrid optimization framework of Pareto optimal theory and genetic algorithm, multi-objective optimization is determined, combined with the weight distribution rules of non-dominated sorting mechanism and target programming method, the optimal governance scheme implementation path is finally determined.

[0135] By the scheme, the dynamic geological parameter set of the target mine is obtained, which helps to eliminate the migration path prediction deviation caused by single parameter sampling. Analyzing the dynamic geological parameter set, the injection data and soil capillary pressure detection data are determined, which helps to avoid the lag error of the static parameter library. According to the injection data and soil capillary pressure detection data, the composite diffusion mode of the treatment agent is determined, which helps to eliminate the misjudgment problem of the traditional linear model to the composite migration path and avoid the excessive prediction of the capillary effect on the deep diffusion. Analyzing the composite diffusion mode, the spatio-temporal evolution law is determined, which helps to eliminate the prediction deviation of the residual amount of the agent caused by the parameter solidification. According to the spatio-temporal evolution law, the AI dynamic simulation model containing the penetration feedback mechanism of the treatment agent is constructed, which helps to early warn the potential risk of the decrease of the vegetation survival rate. Based on the AI dynamic simulation model, a plurality of treatment schemes are simulated by multi-node data growth, and the treatment effect evolution atlas containing the time-varying characteristics of the stratum response is generated, which helps to clearly show the standard time of the pollution removal rate and assist the engineering progress control. Analyzing the treatment effect evolution atlas, the matching degree of the action depth of the treatment agent and the vegetation planting condition in each treatment scheme is determined, which helps to reduce the resource waste rate and improve the standard rate of the comprehensive treatment mine. According to the matching degree, the optimal treatment scheme implementation path is determined, which helps to improve the pollution removal rate while reducing the corresponding agent cost increase.

[0136] In some embodiments, the injection data is analyzed to determine the injection hole distribution; based on the injection hole distribution, an injection hole image is obtained; the injection hole image is analyzed to determine the surface residual liquid feature; based on the surface residual liquid feature, the residual liquid concentration is determined; based on the residual liquid concentration and the soil capillary pressure detection data, the migration rate of the treatment agent is determined; the geological structure is analyzed to determine the stratum fracture trend; based on the migration rate and the stratum fracture trend, a three-dimensional pollution diffusion vector field is constructed; based on the three-dimensional pollution diffusion vector field, the composite diffusion mode of the treatment agent is determined.

[0137] The injection hole distribution can be the three-dimensional space arrangement state of the injection hole in the target mine.

[0138] The injection hole image can be an optical image of the surface of the injection hole.

[0139] The surface residual liquid feature can be the visual attributes such as color, texture, and distribution form of the residual liquid around the injection hole in the image.

[0140] The residual liquid concentration can be the mass percentage concentration of the acid liquid residual in the surface of the injection hole and the shallow soil.

[0141] The migration rate can be the penetration speed of the treatment agent in the vertical or horizontal direction.

[0142] The geological structure can be a physical structure composed of stratification characteristics, fracture network distribution, and pore connectivity of rock-soil bodies in the stratum of the target mine.

[0143] The formation fracture direction can be an included angle of a main fracture extension direction in the formation relative to a geographic north direction.

[0144] The three-dimensional pollution diffusion vector field can be a three-dimensional space model constructed based on the migration rate and the formation fracture direction.

[0145] Specifically, the spatial coordinates of injection holes and the injection flow parameters in the injection data are extracted, and the injection hole distribution is generated by a spatial interpolation algorithm. Then, based on the injection hole distribution, a high-spectrum camera is carried by a UAV to collect the surface image of the injection hole. Subsequently, the HSV color space segmentation algorithm is used to accurately extract the residual liquid area of the acid liquid by hue separation; the residual liquid patch area ratio and the edge fractal dimension are calculated as the surface residual liquid characteristics. The residual liquid concentration of NH4 + -N and SO4 2- is calculated by near-infrared spectrum analysis to determine the absorption peak intensity of the residual liquid, combined with the calibration curve of the ammonium sulfate standard solution. The vertical migration rate is calculated by inputting the soil capillary pressure monitoring data and the permeability coefficient obtained by looking up the soil type table. Further, the three-dimensional geological radar scanning data is loaded, the Hough transform is used to detect the planar fracture trace in the three-dimensional point cloud, and the main fracture spatial azimuth angle is identified by accumulator space peak detection. Thus, the included angle between the fracture direction and the horizontal plane is calculated, and the effective water-conducting fracture with a large inclination angle, i.e., the formation fracture direction, is selected. In the finite element grid, the grid density is dynamically adjusted according to the permeability coefficient of the formation medium, thereby defining the three-dimensional pollution diffusion vector field: first, the shallow grid adopts the vertical rate component dominated by capillary action; then, the deep grid superimposes the horizontal rate component in the direction of the fracture direction; finally, the capillary and fracture migration mechanisms are coupled through the Darcy-Buysinesq equation to generate the three-dimensional pollution diffusion vector field. Subsequently, the spatiotemporal distribution characteristics of the maximum diffusion direction and rate in the three-dimensional pollution diffusion vector field are extracted; then, the diffusion boundary of the capillary zone and the fracture zone is divided according to the rate threshold; finally, the composite diffusion mode containing the range of double mechanism is output.

[0146] By the scheme, the liquid injection data is analyzed, the liquid injection hole distribution is determined, the spatial boundary of the high permeability risk area is determined, and the problem of incomplete area coverage caused by discrete sampling is avoided. Based on the liquid injection hole distribution, the liquid injection hole image is obtained, which helps to eliminate the subjective error problem of visual inspection. Analyzing the liquid injection hole image, the surface residual liquid characteristics are determined, which helps to break through the limitation of single thickness measurement. According to the surface residual liquid characteristics, the residual liquid concentration is determined, which helps to improve the detection efficiency and has no secondary pollution risk. According to the residual liquid concentration and the soil capillary pressure detection data, the migration rate of the remediation agent is determined, and the rate prediction error of the shallow capillary action area is reduced. Analyzing the geological structure, the stratum fracture trend is determined, which helps to eliminate the problem that the two-dimensional profile analysis cannot capture the three-dimensional fracture spatial distribution. According to the migration rate and the stratum fracture trend, a three-dimensional pollution diffusion vector field is constructed, which helps to improve the prediction accuracy of the traditional linear model. According to the three-dimensional pollution diffusion vector field, the composite diffusion mode of the remediation agent is determined, which avoids the adaptation error risk of the traditional experience judgment.

[0147] In some embodiments, based on the geological structure, the soil capillary pressure detection data is analyzed to determine the capillary pressure gradient of different levels; based on the initial liquid injection concentration and the residual liquid concentration, the migration rate of the remediation agent is calculated according to the Darcy law and the capillary pressure gradient of different levels.

[0148] The capillary pressure gradient can be the change rate of the soil capillary pressure per unit depth.

[0149] The initial liquid injection concentration can be the initial mass concentration of the remediation agent injected into the ore body through the liquid injection hole.

[0150] Specifically, based on the geological structure, the capillary pressure original data of different stratum depths is collected by the soil capillary pressure monitoring device; then, the capillary pressure original data is spatially interpolated according to the vertical stratification to generate the capillary pressure gradient of different levels. Subsequently, the spectral absorption peak intensity of the residual liquid on the surface of the liquid injection hole is obtained based on the near-infrared spectrum analyzer; then, the characteristic absorption peak intensity-concentration relationship curve of NH4 + and SO4 2- is calibrated by the ammonium sulfate standard solution; then, the residual liquid concentration is inversely calculated according to the measured absorption peak intensity; the dynamic viscosity is measured by the rotary viscometer and the concentration inhibition coefficient is determined by the soil column experiment; then, the Darcy law formula is constructed by the permeability coefficient, the dynamic viscosity and the concentration inhibition coefficient; and then the migration rate of the remediation agent in the shallow layer, the middle layer and the deep layer is calculated according to the Darcy law formula and the capillary pressure gradient of different levels.

[0151] Through the scheme, based on the geological structure, the soil capillary pressure detection data is analyzed to determine the capillary pressure gradient at different levels, which helps to eliminate the problem of separation of shallow and deep migration mechanisms caused by the fragmentation analysis of static geological parameters. Based on the residual liquid concentration of the initial injection concentration meter, according to the Darcy law and the capillary pressure gradient at different levels, the migration rate of the treatment agent is calculated, which provides the basis for defining the permeation range of the crack extension depth for the comprehensive treatment mine.

[0152] In some embodiments, according to the space-time evolution law, the spatial concentration distribution characteristics of the treatment agent are determined; the spatial concentration distribution characteristics are input into a preset neutralization reaction efficiency prediction model to output time series data of the neutralization efficiency of the treatment agent; according to the neutralization efficiency time series data and the preset mine treatment standard, a self-optimization instruction set of the reagent injection concentration and the injection hole position adjustment is dynamically generated; and the AI dynamic simulation model containing the treatment agent permeation feedback mechanism is constructed by fusing the self-optimization instruction set and the stratum fracture trend.

[0153] The spatial concentration distribution characteristics can be a quantitative representation of the concentration distribution of the treatment agent in three-dimensional space.

[0154] The preset neutralization reaction efficiency prediction model can be a mathematical relationship model for calculating the neutralization efficiency of the treatment agent and the pollutants. It is pre-stored in the server and called when used.

[0155] The time series data of the neutralization efficiency can be the neutralization efficiency decay curve data varying with time.

[0156] The preset mine treatment standard can be a quantitative treatment target set according to different mine types. It is pre-stored in the server and called when used.

[0157] The reagent injection concentration can be the initial mass concentration of the treatment agent injected into the stratum.

[0158] The injection hole position can be a spatial distribution parameter of the drill hole for injecting the treatment agent.

[0159] The self-optimization instruction set can be a dynamic control strategy set generated based on the time series data of the neutralization efficiency and the mine treatment standard.

[0160] Specifically, based on the spatiotemporal evolution law, the spatiotemporal covariance function is constructed through dynamic geological parameters, the spatiotemporal Kriging interpolation method of the three-dimensional concentration field is solved by using the Kriging weight matrix through the weight relationship and time lag effect such as the fracture density and the migration rate, and then the spatial optimal estimation theory of the spatiotemporal Kriging interpolation method is adopted and extended to the time dimension, the migration rate is coupled with the fracture trend, and the spatial concentration distribution characteristics of the treatment agent are calculated. Then, the spatial concentration distribution characteristics at the current time and the dynamic geological parameter set are input, the preset neutralization reaction efficiency prediction model is called based on the neutralization reaction kinetics model, the linear-exponential mixed attenuation law of the soil buffer capacity and the residual liquid concentration is combined, and the preset neutralization reaction efficiency prediction model is constructed through the laboratory calibration parameters; subsequently, the nonlinear regression equation is adopted, the first-order exponential model of the pollutant decay kinetics is used, and the neutralization efficiency time series data is output according to the preset time step. Based on the preset mine treatment standard, the penetration depth target is set according to the three-dimensional geological modeling data of the comprehensive treatment mine, the surface soil pH value target interval and the compliance time window are set according to the real-time monitoring feedback of the pH sensor according to the simple treatment mine, and if the neutralization efficiency time series data is lower than the target interval and does not reach the compliance time window or the penetration depth does not reach the standard, the initial injection concentration and the injection hole coordinate are optimized based on the gradient descent algorithm, and a self-optimization instruction set of the treatment agent injection concentration and the injection hole adjustment is generated. The self-optimization instruction set and the extended Darcy formula are solved, and the migration rate and the concentration distribution are iteratively updated; then, the high-permeability channel is marked in the three-dimensional grid according to the stratum fracture trend; and finally, an AI dynamic simulation model including the treatment agent penetration feedback mechanism is generated.

[0161] According to the spatiotemporal evolution law, the spatial concentration distribution characteristics of the treatment agent are determined, which helps to eliminate the problem of insufficient adaptability of the treatment scheme, and realizes the quantitative expression of the nonlinear decay of the concentration with the depth. The spatial concentration distribution characteristics are input into the preset neutralization reaction efficiency prediction model, the neutralization efficiency time series data of the treatment agent is output, the influence of the soil buffer capacity on the neutralization efficiency is quantified, and the problem of missing dynamic feedback of the neutralization reaction is eliminated. According to the neutralization efficiency time series data and the preset mine treatment standard, the self-optimization instruction set of the treatment agent injection concentration and the injection hole adjustment is dynamically generated, which helps to eliminate the problem of insufficient penetration depth of the treatment agent and meets the adaptation requirements of the vegetation root system. The AI dynamic simulation model including the treatment agent penetration feedback mechanism is constructed by combining the self-optimization instruction set and the stratum fracture trend, which helps to eliminate the problem of missing multi-objective collaborative optimization and improves the modeling robustness of the heterogeneous stratum.

[0162] In some embodiments, the chemical properties of the treatment agent are determined according to the injection data; the soil buffer capacity is determined according to the geological structure and the soil capillary pressure detection data; and the neutralization reaction efficiency prediction model is established according to the chemical properties and the soil buffer capacity.

[0163] The chemical properties can be the component attributes of the treatment agent.

[0164] The soil buffer capacity can be the ability of the soil to resist changes in pH through its own chemical composition.

[0165] Specifically, based on the injection data, the type of the medicament, the initial concentration, and the injection pressure parameter recorded by the injection hole are extracted; then, the molar concentration of the reaction component in the medicament is determined by ion chromatography; and thus the chemical properties of the treatment agent are determined. Based on the drilling exploration data, the fracture direction of the stratum and the capillary pressure detection data are extracted; then, the capillary pressure-saturation curve of the soil at different depths is determined by a pressure membrane instrument, and the pore connectivity is calculated; further, the mass percentage of calcium carbonate is obtained by a calcium carbonate content determination method; subsequently, the soil buffer capacity of the unit mass of soil is determined by a cation exchange capacity detection kit. Then, under the simulated conditions in the laboratory, the treatment agent is mixed with the target pollutant according to the chemical properties; subsequently, the neutralization reaction endpoint is monitored by a pH meter and an ion selective electrode, and the basic neutralization efficiency under static conditions is calculated; finally, the soil buffer parameters and the basic neutralization efficiency are linearly combined to generate a neutralization reaction efficiency prediction model.

[0166] Through the scheme, the chemical properties of the treatment agent are determined according to the injection data, and the action strength of the treatment agent under different geological conditions is clear. The soil buffer capacity is determined according to the geological structure and the capillary pressure detection data of the soil, which helps to reveal the spatial heterogeneity of the buffer capacity with the fracture direction of the stratum and the capillary pressure detection data. According to the chemical properties and the soil buffer capacity, the neutralization reaction efficiency prediction model is established, which helps to eliminate the defects of the lack of dynamic feedback of the neutralization reaction.

[0167] In some embodiments, the soil sampling data is analyzed to determine the calcium carbonate content and the cation exchange capacity; the change relationship between the consumption of the treatment agent and the pH value is determined according to the chemical properties; and the neutralization reaction efficiency prediction model is established according to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity, and the soil buffer capacity.

[0168] The soil sampling data can be a set of physical and chemical properties of the layered soil sample.

[0169] The calcium carbonate content can be the mass percentage of calcium carbonate in the soil.

[0170] The cation exchange capacity can be the total amount of exchangeable cations adsorbed by unit mass of soil.

[0171] The consumption can be the concentration reduction amount of the treatment agent caused by chemical reactions, adsorption, etc. during the neutralization reaction process.

[0172] The pH value can be an acid-base index of the soil.

[0173] The change relationship can be the dynamic correlation between the consumption of the treatment agent and the pH value.

[0174] Specifically, based on the soil sampling data, the layered soil samples obtained by drilling exploration are taken, and the calcium carbonate content is determined by calcium carbonate content determination method: first, the soil sample is reacted with excess hydrochloric acid until no bubbles are generated; then, the mass percentage of calcium carbonate is calculated by mass difference method, that is, the calcium carbonate content.

[0175] Based on the soil sampling data, the cation exchange capacity is determined using a cation exchange capacity detection kit: first, the soil sample is saturated with barium chloride solution and then centrifuged to displace the exchangeable cations; then, the total amount of displaced cations is determined by sulfuric acid titration method, and the cation exchange capacity per unit mass of soil is calculated. Based on the chemical properties, the treatment agent is mixed with the target contaminated soil in the laboratory according to the molar ratio in the chemical properties; then, the pH value of the mixed solution is monitored in real time by a pH meter until the reaction reaches equilibrium; finally, the consumption of the treatment agent and the change of the pH value are recorded. The change relationship, calcium carbonate content and cation exchange capacity are integrated; then, a cooperative correction factor of soil buffer capacity and capillary pressure detection data is introduced to generate the final neutralization reaction efficiency prediction model.

[0176] Through the scheme, the soil sampling data is analyzed to determine the calcium carbonate content and the cation exchange capacity, which helps to quantify the chemical buffering capacity of the soil to the acidic treatment agent, avoids introducing redundant parameters, and reflects the adsorption capacity of the soil to the cations in the treatment agent. According to the chemical properties, the change relationship between the consumption of the treatment agent and the pH value is determined, the actual dynamic process of the consumption of the treatment agent is quantified, and the defects of static parameter fragmentation analysis are avoided. According to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity and the soil buffer capacity, the neutralization reaction efficiency prediction model is established, which helps to eliminate the three-dimensional modeling distortion problem of the interaction between capillary pressure and fracture structure.

[0177] In some embodiments, according to the soil capillary pressure detection data, the soil capillary capacity is determined; the migration rate and the soil capillary capacity are parameterized to obtain the basic parameters and input into the AI dynamic simulation model to make the AI dynamic simulation model simulate the multi-node data growth of a plurality of treatment schemes based on the basic parameters; the simulation results of the AI dynamic simulation model are obtained, the simulation results are analyzed, and the penetration data of each treatment scheme is determined; the penetration depth change and the change timestamp are determined by analyzing the penetration data; and the treatment effect evolution map containing the time-varying characteristics of the stratum response is generated according to the penetration depth change and the change timestamp.

[0178] The soil capillary capacity can be the soil capillary action strength.

[0179] The basic parameters can be a parameter set generated by dimensionless processing of the migration rate and the soil capillary capacity.

[0180] The simulation result can be a data set output by the AI dynamic simulation model.

[0181] The penetration data can be governance agent migration characteristic data extracted from the simulation result.

[0182] The penetration depth change can be a time-varying process of the penetration distance of the governance agent in the vertical direction.

[0183] The change timestamp can be a time node mark when the penetration depth reaches the cumulative time.

[0184] Specifically, according to the soil capillary pressure detection data, the soil capillary capacity is calculated. The migration rate and the soil capillary capacity are processed by dimensionless processing to generate the basic parameters; the basic parameters are input into the AI dynamic simulation model to start the multi-node data growth simulation: first, automatically generate several governance schemes; second, based on the difference between the initial injection concentration and the residual liquid concentration, simulate the time-varying process of the agent migration rate; finally, verify the penetration depth and the surface soil capillary capacity synchronously, and predict the survival rate risk of vegetation. Then, the simulation result is extracted to establish a penetration depth-time sequence data set, and the penetration data of each governance scheme is determined; further, the maximum penetration depth corresponding to each timestamp is extracted from the penetration data to generate a penetration depth change curve; subsequently, the penetration depth is associated with the corresponding timestamp to mark the time node and generate the change timestamp. Finally, the penetration depth change and the change timestamp are spatio-temporally associated to construct a governance effect evolution map.

[0185] According to the soil capillary pressure detection data, the soil capillary capacity is determined, and the problem of modeling distortion caused by the interaction between capillary pressure and fracture structure is eliminated. The migration rate and the soil capillary capacity are parameterized to obtain the basic parameters and input into the AI dynamic simulation model to enable the AI dynamic simulation model to perform multi-node data growth simulation of several governance schemes based on the basic parameters, eliminate model errors caused by dynamic parameter fragmentation analysis, provide high-precision input data for the AI dynamic simulation model, cover the needs of comprehensive and simple governance mines, eliminate concentration distribution errors caused by the lack of spatio-temporal evolution prediction, and break through the limitations of the multi-node data growth mechanism. The simulation result of the AI dynamic simulation model is obtained, the simulation result is analyzed, the penetration data of each governance scheme is determined, which helps to verify whether the coverage of the comprehensive governance mine meets the standard, and whether the simple governance mine meets the time requirement of the target vegetation root adaptation interval. Analyzing the penetration data, determining the penetration depth change and the change timestamp, and providing a basis for decision-making of the planting time of the simple governance mine. According to the penetration depth change and the change timestamp, a governance effect evolution map containing the time-varying characteristics of the stratum response is generated to clarify the dynamic evolution law of the stratum response under different governance schemes and provide quantitative basis for multi-objective optimization.

[0186] In some embodiments, a set of vegetation planting condition parameters corresponding to each treatment scheme is obtained; based on the treatment effect evolution map, the actual action depth of each treatment scheme at different time nodes and the corresponding soil parameters are extracted; the matching degree calculation is performed between the soil parameters within the actual action depth and the set of vegetation planting condition parameters, to determine the matching degree between the action depth of the treatment agent in each treatment scheme and the vegetation planting condition.

[0187] The set of vegetation planting condition parameters can be a set of quantitative parameters established according to vegetation restoration requirements.

[0188] The actual action depth can be the vertical penetration depth of the treatment agent at different time nodes.

[0189] The soil parameters can be core soil properties directly affecting vegetation planting within the actual action depth of the treatment agent.

[0190] Specifically, based on each treatment scheme, the surface soil pH target range, the maximum depth of target vegetation root distribution, and the minimum pore connectivity required for vegetation survival are set according to the mine type to construct the set of vegetation planting condition parameters. Based on the time axis of the treatment effect evolution map, the actual action depth of the treatment agent at different time nodes in each treatment scheme is extracted according to the time step; based on the three-dimensional coordinates corresponding to the actual action depth, the soil parameters are obtained. Subsequently, the coverage rate is calculated by comparing the difference between the actual action depth and the maximum depth of the target vegetation root distribution; then, the duration ratio of the actual pH value within the target interval is calculated, and it is verified whether the porosity is continuously higher than the critical value; the matching degree between the action depth of the treatment agent and the vegetation planting condition in each treatment scheme is output by comprehensive weight allocation.

[0191] By this scheme, the set of vegetation planting condition parameters corresponding to each treatment scheme is obtained, which helps to eliminate the problem of fragmentation between treatment targets and vegetation conditions, and to clearly define the target threshold of the action depth of the treatment agent, avoiding the risk of reduced vegetation survival rate caused by insufficient capillary ability. Based on the treatment effect evolution map, the actual action depth of each treatment scheme at different time nodes and the corresponding soil parameters are extracted, to realize the spatio-temporal evolution tracking of the penetration process of the agent, and to eliminate the problem of unable to reflect the dynamic geological response. The matching degree calculation is performed between the soil parameters within the actual action depth and the set of vegetation planting condition parameters, to determine the matching degree between the action depth of the treatment agent in each treatment scheme and the vegetation planting condition, which helps to eliminate the problem of disconnection between the penetration depth and vegetation requirements, to protect the core influence of pH on vegetation survival, and to ensure the synchronization of the action effect of the treatment agent and the vegetation planting opportunity, and to eliminate the problem of not integrating the time dimension.

[0192] In some embodiments, a governance cost evaluation model is established to calculate the implementation cost parameters of each governance scheme; a multi-objective optimization function is constructed to comprehensively match the degree and implementation cost parameters for weight calculation; and the governance scheme with the highest score and the lowest cost is selected as the optimal implementation path according to the calculation result.

[0193] The governance cost evaluation model can be a calculation model for quantifying the economic input of the mine environmental governance scheme.

[0194] The implementation cost parameter can be a quantifiable economic indicator generated during the execution of the governance scheme.

[0195] The multi-objective optimization function can be a mathematical decision tool for balancing the comprehensive matching degree of the governance scheme.

[0196] The optimal implementation path can be the execution strategy of the governance scheme with the highest score and the lowest cost.

[0197] Specifically, based on the pollution diffusion vector field and the migration rate of the governance agent, the demand for agent per unit area is calculated; then, the dynamic monitoring frequency of the soil parameters is determined according to the accuracy requirements of the space-time evolution prediction; subsequently, the injection frequency is calculated according to the compliance rate of the vegetation planting condition parameter set, and the governance cost evaluation model is constructed; further, based on the governance cost evaluation model, the actual action depth is extracted from the governance effect evolution map, and the injection concentration and frequency are dynamically adjusted to reduce the excessive injection cost; then, the secondary governance cost budget is added to the cross-level risk nodes, so as to determine the implementation cost parameters of each governance scheme. Subsequently, the weight coefficients of the matching degree and the implementation cost parameters are defined, and the matching degree and the implementation cost parameters are normalized to eliminate the dimensional differences; thus, the optimization score is generated to quantify the comprehensive performance of the governance scheme. Then, the governance schemes with a matching degree lower than the minimum or a cost exceeding the minimum are excluded; subsequently, the matching degree-cost Pareto optimal solution set is selected to eliminate the dominated schemes; finally, the governance scheme with the highest score and the lowest cost is selected as the optimal implementation path.

[0198] Through the scheme, the governance cost evaluation model is established to calculate the implementation cost parameters of each governance scheme, which helps to eliminate the problem of excessive injection of agents caused by the lack of dynamic feedback of neutralization reaction, reduce the resource waste rate, reduce the invalid monitoring cost, avoid the repeated monitoring redundancy caused by the blank of multi-node data growth mechanism, and avoid the cross-level correlation risk. The multi-objective optimization function is constructed to comprehensively match the degree and implementation cost parameters for weight calculation, which avoids excessive injection of agents caused by the interaction of capillary pressure and fracture structure, and eliminates the problem of insufficient adaptability of the governance scheme. According to the calculation result, the governance scheme with the highest score and the lowest cost is selected as the optimal implementation path, which helps to eliminate the problem of disconnection between the governance scheme and the vegetation restoration target.

[0199] Figure 3A structural schematic diagram of a mine environment governance system based on an AI virtual simulation technology provided by an embodiment of the present application is shown in FIG. 3. The mine environment governance system 300 based on the AI virtual simulation technology of the present embodiment includes a parameter set analysis module 301, a mode determination module 302, a model establishment module 303, a graph generation module 304, and a path determination module 305. Figure 3

[0200] The parameter set analysis module 301 is configured to obtain a dynamic geological parameter set of a target mine, analyze the dynamic geological parameter set, and determine injection data and soil capillary pressure detection data.

[0201] The mode determination module 302 is configured to determine a compound diffusion mode of a governance agent based on the injection data and the soil capillary pressure detection data.

[0202] The model establishment module 303 is configured to analyze the compound diffusion mode, determine a spatiotemporal evolution rule, and construct an AI dynamic simulation model containing a governance agent penetration feedback mechanism based on the spatiotemporal evolution rule.

[0203] The graph generation module 304 is configured to perform multi-node data growth simulation on a plurality of governance schemes based on the AI dynamic simulation model, and generate a governance effect evolution graph containing stratum response time-varying characteristics.

[0204] The path determination module 305 is configured to analyze the governance effect evolution graph, determine a matching degree of an action depth of the governance agent and a vegetation planting condition in each governance scheme, and determine an optimal governance scheme implementation path based on the matching degree.

[0205] Optionally, the dynamic geological parameter set further includes a geological structure of the target mine. When the mode determination module 302 determines the compound diffusion mode of the governance agent based on the injection data and the soil capillary pressure detection data, the mode determination module 302 is configured to:

[0206] analyze the injection data to determine an injection hole distribution;

[0207] obtain an injection hole image based on the injection hole distribution;

[0208] analyze the injection hole image to determine a surface residual liquid feature;

[0209] determine a residual liquid concentration based on the surface residual liquid feature;

[0210] determine a migration rate of the governance agent based on the residual liquid concentration and the soil capillary pressure detection data;

[0211] analyze the geological structure to determine a stratum fracture strike;

[0212] ​construct a three-dimensional pollution diffusion vector field according to the migration rate and the stratum fracture direction;

[0213] determine a composite diffusion mode of the treatment agent according to the three-dimensional pollution diffusion vector field.

[0214] Optionally, the liquid injection data includes an initial liquid injection concentration; when the mode determination module 302 determines the migration rate of the treatment agent according to the residual liquid concentration and the soil capillary pressure detection data, it is used for:

[0215] based on the geological structure, analyzing the soil capillary pressure detection data to determine the capillary pressure gradient of different levels;

[0216] based on the initial liquid injection concentration and the residual liquid concentration, calculating the migration rate of the treatment agent according to Darcy's law and the capillary pressure gradient of different levels.

[0217] Optionally, when the model establishment module 303 establishes an AI dynamic simulation model containing a treatment agent penetration feedback mechanism according to the spatio-temporal evolution law, it is used for:

[0218] determine the spatial concentration distribution characteristics of the treatment agent according to the spatio-temporal evolution law;

[0219] input the spatial concentration distribution characteristics into a preset neutralization reaction efficiency prediction model, and output the neutralization efficiency time series data of the treatment agent;

[0220] according to the neutralization efficiency time series data and a preset mine treatment standard, dynamically generate a self-optimization instruction set of reagent injection concentration and liquid injection hole position adjustment;

[0221] fuse the self-optimization instruction set and the stratum fracture direction to construct an AI dynamic simulation model containing a treatment agent penetration feedback mechanism.

[0222] Optionally, the mine environment treatment system based on AI virtual simulation technology further includes a prediction model establishment module 306, which is used for:

[0223] determine the chemical properties of the treatment agent according to the liquid injection data;

[0224] determine the soil buffer capacity according to the geological structure and the soil capillary pressure detection data;

[0225] establish a neutralization reaction efficiency prediction model according to the chemical properties and the soil buffer capacity.

[0226] Optionally, the dynamic geological parameter set includes soil sampling data; when the prediction model establishment module 306 establishes a neutralization reaction efficiency prediction model according to the chemical properties and the soil buffer capacity, it is used for:

[0227] analyzing the soil sampling data to determine calcium carbonate content and cation exchange capacity;

[0228] determining a change relationship between consumption of the treatment agent and PH value according to the chemical properties;

[0229] establishing a neutralization reaction efficiency prediction model according to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity, and the soil buffer capacity.

[0230] Optionally, the atlas generation module 304 generates a treatment effect evolution atlas containing time-varying characteristics of stratum responses based on the AI dynamic simulation model and multi-node data growth simulation of a plurality of treatment schemes, for:

[0231] determining soil capillary capacity according to the soil capillary pressure detection data;

[0232] quantifying the migration rate and the soil capillary capacity as parameters to obtain basic parameters and input the basic parameters to the AI dynamic simulation model to enable the AI dynamic simulation model to perform multi-node data growth simulation of a plurality of treatment schemes based on the basic parameters;

[0233] obtaining simulation results of the AI dynamic simulation model, analyzing the simulation results, and determining penetration data of each treatment scheme;

[0234] analyzing the penetration data to determine penetration depth changes and change time stamps;

[0235] generating a treatment effect evolution atlas containing time-varying characteristics of stratum responses according to the penetration depth changes and the change time stamps.

[0236] Optionally, when the path determination module 305 analyzes the treatment effect evolution atlas to determine a matching degree of an action depth of a treatment agent in each treatment scheme and vegetation planting conditions, it is used for:

[0237] obtaining a set of vegetation planting condition parameters corresponding to each treatment scheme;

[0238] extracting actual action depths of each treatment scheme at different time nodes and corresponding soil parameters based on the treatment effect evolution atlas;

[0239] performing matching degree calculation on soil parameters within the actual action depths and the set of vegetation planting condition parameters to determine a matching degree of an action depth of a treatment agent in each treatment scheme and vegetation planting conditions.

[0240] Optionally, the path determination module 305 determines an optimal treatment scheme implementation path according to the matching degree, for:

[0241] determining to establish a governance cost evaluation model to calculate an implementation cost parameter of each governance scheme;

[0242] constructing a multi-objective optimization function to perform weight calculation on the matching degree and the implementation cost parameter;

[0243] selecting a governance scheme with the highest score and the lowest cost as an optimal implementation path according to the calculation result.

[0244] The system of the embodiment can be used to execute the method of any of the above embodiments, and has similar implementation principles and technical effects, which will not be described here again.

Claims

1. A mine environment management method based on an AI virtual simulation technology, characterized in that, The method comprises the following steps: acquiring a set of dynamic geological parameters of a target mine; analyzing the set of dynamic geological parameters to determine injection data and soil capillary pressure detection data; determining a composite diffusion mode of a treatment agent according to the injection data and the soil capillary pressure detection data; analyzing the composite diffusion mode to determine a spatiotemporal evolution rule; and constructing an AI dynamic simulation model comprising a treatment agent penetration feedback mechanism according to the spatiotemporal evolution rule, including: determining spatial concentration distribution characteristics of the treatment agent according to the spatiotemporal evolution rule; inputting the spatial concentration distribution characteristics into a preset neutralization reaction efficiency prediction model to output time series data of neutralization efficiency of the treatment agent; dynamically generating a self-optimization instruction set of reagent injection concentration and injection hole site adjustment according to the time series data of neutralization efficiency and a preset mine treatment standard; fusing the self-optimization instruction set and the stratum fracture trend to construct the AI dynamic simulation model comprising the treatment agent penetration feedback mechanism; dividing time nodes according to treatment stages and space nodes according to depth layers according to the AI dynamic simulation model; based on the AI dynamic simulation model, performing multi-node data growth simulation on a plurality of treatment schemes to generate a treatment effect evolution map comprising stratum response time-varying characteristics, including: determining soil capillary ability according to the soil capillary pressure detection data; quantifying the migration rate and the soil capillary ability to obtain basic parameters and input the basic parameters into the AI dynamic simulation model to enable the AI dynamic simulation model to perform multi-node data growth simulation on the plurality of treatment schemes based on the basic parameters; acquiring simulation results of the AI dynamic simulation model, analyzing the simulation results, and determining penetration data of each treatment scheme; analyzing the penetration data to determine penetration depth variation and variation time stamps; generating the treatment effect evolution map comprising stratum response time-varying characteristics according to the penetration depth variation and the variation time stamps; analyzing the treatment effect evolution map to determine a matching degree of an action depth of the treatment agent and a vegetation planting condition in each treatment scheme; and determining an optimal treatment scheme implementation path according to the matching degree.

2. The method of claim 1, wherein, The set of dynamic geological parameters further comprises a geological structure of the target mine; and the determination of the composite diffusion mode of the treatment agent according to the injection data and the soil capillary pressure detection data comprises: analyzing the injection data to determine injection hole distribution; acquiring an injection hole image based on the injection hole distribution; analyzing the injection hole image to determine surface residual liquid characteristics; determining residual liquid concentration according to the surface residual liquid characteristics; determining a migration rate of the treatment agent according to the residual liquid concentration and the soil capillary pressure detection data; analyzing the geological structure to determine a stratum fracture trend; constructing a three-dimensional pollution diffusion vector field according to the migration rate and the stratum fracture trend; determining the composite diffusion mode of the treatment agent according to the three-dimensional pollution diffusion vector field.

3. The method of claim 2, wherein, The injection data comprises an initial injection concentration; and the determination of the migration rate of the treatment agent according to the residual liquid concentration and the soil capillary pressure detection data comprises: Based on the geological structure, the soil capillary pressure detection data is analyzed to determine the capillary pressure gradient of different levels; Based on the initial injection concentration and the residual liquid concentration, the migration rate of the treatment agent is calculated according to Darcy's law and the capillary pressure gradient of different levels.

4. The method of claim 3, wherein, The construction of the preset neutralization reaction efficiency prediction model includes: According to the injection data, the chemical properties of the treatment agent are determined; According to the geological structure and the soil capillary pressure detection data, the soil buffer capacity is determined; According to the chemical properties and the soil buffer capacity, a neutralization reaction efficiency prediction model is established.

5. The method of claim 4, wherein, The dynamic geological parameter set includes soil sampling data; According to the chemical properties and the soil buffer capacity, a neutralization reaction efficiency prediction model is established, which includes: The soil sampling data is analyzed to determine the calcium carbonate content and the cation exchange capacity; According to the chemical properties, the change relationship between the consumption of the treatment agent and the PH value is determined; According to the change relationship, the soil capillary pressure detection data, the calcium carbonate content, the cation exchange capacity and the soil buffer capacity, a neutralization reaction efficiency prediction model is established.

6. The method of claim 1, wherein, The analysis of the treatment effect evolution map to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme includes: Obtain the vegetation planting condition parameter set corresponding to each treatment scheme; Based on the treatment effect evolution map, the actual action depth of each treatment scheme at different time nodes and the corresponding soil parameters are extracted; The matching degree calculation is performed between the soil parameters within the actual action depth and the vegetation planting condition parameter set to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme.

7. The method of claim 6, wherein, According to the matching degree, the optimal treatment scheme implementation path is determined, which includes: A treatment cost evaluation model is established to calculate the implementation cost parameters of each treatment scheme; A multi-objective optimization function is constructed to perform weight calculation on the matching degree and the implementation cost parameters; According to the calculation result, the treatment scheme with the highest score and the lowest cost is selected as the optimal implementation path.

8. A mine environment governance system based on an AI virtual simulation technology, characterized in that, Applied to the method of any one of claims 1-7, comprising: A parameter set analysis module is configured to obtain the dynamic geological parameter set of the target mine; analyze the dynamic geological parameter set to determine injection data and soil capillary pressure detection data; A mode determination module is configured to determine the composite diffusion mode of the treatment agent according to the injection data and the soil capillary pressure detection data; A model establishment module is configured to analyze the composite diffusion mode to determine the spatio-temporal evolution law; and construct an AI dynamic simulation model containing the treatment agent penetration feedback mechanism according to the spatio-temporal evolution law; A map generation module is configured to generate a treatment effect evolution map containing stratum response time-varying characteristics by performing multi-node data growth simulation on a plurality of treatment schemes based on the AI dynamic simulation model; A path determination module is configured to analyze the treatment effect evolution map to determine the matching degree of the action depth of the treatment agent and the vegetation planting conditions in each treatment scheme; and determine the optimal treatment scheme implementation path according to the matching degree.

Images