A loess gully landform mining subsidence area landslide body deformation monitoring and early warning method based on an airborne laser scanning point cloud

Abstract

The application discloses a kind of methods for monitoring and early warning of landslide caused by mining subsidence under loess gully terrain condition by airborne laser radar scanning, belongs to the technical field of unmanned aerial vehicle measurement, mainly solves the problem that landslide caused by underground mining occurs rapidly under loess gully terrain condition, and it is difficult to monitor quickly and early warning.The steps include: mining subsidence area determination;Unmanned aerial vehicle three-dimensional laser scanning, obtain initial ground point cloud data in mining subsidence range;According to the special form of landslide hidden danger point mark, identify landslide hidden danger area;Landslide hidden danger body range marking;Periodically obtain the ground point cloud data of working face;Establish digital surface model;Compare landslide body deformation monitoring by high-precision digital surface model before and after landslide occurs;Landslide body displacement rate calculation;Establish landslide early warning model.

Description

Technical Field

[0001] This invention discloses a method for monitoring and early warning of landslide deformation in mining subsidence areas of loess gullies based on airborne laser scanning point clouds, belonging to the field of UAV measurement technology. Background Technology

[0002] Domestic and international research on landslide deformation monitoring and early warning shows that current landslide deformation monitoring mainly uses synthetic aperture radar interferometry, and landslide early warning primarily relies on local precipitation models. This research is relatively mature. However, underground mining activities in loess gully terrain are prone to triggering rapid landslides. The initial deformation and isokinetic deformation stages of landslides are short, and accelerated deformation and destructive movements occur suddenly, with almost no warning signs before a collapse. Current methods are insufficient for effective early warning and cannot meet the needs of surface disaster control and prevention in mining areas. To address these issues, this invention studies the occurrence patterns of landslides caused by underground mining activities in loess gully terrain. It uses airborne lidar for high-frequency, high-precision landslide deformation monitoring. By summarizing landslide occurrence patterns, a landslide early warning model for mining areas is established, achieving landslide monitoring and early warning that conforms to the geomorphological characteristics and mining conditions of the region. Summary of the Invention

[0003] This invention discloses a method for monitoring and early warning of landslide deformation in mining subsidence areas of loess gullies based on airborne laser scanning point clouds, in order to solve the problem in the prior art that landslides caused by coal mining in loess gullies occur rapidly and cannot be accurately monitored and warned.

[0004] To solve the above-mentioned technical problems, the present invention adopts the following technical solution. S1. Determine the scope of the mining subsidence area. Based on mining subsidence prediction models such as the probability integral method, determine the scope of surface deformation caused by the current mining face. S2. Initial data acquisition: Based on the impact range of mining subsidence, plan the flight path to conduct UAV three-dimensional laser scanning to obtain initial surface point cloud data within the mining subsidence area; S3. Early identification of potential landslide sites: Based on the special morphological features of potential landslide sites in the surface point cloud, an automatic landslide site identification algorithm is used to identify potential landslide-prone areas. S4. Landslide marking: Analyze the topographic structure of landslide hazard points and mark the extent of landslide hazard bodies; S5. Periodic monitoring by airborne lidar: The UAV laser scanning system is used to periodically monitor and scan the potential landslide points in the mining subsidence area. S6. Digital surface model establishment: Denoising, filtering, interpolation and other processing are performed on the monitoring point cloud data of each period to construct a high-precision digital surface model of the landslide body. S7. Landslide deformation monitoring: By comparing high-precision digital surface models before and after a landslide, the volume of the landslide, the area affected by the landslide damage activity, and the damage caused by the landslide damage activity are calculated and assessed. S8. Calculation of landslide displacement rate: In the periodic monitoring data, the horizontal and vertical displacement of the landslide body is calculated using laser point cloud data, and the average horizontal and vertical displacement rate of the landslide body during the period is calculated. S9. Landslide early warning model establishment: By analyzing the correlation between different structural features of landslide hazard areas within the mining subsidence zone and landslide displacement rate and occurrence probability, a landslide early warning model is established to provide real-time early warning for each landslide hazard point.

[0005] Preferably, in step S1, when determining the extent of the mining subsidence area, it can be determined based on the measured data of rock movement observation in the mining area through a probability integral prediction model or a simulation prediction model, and the actual monitoring range can be expanded accordingly.

[0006] Preferably, in step S2, during the initial data collection, the flight range should cover the entire mining subsidence area. When planning the flight route, the flight altitude should be set according to the actual terrain on the route so that the UAV can collect data in a terrain-following flight mode. Correspondingly, the flight speed should be reduced in dense forest areas to obtain benchmark data with uniform density and accuracy.

[0007] Preferably, in step S3, local landslide data should also be collected as empirical data and added manually to the hazard identification process to ensure that the identification of landslide hazard points is comprehensive and without omission.

[0008] Preferably, in step S4, the movement range of the landslide body is analyzed, and the range of the landslide body is marked by coordinates.

[0009] Preferably, in step S5, during the periodic monitoring data acquisition process, the focus should be on scanning the landslide hazard area. For landslide hazard points under dense forest cover, if a single flight cannot obtain a good ground point cloud, a round trip flight should also be conducted.

[0010] Preferably, in step S6, a denoising, filtering, and interpolation algorithm with good adaptability to loess gully terrain is selected to obtain a high-precision digital surface model of the landslide body.

[0011] Preferably, in step S7, the difference calculation is performed on the digital surface models of the landslide body before and after the landslide to generate a difference digital surface model (D-DSM). The volume of the landslide body and the range of influence are calculated by the D-DSM to assess the impact of the landslide disaster.

[0012] Preferably, in step S8, the same regular grid is established in the monitoring data of different cycles. The grid size can be set according to the accuracy requirements. The landslide range is marked in the grid, each grid is numbered, and the correspondence between the grid and the potential landslide is recorded. Furthermore, in step S8, for the marked grid in the monitoring data of different periods, the grid elevation is calculated by the point cloud elevation of the ground points. If the grid setting is large, the grid elevation can be calculated by averaging the point cloud elevations scattered in the grid. If the grid setting is small and some grids have no scattered point clouds, the grid elevation can be calculated by interpolation.

[0013] Further, in step S8, the grid elevation difference between adjacent monitoring periods and the monitoring period duration are used to calculate the average rate of grid vertical displacement within this period. The average value of the grid vertical displacement rate of the same landslide body is taken as the vertical displacement rate of the landslide body within this period.

[0014] Further, in step S8, the same feature points are identified in the landslide point cloud of adjacent monitoring periods in the grid, the horizontal displacement and direction of the same feature points at different times are calculated, and the horizontal displacement rate of the feature points is calculated accordingly. As many feature points as possible are identified on the same landslide body, and the horizontal displacement rate of the landslide body is calculated accordingly.

[0015] Preferably, in step S9, a landslide early warning model is established for each landslide hazard body monitored within the mining subsidence area, and the horizontal and vertical displacement rates of each landslide body are transmitted in real time after monitoring. The displacement rates of landslide bodies that have already undergone destructive movement and those before the landslide are added to the model as thresholds, and early warning information is issued for landslide bodies that are close to or have reached the displacement rate threshold.

[0016] Compared with existing technologies, this invention utilizes lidar point cloud data for early landslide identification in data processing, solving the problem of the suddenness and unpredictability of landslides caused by mining subsidence in loess gully areas. Compared with current mainstream methods for landslide deformation monitoring, this invention is more predictive in landslide deformation monitoring, breaking through the limitation of traditional monitoring methods that can only monitor after a landslide. Through periodic high-precision monitoring, it analyzes and studies the formation mechanism, evolution law, and destructive effects of landslides caused by mining subsidence under loess gully terrain conditions, and incorporates this law into the early warning model to provide early warning of landslide hazards in mining subsidence areas. By establishing a fixed regular grid and using the interpolation of monitoring data from different periods to obtain grid elevation, the numerical displacement rate of the landslide body is calculated. By calculating and tracking the displacement rate of corresponding points in the regular grid, the horizontal displacement rate of the landslide body is calculated, solving the problem of difficulty in comparing point clouds. At the same time, it greatly optimizes the efficiency, accuracy, and early warning of landslide deformation monitoring in mining subsidence areas under loess gully terrain conditions. Attached Figure Description

[0017] Appendix Figure 1 This is a process flow diagram of the present invention. Appendix Figure 2 This is a schematic diagram of landslide monitoring using airborne lidar. (Attached) Figure 2 In the diagram, 1 represents the airborne lidar system, 2 represents the cracks in the landslide body, 3 represents the displacement direction of the landslide body during the initial deformation stage, 4 represents the displacement direction of the landslide body during the isostatic deformation stage and the accelerated deformation stage, 5 represents the landslide body after displacement, and 6 represents the landslide body before displacement.

[0018] Appendix Figure 3 This is a schematic diagram illustrating the principle of calculating the vertical displacement of a landslide. (Attached) Figure 3 In the diagram, 7 represents the grid fitting elevation before displacement, 8 represents the grid fitting elevation after displacement, 9 represents the grid vertical displacement value, and 10 represents a fixed grid.

[0019] Appendix Figure 4 This is a diagram illustrating the principle of calculating the horizontal displacement of a landslide. (Attached) Figure 4 In the diagram, 11 represents the corresponding point on the landslide body, 12 represents the horizontal displacement rate of the corresponding point, and 13 represents the fixed grid. Detailed Implementation

[0020] The present invention will be further described in detail below with reference to specific embodiments: A method for monitoring and early warning of landslide deformation in mining subsidence areas of loess gullies based on airborne laser scanning point clouds, comprising: S1. Conduct site reconnaissance through on-site investigation and working face analysis. Then, calculate the probability integral prediction model of working face mining based on existing data in the mining area to determine the expected subsidence impact range of the working face mining. This expected impact range will be designated as the key area for landslide monitoring. Site layout will then be carried out, such as... Figure 1 The drone 1 in the figure represents its three locations and flight path. Based on the location and orientation of the mining face, the flight path of drone 1 is determined. Permanent monitoring device 3 is deployed outside the mining subsidence area 2 along the flight path, specifically located on the front and back sides of the mining subsidence area 2. The permanent monitoring device 3 is continuously monitored for 48 hours to measure the absolute coordinates of each feature point and the azimuth of the feature line. S2. Initial data acquisition: Plan the flight path and conduct UAV 3D laser scanning measurement to obtain initial surface point cloud data within the mining subsidence area, such as... Figure 2 In the figure, 1 represents the airborne lidar system. During scanning, it sets up terrain-following flight according to the surface undulation of the gully terrain to obtain surface point cloud data with uniform density. S3. Early identification of potential landslide sites: Identify ridge-shaped, tongue-shaped, armchair-shaped, and horseshoe-shaped features in gully areas; landslide bodies such as nose-shaped mounds and multi-level staggered platforms; various fan-shaped and arc-shaped block structures; and gully edges with cracks. All of the above-mentioned topographical features constitute potential landslide sites; identify their morphology and mark their locations. S4. Establish a fixed regular grid of 0.3m × 0.3m within the mining subsidence area, such as... Figure 3 10 represents a regular grid. The grid is numbered, the topographic structure of each landslide hazard point is analyzed, the sliding range of each landslide body is determined, the range of the landslide hazard body is marked in the grid, the connection between a single landslide body and the grid is established, and the grid occupied by each landslide body is numbered. S5. Periodic monitoring by airborne lidar: During the mining process, the UAV laser scanning system is used periodically to conduct laser scanning measurements on the ground surface, with a focus on monitoring and scanning various landslide hazard points. S6. Digital surface model establishment: Point cloud denoising, progressive triangulation filtering to extract ground points, Kriging interpolation and other processing are performed on the monitoring point cloud data of each period to construct a high-precision digital surface model of the landslide body. S7. Landslide deformation monitoring: By comparing high-precision digital surface models before and after a landslide, a difference DSM is constructed through the difference calculation of the DSM. The difference DSM is used to calculate the landslide volume, the area affected by the landslide failure, and to assess the damage caused by the landslide failure. S8. Calculation of landslide displacement rate: In the periodic monitoring data, the horizontal and vertical displacement of the landslide body is calculated using laser point cloud data, and the average horizontal and vertical displacement rate of the landslide body during the period is calculated. S9. Landslide early warning model establishment: By analyzing the correlation between different structural features of landslide hazard areas within the mining subsidence zone and landslide displacement rate and occurrence probability, a landslide early warning model is established to provide real-time early warning for each landslide hazard point.

[0021] In step S8, the grid elevation is calculated by averaging the point cloud elevations of the marked grid points. Figure 3 In the diagram, 7 and 8 represent the grid elevation calculated by averaging the two measurements, 9 represents the vertical displacement of the grid during this period, the periodic displacement rate of a single grid is calculated using this displacement and the monitoring period, the displacement rate of all grids in a single landslide is calculated, and the average value is taken to obtain the average vertical displacement rate of the landslide.

[0022] In step S8, corresponding feature points are identified in the landslide point cloud of two monitoring cycles in the grid, the horizontal displacement and direction of the corresponding feature points at different times are calculated, and the horizontal displacement rate of the feature points is calculated accordingly. As many feature points as possible are identified on the same landslide, their horizontal displacement rates are calculated, and the average value is taken to calculate the horizontal displacement rate of the landslide.

[0023] In step S9, the first landslide in the mining subsidence area is used as a reference. The horizontal and vertical displacement rates calculated in the previous cycle before the landslide failure movement are used as thresholds. Other landslides with displacement rates reaching the thresholds are marked as dangerous areas, and early warning information is issued. Since the threshold obtained from one set of data is difficult to accurately reflect the actual displacement rate before the landslide failure movement, all subsequent landslide data should be added to the landslide early warning model. The landslide early warning model for this area should be improved during the monitoring process.

[0024] It will be understood by those skilled in the art that, unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. It should also be understood that terms such as those defined in general dictionaries should be understood to have the same meaning as in the context of the prior art, and should not be interpreted in an idealized or overly formal sense unless defined as herein.

[0025] Of course, the above description is not intended to limit the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A method for monitoring and early warning of landslide deformation in mining subsidence areas of loess gullies based on airborne laser scanning point clouds, characterized in that, include: S1. Determine the extent of mining subsidence area: Based on the probability integral method mining subsidence prediction model, determine the extent of surface deformation caused by the current mining face; S2. Initial data acquisition: Based on the extent of the mining subsidence area, plan the flight path to conduct UAV three-dimensional laser scanning to obtain initial surface point cloud data within the mining subsidence area; S3. Early identification of potential landslide sites: Based on the special morphological features of landslide potential sites in the acquired surface point cloud data, an automatic landslide potential site identification algorithm is used to identify potential landslide-prone areas. S4. Landslide marking: Analyze the topographic structure of landslide hazard points and mark the extent of landslide hazard bodies; S5. Monitoring Data Acquisition: Use a drone laser scanning system to periodically monitor and scan each landslide hazard point in the mining subsidence area and acquire regular monitoring data; S6. Digital surface model establishment: Denoising, filtering, and interpolation processing are performed on the monitoring point cloud data to construct a high-precision digital surface model of the landslide body; S7. Landslide deformation monitoring: By comparing high-precision digital surface models before and after a landslide, the volume of the landslide, the area affected by the landslide damage activity, and the damage caused by the landslide damage activity are calculated and assessed. S8. Calculation of landslide displacement rate: In the periodic monitoring data, the horizontal and vertical displacement of the landslide body is calculated using laser point cloud data, and the average horizontal and vertical displacement rate of the landslide body during the period is calculated. S9. Landslide early warning model establishment: By analyzing the correlation between different structural features of landslide hazard areas within the mining subsidence zone and landslide displacement rate and occurrence probability, a landslide early warning model is established to provide real-time early warning for each landslide hazard point.

2. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S1, based on the previous measured or simulated data of the mining area, the surface impact range that may be generated by the mining face is determined by the probability integral method mining subsidence prediction model, and this impact range is used as the monitoring area.

3. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In steps S2 and / or S5, when planning the flight path of the UAV, the flight altitude is set according to the digital ground model within the mining subsidence area, so that the UAV can perform terrain-following flight to obtain ground point cloud data with uniform density.

4. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S3, in the mining subsidence area of ​​the laser point cloud, the following are identified as potential landslide hazard points: basket-shaped, tongue-shaped, armchair-shaped, and horseshoe-shaped steps in the gully area; nose-shaped mounds, multi-level staggered platform sliding bodies; and various fan-shaped and arc-shaped block structures.

5. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S4, a structural analysis is performed on each landslide hazard point to estimate and mark the structure and extent of the landslide hazard body.

6. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S6, the airborne laser point cloud data for each period is denoised, filtered, and high-precision surface point cloud data is extracted. Then, a point cloud interpolation algorithm is used to generate a high-precision digital surface model of the landslide body. The filtering and denoising parameters should be set based on the on-site survey to obtain accurate landslide body point cloud data.

7. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S7, the volumetric digital surface model before and after the landslide failure movement is processed. The models are accurately registered and compared. Based on this, the damage caused by the landslide is analyzed, the volume of the landslide body is calculated, and the safety hazards of the morphological structure formed after the landslide are assessed.

8. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S8, the horizontal and vertical displacement and displacement rate of the landslide body at different times are calculated by comparing laser point cloud interpolation fitting.

9. The method for monitoring and early warning of landslide deformation in loess gully terrain mining subsidence areas based on airborne laser scanning point clouds, as described in claim 1, is characterized in that... In step S9, a landslide early warning model is established. By collecting high-frequency data on the displacement of each landslide body, the model monitors and analyzes the initial deformation stage, constant velocity deformation stage, accelerated deformation stage, duration of the destructive movement, and statistical displacement rate of the landslide body. The average displacement rate before the destructive movement of the landslide body is used as a threshold and added to the early warning model. Early warning information is promptly issued for landslide bodies whose displacement rate reaches the threshold.

Images