Overhead line construction potential safety hazard point positioning method, system, equipment and medium

Abstract

The invention discloses an overhead line construction potential safety hazard point positioning method, system and device and a medium, and belongs to the technical field of line construction, and the method comprises the following steps: carrying out image collection on a predetermined construction area to obtain image data, and constructing a three-dimensional topographic map based on the image data; analyzing the distribution condition of the features through the three-dimensional topographic map, and screening a landform change position according to an analysis result; and deploying ground monitoring equipment at the landform change position, evaluating a monitoring result, and marking a risk point based on an evaluation result. According to the invention, comprehensive coverage of the predetermined construction area is realized through the multi-angle image acquisition device, the detailed three-dimensional topographic map is constructed, the ground cover distribution condition can be analyzed on the basis of the three-dimensional topographic map, key areas influencing the construction safety are screened out, and the construction safety is improved. By deploying ground monitoring equipment to track the change condition of potential risk factors in real time, the identification and coping strategies of potential safety hazard points are dynamically adjusted, and it is ensured that construction activities avoid high-risk sites.

Description

Technical Field

[0001] The present invention relates to the technical field of line construction, and in particular to a method, system, equipment and medium for locating safety hazard points in overhead line construction. Background Art

[0002] In overhead line construction, ensuring the safety of the construction area is one of the key factors for project success. Traditional overhead line construction safety assessments typically rely on manual on-site surveys and static data analysis, which has certain limitations. First, manual surveys are time-consuming and costly, and their efficiency and accuracy are significantly reduced for large or difficult-to-reach areas (such as mountainous areas or areas with dense vegetation cover). Second, traditional methods mainly conduct risk assessments based on historical data and cannot reflect changes in the construction site in real time, especially in dynamic environments where the distribution of surface cover and landform characteristics may change rapidly, thus affecting construction safety.

[0003] Existing methods struggle to accurately locate and promptly update safety hazards during overhead line construction. Due to a lack of effective dynamic monitoring, traditional methods cannot accurately capture new hazards that may emerge at any time during construction. This can lead to construction plans failing to fully incorporate the latest safety information, increasing the risk of accidents. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is that the traditional method cannot accurately capture new hidden dangers that may appear at any time during the construction process.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for locating safety hazards in overhead line construction, which comprises the following steps:

[0007] Collect images of the planned construction area to obtain image data, and construct a three-dimensional topographic map based on the image data;

[0008] Analyzing the distribution of features using the three-dimensional topographic map, and screening locations of landform changes based on the analysis results;

[0009] Deploy ground monitoring equipment at the location of the landform change, evaluate the monitoring results, and mark risk points based on the evaluation results.

[0010] As a preferred solution of the method for locating safety hazards in overhead line construction described in the present invention, the method for obtaining the image data includes:

[0011] The image acquisition device is mounted on an aerial platform and moved along a preset path;

[0012] The image acquisition device is preset with a time interval, the image acquisition device shoots images according to the time interval, arranges the images according to the shooting time sequence, and obtains an image sequence;

[0013] The displacement value between each two adjacent images is calculated by using the obtained image sequence;

[0014] The images meeting the comparison requirement are output as image data by comparing the displacement value with a preset threshold value.

[0015] The beneficial effects of the preferred technical scheme are as follows: moving along the preset path ensures the complete coverage of the construction area, avoids missing dangerous points, and through the calculation of the displacement value of adjacent images and the comparison with the threshold value, the quality of the obtained images meets the analysis requirement, and the images with insufficient blur or overlap do not affect the subsequent analysis.

[0016] As a preferred scheme of the overhead line construction safety hidden danger point positioning method, the three-dimensional topographic map is constructed in the following manner,

[0017] The collected image data is preprocessed to obtain the geographic coordinates corresponding to each image;

[0018] All images are registered by using the geographic coordinates, the overlapping areas of each image are calculated, and a registered collective map is obtained;

[0019] The parameters in each image are analyzed by using the registered collective map, and the image height value is obtained based on the analysis result;

[0020] The image height value is combined with the registered collective map to generate a three-dimensional topographic map.

[0021] As a preferred scheme of the overhead line construction safety hidden danger point positioning method, the step of analyzing the distribution of the features by using the three-dimensional topographic map comprises,

[0022] The three-dimensional topographic map is classified and processed;

[0023] The first feature under each category is calculated, and the calculated first feature result is sorted, and the region ranked first is output as a first region;

[0024] The second feature of the first region is evaluated, and a second feature result is obtained;

[0025] The first feature result and the second feature result are output to the three-dimensional topographic map.

[0026] The beneficial effects of the preferred technical scheme are as follows: different ground features are automatically identified by classification processing, the subjectivity of manual judgment is reduced, the most important region is identified by the sorting mechanism, and the analysis is more targeted.

[0027] As a preferred scheme of the overhead line construction safety hidden danger point positioning method, wherein: the screening method of the landform change position comprises,

[0028] The stability of the first region is evaluated;

[0029] Using the stability evaluation result, the position where the terrain inclination exceeds the preset threshold is marked as an abnormal area by calculating the terrain inclination;

[0030] The soil compaction degree of the first region marked as the abnormal area is evaluated, and when the soil compaction degree is lower than the set compaction threshold, the output is a risk area;

[0031] The landform change position is output in combination with the risk area and the abnormal area.

[0032] The beneficial effect of the preferred technical scheme is that the potential risk area is more accurately identified through double evaluation of the inclination and the soil compaction degree, the abnormal area and the risk area are distinguished, the risk is managed in a hierarchical manner, and the resource allocation is more reasonable.

[0033] As a preferred scheme of the overhead line construction safety hidden danger point positioning method, wherein: the step of evaluating the monitoring result of the landform change position comprises,

[0034] The preliminary stability analysis is performed on each data of the landform change position monitoring, and the stability change trend graph is constructed based on the preliminary stability analysis result;

[0035] The risk point is marked when the change speed of a certain hidden danger point is greater than the preset threshold by using the stability change trend graph to identify the hidden danger point with deteriorating stability.

[0036] The beneficial effect of the preferred technical scheme is that the stability change trend graph is constructed, the dynamic monitoring of the risk point is realized, and the early warning mechanism is established through the change speed threshold setting.

[0037] As a preferred scheme of the overhead line construction safety hidden danger point positioning method, wherein: further comprising the following steps,

[0038] The evaluation period is set, the current time and the evaluation period are calculated, and it is determined whether to start the next round of evaluation;

[0039] Based on the evaluation result, the image data is updated.

[0040] Another object of the present application is to provide an overhead line construction safety hidden danger point positioning system.

[0041] To solve the above technical problems, the present application provides the following technical scheme: an overhead line construction safety hidden danger point positioning system, comprising:

[0042] The present invention provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of the method for locating safety hazard points in overhead line construction are implemented.

[0043] The present invention provides 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 for locating safety hazard points in overhead line construction are implemented.

[0044] The beneficial effects of the present invention are as follows: comprehensive coverage of the predetermined construction area is achieved through a multi-angle image acquisition device, a detailed three-dimensional topographic map is constructed, and the distribution of surface cover can be analyzed on the basis of the three-dimensional topographic map to screen out key areas that affect construction safety. Furthermore, by deploying ground monitoring equipment to track changes in potential risk factors in real time, the identification and response strategies for potential safety hazards can be dynamically adjusted to ensure that construction activities avoid high-risk locations. This method greatly improves the accuracy and timeliness of identifying potential safety hazards, allowing construction teams to more effectively plan and adjust construction routes, avoid unnecessary risks, and ensure the safety of construction personnel and the smooth progress of the project. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] 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.

[0046] Figure 1 The present invention provides an overall flow chart of a method for locating safety hazards in overhead line construction according to an embodiment of the present invention.

[0047] Figure 2 A flowchart of constructing a three-dimensional topographic map for a method of locating safety hazard points in overhead line construction provided by one embodiment of the present invention.

[0048] Figure 3 A flowchart of an abnormality determination method for locating safety hazard points in overhead line construction provided by one embodiment of the present invention.

[0049] Figure 4 A flowchart of image data updating for a method for locating safety hazard points in overhead line construction provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0050] 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.

[0051] Example 1, with reference to Figure 1 , which is the first embodiment of the present invention, provides a method for locating safety hazard points in overhead line construction, comprising the following steps:

[0052] S1. Collect images of the planned construction area to obtain image data, and construct a three-dimensional topographic map based on the image data.

[0053] S2. Analyze the distribution of features using the three-dimensional topographic map, and select locations of landform changes based on the analysis results.

[0054] S3. Deploy ground monitoring equipment at the location of the landform change, evaluate the monitoring results, and mark risk points based on the evaluation results.

[0055] This embodiment provides a systematic method for locating safety hazards to address specific issues such as complex terrain, uncertain distribution of risk points, and low efficiency of traditional manual surveys during overhead line construction. Specifically, in the S1 stage, drones equipped with multispectral cameras are used to capture multi-angle images of the construction area along a preset trajectory, acquiring a high-definition image with a resolution of 0.1 m / pixel every T = 5 seconds, and constructing a three-dimensional terrain model with centimeter-level accuracy through stereo photogrammetry technology. In the S2 stage, machine learning algorithms are used to automatically analyze the three-dimensional terrain map to identify steep areas with slopes exceeding 25°, exposed areas with vegetation coverage below 30%, and abrupt terrain changes with elevation changes exceeding 5 m / 100 m, and screen out potential risk locations through multi-factor overlay analysis. In the S3 stage, different types of monitoring equipment are deployed at key locations according to the risk level, including high-precision GPS monitoring points, soil moisture sensors, and ground subsidence meters. Real-time monitoring thresholds are set (such as displacement speed > 2 mm / day, moisture content change > 15%) to achieve 24-hour continuous monitoring of risk points. When the monitoring parameters exceed the preset thresholds, the early warning mechanism is automatically triggered, providing dynamic protection for construction safety.

[0056] Through the steps S1-S3, the overhead line construction project has achieved remarkable results in safety management. Compared with the traditional manual survey method, the implementation of this scheme has improved the coverage rate of hidden danger identification and significantly reduced the omission of safety hazards; through accurate risk point positioning, the project successfully avoided 5 potential geological disaster risks, including 3 landslides caused by heavy rain and 2 soft soil foundation settlement hazards; the construction efficiency has been significantly improved. At the same time, the real-time monitoring network established provides valuable data basis for subsequent equipment maintenance and operation management. Through historical data analysis, the equipment operating condition can be predicted, maintenance plan can be made in advance, equipment service life can be prolonged, and the safety management mode is changed from passive response to active prevention, which brings double value to the safety and economy of power engineering construction.

[0057] Embodiment 2, refer to Figures 1 to 4 For an embodiment of the present application, based on the above embodiment, an overhead line construction safety hazard point positioning method is provided.

[0058] In the present application embodiment, the method for obtaining image data in step S1 includes the steps of:

[0059] S1.1, carrying an image acquisition device on an aerial platform, and moving the image acquisition device according to a preset path;

[0060] S1.2, the image acquisition device presets a time interval, the image acquisition device shoots images according to the time interval, and arranges the images according to the shooting time sequence to obtain an image sequence;

[0061] S1.3, using the obtained image sequence, calculating the displacement value between each two adjacent images;

[0062] S1.4, comparing the displacement value with a preset threshold value, and outputting the images meeting the comparison requirements as image data.

[0063] In S1.1-S1.4, when collecting image data in the construction area, the aerial platform moves according to the preset path, and maintains a constant height H and speed V in the process to ensure that the image acquisition device can uniformly cover every part of the predetermined construction area; by setting a constant height and speed, the stability and consistency of image acquisition are ensured, the data error caused by changes in height or speed is reduced, and the quality and reliability of the image data are improved.

[0064] The image acquisition device shoots a series of images I at a fixed time interval T, wherein each image I covers a ground area A, and there is an overlap ratio P between adjacent images, i.e. the mathematical expression is:

[0065] P=(A next ∩A prev ) / Aprev ;

[0066] Where A next is the coverage area of ​​the next image, A prev is the coverage area of ​​the previous image;

[0067] The continuity and integrity between images are guaranteed; the fixed time interval T ensures the continuity of the image, and the appropriate overlap ratio P enhances the connection between different images, avoids information loss, and provides a more complete and accurate data basis for subsequent analysis.

[0068] Using the obtained image sequence S, the relative displacement D between each pair of adjacent images is calculated. The calculation formula is expressed as:

[0069]

[0070] X next Is the central horizontal coordinate of the next image, X prev Is the center horizontal coordinate of the previous image, Y next Is the center vertical coordinate of the next image, Y prev is the center vertical coordinate of the previous image;

[0071] By calculating the relative displacement D, the changes in surface features can be quantified, helping to identify areas with complex terrain or potential safety hazards, and improving the ability to capture the sensitivity of landform changes.

[0072] Based on the calculated displacement value D, image pairs that meet the threshold condition D > Dmin are selected for detailed analysis. Dmin is a pre-set value smaller than all normal displacement values, intended to highlight image features in areas with complex terrain or potential safety hazards. By setting the displacement threshold Dmin, image pairs with significant changes are effectively screened out, focusing on potentially problematic areas. This improves the relevance and efficiency of data analysis and reduces unnecessary processing workload.

[0073] In an optional embodiment, when an overhead line construction project is being carried out in a mountainous area, the area has complex terrain and dense vegetation, and traditional manual survey methods are difficult to fully cover.

[0074] A drone was selected as the aerial platform, equipped with a high-resolution camera as the image acquisition device. The drone flew within the construction area according to a pre-programmed flight path, maintaining a constant altitude (H) of 100 meters and a speed (V) of 5 meters per second, ensuring uniform coverage of the entire area.

[0075] The camera captures an image at a fixed time interval T = 5 seconds. Each capture covers a ground area A of approximately 100 square meters. The overlap ratio P between adjacent images is set to 60% to ensure sufficient overlap between images to maintain continuity.

[0076] Computer vision software is used to process the image sequence S and calculate the relative displacement D between each pair of adjacent images. For each image pair, the displacement is calculated based on the image center coordinates to obtain accurate surface feature change data.

[0077] A displacement threshold of Dmin = 5 meters was set, and image pairs exceeding this threshold were selected for detailed analysis. These image pairs typically correspond to areas with complex terrain or potential safety hazards, such as steep slopes and loose soil areas.

[0078] Through this process, the project team successfully identified and flagged multiple potential safety hazards, including several previously unnoticed areas of geological instability. Based on this information, the team adjusted the construction plan to avoid high-risk areas, ensuring a safe and efficient construction process. Furthermore, the deployment of real-time monitoring equipment further strengthened dynamic monitoring of the construction site, enabling timely response to any new risks.

[0079] In another optional embodiment, when an overhead transmission line construction project is being carried out in a coastal area, the terrain in the area is diverse, with both hilly areas and coastal alluvial plains. The geological conditions are complex, the groundwater level changes frequently, and it is often affected by sea breezes and rainfall. Traditional manual surveys have safety risks and are inefficient.

[0080] A large quadrotor drone was selected as the aerial platform, equipped with a combination of LiDAR and a high-resolution infrared camera for image acquisition. The drone followed a pre-set zigzag flight path to cover the construction area, maintaining a constant altitude (H) of 80 meters and a speed (V) of 8 meters per second to ensure effective coverage of the undulating terrain.

[0081] The camera captures images at intervals of T = 3 seconds, with each capture covering approximately 120 square meters of ground area A. Considering the large variations in terrain, the overlap ratio of adjacent images was set to P = 70% to ensure good image connectivity even in complex terrain conditions.

[0082] The image sequence S was processed using professional photogrammetry software, and the relative displacement D between adjacent image pairs was calculated. Based on the geological characteristics of coastal areas, a displacement threshold Dmin was set at 3.5 meters to identify areas with drastic topographic changes. This threshold takes into account potential geological phenomena such as coastal erosion and soft soil settlement.

[0083] By screening image pairs with D>Dmin, the project team identified the following key areas: within 500 meters of the coast, three areas with large terrain changes were found, with displacement values ​​of D = 4.2-6.8 meters; in the alluvial plain, five locations with possible soft soil settlement were detected, with displacement values ​​of D = 3.7-5.5 meters; and two areas of abnormal moisture distribution were identified through infrared images, with displacement values ​​of D = 4.0-4.8 meters.

[0084] In the embodiment of the present application, the method for constructing the three-dimensional topographic map in step S1 includes the following steps S1.5-S1.8:

[0085] S1.5. Preprocess the collected image data to obtain the geographic coordinates corresponding to each image;

[0086] S1.6. Register all images using geographic coordinates and calculate the overlapping area of ​​each image to obtain a registration set map.

[0087] S1.7. Analyze parameters in each image using the registration set graph, and obtain image height values ​​based on the analysis results;

[0088] S1.8. Combine the image height values ​​with the registration set map to generate a three-dimensional topographic map.

[0089] Specifically, the image data collected by S1 is preprocessed to ensure that each image I has corresponding geographic coordinates G(X,Y), where X and Y represent the horizontal position on the Earth's surface. By assigning accurate geographic coordinates to each image, spatial accuracy is ensured during subsequent analysis and overlay. This step is the foundation for constructing accurate 3D topographic maps, improving the reliability and application value of the final results.

[0090] All images I are registered based on geographic coordinates G to ensure they accurately overlay in the same coordinate system. The overlap area R between each image is calculated using the formula R = (Aintersect / Aimage), where Aintersect represents the intersection area of ​​the two images and Aimage is the total area covered by the individual images. Image registration ensures that all images can be seamlessly stitched together in the same coordinate system, eliminating misalignment caused by differences in shooting angle or position. Calculating the overlap area helps assess the degree of match between images, ensuring the quality and integrity of the stitched image and reducing the possibility of information loss.

[0091] Using the registered image set, the ground height Z is estimated by analyzing the gray level changes or color differences in each image. For each pixel point P, the height estimation formula Z(P) = f(I(P)) is used, where f is a mapping function that converts the image intensity I into the corresponding height value Z. Estimating the ground height based on the gray level or color changes of the image can capture subtle topographic features and provide more detailed terrain information. This method not only improves the accuracy of height estimation, but also enhances the adaptability to complex terrain (such as vegetation-covered areas), making the generated three-dimensional terrain map more realistic and detailed.

[0092] Combined with the obtained height information Z, a continuous three-dimensional terrain grid M is generated. For any two adjacent pixel points P1 and P2, if the height difference ΔZ between them exceeds the set threshold ΔZmax, an additional grid node N is inserted between the two points to more accurately depict the topographic features, that is, when |Z(P1)-Z(P2)|>ΔZmax, N is added. By generating a continuous three-dimensional terrain grid and dynamically adjusting the grid density according to the height difference, the actual topographic features can be more accurately reflected, especially for areas with dramatic changes in terrain. This adaptive grid division method improves the fineness and accuracy of terrain description, providing solid data support for subsequent safety hazard analysis.

[0093] In an optional embodiment, a mountainous overhead line construction project is constructed in an area with complex terrain and dense vegetation, and a high-precision three-dimensional terrain map is needed for construction planning and safety evaluation. Collect multi-angle image data taken by a UAV, and use a geographic information system (GIS) software to assign accurate geographic coordinates (X, Y) to each image to ensure that the image data has spatial positioning capability.

[0094] Use professional image processing software to register all images according to geographic coordinates to ensure that they are accurately overlaid in the same coordinate system. Calculate the overlapping area between each image, and select images with higher overlap ratio for priority processing to improve the quality of image stitching.

[0095] Apply computer vision algorithms to estimate the ground height by analyzing the gray level changes or color differences of the images. For each pixel point, use a pre-trained mapping function f to convert the image intensity into a height value, generating detailed elevation data. Combined with the estimated height information, use a special three-dimensional modeling software to generate a continuous terrain grid. Set the height difference threshold ΔZmax = 5 meters, and for adjacent pixel points with a height difference exceeding this threshold, insert additional grid nodes between them to ensure that the topographic features are accurately depicted.

[0096] In another optional embodiment, a 110kV overhead transmission line construction project is planned for a coastal plain area. The area has a flat terrain but contains multiple micro-topography formed by river alluvial deposits, a high groundwater level, and mostly silty clay soil. Therefore, a precise three-dimensional topographic map is required to accurately identify subtle landform changes and potential soft soil areas, providing a reliable basis for line design and foundation construction.

[0097] A total of 680 images were collected using a dual-spectral camera mounted on a rotary-wing drone at an altitude of 60 meters and a speed of 3 meters per second. RTK-GPS technology was used to assign precise geographic coordinates (G(X, Y)) to each image, with an accuracy of ±2 centimeters. Specifically, for coastal areas, the coordinate system was converted from WGS84 to the local Gauss-Krüger projection to ensure that the image data was consistent with local surveying and mapping benchmarks.

[0098] Based on the precise geographic coordinates obtained, all images were batch-registered. The overlap ratio, R, of adjacent images was calculated. The average overlap ratio of adjacent images was calculated to be 65%, meeting the requirements for stereo photogrammetry. The SIFT feature matching algorithm was used to identify common feature points within the overlapping region, with an average of 850 ± 120 matching points identified for each image pair. The least squares method was used for coordinate transformation during the registration process, with the transformation error controlled within ±0.5 pixels. Depth analysis was performed on the registered image set, and the ground elevation was calculated for each pixel using a stereo matching algorithm combined with the disparity information of the image pairs.

[0099] In this project, the average baseline distance B = 45 meters, the shooting height h = 60 meters, and the parallax accuracy achieved through stereo matching reached 0.2 pixels, resulting in an elevation accuracy of approximately ±8 centimeters. The obtained elevation information was combined to generate an adaptive 3D terrain mesh. The height difference threshold ΔZ_max was set to 0.5 meters (to account for the micro-topography characteristics of the plains).

[0100] On-site GPS-RTK measurements verified that the constructed 3D topographic map had an elevation accuracy of ±12 cm and a horizontal position accuracy of ±8 cm, meeting engineering design requirements. A total of 18 micro-depressions (possible soft soil distribution zones) and 12 micro-uplifts (relatively stable soil areas) were identified.

[0101] In summary, this invention constructs a three-dimensional topographic map covering the entire construction area, clearly displaying complex terrain features, including steep slopes, valleys, and vegetation-covered areas. Based on this map, the team can accurately identify potential safety hazards, optimize construction paths, avoid high-risk areas, and provide a scientific basis for the subsequent deployment of monitoring equipment. In addition, the real-time updated map helps the team adjust construction plans in a timely manner to cope with the ever-changing on-site environment, ensuring the safety and efficiency of the construction process.

[0102] S2. Analyzing the distribution of features using the three-dimensional topographic map, and selecting locations of landform changes based on the analysis results;

[0103] In the embodiment of the present application, in step S2, the step of analyzing the distribution of features using the three-dimensional topographic map includes:

[0104] S2.1. Classify and process the three-dimensional topographic map;

[0105] Specifically, the 3D topographic map generated in step 2 is classified, with each area on the map divided into different categories based on color and texture characteristics. Let C represent the classification result, where each pixel point P is assigned a class label L(P), identifying the surface type to which it belongs. This classification process allows for intuitive distinction between different types of surface cover, such as vegetation, bare soil, water bodies, and artificial structures. This step provides the foundation for subsequent, more detailed analysis, enabling the study of specific types of cover and enhancing the relevance and accuracy of data analysis.

[0106] S2.2. Calculate the first feature of each category, sort the calculated first feature results, and output the region ranked first as the first region;

[0107] In an optional embodiment, the first feature is vegetation coverage, so the first area output is a dense vegetation area. Based on the classification result C, the dense vegetation area V is identified. The vegetation index VI under each category L is calculated using the formula VI = (NIR-R) / (NIR+R), where NIR represents near-infrared reflectance and R represents red light reflectance. The area with a higher VI value is selected as the dense vegetation area V.

[0108] The Vegetation Index (VI) is used to quantify vegetation density. This method effectively highlights areas with abundant vegetation, even in complex and changing environments. Areas with a high VI typically indicate greater plant biomass or denser vegetation cover, which is valuable for ecological research, agricultural monitoring, and environmental protection.

[0109] S2.3. Evaluate a second feature of the first region and obtain a second feature result;

[0110] In this embodiment, the second feature uses moisture content. Using the identified dense vegetation areas V, the potential moisture content W within these areas is estimated. The moisture estimation formula W(P) = g(VI(P)) is employed, where g is a conversion function that infers the corresponding moisture content W based on the vegetation index VI, ensuring detailed analysis of dense vegetation areas. By modeling the relationship between the vegetation index and moisture content, moisture conditions within vegetation areas can be estimated without direct measurement. This non-invasive moisture content assessment method not only improves efficiency but also enables long-term monitoring of moisture trends, which is crucial for water resource management, drought warnings, and vegetation health assessments.

[0111] S2.4. Output the first feature result and the second feature result to the three-dimensional topographic map.

[0112] Detect the location and extent of artificial structures S. On the 3D terrain grid M, compare the elevation changes Z and texture characteristics T of each category L. When significant discontinuities in Z or T are found, mark the location as a potential artificial structure S. This process uses the structure recognition formula S = h(Z, T), where h is a judgment rule that comprehensively considers height and texture differences.

[0113] By combining elevation changes and texture features to detect artificial structures, this method can effectively distinguish traces of human activity, such as buildings, roads, and other infrastructure, from natural landscapes. This can help support decision-making in urban planning, disaster assessment, and monitoring illegal construction in protected areas.

[0114] In this example, an overhead line construction project in a mountainous area requires a detailed analysis of the distribution of ground cover in the area, with particular attention paid to areas of dense vegetation and their moisture content. The locations of all artificial structures must also be identified and recorded to assist in ecological protection and management decisions.

[0115] The three-dimensional topographic map generated in step 2 is automatically classified using a machine learning algorithm. The surface is divided into multiple categories based on the color and texture characteristics of the image, such as forest, grassland, water, bare land, and artificial structures, and a category label L(P) is assigned to each pixel.

[0116] Remote sensing technology is used to calculate the vegetation index (VI) to identify the areas with the densest vegetation within the mountainous area. For those areas with high VI values, further detailed analysis is conducted to understand their specific vegetation composition and distribution patterns.

[0117] For identified dense vegetation areas, a pre-established conversion function g is used to estimate the moisture content W at each pixel based on the vegetation index VI. This step helps managers understand which areas may be at risk of drought or where vegetation growth conditions are more favorable.

[0118] On the three-dimensional terrain grid, by analyzing the elevation change Z and the texture characteristics T, the positions of any significant discontinuities are identified, which can be a sign of the presence of artificial structures S. Once confirmed, these structures are recorded in detail, including their location, size and type, for subsequent management and protection work.

[0119] Through the above analysis, the project team obtains detailed information on the distribution of surface coverings, especially data on densely vegetated areas and artificial structures. These information provides valuable decision support for managers of the country's mountainous areas, for example, in densely vegetated areas, scientific irrigation plans or necessary drought mitigation measures can be developed based on water content assessment results; for artificial structures, precise map updates can be used to strengthen monitoring of illegal buildings and maintain the integrity and original appearance of nature reserves.

[0120] In this embodiment, in step S2, the screening method of the landform change position includes,

[0121] S2.5, evaluate the stability of the first region;

[0122] S2.6, using the stability evaluation result, by calculating the terrain slope, mark the position whose terrain slope exceeds the preset threshold as an abnormal area;

[0123] S2.7, soil compaction evaluation is performed on the first region marked as an abnormal area, and when the soil compaction is lower than the set compaction threshold, it is output as a risk area;

[0124] S2.8, output the landform change position in combination with the risk area and the abnormal area.

[0125] In an optional embodiment, a 220kV overhead transmission line construction project in a mountainous area with an elevation of 500-1200 meters, the terrain is relatively large, the soil layer is relatively thin, and the degree of rock weathering is different, and it is necessary to accurately identify the unstable landforms that may affect the safety of construction.

[0126] For the first region, the determined densely vegetated area and the surface feature change area in this embodiment are used for stability evaluation. A multi-factor comprehensive evaluation model is used:

[0127] 32 first regions are identified, with a total area of about 8.5 square kilometers, and the stability index ranges from 0.32 to 0.85 (0 is the most unstable, and 1 is the most stable). The region with a stability index lower than 0.6 is marked as a key attention area.

[0128] Using a 3D terrain model, the slope of each grid point was calculated: a first-level threshold of 25° corresponds to sparse vegetation; a second-level threshold of 35° corresponds to dense vegetation; and a third-level threshold of 45° corresponds to extremely steep areas. Eighteen outliers with terrain slopes exceeding 25° were identified, including five exceeding 35° and two approaching 45°. The total area of ​​these outliers is approximately 1.2 square kilometers.

[0129] The soil compaction of the 18 marked abnormal areas was assessed using the dynamic penetration test (DPT) method, where N63.5 is the standard penetration number of the standard penetration test (SPT). The compaction thresholds were set as follows: danger threshold: N63.5 < 10 (very soft soil); warning threshold: 10 ≤ N63.5 < 20 (soft soil); and safety threshold: N63.5 ≥ 20 (medium-hard soil).

[0130] Four outlier areas had soil compaction levels below the danger threshold and were designated high-risk areas; seven outlier areas had soil compaction levels within the warning range and were designated medium-risk areas; and seven outlier areas had soil compaction levels that met safety standards. Ultimately, 18 locations of landform changes were identified and managed according to risk level.

[0131] In another optional embodiment, a 500kV overhead transmission line construction project is being carried out in a collapsible loess area in a northern plain. The terrain in this area is relatively flat, but there is a large area of ​​collapsible loess, which is prone to uneven settlement when exposed to water, posing a special threat to construction safety.

[0132] Forty-six first zones were identified, mainly concentrated in paleo-river channels and low-lying areas, with stability indices ranging from 0.15 to 0.75. Areas with stability indices below 0.5 were marked as subsidence-sensitive areas.

[0133] In this embodiment, the characteristics of plain areas are also taken into consideration, and the slope threshold is adjusted. The final marking result is: 31 abnormal areas with a slight slope exceeding 2° are identified; 12 of them exceed 5°, and 3 are close to 8°; the total area of ​​the abnormal areas is about 2.8 square kilometers.

[0134] The soil compaction was evaluated using a bearing capacity test method specifically for collapsible loess: collapsibility coefficient = (h0-h) / h0; where h0 is the original height of the specimen and h is the height of the specimen after immersion in water. The compaction threshold (for collapsibility) was set as follows: high collapsibility: δ≥0.030; moderate collapsibility: 0.015≤δ<0.03; slight collapsibility: δ<0.015.

[0135] Ultimately, eight outlier areas exhibited high collapsibility and were marked as extremely high-risk areas; 15 outlier areas exhibited moderate collapsibility and were marked as high-risk areas; and eight outlier areas exhibited only slight collapsibility and were marked as medium-risk areas. 45 locations of geomorphic changes were output, with a particular focus on the stability changes of collapsible loess under different water conditions.

[0136] In this embodiment, in step S3, the step of evaluating the monitoring results of the landform change location includes:

[0137] S3.1. Conduct preliminary stability analysis on each data point monitored for geomorphic change locations, and construct a stability change trend graph based on the preliminary stability analysis results;

[0138] S3.2. Use the stability change trend chart to identify potential risk points of stability deterioration. When the change rate of a potential risk point is greater than the preset threshold, it is marked as a risk point.

[0139] In one alternative implementation, a mountainous overhead line project traverses multiple slopes with complex geological conditions, presenting 15 potential landslide risk points requiring focused monitoring. The project team employed a comprehensive monitoring system consisting of distributed fiber optic sensors, surface displacement monitors, and inclinometers to continuously monitor surface displacement, tilt angle changes, and soil strain at each risk point. Data collection was set to occur every six hours to ensure timely capture of dynamic landform changes. Over 60 days of continuous monitoring, the system collected a cumulative 36,000 valid data points, providing a robust data foundation for subsequent stability analysis.

[0140] In the S3.1 phase, the project team conducted a preliminary stability analysis of the monitoring data and used a comprehensive assessment model to calculate the stability index of each potential hazard point. This index comprehensively considers the change in surface displacement, inclination angle, and soil strain, and assigns these three parameters an importance of 0.4, 0.3, and 0.3, respectively, through weight coefficients. Based on the monitoring data every 6 hours, the system constructed a dynamic stability change trend chart, used the sliding window mean method to eliminate noise interference, and combined with the trend analysis algorithm to draw the fitting line. Taking into account the possible periodic impact of the day and night temperature difference in mountainous areas, a sine function term was introduced in the trend analysis to improve the prediction accuracy. Through the trend chart analysis, it was found that the stability index of 12 of the 15 potential hazard points remained within the safe range of above 0.7, 2 points showed a clear downward trend, and 1 point fluctuated greatly but was generally stable.

[0141] During the S3.2 phase, the project established a three-level warning system for landslides, targeting the sudden nature of mountain landslides: Level 1 for a rate of change exceeding 0.5 mm / day, Level 2 for a rate exceeding 2.0 mm / day, and Level 3 for a rate exceeding 5.0 mm / day. By analyzing the average rate of change and the acceleration of the rate of change over the past seven days, the system identified potential site #3 as having a surface displacement rate of 6.2 mm / day, marking it as a Level 3 risk site. The tilt angle of potential site #9 was changing at a rate of 2.8° / day, marking it as a Level 2 risk site. The tilt angle of potential site #12 was changing at a rate of 0.8 mm / day, marking it as a Level 1 risk site. Based on these risk assessment results, the project team immediately implemented emergency reinforcement measures at the Level 3 risk sites, including adding anti-slide piles and improving the drainage system. Monitoring at the Level 2 risk sites was increased to every two hours, and manual inspections were increased at the Level 1 risk sites to ensure safe construction.

[0142] In another optional implementation, in a 500kV overhead line construction project in a plain area, the tower foundations faced the risk of uneven settlement due to the large area of ​​collapsible loess in the area. The project team established a long-term monitoring system for the foundations of 28 towers. The monitoring parameters included key indicators such as vertical settlement, horizontal displacement, groundwater level changes, and water content. Considering that collapsible loess is particularly sensitive to rainfall, the system collects data every 12 hours under normal circumstances, and automatically intensifies the data collection to every 3 hours during rainfall to ensure that changes in foundation performance caused by changes in soil moisture content can be accurately captured. During the 120-day continuous monitoring period, the system collected a large amount of monitoring data, including environmental factors, providing a reliable basis for accurately assessing foundation stability.

[0143] During the S3.1 phase, the project team developed a specialized stability assessment model tailored to the unique properties of collapsible loess. This model incorporates three dimensions: settlement control index, moisture sensitivity index, and uniformity index. The settlement control index measures the bearing capacity of foundations by comparing the measured settlement to the allowable settlement. The moisture sensitivity index assesses collapsibility risk by comparing the change in water content to the change in critical water content. The uniformity index determines the compatibility of foundation groups by comparing the difference in settlement between adjacent foundations to the allowable difference. The system assigned weights of 0.5, 0.3, and 0.2 to these three indicators, respectively, to construct a collapsible stability index. To construct the stability trend graph, a multivariate regression analysis was used, incorporating factors such as time, rainfall, temperature, and groundwater level fluctuations into the analysis model. The model achieved a goodness-of-fit of at least 0.85, providing a scientific basis for predicting changes in foundation performance. The analysis results showed that the stability index of 20 foundations remained within the normal range of above 0.8, while five foundations showed a downward trend and three foundations exhibited significant accelerated settlement after rainfall.

[0144] During the S3.2 phase, the project adopted a risk assessment strategy distinct from that used for landslide monitoring in mountainous areas, taking into account the long-term, progressive deformation characteristics of collapsible loess. A threshold system based on average monthly settlement velocity was established: less than 1 mm / month during the normal period, 1-5 mm / month during the alert period, and greater than 5 mm / month during the danger period. Concern thresholds were also set for changes in water content exceeding 3% and groundwater levels exceeding 1 meter. Using a multi-factor comprehensive evaluation approach, the system assessed the risk level of each foundation, taking into account settlement velocity, moisture content change, and compatibility with adjacent foundations. Ultimately, three foundations were identified as extremely high-risk: Foundation #7 (monthly settlement of 7.8 mm and a 4.2% increase in water content), Foundation #15 (monthly settlement of 6.1 mm and a 1.8-meter rise in the groundwater level), and Foundation #23 (monthly settlement of 8.5 mm and a 12-mm difference in settlement with adjacent foundations). For these risky foundations, the project team adopted a graded treatment strategy: grouting reinforcement and drainage and dehumidification measures were immediately implemented for extremely high-risk foundations, and the monitoring frequency was increased to 4 times a day; a phased reinforcement plan was formulated for high-risk foundations and additional observation points were added; and for medium-risk foundations, intensified observations were maintained and a rainfall warning response mechanism was established to ensure the long-term stable operation of the entire line foundation.

[0145] This embodiment also includes the following steps:

[0146] Set the evaluation cycle, calculate the current time and evaluation cycle, and determine whether to start the next round of evaluation;

[0147] Based on the evaluation results, the imaging data is updated.

[0148] In this embodiment, during the construction process, steps 1 to 7 are repeatedly executed according to a predetermined period TC. A periodic evaluation function EA(T) = a(T, TC) is defined, where a is a rule that determines whether to start a new round of evaluation based on the current time T and the evaluation period TC;

[0149] By setting an evaluation cycle TC and applying the periodic evaluation function EA(T), we can ensure regular identification of potential safety hazards and update of response strategies during the construction process. This periodic evaluation mechanism increases sensitivity to changes on the construction site, enabling potential risks to be discovered and addressed promptly, thereby enhancing the safety and controllability of the construction process.

[0150] Here's what a means and how it works:

[0151] The current time T is the specific point in time at which the evaluation is being conducted, typically a timestamp or date. The evaluation period TC is a pre-defined interval that indicates how often a new evaluation should be conducted. For example, this could be daily, weekly, or monthly.

[0152] Rule a checks whether the current time T matches the time of the most recent assessment plus a complete assessment cycle TC. If so, a new round of assessment should be initiated. In addition to simple cycle matching, a can also consider other trigger conditions, such as the occurrence of specific events (for example, a small landslide) or changes in certain key indicators (such as monitoring data showing a significant deterioration trend at a risk point). These situations may also prompt the early initiation of a new round of assessment. In order to accurately manage the assessment cycle, a may need to maintain an internal state machine or logging system to track the time of the last assessment and the current assessment stage.

[0153] Ultimately, rule a outputs an EA(T) indicating whether to initiate a new round of evaluation. This is typically a Boolean value (true / false) or a more complex signal, such as a specific evaluation instruction or task list. If the output is true, a new round of evaluation should be initiated immediately; if false, the process waits until the next cycle point or until other triggering conditions are met.

[0154] Example 3 is the third embodiment of the present invention, which is different from the previous embodiment in that:

[0155] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0156] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0157] More specific examples (a non-exhaustive list) of the computer-readable medium include the following: an electrical connection (electronic) having one or more wires, a portable computer diskette (magnetic), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer readable medium can be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, for example via an optical scanner, then compiled, interpreted, or otherwise processed, using a suitable medium, into a computer readable medium.

[0158] It should be understood that aspects of the application can be implemented in hardware, software, firmware or combinations thereof. In the above embodiments, various steps or methods can be implemented in software or firmware which is stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any of the following technologies, known in the art, or combinations thereof, can be used: a discrete logic circuit having logic gates for implementing logic functions upon data signals, an application specific integrated circuit having appropriate combinational logic gates, a programmable gate array (PGA), a field programmable gate array (FPGA), and the like.

[0159] Embodiment 4, which is a fourth embodiment of the present application, provides an overhead line construction safety hidden danger point positioning system, comprising a three-dimensional topographic map construction module, a feature analysis module, and a risk point marking module;

[0160] The three-dimensional topographic map construction module is responsible for collecting images of a predetermined construction area, obtaining relevant image data, and constructing a three-dimensional topographic map based on the image data;

[0161] The feature analysis module analyzes the feature distribution condition using the three-dimensional topographic map, and filters out positions with changes in topography according to the obtained analysis result, so as to determine the range of the area with safety hazards;

[0162] The risk point marking module collects monitoring data and evaluates the monitoring result, and marks specific risk points according to the evaluation conclusion.

[0163] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A method for locating safety hazards in overhead line construction, characterized by: The following steps are included: Collect images of the planned construction area to obtain image data, and construct a three-dimensional topographic map based on the image data; Analyzing the distribution of features using the three-dimensional topographic map, and screening locations of landform changes based on the analysis results; Deploy ground monitoring equipment at the location of the landform change, evaluate the monitoring results, and mark risk points based on the evaluation results.

2. A method for locating safety hazards in overhead line construction according to claim 1, characterized in that: The method for obtaining the image data includes: The image acquisition device is mounted on an aerial platform and moved along a preset path; The image acquisition device presets a time interval, captures images according to the time interval, and arranges the images according to the time sequence of the capture to obtain an image sequence; Using the obtained image sequence, the displacement value between every two adjacent images is calculated; By comparing the displacement value with the preset threshold, the image that meets the comparison requirements is output as image data.

3. A method for locating safety hazards in overhead line construction according to claim 2, characterized in that: The three-dimensional topographic map is constructed in the following manner: Pre-process the collected image data to obtain the geographic coordinates corresponding to each image; All images are registered using geographic coordinates, and the overlapping area of ​​each image is calculated to obtain a registration set map; Analyze the parameters in each image through the registration set graph, and obtain the image height value based on the analysis results; The image height values ​​are combined with the registration set map to generate a three-dimensional topographic map.

4. A method for locating safety hazards in overhead line construction according to claim 3, characterized in that: The step of analyzing the distribution of features through the three-dimensional topographic map includes: Classify and process the three-dimensional topographic map; By calculating the first feature under each category and sorting the calculated first feature results, the area ranked first is output as the first area; evaluating a second characteristic of the first region and obtaining a second characteristic result; The first feature result and the second feature result are outputted onto a three-dimensional terrain map.

5. A method for locating safety hazards in overhead line construction according to claim 4, characterized in that: The method for screening the location of landform changes includes: Assess the stability of the first area; Using the stability assessment results, the terrain inclination is calculated and the locations where the terrain inclination exceeds the preset threshold are marked as abnormal areas; The soil compaction of the first area marked as an abnormal area is evaluated. When the soil compaction is lower than the set compaction threshold, it is output as a risk area. The location of landform changes is output by combining risk areas and abnormal areas.

6. A method for locating safety hazards in overhead line construction according to claim 5, characterized in that: The steps for evaluating the monitoring results of the location of landform changes include: Conduct preliminary stability analysis on each data point monitored for geomorphic changes, and construct a stability change trend graph based on the preliminary stability analysis results; The stability change trend chart is used to identify potential risk points of stability deterioration. When the change rate of a potential risk point is greater than the preset threshold, it is marked as a risk point.

7. A method for locating safety hazards in overhead line construction according to claim 6, characterized in that: The following steps are also included: Set the evaluation cycle, calculate the current time and evaluation cycle, and determine whether to start the next round of evaluation; Based on the evaluation results, the imaging data is updated.

8. A system for locating safety hazard points in overhead line construction, using a method for locating safety hazard points in overhead line construction as claimed in any one of claims 1 to 7, characterized in that: It includes a 3D topographic map construction module, a feature analysis module, and a risk point marking module; The three-dimensional topographic map construction module is responsible for collecting images of the scheduled construction area, obtaining relevant image data, and constructing a three-dimensional topographic map based on these image data; The feature analysis module uses a three-dimensional topographic map to analyze the feature distribution, and based on the analysis results, screens out locations where landform changes occur, thereby determining the scope of areas where potential safety hazards exist; The risk point marking module collects monitoring data and evaluates the monitoring results, and marks specific risk points based on the conclusions drawn from the evaluation.

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 a method for locating safety hazard points in overhead line construction 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 a method for locating safety hazard points in overhead line construction according to any one of claims 1 to 7 are implemented.

Images