Mine ecological problem identification method based on artificial intelligence

Abstract

The present invention discloses a method for identifying mine ecological problems based on artificial intelligence, which belongs to the technical field of ecological problem identification. It obtains multi-source data by integrating remote sensing technology and Internet of Things sensors, and realizes zoning identification of mines with wide distribution and complex terrain through fusion analysis of remote sensing images, sensor data, and meteorological data. Then, mining area ecological information is obtained for the mining area, and characteristic ecological data related to the corresponding characteristic ecological problem model is extracted based on the mining area ecological information. The characteristic ecological data is substituted into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for evaluating characteristic risks, thereby identifying the type of ecological problem in the target mine. In addition, ecological restoration information is obtained for land degradation areas, and the ecological restoration hidden dangers in land degradation areas are determined based on this. This not only provides a scientific basis for mine ecological protection and restoration, but also effectively solves the problems of low efficiency and incomplete coverage of traditional manual monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of ecological problem identification, and more specifically, to a method for identifying mine ecological problems based on artificial intelligence. Background Art

[0002] Since mining causes problems such as land subsidence, vegetation destruction, soil erosion, and biodiversity decline, it is necessary to survey the ecological conditions of the mines and the recovery of the mines after mining.

[0003] Traditional manual surveys are inefficient and have limited coverage. Inspections require a lot of time and manpower costs. In addition, due to the wide distribution of mines and complex terrain, some places are difficult for equipment and manpower to reach. The time and labor costs of surveys exceed the costs of general projects. In addition, the data obtained on mine ecological problems is not complete and accurate enough, and the accuracy of identifying the types of mine ecological problems is not high.

[0004] Therefore, in view of the above technical problems, the following technical solution is proposed. Summary of the Invention

[0005] The purpose of the present invention is to solve the existing problems and provide an artificial intelligence-based mine ecological problem identification method compared with the existing technology.

[0006] The purpose of the present invention can be achieved through the following technical solution: A method for identifying mine ecological problems based on artificial intelligence, comprising the following steps:

[0007] Step 1: Divide the target mine into regions and collect mine remote sensing images in different areas. Perform color feature analysis on the mine remote sensing images to identify exposed areas of bare soil.

[0008] Step 2: Capture images of the exposed bare soil area to obtain bare soil images, perform bare soil feature analysis on the bare soil images, and divide the exposed bare soil area into exposed rock areas, land degradation areas, and mining areas based on the analysis results;

[0009] Step 3: Extract features from existing mine ecological problem data to obtain the features of actual mine ecological problem data. Use the features of the actual mine ecological problem data to establish a corresponding characteristic ecological problem model. Collect mining area ecological information from the mining area. Extract characteristic ecological data related to the corresponding characteristic ecological problem model based on the mining area ecological information. Substitute the characteristic ecological data into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for characteristic risk assessment. Determine the target mine ecological problem type based on the assessment results.

[0010] Step 4: Set a monitoring period, collect ecological restoration information of land degradation areas during the monitoring period, and determine the ecological restoration risks of land degradation areas based on the ecological restoration information.

[0011] As a preferred embodiment of the present invention, the process of performing bare soil feature analysis on a bare soil image includes:

[0012] The bare soil image is enlarged into a pixel grid image and grayscale transformation is performed to obtain the grayscale value of each pixel grid. The grayscale values ​​of adjacent pixel grids are subtracted and the absolute value is taken to obtain the grayscale floating value. Multiple grayscale floating values ​​are summed to obtain the grayscale jump value; the grayscale jump value is compared with the preset grayscale jump threshold, and the bare soil exposure area corresponding to the bare soil image is determined as a bare rock area or a bare soil area.

[0013] As a preferred embodiment of the present invention, continuous image acquisition is performed on the exposed soil area at preset monitoring intervals to obtain multiple groups of exposed soil images, and peripheral range comparison and analysis are performed on the multiple groups of exposed soil images before and after;

[0014] When the peripheries of the current and subsequent groups of bare soil images overlap or the periphery range of the bare soil images gradually shrinks with the update of the acquisition time, a static image reduction signal is generated. When the current and subsequent groups of bare soil images do not overlap and the periphery range of the bare soil images gradually expands with the update of the acquisition time, a dynamic image expansion signal is generated. The bare soil area corresponding to the static image reduction signal is marked as a land degradation area, and the bare soil area corresponding to the dynamic image expansion signal is marked as a mining area.

[0015] As a preferred embodiment of the present invention, the characteristic ecological problem model includes a soil erosion model, a goaf collapse model, a dust pollution diffusion model, and a heavy metal diffusion model.

[0016] As a preferred embodiment of the present invention, the ecological information of the mining area includes topographic data, meteorological data, and pollution data, wherein the topographic data includes vegetation coverage, fault density, slope angle, and thickness of the underground goaf roof; the meteorological data includes rainfall, wind speed, and humidity; and the pollution data includes the amount of open-air slag storage, dust volume, soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient.

[0017] As a preferred embodiment of the present invention, the process of determining the target mine ecological problem type includes:

[0018] Substitute vegetation coverage, slope angle, and rainfall as characteristic ecological data into the soil erosion model. The characteristic ecological risk data obtained is soil loss. The soil loss is compared with the preset soil loss threshold to assess the characteristic risk and determine whether the target mine ecological problem is identified as a soil erosion problem.

[0019] Substituting fault density, underground goaf roof thickness, and rainfall as characteristic ecological data into the goaf collapse model to obtain the collapse probability, which is then compared with a preset collapse probability threshold to determine whether the target mine ecological problem is identified as a goaf collapse problem.

[0020] Substituting the amount of open-air slag storage, dust volume, wind speed, and humidity as characteristic ecological data into the dust pollution diffusion model, the dust pollution coefficient is obtained for comparison to determine whether the target mine ecological problem is identified as a dust pollution problem;

[0021] The soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient are substituted into the heavy metal diffusion model as characteristic ecological data to obtain the heavy metal pollution coefficient for comparison to determine whether the target mine ecological problem should be identified as a heavy metal pollution problem.

[0022] As a preferred embodiment of the present invention, the process of determining ecological restoration hazards in land degradation areas is as follows: collecting ecological restoration information of the land degradation areas during a monitoring period, the ecological restoration information including vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate; the vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate are respectively formed into a set A, a set B, and a set C; and plotting a vegetation cover growth rate change curve, a landslide decrease rate change curve, and a soil heavy metal content decrease rate change curve, respectively;

[0023] Obtain the number of times in the vegetation cover growth rate change curve that is lower than the preset vegetation cover growth rate threshold. If the number exceeds the preset upper limit, it is determined that there is an abnormal vegetation recovery problem in the land degradation area;

[0024] Obtain the number of landslides in the landslide reduction rate change curve that is lower than the preset landslide reduction rate threshold. If the number exceeds the preset upper limit, it is determined that there is a landslide risk problem in the land degradation area.

[0025] Obtain the number of soil heavy metal content reduction rate change curves that are lower than the preset soil heavy metal content reduction rate threshold. If the number exceeds the preset upper limit, it is determined that there is a problem of soil pollution in the land degradation area.

[0026] Compared with the prior art, the advantages of the present invention are:

[0027] 1. This solution is to obtain multi-source data by integrating remote sensing technology and Internet of Things sensors, and realize the zoning identification of mines with wide distribution and complex terrain through the fusion analysis of remote sensing images, sensor data, and meteorological data. Then, the mining area ecological information is obtained for the mining area. Based on the mining area ecological information, characteristic ecological data related to the corresponding characteristic ecological problem model is extracted. The characteristic ecological data is substituted into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for assessing characteristic risks, thereby identifying the type of ecological problem in the target mine. Based on the collaborative judgment of multiple models, the accuracy of identifying mine ecological problems is improved, solving the problems of low efficiency and incomplete coverage of traditional manual monitoring.

[0028] 2. This plan also collects ecological restoration information for land degradation areas, and identifies potential ecological restoration risks in land degradation areas based on the ecological restoration information, providing a scientific basis for mine ecological protection and restoration, effectively promoting the intelligentization of the entire "monitoring-early warning-restoration" chain, and contributing to the development of green mining. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 is a flow chart of the method of the present invention;

[0030] Figure 2 A flow chart of the method for determining the type of ecological problem in a target mine according to the present invention;

[0031] Figure 3 This is a flow chart of the method for identifying hidden dangers of ecological restoration in land degradation areas according to the present invention. DETAILED DESCRIPTION

[0032] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making creative work shall fall within the scope of protection of the present invention.

[0033] Example 1: The present invention discloses a method for identifying mine ecological problems based on artificial intelligence. Figure 1 , including the following steps:

[0034] Step 1: Divide the target mine into regions and collect mine remote sensing images using remote sensing technology. Perform color feature analysis on the mine remote sensing images to identify exposed areas of bare soil.

[0035] Step 2: Capture images of the exposed bare soil area to obtain bare soil images, perform bare soil feature analysis on the bare soil images, and divide the exposed bare soil area into exposed rock areas, land degradation areas, and mining areas based on the analysis results;

[0036] Step 3: Extract features from existing mine ecological problem data to obtain the features of actual mine ecological problem data. Use the features of the actual mine ecological problem data to establish a corresponding characteristic ecological problem model. Collect mining area ecological information from the mining area. Extract characteristic ecological data related to the corresponding characteristic ecological problem model based on the mining area ecological information. Substitute the characteristic ecological data into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for characteristic risk assessment. Determine the target mine ecological problem type based on the assessment results.

[0037] Step 4: Set a monitoring period, collect ecological restoration information of land degradation areas during the monitoring period, and determine the ecological restoration risks of land degradation areas based on the ecological restoration information.

[0038] In step 2, the process of performing bare soil feature analysis on the bare soil image includes:

[0039] The bare soil image is enlarged into a pixel grid image and grayscale transformation is performed to obtain the grayscale value of each pixel grid. The grayscale values ​​of adjacent pixel grids are subtracted and the absolute value is taken to obtain the grayscale floating value. Multiple grayscale floating values ​​are summed to obtain the grayscale jump value.

[0040] The grayscale jump value is compared with the preset grayscale jump threshold. When the grayscale jump value is greater than or equal to the preset grayscale jump threshold, the bare soil exposure area corresponding to the bare soil image is judged as a bare rock area. The rock surface is rough and angular, the local grayscale changes dramatically, and the texture contrast is high. Otherwise, the bare soil exposure area corresponding to the bare soil image is judged as a bare soil area. The surface of degraded soil is relatively smooth, especially after mechanical turning, the grayscale transition is uniform, and the texture periodicity is obvious. Through image grayscale texture analysis, the exposed rock area is selected from the bare soil exposure area and excluded. The exposed rock area is not considered as a man-made mining ecological problem.

[0041] Continuously collect images of the exposed soil area at preset monitoring intervals to obtain multiple sets of exposed soil images, and perform perimeter comparison and analysis on the previous and subsequent sets of exposed soil images;

[0042] When the perimeter comparisons of the current and subsequent groups of bare soil images overlap or the perimeter range of the bare soil images gradually shrinks with the update of the acquisition time, a static image reduction signal is generated. When the current and subsequent groups of bare soil images do not overlap and the perimeter range of the bare soil images gradually expands with the update of the acquisition time, a dynamic image expansion signal is generated. The exposed soil area corresponding to the static image reduction signal is marked as a land degradation area, and the exposed soil area corresponding to the dynamic image expansion signal is marked as a mining area, indicating that with the passage of monitoring and acquisition time, the scope of the bare soil area continues to expand due to artificial mining activities.

[0043] In step three, the characteristic ecological problem models include soil erosion model, goaf collapse model, dust pollution diffusion model, and heavy metal diffusion model;

[0044] Collect mining area ecological information through IoT sensors, including topographic data, meteorological data, and pollution data;

[0045] Among them, topographic data includes vegetation coverage, fault density, slope angle, and underground goaf roof thickness; meteorological data includes rainfall, wind speed, and humidity; pollution data includes open-air slag storage volume, dust volume, soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient;

[0046] The process for determining the types of ecological issues to target at a mine includes:

[0047] See also Figure 2 , the vegetation coverage rate, slope angle, and rainfall are substituted into the soil erosion model as characteristic ecological data, and the characteristic ecological risk data obtained is the soil loss amount. The soil loss amount is compared with the preset soil loss amount threshold. When the soil loss amount is greater than the maximum value of the preset soil loss amount threshold range, a high-risk loss signal is generated, and the ecological problem of the target mine is determined to be a soil erosion problem. When the soil loss amount is within the preset soil loss amount threshold range, a low-risk loss signal is generated, and operations are restricted and monitoring is intensified. Otherwise, a normal signal is generated and inspections are strengthened;

[0048] The fault density, underground goaf roof thickness and rainfall are used as characteristic ecological data to substitute into the goaf collapse model. The characteristic ecological risk data obtained is the collapse probability. The collapse probability is compared with the preset collapse probability threshold. When the collapse probability is greater than the maximum value of the preset collapse probability threshold range, a high-risk collapse signal is generated. The target mine ecological problem is determined to be a goaf collapse problem. When the collapse probability is within the preset collapse probability threshold range, a high-risk collapse signal is generated, operations are restricted, and monitoring is intensified. Otherwise, a normal signal is generated and inspections are strengthened.

[0049] The amount of open-air slag storage, dust volume, wind speed, and humidity are used as characteristic ecological data to be substituted into the dust pollution diffusion model. The characteristic ecological risk data obtained is the dust pollution coefficient. The dust pollution coefficient is compared with the preset dust pollution coefficient threshold. The risk level is determined in the same way. When the dust pollution coefficient is greater than the maximum value of the preset dust pollution coefficient threshold range, the target mine ecological problem is determined to be a dust pollution problem.

[0050] The soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient are substituted into the heavy metal diffusion model as characteristic ecological data. The characteristic ecological risk data obtained is the heavy metal pollution coefficient. The heavy metal pollution coefficient is compared with the preset heavy metal pollution coefficient threshold. The risk level is judged in the same way. When the heavy metal pollution coefficient is greater than the maximum value of the preset heavy metal pollution coefficient threshold range, the target mine ecological problem is determined to be a heavy metal pollution problem.

[0051] Acquire mining area ecological information for the mining area, extract characteristic ecological data related to the corresponding characteristic ecological problem model based on the mining area ecological information, substitute the characteristic ecological data into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data, which is used to assess characteristic risks, thereby identifying the type of ecological problems in the target mine, and improving the accuracy of identifying mine ecological problems based on collaborative judgment of multiple models.

[0052] Example 2: In step 4, the process of determining the ecological restoration hidden dangers in land degradation areas is as follows: Figure 3 During the monitoring period, ecological restoration information of land degradation areas was collected. The ecological restoration information included vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate. The vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate were respectively formed into Set A, Set B, and Set C. The vegetation cover growth rate change curve of Set A, the landslide decrease rate change curve of Set B, and the soil heavy metal content decrease rate change curve of Set C were drawn respectively.

[0053] If the vegetation cover growth rate is higher than the preset vegetation cover growth rate threshold or the number of times it is lower than the preset vegetation cover growth rate threshold is very small, it indicates that a virtuous cycle has been entered. The number of times in the vegetation cover growth rate change curve that is lower than the preset vegetation cover growth rate threshold is obtained. If the number exceeds the preset upper limit, a slow vegetation cover growth signal is generated, and it is determined that there is an abnormal vegetation recovery problem in the land degradation area.

[0054] If the landslide decrease rate is higher than the preset landslide decrease rate threshold or the number of landslides below the preset landslide decrease rate threshold is very small, it indicates that the number of landslides is continuously decreasing and the soil erosion resistance is enhanced. The number of landslides below the preset landslide decrease rate threshold in the landslide decrease rate change curve is obtained. If the number exceeds the preset upper limit, a potential landslide risk signal is generated, and it is determined that there is a landslide hazard problem in the land degradation area.

[0055] If the soil heavy metal content decrease rate exceeds the preset soil heavy metal content decrease rate threshold or the number of times it is lower than the preset soil heavy metal content decrease rate threshold is very small, it indicates that the soil heavy metal content continues to decline and soil pollution is developing towards a controllable trend. The number of times in the soil heavy metal content decrease rate change curve that is lower than the preset soil heavy metal content decrease rate threshold is obtained. If the number exceeds the preset upper limit, a soil pollution persistence signal is generated, and it is determined that there is a soil pollution persistence problem in the land degradation area.

[0056] In summary, by integrating remote sensing technology and IoT sensors to obtain multi-source data, the fusion analysis of remote sensing images, sensor data, and meteorological data is used to achieve zoning identification of mines with wide distribution and complex terrain. Mining area ecological information is then obtained for each mining area. Based on this mining area ecological information, characteristic ecological data related to the corresponding characteristic ecological problem model is extracted. The characteristic ecological data is substituted into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for characteristic risk assessment. This is used to identify the type of ecological problem in the target mine, and through multi-model collaborative judgment, the accuracy of mine ecological problem identification is improved.

[0057] We also collect ecological restoration information from land degradation areas, and use this information to identify potential ecological restoration risks in land degradation areas, providing a scientific basis for mine ecological protection and restoration.

[0058] This solution involves multiple parameter thresholds. It should be noted that the thresholds or preset values, preset ranges, etc. are set for result comparison and analysis in order to determine whether they are good or bad. The values ​​of these thresholds are set for entry and storage based on a combination of large-scale model analysis of sample data and manual experience. Appropriate adjustments can also be made based on seasonal or common-sense influencing conditions.

[0059] The above description is only a preferred specific embodiment of the present invention; however, the protection scope of the present invention is not limited thereto; any technician familiar with the technical field within the technical scope disclosed by the present invention; any equivalent replacement or change based on the technical solution and improved conception of the present invention shall be covered within the protection scope of the present invention.

Claims

1. The method for identifying mine ecological problems based on artificial intelligence is characterized by: The steps include: Step 1: Divide the target mine into regions and collect mine remote sensing images in different areas. Perform color feature analysis on the mine remote sensing images to identify exposed areas of bare soil. Step 2: Capture images of the exposed bare soil area to obtain bare soil images, perform bare soil feature analysis on the bare soil images, and divide the exposed bare soil area into exposed rock areas, land degradation areas, and mining areas based on the analysis results; Step 3: Extract features from existing mine ecological problem data to obtain the features of actual mine ecological problem data. Use the features of the actual mine ecological problem data to establish a corresponding characteristic ecological problem model. Collect mining area ecological information from the mining area. Extract characteristic ecological data related to the corresponding characteristic ecological problem model based on the mining area ecological information. Substitute the characteristic ecological data into the corresponding characteristic ecological problem model to obtain characteristic ecological risk data for characteristic risk assessment. Determine the target mine ecological problem type based on the assessment results. Step 4: Set a monitoring period, collect ecological restoration information of land degradation areas during the monitoring period, and identify potential ecological restoration risks in land degradation areas based on the ecological restoration information; The process of performing bare soil feature analysis on bare soil images includes: The bare soil image is enlarged into a pixel grid image and grayscale transformed to obtain the grayscale value of each pixel grid. The grayscale values ​​of adjacent pixel grids are subtracted and the absolute value is taken to obtain a grayscale floating value. Multiple grayscale floating values ​​are summed to obtain a grayscale jump value. The grayscale jump value is compared with a preset grayscale jump threshold. The bare soil exposed area corresponding to the bare soil image is determined to be a bare rock area or a bare soil area. Image acquisition and analysis of the bare soil area is performed. The process of collecting and analyzing images of the exposed soil area includes: continuously collecting images of the exposed soil area at preset monitoring intervals to obtain multiple sets of exposed soil images, and performing peripheral range comparison and analysis on the multiple sets of exposed soil images before and after; When the perimeters of the current and subsequent groups of exposed soil images overlap or the perimeter range of the exposed soil images gradually shrinks with the update of the acquisition time, a static image reduction signal is generated; when the current and subsequent groups of exposed soil images do not overlap and the perimeter range of the exposed soil images gradually expands with the update of the acquisition time, a dynamic image expansion signal is generated, and the exposed soil area corresponding to the static image reduction signal is marked as a land degradation area, and the exposed soil area corresponding to the dynamic image expansion signal is marked as a mining area; Characteristic ecological problem models include soil erosion models, goaf collapse models, dust pollution diffusion models, and heavy metal diffusion models. Mining area ecological information includes topographic data, meteorological data, and pollution data. Topographic data includes vegetation coverage, fault density, slope angle, and underground goaf roof thickness. Meteorological data includes rainfall, wind speed, and humidity. Pollution data includes open-air slag storage volume, dust volume, soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient. The process for determining the types of ecological issues to target at a mine includes: Substitute vegetation coverage, slope angle, and rainfall as characteristic ecological data into the soil erosion model. The characteristic ecological risk data obtained is soil loss. The soil loss is compared with the preset soil loss threshold to assess the characteristic risk and determine whether the target mine ecological problem is identified as a soil erosion problem. Substituting fault density, underground goaf roof thickness, and rainfall as characteristic ecological data into the goaf collapse model to obtain the collapse probability, which is then compared with a preset collapse probability threshold to determine whether the target mine ecological problem is identified as a goaf collapse problem. Substituting the amount of open-air slag storage, dust volume, wind speed, and humidity as characteristic ecological data into the dust pollution diffusion model, the dust pollution coefficient is obtained for comparison to determine whether the target mine ecological problem is identified as a dust pollution problem; The soil heavy metal concentration, groundwater flow rate, and soil permeability coefficient are substituted into the heavy metal diffusion model as characteristic ecological data to obtain the heavy metal pollution coefficient for comparison to determine whether the target mine ecological problem should be identified as a heavy metal pollution problem.

2. The method for identifying mine ecological problems based on artificial intelligence according to claim 1, characterized in that: The process for identifying potential ecological restoration risks in land degradation areas is as follows: During the monitoring period, ecological restoration information of land degradation areas was collected. This information included vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate. The vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate were grouped into Set A, Set B, and Set C, respectively. Curves for vegetation cover growth rate, landslide decrease rate, and soil heavy metal content decrease rate were drawn, respectively. Obtain the number of times in the vegetation cover growth rate change curve that is lower than the preset vegetation cover growth rate threshold. If the number exceeds the preset upper limit, it is determined that there is an abnormal vegetation recovery problem in the land degradation area; Obtain the number of landslides in the landslide reduction rate change curve that is lower than the preset landslide reduction rate threshold. If the number exceeds the preset upper limit, it is determined that there is a landslide risk problem in the land degradation area. Obtain the number of times in the soil heavy metal content reduction rate change curve that is lower than the preset soil heavy metal content reduction rate threshold. If it exceeds the preset upper limit, it is determined that there is a problem of soil pollution in the land degradation area.

Images