Transmission line geological disaster identification method, system and device and medium

Abstract

The invention relates to the technical field of geological disaster recognition, in particular to a power transmission line geological disaster recognition method, system and device and a medium, and the method comprises the steps: regularly obtaining satellite InSAR data and unmanned aerial vehicle LiDAR point cloud data of a target area; performing first preprocessing on the satellite InSAR data, and extracting earth surface deformation information; performing second preprocessing on the unmanned aerial vehicle LiDAR point cloud data to generate a digital elevation model; carrying out geological disaster feature extraction based on a digital elevation model to obtain a disaster type classification result; and analyzing the surface deformation information and the disaster type classification result, and outputting geological disaster state information. The beneficial effects of the invention are that data support and technical guarantee are provided for safe and stable operation of the power transmission line, operation and maintenance cost is reduced, monitoring efficiency is improved, timely prevention measures are ensured, and influence of geological disasters on the power transmission line is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disaster identification, and in particular to a method, system, equipment and medium for identifying geological disasters on power transmission lines. Background Art

[0002] In recent years, geological disaster monitoring technology has continued to develop, and many methods have been widely used in the identification and monitoring of geological disasters. These technologies have improved the accuracy and efficiency of geological disaster monitoring to a certain extent, providing support for the prevention and control of geological disasters.

[0003] However, existing technologies have many limitations: (1) Satellite InSAR technology is susceptible to signal attenuation and atmospheric delay in complex terrain and densely vegetated areas, and data accuracy is limited; (2) UAV LiDAR technology has limited measurement distance, and the quality of data collection is easily affected in complex environments, and has high requirements for flight path planning and data processing; (3) DEM-based monitoring methods can only perform static analysis and are difficult to meet the needs of long-term geological disaster trend prediction and risk assessment. Summary of the Invention

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0005] In a first aspect, the present invention provides a method for identifying geological hazards along a transmission line, comprising regularly acquiring satellite InSAR data and UAV LiDAR point cloud data of a target area;

[0006] Perform the first preprocessing on the satellite InSAR data to extract the surface deformation information;

[0007] Perform the second preprocessing on the UAV LiDAR point cloud data to generate a digital elevation model;

[0008] Extract geological disaster characteristics based on digital elevation models and obtain disaster type classification results;

[0009] Analyze the surface deformation information and disaster type classification results, and output geological disaster status information.

[0010] As a preferred solution of the method for identifying geological hazards of power transmission lines of the present invention, the first preprocessing of satellite InSAR data to extract surface deformation information includes:

[0011] Perform first-target processing on satellite InSAR data to generate interferograms;

[0012] The surface deformation information in the radar line of sight is extracted by performing second target processing on the interference pattern.

[0013] As a preferred solution of the transmission line geological hazard identification method of the present invention, wherein: the UAV LiDAR point cloud data is subjected to a second preprocessing to generate a digital elevation model, including:

[0014] Perform third-object processing on the UAV LiDAR point cloud data to obtain ground point cloud data;

[0015] The ground point cloud data is processed for the fourth target to generate a digital elevation model.

[0016] As a preferred solution of the method for identifying geological hazards of transmission lines of the present invention, the geological hazard features are extracted based on the digital elevation model to obtain the hazard type classification results, including:

[0017] Perform the fifth target processing on the single-period digital elevation model to extract terrain factors;

[0018] Perform differential analysis on multi-period digital elevation models to generate terrain change raster maps;

[0019] Integrate multi-source remote sensing image feature information and determine the disaster type through deep learning models.

[0020] As a preferred solution of the method for identifying geological hazards of power transmission lines of the present invention, the surface deformation information and the hazard type classification results are analyzed to output geological hazard status information, including:

[0021] Calculations based on multi-period digital elevation models were performed to obtain the total erosion or accumulation volume;

[0022] Combining multiple source parameters to build a risk indicator system and perform weighted analysis to obtain a weighted risk index;

[0023] Probabilistic risk output is performed based on Monte Carlo simulation to determine the quantitative degree and risk probability of geological hazards.

[0024] As a preferred solution of the method for identifying geological hazards of power transmission lines of the present invention, the first target processing is performed on the satellite InSAR data to generate an interference pattern, including:

[0025] Perform radiometric calibration and geometric correction on the satellite InSAR data to obtain the first target satellite InSAR data;

[0026] Performing multi-view processing on the first target satellite InSAR data to obtain the second target satellite InSAR data;

[0027] Registering the InSAR data of the second target satellite to obtain the InSAR data of the third target satellite;

[0028] Phase extraction is performed on the InSAR data of the third target satellite to generate an interferogram.

[0029] As a preferred solution of the transmission line geological disaster identification method of the present invention, the method further comprises: performing a second preprocessing on the UAV LiDAR point cloud data to generate a digital elevation model, and also planning the UAV flight path based on the transmission line coverage and setting the overlap rate.

[0030] In a second aspect, the present invention provides a transmission line geological hazard identification system, comprising: a first processing module: configured to perform a first preprocessing on satellite InSAR data to extract surface deformation information;

[0031] The second processing module is used to perform a second pre-processing on the UAV LiDAR point cloud data to generate a digital elevation model;

[0032] Extraction module: used to extract geological disaster features based on digital elevation models and obtain disaster type classification results;

[0033] Output module: used to analyze surface deformation information and disaster type classification results, and output geological disaster status information.

[0034] In a third aspect, the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0035] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above method when the computer program is executed by a processor.

[0036] Compared with the existing technology, the beneficial effects of the present invention are as follows: by fusing satellite InSAR and UAV LiDAR data, the geological environment characteristics along the transmission line can be accurately grasped. By taking advantage of the large-scale monitoring advantages of satellite InSAR data and combining it with the high-precision terrain details of UAV LiDAR data, a comprehensive geological disaster monitoring model is constructed. High-precision extraction of surface deformation information and intelligent identification of geological disaster characteristics are achieved, effectively solving the problem of insufficient monitoring accuracy in complex terrain and vegetation-covered areas. Based on multi-source data fusion and deep learning models, geological disaster risks are dynamically monitored and potential disaster threats are warned in advance. Data support and technical guarantees are provided for the safe and stable operation of transmission lines, operation and maintenance costs are reduced, monitoring efficiency is improved, and preventive measures are ensured to reduce the impact of geological disasters on transmission lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0038] Figure 1 The figure is a flow chart of a method for identifying geological hazards of transmission lines.

[0039] Figure 2 Schematic diagram of the geometric structure of InSAR imaging observation.

[0040] Figure 3 Schematic diagram for solving UAV LiDAR coordinates.

[0041] Figure 4 Schematic diagram of DEM difference theory. DETAILED DESCRIPTION

[0042] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0043] Example 1, with reference to Figures 1 to 4 , which is the first embodiment of the present invention, provides a method for identifying geological hazards on a transmission line, comprising:

[0044] S100: Regularly acquires satellite InSAR data and UAV LiDAR point cloud data of the target area;

[0045] S200: performing a first preprocessing on the satellite InSAR data to extract surface deformation information;

[0046] S300: Perform a second preprocessing on the UAV LiDAR point cloud data to generate a digital elevation model;

[0047] S400: Extract geological disaster features based on the digital elevation model and obtain disaster type classification results;

[0048] S500: Analyze the surface deformation information and disaster type classification results, and output geological disaster status information.

[0049] It should be noted that the transmission line corridors and surrounding areas have a complex geological environment, primarily mountainous and vegetated. While satellite-based InSAR technology can monitor surface deformation over large areas, it is susceptible to signal attenuation and atmospheric delay in areas with vegetated and undulating terrain, limiting data accuracy. While drone-based LiDAR can acquire high-precision terrain data, it is inefficient and suffers from poor data quality over long distances and in complex environments, and requires high flight path planning and data processing. Traditional DEM-based monitoring methods are limited to static analysis and are unable to meet the needs of long-term geological hazard trend prediction and risk assessment.

[0050] Therefore, in order to solve the problem of geological disaster monitoring in the above-mentioned complex environment, through steps S100-S500, satellite InSAR data and UAV LiDAR point cloud data are regularly obtained, and pre-processed to extract surface deformation information and generate digital elevation models respectively; combined with multi-source data fusion strategy, digital elevation models are used to extract and classify geological disaster characteristics, and comprehensively analyze them with surface deformation information to achieve comprehensive monitoring and accurate assessment of geological disaster status; through dynamic monitoring and risk prediction, the accuracy of geological disaster identification is improved, the impact of complex environment on monitoring data quality is reduced, and reliable geological disaster monitoring technology support is provided for the safe and stable operation of transmission lines.

[0051] Example 2, reference Figures 1 to 4 , which is an embodiment of the present invention, provides a method for identifying geological hazards on transmission lines based on the above embodiments.

[0052] In the embodiment of the present application, in step S100 , satellite InSAR data and UAV LiDAR point cloud data of the target area are regularly acquired.

[0053] It should be noted that when acquiring satellite InSAR data for a target area, it is important to select SAR satellites appropriate for the area, such as the C-band Sentinel-1 or the L-band ALOS-2. C-band satellites offer higher temporal resolution (global coverage of approximately six days) and are suitable for monitoring areas that require frequent updates, such as power transmission lines. L-band satellites, on the other hand, have greater penetration into vegetation, providing more reliable data in forested areas. When acquiring data, it is important to ensure a reasonable temporal baseline between the two imagery phases (typically less than one year) to avoid phase decorrelation due to significant surface dynamics. VV polarization should also be selected to enhance surface reflection signals.

[0054] In an optional embodiment, the satellite InSAR data and UAV LiDAR point cloud data for the target area in step S100 can also be obtained based on commercial data services and autonomous UAV mapping. Specifically, satellite InSAR data for the target area can be regularly purchased from professional commercial remote sensing data providers, and an in-house UAV mapping team equipped with a high-performance UAV LiDAR system can be established. Based on the distribution and monitoring requirements of the transmission lines, detailed flight plans can be developed to regularly map the target area.

[0055] In another optional embodiment, the satellite InSAR data and UAV LiDAR point cloud data of the target area obtained in step S100 can also be obtained through open source data and cooperative sharing mechanisms. Specifically, open source satellite remote sensing data resources are utilized. These data are free and open to the public. By accessing relevant data platforms, the required InSAR data can be downloaded according to set time intervals and regional scopes. Furthermore, a cooperative sharing mechanism can be established with local surveying and mapping departments, scientific research institutions, or universities. These institutions often have professional UAV LiDAR equipment and technicians. Through cooperative sharing projects, the collected UAV LiDAR point cloud data can be obtained.

[0056] In the embodiment of the present application, step S200 performs a first preprocessing on the satellite InSAR data to extract surface deformation information, including the following steps A1-A2:

[0057] A1: Perform the first target processing on the satellite InSAR data to generate the interferogram;

[0058] Furthermore, in the embodiment of the present application, performing the first target processing on the satellite InSAR data in step A1 to generate an interferogram includes the following steps A11-A14:

[0059] A11: Perform radiometric calibration and geometric correction on the satellite InSAR data to obtain the first target satellite InSAR data;

[0060] It can be understood that radiation calibration refers to converting the image radiation value of satellite InSAR data into backscatter coefficients to eliminate the influence of factors such as the sensor's own characteristics and observation geometric conditions on the radiation measurement value, so that the data acquired at different times are comparable; geometric correction refers to using the satellite's precise orbit file (such as Sentinel-1's POEORB file) to perform orbit correction on the original data (SAR satellite data) to eliminate the geometric distortion caused by factors such as satellite orbit deviation, so that the geographic coordinates of the image accurately correspond to the actual geographic location.

[0061] A12: Perform multi-view processing on the first target satellite InSAR data to obtain the second target satellite InSAR data;

[0062] It can be understood that multi-view processing reduces the impact of speckle noise by merging multiple adjacent pixels into one pixel and calculating the average value of these pixels, thereby making the image clearer and reducing the amount of data for subsequent processing.

[0063] For example, the satellite InSAR data is divided into blocks according to the set number of pixels in the azimuth and range directions. For example, 5 pixels are taken in the azimuth direction and 1 pixel is taken in the range direction to form a 5×1 pixel block. The pixel values ​​in each 5×1 pixel block are averaged to obtain a new pixel value. These new pixel values ​​are used to replace the original pixel block to generate a new multi-view image.

[0064] A13: Register the InSAR data of the second target satellite to obtain the InSAR data of the third target satellite;

[0065] It should be noted that when repeatedly observing the same area on the Earth's surface, the radar antenna carried by the satellite moves along the orbit and obtains two images S1 and S2 (see the attached image). Figure 2 (as shown). The line connecting the center points of the radar antennas in the two images is called the baseline. The satellite positions were different when the two images were acquired, resulting in slight variations in viewing distance and angle. Therefore, accurate image registration is required before interferometric processing of the S1 and S2 images.

[0066] It should be further explained that the registration of the secondary target satellite InSAR data involves performing a preliminary coarse registration of the master and slave images using a reference DEM (e.g., SRTM 30-meter resolution terrain data) to establish a preliminary mapping relationship and constrain the registration error to within 1-2 pixels. A dense displacement field is calculated using amplitude cross-correlation or phase gradient methods within a local window (e.g., 256×256 pixels) to further improve the registration accuracy to within 0.1 pixel. Amplitude cross-correlation is used in areas covered by stable scatterers, while a phase gradient algorithm is used in areas covered by dynamic targets. A control point screening strategy is combined to eliminate outlier windows. Finally, the offsets of the remaining control points are fitted to a two-dimensional polynomial surface, and the model coefficients are iteratively optimized until the overall residual converges to 0.05-0.1 pixels.

[0067] A14: Perform phase extraction on the InSAR data of the third target satellite to generate an interferogram.

[0068] It should be noted that phase extraction and interferogram generation from the third-target satellite InSAR data involves pixel-by-pixel complex conjugate multiplication of the master image and the registered slave image to generate the initial interferogram. Goldstein filtering (filter coefficient 0.6) is then used to smooth the phase noise, extract the interferometric phase information, and generate the interferogram. Furthermore, it should be noted that vertical baseline control is crucial and must be less than 30% of the critical baseline (the Sentinel-1 baseline must not exceed 200 meters). Otherwise, the signal will completely lose correlation.

[0069] A2: By performing second-target processing on the interferogram, the surface deformation information in the direction of the radar line of sight is extracted.

[0070] It should be noted that the second-target processing of the interferogram involves phase unwrapping using the SNAPHU algorithm combined with the coherence map (threshold > 0.3) to remove 2π ambiguity. The terrain phase is then simulated and subtracted from the high-precision LiDAR-generated DEM (1-meter resolution) to obtain the deformation phase component. The atmospheric phase typically has low-frequency characteristics and needs to be separated using a Gaussian low-pass filter (1-5 km window) and corrected by integrating with a meteorological model (such as ERA5). Ultimately, the deformation phase is converted into a displacement in the radar line of sight (with an accuracy of ±3 mm), representing surface deformation information.

[0071] Specifically, the initial interference pattern I is obtained by the complex conjugate product between the two images after precise registration, and its mathematical expression can be expressed as:

[0072]

[0073] Among them, S1 and S2 are two images of repeated observations of the same area on the Earth's surface when the radar antenna carried by the satellite moves along the orbit, and I is the initial interferogram.

[0074] Furthermore, the interference phase of the interference pattern I can be expressed as follows:

[0075]

[0076] in, is the interference phase, is the terrain phase component that can be used to recover terrain information, is the flat phase component, is the deformation phase component, is the target scattering characteristic, is the atmospheric delay phase, The observation noise is caused by the SAR system. If the scattering characteristics of the target within the resolution unit do not change between two imaging times, then Ignore atmospheric propagation and the noise caused by the SAR system but and The simplified interferometric phase (this simplification helps to focus more on the extraction of surface deformation information) can be expressed as:

[0077]

[0078] Depending on the parameters to be extracted, different components in the formula are eliminated or suppressed, resulting in two main applications: generating digital surface models and extracting surface deformation information. The absolute elevation information of the ground target expressed in arbitrary resolution units of the SAR image can be expressed as

[0079]

[0080] Where H is the flight altitude of the satellite platform, θ is the angle of incidence, α is the angle between the baseline and the horizontal plane, and R is the slant distance from the image to the ground observation point. This formula describes how to extract surface elevation information from the interferometric phase. is the terrain phase component, and λ is the radar wavelength.

[0081] In an optional embodiment, the extraction of surface deformation information from satellite InSAR data in step S200 can also be performed by fusing multi-source data. This involves fusing the satellite InSAR data with other relevant data (such as optical remote sensing imagery and meteorological data). By establishing a correlation model between the data, the impact of multiple factors on surface deformation is comprehensively considered, thereby improving the accuracy and reliability of surface deformation information extraction.

[0082] In another optional embodiment, the surface deformation information extracted from satellite InSAR data in step S200 can also be extracted by combining synthetic aperture radar interferometry (InSAR) with synthetic aperture radar multi-baseline technology (MB-InSAR). In combination with the multi-baseline processing results, a time series InSAR analysis method is used to extract the time series of surface deformation information and analyze the long-term trend and short-term changes of the surface deformation.

[0083] In the embodiment of the present application, the second pre-processing of the UAV LiDAR point cloud data in step S300 to generate a digital elevation model (DEM) includes the following steps B1-B3:

[0084] B1: Plan the drone’s flight path based on the coverage of the transmission lines and set the overlap rate;

[0085] It is understandable that based on the SAR image coverage and monitoring targets, the flight area is delineated and expanded outward by 10% to 15% as a redundant boundary to avoid missing edge data. The flight altitude (usually 200 to 500 meters) is calculated based on the UAV LiDAR field of view, point density requirements and equipment performance. The horizontal overlap rate is ≥70% (to prevent gaps between flight strips), and the heading overlap rate is ≥80% (to ensure continuous coverage of the terrain). A reciprocating route (mode) is adopted, combined with a LiDAR scanning angle of ±30°, to ensure point cloud coverage of lateral features (such as steep slopes of mountains and facades of buildings). 3 to 5 GNSS ground base stations are evenly distributed along the survey area to collect static observation data and provide centimeter-level absolute coordinate reference for the point cloud data.

[0086] B2: Perform third-target processing on the UAV LiDAR point cloud data to obtain ground point cloud data;

[0087] It should be noted that this step is to separate ground points and non-ground points (such as vegetation, buildings, etc.) from the original LiDAR point cloud to generate pure terrain surface point cloud data. Specifically, it includes the use of statistical filtering (such as threshold screening based on elevation distribution) and density clustering algorithms (such as DBSCAN) to remove outliers (such as flying birds, noise points); apply regularized filtering (such as slope filtering, surface fitting) to separate ground points from non-ground points such as vegetation and buildings, and further accurately extract terrain surface point cloud data through methods.

[0088] B3: Perform the fourth target processing on the ground point cloud data to generate a digital elevation model.

[0089] It should be noted that this step is to convert the ground point cloud data into a regular rasterized elevation model for geological hazard feature analysis. Specifically, it includes the inverse distance weighted method (IDW) or Kriging interpolation algorithm to convert discrete point clouds into continuous elevation raster data to generate a DEM with a resolution of 0.2 to 1 meter; further, the offset between multiple flight strips is eliminated through flight strip splicing error correction (such as the ICP algorithm), and local surface fitting or edge sharpening is performed for areas with complex terrain (such as ridges and gullies) to improve the consistency between the DEM and the actual terrain.

[0090] Specifically, calculate the distance from the laser foot point to the reference point. According to the principle of spatial triangulation, the three-dimensional coordinate information of the target point can be obtained. The schematic diagram of the coordinate solution is shown in the attached figure. Figure 3 As shown. Point G is the projection center of the laser scanner, and the position information is provided by differential GPS; ( ω, κ) are flight attitude parameters, representing pitch, roll, and yaw respectively; point M is the target point position to be acquired. The calculation formula for the target point position according to the graph is as follows:

[0091]

[0092] Where b = sinφcosωcosκ+cosωsinκ. Where (X M ,Y M ,Z M ) is the coordinate of the target point M to be obtained, (X G ,Y G ,Z G ) are the coordinates of the projection center point G of the laser scanner, j, w, and k represent the pitch angle, roll angle, and yaw angle, respectively, q is the emission angle of the laser beam relative to the vertical downward direction, b is the correction factor to compensate for the deviation of the laser beam direction from the expected direction due to the change of the drone's posture, and d is the length of the line segment NM in the figure.

[0093] In an optional embodiment, the terrain surface point cloud data extracted in step S300 may be extracted by adopting an iterative classification algorithm (such as cloth simulation filtering).

[0094] In another optional embodiment, the terrain surface point cloud data extracted in step S300 can also be extracted using a machine learning model (such as random forest classification).

[0095] In the embodiment of the present application, step S400 extracts geological disaster features based on the digital elevation model to obtain disaster type classification results, including the following steps C1-C3:

[0096] C1: Perform the fifth target processing on the single-period digital elevation model to extract terrain factors;

[0097] It should be noted that the single-period digital elevation model is first preprocessed through geometric correction (eliminating DEM coordinate deviations and ensuring spatial alignment with remote sensing images), noise filtering (using median filtering or Gaussian filtering to suppress elevation outliers), and temporal registration (if multiple periods of data are involved, the coordinate system and resolution are unified). Then, by calculating terrain factors such as slope, aspect, and curvature from the single-period DEM, geological hazards and environmental changes are extracted, identified, located, and characterized. Slope represents the inclination of the surface, aspect indicates the direction of surface runoff, and curvature reflects the degree of terrain concavity. Further analysis of terrain factors is aided by hillshade analysis.

[0098] C2: Perform differential analysis on multi-period digital elevation models to generate a terrain change raster map;

[0099] It should be noted that the DEMs of different periods were subtracted pixel by pixel to generate a raster map of terrain changes, in which positive areas represent accumulation and negative areas represent erosion.

[0100] C3: Integrate multi-source remote sensing image feature information and determine the disaster type through deep learning models.

[0101] It should be noted that the fusion of multi-source remote sensing imagery features involves combining optical imagery, SAR imagery, and thermal infrared data to construct a multidimensional feature vector. Optical imagery is used to identify vegetation damage and loose deposits, SAR imagery is used to capture the creep process of landslides, and thermal infrared data is used to identify temperature anomalies caused by water seepage. Deep learning models are used to determine disaster types, including deep learning models such as deep forests or graph convolutional networks, to analyze multidimensional feature vectors and achieve probabilistic classification of disaster types. For example, a landslide appears as a "tongue-shaped" shadow in optical imagery and a retreating terrain drop in the DEM. Debris flows require the simultaneous uplift of the gully bed and erosion along its sides.

[0102] Specifically, in C2, refer to the attached Figure 4 As shown, the difference of DEM is the subtraction between two pairs of multi-period DEMs. After the multi-period DEM spatial matching is completed, the value of the grid cell of the other period is subtracted from the value of the grid cell of the one period DEM pixel by pixel. Furthermore, the dangerous area is detected by DEM difference analysis. When the difference result is positive, the positive area is the accumulation area; when the difference result is negative, the negative area is the erosion area. The difference result is superimposed with basic information such as remote sensing images. When the accumulation area is adjacent to the erosion area and is thin, it is a collapse; when the accumulation area is adjacent to the erosion area and is large in size, it is a landslide; when the erosion area is close to the river and there is no accumulation area, it indicates a debris flow; if the negative area is large, it indicates that there may be a ground subsidence disaster.

[0103] In an optional embodiment, the geological hazard feature extraction based on the digital elevation model in step S400 can also use advanced 3D modeling software to convert the DEM data into a 3D terrain model to intuitively display the terrain undulations and geological structure, that is, to fuse the digital elevation model (DEM) with high-resolution optical images, geological exploration data, etc., and use feature extraction algorithms on the basis of the 3D terrain model to identify geological structural features such as faults and folds.

[0104] In another optional embodiment, the geological hazard feature extraction based on the digital elevation model in step S400 can also adopt the DEM enhancement technology of the generative adversarial network (GAN), that is, constructing a generative adversarial network, in which the generator is used to generate enhanced DEM data, and the discriminator is used to distinguish the generated DEM from the real DEM. Through training, the DEM generated by the generator is close to the real data in details and authenticity, and the trained GAN is used to enhance the new DEM data to improve the data quality and extract clearer and more accurate geological hazard features from the enhanced DEM.

[0105] In the embodiment of the present application, step S500 analyzes the surface deformation information and the disaster type classification results to output geological disaster status information, including the following steps D1-D3:

[0106] D1: Calculate the total erosion or accumulation volume based on the multi-period digital elevation model;

[0107] D2: Construct a risk indicator system based on multi-source parameters and conduct weighted analysis to obtain a weighted risk index. The weights of each indicator are assigned using the analytic hierarchy process (AHP) (e.g., deformation weight 0.4, volume weight 0.3, dynamic evolution weight 0.2, and environmental covariate weight 0.1).

[0108] It is understandable that the construction of risk indicator system involves the deformation strength index (S d ), scale threshold index (V), dynamic evolution index (E m ) and environmental covariates (C r ), furthermore, the deformation intensity index is based on the annual deformation rate obtained by InSAR technology and is Z-score normalized; the scale threshold index is the ratio of the total volume change to the critical volume calculated based on DEM difference; the dynamic evolution index is based on the root mean square error (RMSE) of the time series fitting residual, reflecting the nonlinear acceleration characteristics; the environmental covariates are 24-hour rainfall (from meteorological station data) and rock friction angle (based on the lithology classification map of multispectral remote sensing inversion).

[0109] D3: Probabilistic risk output based on Monte Carlo simulation to determine the quantitative degree and risk probability of geological hazards.

[0110] Specifically, in D1, the specific calculation formula for the total erosion or accumulation volume is:

[0111]

[0112] Where V is the total erosion / deposition volume (m 3 ), Δh i is the elevation change value of the i-th grid (accuracy ±0.15m), A is the grid projection area (e.g. 5×5m 2 ), cosθ i is the slope correction factor (based on a high-precision slope map, error <3% when θ≤45°).

[0113] It should be noted that for the non-uniform deformation area, an adaptive grid division strategy is adopted. In the area with large deformation gradient (such as the trailing edge of the landslide), the grid is encrypted to 1×1m, which improves the volume calculation accuracy to ±5%.

[0114] In D2, the specific expression of the weighted risk index is:

[0115] R=0.4·S d +0.3·log(V / V c )+0.2·E m +0.1·C r

[0116] Where: S d is the normalized deformation rate, E m is the environmental factor, C r is the surface cover coefficient (bare land = 1, vegetation = 0.7), V is the total volume change calculated by DEM difference, and V c is the critical volume, log(V / V c ) is used to measure the scale risk of terrain change.

[0117] In D3, the probabilistic risk output based on Monte Carlo simulation includes the following steps D31-D33:

[0118] D31: Uncertainty modeling of input parameters, including determining the probability distribution of input parameters, such as DEM error (±0.2m normal distribution), rock friction angle (±5° triangular distribution), rainfall prediction error (±20% uniform distribution);

[0119] D32: Perform a large number of iterative simulations (e.g., 10,000 times), each time randomly sampling from the uncertainty distribution of the input parameters, calculating the risk index, and summarizing the results;

[0120] D33: Determine the probability of occurrence of different geological disaster risk levels based on the cumulative distribution function (CDF) of the simulation results.

[0121] Exemplary risk level thresholds:

[0122] Low risk (R<0.35): probability of occurrence <10%;

[0123] Medium risk (0.35≤R<0.65): probability of occurrence 10%-40%;

[0124] High risk (R≥0.65): probability of occurrence >40%.

[0125] In an optional implementation, the output of the geological disaster status information in step S500 may also be performed based on an early warning system for geological disaster risk assessment.

[0126] In another optional embodiment, the output of the geological disaster status information in step S500 can also be performed by utilizing a visualization and analysis platform of a geographic information system (GIS).

[0127] In summary, the present invention accurately grasps the geological environment characteristics along the transmission line by fusing satellite InSAR and UAV LiDAR data. Utilizing the large-scale monitoring advantages of satellite InSAR data and combining it with the high-precision terrain details of UAV LiDAR data, a comprehensive geological disaster monitoring model is constructed. High-precision extraction of surface deformation information and intelligent identification of geological disaster characteristics are achieved, effectively solving the problem of insufficient monitoring accuracy in complex terrain and vegetation-covered areas. Based on multi-source data fusion and deep learning models, geological disaster risks are dynamically monitored and potential disaster threats are warned in advance. Data support and technical guarantees are provided for the safe and stable operation of transmission lines, operation and maintenance costs are reduced, monitoring efficiency is improved, and preventive measures are ensured to reduce the impact of geological disasters on transmission lines.

[0128] Example 3. The above is a schematic scheme of a method for identifying geological hazards on transmission lines. It should be noted that the technical scheme of the system for identifying and monitoring geological hazards on transmission lines is the same as the technical scheme of the above-mentioned method for identifying geological hazards on transmission lines. For details not described in detail in the technical scheme of the system for identifying geological hazards on transmission lines in this embodiment, please refer to the description of the technical scheme of the above-mentioned method for identifying geological hazards on transmission lines.

[0129] This embodiment also provides a transmission line geological disaster identification system, including:

[0130] The acquisition module is used to regularly acquire satellite InSAR data and UAV LiDAR point cloud data of the target area;

[0131] The first processing module is used to perform the first preprocessing on the satellite InSAR data and extract the surface deformation information;

[0132] The second processing module is used to perform a second pre-processing on the UAV LiDAR point cloud data to generate a digital elevation model;

[0133] Extraction module: used to extract geological disaster features based on digital elevation models and obtain disaster type classification results;

[0134] Output module: used to analyze surface deformation information and disaster type classification results, and output geological disaster status information.

[0135] This embodiment also provides an electronic device suitable for identifying geological hazards on transmission lines, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement the method for identifying geological hazards on transmission lines proposed in the above embodiment.

[0136] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for identifying geological hazards in transmission lines proposed in the above embodiment is implemented.

[0137] The storage medium proposed in this embodiment and the method for identifying geological hazards in transmission lines proposed in the above embodiment belong to the same inventive concept. For technical details not described in detail in this embodiment, please refer to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0138] Through the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented with the help of software and necessary general hardware, and of course can also be implemented by hardware. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0139] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for identifying geological hazards on power transmission lines, characterized by: include, Regularly acquire satellite InSAR data and UAV LiDAR point cloud data of the target area; performing a first preprocessing on the satellite InSAR data to extract surface deformation information; Performing a second preprocessing on the UAV LiDAR point cloud data to generate a digital elevation model; Extracting geological disaster features based on the digital elevation model to obtain disaster type classification results; The surface deformation information and the disaster type classification results are analyzed to output geological disaster status information.

2. A method for identifying geological hazards on power transmission lines according to claim 1, characterized in that: The satellite InSAR data is first preprocessed to extract surface deformation information, include, performing a first target processing on the satellite InSAR data to generate an interferogram; By performing second target processing on the interference pattern, surface deformation information in the radar line of sight direction is extracted.

3. A method for identifying geological hazards on power transmission lines according to claim 2, characterized in that: The second preprocessing of the UAV LiDAR point cloud data to generate a digital elevation model includes: Performing third-target processing on the UAV LiDAR point cloud data to obtain ground point cloud data; The ground point cloud data is subjected to a fourth target processing to generate a digital elevation model.

4. A method for identifying geological hazards on power transmission lines according to claim 3, characterized in that: The geological disaster feature extraction based on the digital elevation model is performed to obtain the disaster type classification result, including: Perform the fifth target processing on the single-period digital elevation model to extract terrain factors; Perform differential analysis on multi-period digital elevation models to generate terrain change raster maps; Integrate multi-source remote sensing image feature information and determine the disaster type through deep learning models.

5. A method for identifying geological hazards on power transmission lines according to claim 4, characterized in that: The surface deformation information and the disaster type classification results are analyzed to output geological disaster status information. include, Calculations based on multi-period digital elevation models were performed to obtain the total erosion or accumulation volume; Combining multiple source parameters to build a risk indicator system and perform weighted analysis to obtain a weighted risk index; Probabilistic risk output is performed based on Monte Carlo simulation to determine the quantitative degree and risk probability of geological hazards.

6. A method for identifying geological hazards on power transmission lines according to claim 2, characterized in that: The performing first target processing on the satellite InSAR data to generate an interference pattern includes: Performing radiometric calibration and geometric correction on the satellite InSAR data to obtain first target satellite InSAR data; performing multi-look processing on the first target satellite InSAR data to obtain second target satellite InSAR data; performing registration on the second target satellite InSAR data to obtain third target satellite InSAR data; Phase extraction is performed on the InSAR data of the third target satellite to generate the interference pattern.

7. A method for identifying geological hazards on power transmission lines according to claim 3, characterized in that: The second preprocessing of the UAV LiDAR point cloud data to generate a digital elevation model also includes planning the UAV flight path based on the coverage of the transmission line and setting the overlap rate.

8. A transmission line geological disaster identification system, applying the method according to any one of claims 1 to 7, characterized in that: include: The acquisition module is used to regularly acquire satellite InSAR data and UAV LiDAR point cloud data of the target area; A first processing module: configured to perform a first preprocessing on the satellite InSAR data to extract surface deformation information; The second processing module is used to perform a second preprocessing on the UAV LiDAR point cloud data to generate a digital elevation model; Extraction module: used for extracting geological disaster features based on the digital elevation model to obtain disaster type classification results; Output module: used to analyze the surface deformation information and the disaster type classification results, and output geological disaster status information.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images