Power construction unmanned aerial vehicle path planning method and system

Abstract

The invention discloses a power construction unmanned aerial vehicle path planning method and system, and relates to the technical field of space calculation, and the method comprises the following steps: determining a target region needing power construction and a preset construction point coordinate; obtaining a topographic map of the target area based on the geographic information database, constructing a three-dimensional map model, and marking the preset construction point coordinates; generating an initial construction path of the unmanned aerial vehicle; the method comprises the following steps: acquiring meteorological data in real time through meteorological data of a meteorological station in a target area, establishing a meteorological obstacle three-dimensional model, and dividing a preset radius threshold with a meteorological obstacle as a circle center into non-flying areas according to the severity level of the meteorological obstacle; adjusting the initial construction path based on the non-flying area; the cruise height is adjusted according to the meteorological data and topographic relief, a relief tracking path is generated, and the constant relative height is kept; and combining the battery capacity, the flight speed and the task priority, dynamically distributing inspection road sections, and generating a corresponding power construction unmanned aerial vehicle flight path.

Description

Technical Field

[0001] The present invention relates to the field of spatial computing technology, and in particular to a method and system for planning a path for a power construction drone. Background Art

[0002] Currently, the use of drones for power construction is primarily focused on addressing inspection efficiency and safety challenges in complex environments. Traditional manual inspection methods are limited by difficult terrain, harsh climates, and the risks of working at height, and suffer from drawbacks such as high labor intensity, long time consumption, and numerous blind spots. With the development of wireless communication technology, aerial remote sensing and mapping technology, GPS navigation and positioning technology, and automatic control technology, the newly emerging drone telemetry and mapping technology has effectively solved these problems. However, path planning relies on manual control, making it difficult to adapt to the refined inspection and construction needs of large-scale transmission lines. For example, in the field of engineering line stringing, the most primitive method of manually deploying traction ropes is not only inefficient, but also difficult to complete when crossing special terrain encountered during construction. Paramotor is currently the most commonly used method for deploying traction ropes, but it requires pilot control, carries construction risks, and has poor flight stability. Based on the above situation, the following problems still exist in the path planning of drones for power construction: 1. The drone's ability to maintain its attitude in complex terrain and weather disturbances is insufficient, which can easily lead to trajectory deviation and reduced operational accuracy.

[0003] 2. The real-time perception and avoidance mechanisms for multi-source risks, such as sudden obstacles and weather changes, are not yet mature. There is an urgent need to build an active protection system that integrates environmental prediction and dynamic obstacle avoidance.

[0004] 3. High dependence on manual labor, broken data loops, and lack of task flexibility restrict large-scale applications. Summary of the Invention

[0005] In response to the above-mentioned problems, the present invention provides a method and system for path planning of a power construction UAV to solve the above-mentioned problems.

[0006] A method for planning a path for a UAV in electric power construction, comprising the following steps: Determine the target area where electrical construction needs to be carried out and the coordinates of the preset construction points; Acquiring a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model; Generate the initial construction path of the UAV based on the annotated 3D map model; Real-time wind speed, precipitation, and temperature data are obtained from weather stations in the target area. A three-dimensional model of meteorological obstacles is established. Based on the severity of the meteorological obstacle, a preset radius threshold with the meteorological obstacle as the center is divided into a no-fly zone. Adjusting the initial construction path based on the no-flyable area using an improved artificial potential field method; Dynamically adjust the cruising altitude based on meteorological data and terrain undulations, using the artificial hummingbird algorithm to generate an undulating tracking path to maintain a constant relative altitude; A load balancing algorithm is introduced, combining battery capacity, flight speed and task priority to dynamically allocate inspection sections and generate corresponding flight paths for power construction drones.

[0007] Preferably, the determining of the target area where electric power construction is required and the coordinates of the preset construction points includes: Obtaining the coordinates of specific points where power construction needs to be carried out, and determining the power construction area based on the specific coordinates of the points; Gridding the power construction area, extracting the specific point coordinates in each grid within the power construction area, and taking the area where the average number of points exceeds a preset threshold as the core area; Divide the area where the vertical distance between the specific point coordinates and the core area is less than a preset distance threshold into an available area to form an available area combination; Based on the core area, secondary power construction areas are set according to the available area combination of each core area to form a secondary control network covering the power construction area, and the coordinates of each specific point are located in the secondary control network; Based on the topological structure of the power construction points in the secondary control network and the construction coverage range of each drone, the secondary control network is divided into regions through spatial calculation to form a third-level construction area network that can be fully covered by each drone.

[0008] Preferably, the acquiring of a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model include: Obtain terrain data through satellite remote sensing data and generate point cloud data by combining it with lidar scanning; Obtaining the target area elevation data through a standard map library and superimposing the terrain data to generate a textured three-dimensional terrain surface; rasterizing the elevation data of the target area to generate a triangulated network model, combining the point cloud data to generate a refined three-dimensional structure, and rendering the three-dimensional map model using the textured three-dimensional terrain surface; Mapping the two-dimensional map coordinates of the preset construction point to the three-dimensional map model, assigning an elevation value to the preset construction point, and generating three-dimensional space coordinates; The three-dimensional map model is rendered according to the three-level construction area network to form a three-dimensional map of the construction area division of each drone.

[0009] Preferably, the generating of the initial construction path of the UAV according to the annotated three-dimensional map model includes: Divide the construction area of ​​each drone in the 3D map model into a uniform grid, and mark the flyable space and obstacles using grid attributes; Taking the drone launch point as the coordinate origin, the drone flight constraint parameters are set based on the drone working parameters, and the RRT algorithm is used to traverse all construction points to generate the initial path; A pruning algorithm is used to remove continuous collinear nodes in the initial path, and the initial path is smoothed based on a curve smoothing technology to generate an initial construction path for the UAV.

[0010] Preferably, the method of acquiring wind speed, precipitation, and temperature data in real time through meteorological data from a meteorological station in the target area, establishing a three-dimensional model of meteorological obstacles, and dividing a preset radius threshold with the meteorological obstacle as the center into a no-fly zone according to the severity level of the meteorological obstacle includes: Determine a target weather station site according to the target area, and obtain real-time weather radar data including wind speed, precipitation, and temperature of the target weather station site; Determine the polar coordinate format data of the weather radar, convert the polar coordinates into a rectangular coordinate system, and establish a three-dimensional weather data field; The interference intensity of UAV flight is graded according to meteorological targets. Meteorological targets without interference are classified as safe level. Meteorological targets with interference intensity that makes the UAV work efficiency lower than the rated efficiency are classified as low risk level. Meteorological targets with interference intensity that makes the UAV unable to work are classified as high risk level. The high risk level meteorological targets are defined as meteorological obstacles. Reconstructing the meteorological obstacle in three dimensions, synthesizing two-dimensional data at different heights of the meteorological obstacle to obtain a three-dimensional signal, and combining the spatial coordinates of the three-dimensional signal with the three-dimensional map model to generate a textured three-dimensional environment; The safety radius threshold is adjusted based on the severity level, and a spherical no-fly zone is generated using the center of the meteorological obstacle circle and the dynamic safety radius. The no-fly zone is then polygonized using a contour simplification algorithm. The no-fly zone is marked as a no-fly area in the three-dimensional map model.

[0011] Preferably, the adjusting the initial construction path based on the no-fly zone using the improved artificial potential field method includes: The center of the meteorological obstacle is set as a repulsive force source, and a spherical repulsive force field is generated with the dynamic safety radius; Obtaining a no-fly zone in the three-dimensional map model, and marking the no-fly zone with a red potential energy field, wherein a potential energy value of the potential energy field decreases as the distance from the no-fly zone increases; Calculate the resultant force of the drone in the potential energy field. When the direction of the resultant force deviates from the original path by more than a threshold, path replanning is triggered. The collinear waypoints are deleted to obtain an adjusted UAV flight path, and the adjusted UAV flight path is processed so that a curvature radius of the UAV flight path is greater than a preset value.

[0012] Preferably, the method of dynamically adjusting the cruising altitude according to meteorological data and terrain undulations, using an artificial hummingbird algorithm to generate an undulating tracking path, and maintaining a constant relative altitude includes: Calculate the constant relative altitude that the drone needs to maintain based on the relative height difference of the terrain and the meteorological safety altitude; The relative height difference between the path point and the terrain is used as a heuristic value to calculate the terrain fit weight. When the path point falls into the no-fly zone, a weather risk penalty strategy is implemented to forcibly avoid the no-fly zone. Obtain the adjusted 3D UAV flight path and perform reinforcement learning on the path that successfully avoids obstacles and has minimal altitude fluctuations; The objective function of minimizing the path length and height variance is used as a constraint to optimize the parameters of the artificial hummingbird algorithm results.

[0013] Preferably, the load balancing algorithm is introduced to dynamically allocate inspection sections based on battery capacity, flight speed, and task priority to generate a corresponding flight path for the power construction UAV, including: Importing the three-dimensional map model data and the data of the power facilities that need to be constructed, and marking the no-fly areas; Split the inspection section into independent task segments according to the preset construction point coordinates. Each task segment contains the start and end coordinates and length attributes. Obtaining the battery capacity and flight speed of multiple drones to be launched, and calculating the load rate of each of the drones to be launched based on the real-time data; Using a minimum load priority algorithm, the independent mission segment is matched with the load rate of each of the UAVs to be launched; Dynamically assign independent mission segments to each of the unmanned aerial vehicles to be launched based on the matching results, and generate corresponding flight paths for the power construction unmanned aerial vehicles; The trajectory deviation and energy consumption data of each flight are recorded, and the load balancing parameters of the minimum load priority algorithm are adjusted through reinforcement learning.

[0014] Preferably, the method further includes, based on artificial intelligence visual recognition and multi-source data fusion, the drone automatically detects equipment defects during power inspections and generates a report, judges the equipment defects, and triggers a repair verification process when the judgment result is a directly repairable defect; if the judgment result is a defect requiring manual intervention, generates a maintenance work order and uploads it to the power management system, including: During power construction, drones are equipped with cameras and infrared thermal imagers to obtain visible light imaging data and infrared imaging data on the surface of power facilities; Identify surface damage of the power facility using visible light imaging data, and locate temperature anomalies of the power facility using infrared imaging data; The intelligent recognition system carried by the drone is used to judge equipment defects based on the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area. When the probability of a defect exceeds a preset threshold, the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area are uploaded to the server; Identify equipment defects and generate a report using an AI model using surface damage of the power facility, temperature anomalies of the power facility, and meteorological data of the target area; The AI ​​makes an autonomous decision to determine the equipment defect. If a defect can be repaired by a drone equipped with a robotic arm, it will be considered a directly repairable defect. Otherwise, it will be considered a defect requiring manual intervention. For defects that can be repaired directly, the drone's robotic arm is triggered to perform tightening and cleaning operations based on the defect type, and the laser calibration equipment is used to fine-tune the component position. The drone then immediately conducts a re-inspection and verifies the repair results using an image comparison algorithm. Defects requiring manual intervention automatically generate maintenance work orders containing location coordinates, defect images, and risk levels, and push them to the power management system; After manual maintenance is completed, the drone is called for a second inspection, and the acceptance report is automatically generated after AI review.

[0015] A path planning system for a UAV in electric power construction, characterized by comprising: A determination module is used to determine the target area where power construction needs to be carried out and the coordinates of the preset construction points; A map construction module is used to obtain a topographic map of the target area based on a geographic information database, construct a three-dimensional map model of the target area, and mark the coordinates of the preset construction points on the three-dimensional map model; The path generation module is used to generate the initial construction path of the UAV based on the annotated 3D map model; The weather adjustment module is used to obtain real-time wind speed, precipitation, and temperature data from the target area weather station, establish a three-dimensional model of weather obstacles, and divide the preset radius threshold with the weather obstacle as the center into a no-fly zone based on the severity level of the weather obstacle; A first adjustment module is configured to adjust the initial construction path based on the no-fly zone using an improved artificial potential field method; The second adjustment module is used to dynamically adjust the cruising altitude according to meteorological data and terrain undulations, using an ant colony algorithm to generate an undulation tracking path to maintain a constant relative altitude; The allocation module is used to introduce a load balancing algorithm, combine battery capacity, flight speed and task priority, dynamically allocate inspection sections, and generate corresponding flight paths for power construction drones.

[0016] Through the above technical means, the present invention achieves the following beneficial effects: 1) By improving the control algorithm, the UAV can maintain a stable attitude in complex terrain and weather disturbances, and solve the problem of adaptability to dynamic environments by enhancing the robustness of the control algorithm.

[0017] 2) By integrating multiple sensors with real-time meteorological data, a dynamic no-fly zone model is constructed. Through intelligent obstacle avoidance, risks such as electromagnetic interference and sudden obstacles are avoided in advance, which greatly reduces the risk of manual intervention and operation, while improving system safety.

[0018] 3) A full-process system that relies on route planning, automatic task allocation, and defect identification, combined with AI decision-making, realizes dynamic task allocation, automatic defect diagnosis and closed-loop disposal, while simplifying operational processes and improving safety.

[0019] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0020] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0022] Figure 1 This is a schematic diagram of a path planning method for a UAV in power construction provided by the present invention; Figure 2 This is another schematic diagram of a method for planning a path for a UAV in power construction provided by the present invention; Figure 3 This is a schematic diagram of a path planning system for a power construction drone provided by the present invention. DETAILED DESCRIPTION

[0023] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0024] Currently, the use of drones for power construction is primarily focused on addressing inspection efficiency and safety challenges in complex environments. Traditional manual inspection methods are limited by difficult terrain, harsh climates, and the risks of working at height, and suffer from drawbacks such as high labor intensity, long time consumption, and numerous blind spots. With the development of wireless communication technology, aerial remote sensing and mapping technology, GPS navigation and positioning technology, and automatic control technology, the newly emerging drone telemetry and mapping technology has effectively solved these problems. However, path planning relies on manual control, making it difficult to adapt to the refined inspection and construction needs of large-scale transmission lines. For example, in the field of engineering line stringing, the most primitive method of manually deploying traction ropes is not only inefficient, but also difficult to complete when crossing special terrain encountered during construction. Paramotor is currently the most commonly used method for deploying traction ropes, but it requires pilot control, carries construction risks, and has poor flight stability. Based on the above situation, the following problems still exist in the path planning of drones for power construction: 1. The drone's ability to maintain its attitude in complex terrain and weather disturbances is insufficient, which can easily lead to trajectory deviation and reduced operational accuracy.

[0025] 2. The real-time perception and avoidance mechanisms for multi-source risks, such as sudden obstacles and weather changes, are not yet mature. There is an urgent need to build an active protection system that integrates environmental prediction and dynamic obstacle avoidance.

[0026] 3. High dependence on manual labor, broken data loops, and lack of task flexibility restrict large-scale applications.

[0027] A path planning method for UAVs in power construction, such as Figure 1 As shown, it is characterized in that it includes the following steps: Step S101: Determine the target area where power construction needs to be carried out and the coordinates of the preset construction points; Step S102: obtaining a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model; In some embodiments, this embodiment obtains elevation and surface feature data of the target area based on a geographic information database, constructs a three-dimensional terrain surface with surface texture by fusing satellite remote sensing images with airborne lidar point clouds; generates a three-dimensional map model using a rasterization algorithm and triangulation reconstruction technology, and renders the real environment by combining terrain structural features; dynamically maps the plane coordinates of preset construction points to three-dimensional space through a geographic coordinate system conversion engine, assigns corresponding elevation values ​​to achieve three-dimensional spatial calibration, and forms a topological network of construction benchmark points that fully corresponds to the physical environment; Step S103: generating an initial construction path for the UAV based on the annotated three-dimensional map model; In some embodiments, this embodiment generates an initial construction path for a drone based on a labeled three-dimensional map model. First, based on the coordinates of the construction points calibrated in the map, the spatial distribution of obstacles, and the boundary information of the no-fly zone, an intelligent optimization algorithm is used to search for a global optimal solution in the three-dimensional rasterized space. Then, a fitness function is constructed with the dual objective constraints of minimizing the total path length and minimizing the threat avoidance cost. An initial waypoint sequence that satisfies the drone's minimum track segment and maximum steering angle constraints is iteratively calculated. Step S104: Real-time wind speed, precipitation, and temperature data are acquired from the target area weather station. A three-dimensional model of the weather obstacle is established. Based on the severity of the weather obstacle, a preset radius threshold with the weather obstacle as the center is divided into a no-fly zone. In some embodiments, this embodiment collects multi-dimensional data streams such as wind speed, precipitation, and temperature in real time based on a multi-source meteorological sensor network deployed in the target area. A three-dimensional meteorological obstacle model with an intensity gradient is constructed using a spatial interpolation algorithm for meteorological elements. Risk levels are dynamically divided according to a meteorological disaster level assessment system. The spatial location of high-risk meteorological entities, such as severe convective clouds and rainstorm core areas, is used as the geometric center to generate a dynamic no-fly buffer zone whose radius increases exponentially with the disaster level. This buffer zone is superimposed on the terrain obstacle model and marked as a no-fly zone in real time using dynamic geo-fencing technology. This is then synchronously updated to the drone's navigation and obstacle avoidance system. Step S105: using an improved artificial potential field method to adjust the initial construction path based on the no-fly zone; In some embodiments, this embodiment uses an improved artificial potential field method to dynamically optimize and adjust the initial construction path based on a repulsive field model of the no-fly zone. This method updates the gravitational and repulsive field parameters in real time, introduces a fuzzy repulsive control mechanism to accurately quantify the impact of obstacles, and combines a virtual target point strategy to solve the local minimum problem. Step S106: Dynamically adjust the cruising altitude based on meteorological data and terrain undulations, and use the artificial hummingbird algorithm to generate an undulating tracking path to maintain a constant relative altitude; In some embodiments, this embodiment dynamically optimizes the drone's cruising altitude and flight trajectory based on real-time meteorological monitoring data, such as wind speed and precipitation intensity, and high-precision terrain elevation information, using an artificial hummingbird algorithm. This algorithm simulates the three-dimensional foraging behavior of a hummingbird and generates an adaptive path based on the terrain's undulating characteristics. Step S107: Introduce a load balancing algorithm, combine battery capacity, flight speed and task priority, dynamically allocate inspection sections, and generate the corresponding flight path of the power construction drone.

[0028] In some embodiments, this embodiment introduces a multi-objective load balancing algorithm based on the cluster collaborative operation requirements of power construction drones, establishes a battery capacity attenuation model, and dynamically calculates the maximum cruising range by real-time monitoring of the remaining power and flight efficiency of each drone; sets task priority weights based on the urgency of the power grid failure, and implements priority allocation of high-risk sections through a weighted decision matrix.

[0029] The working principle of the above technical solution is: first, determine the target area where power construction is required and the coordinates of the preset construction points; secondly, obtain the topographic map of the target area based on the geographic information database, construct a three-dimensional map model of the target area, and mark the coordinates of the preset construction points on the three-dimensional map model; again, obtain meteorological data in real time through the meteorological data of the meteorological station in the target area, establish a three-dimensional model of meteorological obstacles, and divide the preset radius threshold with the meteorological obstacle as the center into a no-fly area according to the severity level of the meteorological obstacle; adjust the initial construction path based on the no-fly area; adjust the cruising altitude according to the meteorological data and terrain undulations, generate an undulating tracking path, and maintain a constant relative altitude; finally, combine the battery capacity, flight speed and task priority, dynamically allocate inspection sections, and generate the corresponding power construction drone flight path.

[0030] The beneficial effects of the above technical solution are: based on multi-source environmental perception and dynamic path collaborative optimization technology, through three-dimensional space precise modeling, active defense against meteorological risks, terrain adaptive cruise and multi-target resource collaborative decision-making, the full process of drone power construction under complex working conditions can be realized with autonomy and anti-interference capabilities improved, promoting the evolution of drones from execution terminals to intelligent construction nodes with environmental perception, decision-making and execution capabilities.

[0031] In one embodiment, determining the target area where power construction is required and the coordinates of the preset construction points includes: Step S201: Acquire the specific coordinates of the points where power construction needs to be carried out, and determine the power construction area based on the specific coordinates of the points; Step S202: Gridding the power construction area, extracting the specific point coordinates in each grid within the power construction area, and taking the area where the average number of points exceeds a preset threshold as the core area; In some embodiments, this embodiment divides the construction area into several independent grid units according to geographic coordinates and construction stage characteristics to achieve refined zoning management, collects and extracts three-dimensional coordinate data of all key points in each grid in real time, including power facility locations, inspection points, and risk source coordinates, statistically analyzes the average number density of points in each grid, and intelligently identifies areas exceeding a preset threshold as core areas; Step S203: dividing the area where the vertical distance between the specific point coordinates and the core area is less than a preset distance threshold into available areas to form an available area combination; In some embodiments, this embodiment uses grid management of power construction areas to extract three-dimensional spatial data of key points, such as equipment coordinates and inspection points, and calculates the spatial Euclidean distance to the identified core area. Adjacent areas with distance values ​​less than a preset dynamic threshold (reflecting the priority of the operation) are intelligently marked as available areas, and these discrete areas are aggregated through geo-fencing technology to form a structured available area combination. Step S204: Based on the core area, secondary power construction areas are set according to the available area combination of each core area to form a secondary control network covering the power construction area, and each of the specific point coordinates is located within the secondary control network; In some embodiments, this embodiment is based on a grid management framework for power construction areas, takes the identified core area as a benchmark, and intelligently combines adjacent available areas through spatial topology analysis to form secondary power construction area units; Step S205: Based on the topological structure of the power construction points in the secondary control network and the construction coverage range of each drone, the secondary control network is divided into regions through spatial calculation to form a third-level construction area network that can be fully covered by each drone.

[0032] In some embodiments, this embodiment is based on the secondary control network of the power construction area and the spatial topological relationship of the construction points within the network, combined with the effective construction coverage radius and operating capacity boundary of each drone, the secondary control network is adaptively rasterized according to the density of construction points, and the affiliation of the construction points at the dividing boundary is redistributed to ensure that high-priority task points fall into the core coverage area.

[0033] The beneficial effects of the above technical solution are: based on grid processing to extract the density characteristics of construction points, intelligently identify the core area, build a secondary control network covering the entire area, form a spatial reference framework, and provide accurate geographic reference for drone path planning; dynamically delineate the available area combination through the radiation range of the core area, and reconstruct the construction area topology in real time based on variables such as sudden weather changes and terrain undulations to ensure the robustness of the construction network in complex environments; based on the spatial topological structure of the secondary control network and the drone construction coverage parameters, automatically generate a three-level construction area network through spatial calculation to achieve load balancing of drone cluster tasks, transform discrete construction points into a hierarchical spatial topology model, and promote the evolution of drones from single-point operations to group intelligent decision-making; achieve the optimal matching of construction resources and spatial requirements through a three-level network architecture.

[0034] In one embodiment, the obtaining of a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model include: Obtain terrain data through satellite remote sensing data and generate point cloud data by combining it with lidar scanning; Obtaining the target area elevation data through a standard map library and superimposing the terrain data to generate a textured three-dimensional terrain surface; rasterizing the elevation data of the target area to generate a triangulated network model, combining the point cloud data to generate a refined three-dimensional structure, and rendering the three-dimensional map model using the textured three-dimensional terrain surface; In some embodiments, this embodiment performs rasterization processing based on the elevation dataset of the target area to construct an initial digital elevation model. The raster units are then converted into an irregular triangulated network model to accurately express the terrain undulation characteristics. The high-density point cloud data acquired by the lidar is integrated to optimize the distribution of triangulated network vertices through point cloud registration and surface reconstruction algorithms to generate a refined three-dimensional structure that integrates surface details. Mapping the two-dimensional map coordinates of the preset construction point to the three-dimensional map model, assigning an elevation value to the preset construction point, and generating three-dimensional space coordinates; In some embodiments, this embodiment implements the spatial elevation assignment of construction points through two-dimensional-three-dimensional coordinate system mapping technology based on the digital management requirements of the power construction area; The three-dimensional map model is rendered according to the three-level construction area network to form a three-dimensional map of the construction area division of each drone.

[0035] The beneficial effects of the above technical solution are: constructing a textured high-precision three-dimensional terrain surface based on satellite remote sensing and lidar point cloud data, and forming a visual geographic base consistent with the real environment through rasterization processing and triangulation modeling; dynamically mapping the two-dimensional coordinates of preset construction points and assigning elevation values ​​to the three-dimensional model to generate a dynamically rendered spatial coordinate system; relying on the three-level construction area network to adaptively render the three-dimensional map, forming a three-dimensional construction partition that accurately matches the drone cluster mission, thereby realizing the coordinated scheduling of construction resources and seamless coverage of spatial resources under complex terrain conditions.

[0036] In one embodiment, generating an initial construction path for the drone based on the annotated three-dimensional map model includes: Divide the construction area of ​​each drone in the 3D map model into a uniform grid, and mark the flyable space and obstacles using grid attributes; Taking the drone launch point as the coordinate origin, the drone flight constraint parameters are set based on the drone working parameters, and the RRT algorithm is used to traverse all construction points to generate the initial path; A pruning algorithm is used to remove continuous collinear nodes in the initial path, and the initial path is smoothed based on a curve smoothing technology to generate an initial construction path for the UAV.

[0037] In some embodiments, this embodiment is based on the three-level construction area network of the power construction area, and uses multi-channel texture mapping technology and physical lighting model to render the three-dimensional map model; through the spatial index mechanism, each three-level grid unit is associated with the corresponding drone to generate a highly visual three-dimensional map of the construction area division, supporting the precise path planning and obstacle avoidance decisions of the drone cluster.

[0038] The beneficial effects of the above technical solution are: dividing the three-dimensional construction area into a uniform grid and marking the flyable space and obstacle attributes to construct a refined navigation base; integrating the flight constraint parameters with the launch point as the origin to drive the RRT algorithm to traverse the construction points to generate a fully covered initial path; using the pruning algorithm to remove redundant nodes to eliminate path collinearity redundancy, and forming an optimal construction path with continuous curvature through curve smoothing, thereby ensuring the safety of obstacle avoidance while achieving a coordinated leap in flight efficiency and trajectory smoothness, providing a high-precision reference path for subsequent dynamic tuning.

[0039] In one embodiment, the method of acquiring wind speed, precipitation, and temperature data in real time through meteorological data from a meteorological station in the target area, establishing a three-dimensional model of meteorological obstacles, and dividing a predetermined radius threshold centered on the meteorological obstacle into a no-fly zone based on the severity level of the meteorological obstacle includes: Determine a target weather station site according to the target area, and obtain real-time weather radar data including wind speed, precipitation, and temperature of the target weather station site; Determine the polar coordinate format data of the weather radar, convert the polar coordinates into a rectangular coordinate system, and establish a three-dimensional weather data field; The interference intensity of UAV flight is graded according to meteorological targets. Meteorological targets without interference are classified as safe level. Meteorological targets with interference intensity that makes the UAV work efficiency lower than the rated efficiency are classified as low risk level. Meteorological targets with interference intensity that makes the UAV unable to work are classified as high risk level. The high risk level meteorological targets are defined as meteorological obstacles. In some embodiments, the meteorological risk classification standard of this embodiment is as follows: safe level: meteorological targets (wind speed ≤ level 5, no precipitation, visibility > 5km) have no significant interference with flight, and the UAV can maintain rated working efficiency and control accuracy; low risk level: meteorological targets (wind speed 6-7, light to moderate precipitation, visibility 1-5km) cause aerodynamic stability degradation, battery life degradation or positioning accuracy deviation, but basic flight can still be maintained; high risk level: meteorological targets (strong wind ≥ level 8, heavy rain / thunderstorm, visibility < 1km) cause communication interruption, sensor failure or body structure damage, resulting in the complete loss of operational capability of the UAV, which is defined as a meteorological obstacle and needs to be actively avoided in route planning; Reconstructing the meteorological obstacle in three dimensions, synthesizing two-dimensional data at different heights of the meteorological obstacle to obtain a three-dimensional signal, and combining the spatial coordinates of the three-dimensional signal with the three-dimensional map model to generate a textured three-dimensional environment; In some embodiments, this embodiment integrates vertical profile information of radar-based data to generate continuous three-dimensional isosurfaces, constructs a geometric skeleton of the meteorological obstacle, and aligns the spatial coordinates of the reconstructed three-dimensional point cloud with a preset three-dimensional map model; The safety radius threshold is adjusted based on the severity level, and a spherical no-fly zone is generated using the center of the meteorological obstacle circle and the dynamic safety radius. The no-fly zone is then polygonized using a contour simplification algorithm. In some embodiments, this embodiment generates a spherical restricted area based on the risk level of meteorological obstacles, using the centroid of the 3D reconstruction of the meteorological target as the center of the spherical no-fly zone and combining it with a real-time evolving dynamic safety radius. The spherical projection is polygonized using a contour simplification algorithm, ultimately outputting a convex polygonal no-fly zone vector boundary with a controllable number of edges. The no-fly zone is marked as a no-fly area in the three-dimensional map model.

[0040] The beneficial effects of the above technical solution are as follows: specifically, a three-dimensional meteorological field is constructed based on real-time data from meteorological radars and a spatial visualization data field is formed through polar coordinate conversion, a graded warning mechanism is established based on the intensity of meteorological interference to accurately define high-risk obstacles; meteorological obstacles are reconstructed in multiple dimensions and integrated into a three-dimensional geographic model to generate a textured risk environment map; the safety radius threshold is dynamically adjusted according to the severity level to generate an adaptive spherical no-fly zone and polygon calibration is achieved through a contour simplification algorithm, and non-flyable areas are marked in real time on the three-dimensional map, enabling drones to autonomously avoid risk areas under extreme weather conditions such as storms and thunderstorms.

[0041] In one embodiment, the adjusting the initial construction path based on the no-fly zone using the improved artificial potential field method includes: The center of the meteorological obstacle is set as a repulsive force source, and a spherical repulsive force field is generated with the dynamic safety radius; Obtaining a no-fly zone in the three-dimensional map model, and marking the no-fly zone with a red potential energy field, wherein a potential energy value of the potential energy field decreases as the distance from the no-fly zone increases; In some embodiments, this embodiment uses a gradient color field mechanism to achieve dynamic risk quantification based on no-fly zones in a three-dimensional map model. With the geometric center of the no-fly zone as the potential energy peak point, a gradient potential energy field radiating outward is constructed using a nonlinear attenuation function, forming a spatial risk thermal distribution that changes from deep red in the center to transparent outward. The field intensity decays exponentially with increasing Euclidean distance between the location point and the threat source, providing the drone path planning algorithm with an intuitive basis for autonomously generating spatial risk perception barriers and avoidance strategies. Calculate the resultant force of the drone in the potential energy field. When the direction of the resultant force deviates from the original path by more than a threshold, path replanning is triggered. The collinear waypoints are deleted to obtain an adjusted UAV flight path, and the adjusted UAV flight path is processed so that a curvature radius of the UAV flight path is greater than a preset value.

[0042] In some embodiments, this embodiment is based on the demand for optimizing the flight path of the UAV. After deleting collinear redundant waypoints to streamline the path, the track is smoothly reconstructed: a parametric curve fitting method is used to interpolate and correct the path curvature mutation area while retaining the topological relationship of key waypoints, forcing the curvature radius of each point on the path to be greater than the preset threshold of the UAV's minimum turning radius, while maintaining the path's continuous divisibility, eliminating heading angle jumps, and ensuring flight stability and controllability.

[0043] The beneficial effects of the above technical solution are: constructing a spherical dynamic repulsive field with the center of the meteorological obstacle as the repulsive source, and generating an adaptive no-fly barrier in combination with the dynamic safety radius; accurately marking the non-flyable area through the red potential field and establishing a potential gradient attenuation model, so that the UAV can perceive the intensity of meteorological risks in real time in three-dimensional space; realizing intelligent diagnosis and autonomous replanning of path abnormalities based on the resultant force deviation threshold trigger mechanism, effectively avoiding the local minimum trap of the traditional artificial potential field method; and strengthening the dual-track optimization through waypoint collinearity elimination and curvature constraint to generate a smooth flyable path that conforms to the dynamic characteristics of the UAV, thereby ensuring construction continuity and trajectory stability in a strong meteorological interference environment.

[0044] In one embodiment, dynamically adjusting the cruising altitude based on meteorological data and terrain undulations, using an artificial hummingbird algorithm to generate an undulating tracking path, and maintaining a constant relative altitude includes: Calculate the constant relative altitude that the drone needs to maintain based on the relative height difference of the terrain and the meteorological safety altitude; The relative height difference between the path point and the terrain is used as a heuristic value to calculate the terrain fit weight. When the path point falls into the no-fly zone, a weather risk penalty strategy is implemented to forcibly avoid the no-fly zone. In some embodiments, this embodiment is based on the terrain adaptation mechanism of the UAV's three-dimensional path planning. When calculating the terrain fit weight of the path point, the elevation deviation between the path point and the surface is used as the core heuristic value to quantify the matching degree between the flight altitude and the terrain. When the path point falls into the meteorological obstacle no-fly zone or the static no-fly zone, the meteorological risk penalty strategy is triggered, and ultimately the autonomous flight outside the safety boundary is achieved through path replanning. Obtain the adjusted 3D UAV flight path and perform reinforcement learning on the path that successfully avoids obstacles and has minimal altitude fluctuations; The objective function of minimizing the path length and height variance is used as a constraint to optimize the parameters of the artificial hummingbird algorithm results.

[0045] In some embodiments, in the drone path optimization framework of the artificial hummingbird algorithm (AHA), this embodiment takes path length minimization as the core optimization goal, and incorporates flight altitude variance as a key constraint into the algorithm parameter tuning process. The total track length is globally optimized through the spiral flight model of AHA, and the altitude variance is forced to be lower than the preset threshold to ensure flight stability; combined with the adaptive weight adjustment mechanism, the optimization priority of path length and flight altitude is dynamically balanced, and finally a low-energy, high-smoothness three-dimensional track that meets dynamic constraints is generated.

[0046] The beneficial effects of the above technical solution are: intelligent calculation of a constant relative height benchmark based on terrain undulations and meteorological safety thresholds, and conversion of the terrain fit of path points into heuristic weights driven by the artificial hummingbird algorithm to generate undulation tracking paths; forced avoidance of no-fly zones through a meteorological risk penalty mechanism to ensure absolute obstacle avoidance reliability under extreme weather conditions; autonomous strategy optimization of paths with successful obstacle avoidance and minimal altitude fluctuations based on reinforcement learning, and global optimization of algorithm parameters with dual-objective constraints of path length and altitude variance, thereby achieving coordinated control of terrain following and altitude stability in traditional flight restricted areas such as rugged mountainous areas and strong convective areas.

[0047] In one embodiment, the load balancing algorithm is introduced to dynamically allocate inspection sections based on battery capacity, flight speed, and task priority, and generate corresponding flight paths for power construction drones, including: Importing the three-dimensional map model data and the data of the power facilities that need to be constructed, and marking the no-fly areas; Split the inspection section into independent task segments according to the preset construction point coordinates. Each task segment contains the start and end coordinates and length attributes. Obtaining the battery capacity and flight speed of multiple drones to be launched, and calculating the load rate of each of the drones to be launched based on the real-time data; In some embodiments, this embodiment is based on the requirements of multi-UAV collaborative task scheduling, and requires real-time collection of key parameters such as battery capacity, flight speed, and maximum load capacity of each UAV to be launched, and a dynamic load rate calculation model is constructed in combination with a preset dynamic constraint model; Using a minimum load priority algorithm, the independent mission segment is matched with the load rate of each of the UAVs to be launched; In some embodiments, this embodiment uses a task allocation mechanism based on the minimum load first algorithm. When matching independent task segments with the drones to be dispatched, the drone with the lowest current dynamic load rate is preferentially selected as the task execution carrier. When a new task arrives, the scheduling system automatically assigns the task segment to the available drone node with the lowest load rate based on the load rate sorting result. Dynamically assign independent mission segments to each of the unmanned aerial vehicles to be launched based on the matching results, and generate corresponding flight paths for the power construction unmanned aerial vehicles; The trajectory deviation and energy consumption data of each flight are recorded, and the load balancing parameters of the minimum load priority algorithm are adjusted through reinforcement learning.

[0048] In some embodiments, this embodiment is based on a closed-loop optimization mechanism for the execution of drone cluster tasks, and uses airborne sensors and ground stations to collaboratively record the three-dimensional trajectory tracking error and equivalent energy consumption of each flight to construct a dynamic evaluation system for the load balancing algorithm.

[0049] The beneficial effects of the above technical solution are: building a spatial operation model with no-fly markings based on three-dimensional maps and power facility data, and decomposing the inspection section into parameterized independent task segments to realize the encapsulation of construction elements; integrating battery capacity and flight speed to calculate the drone load rate in real time, and matching the task segments with drone resources through the minimum load priority algorithm; combining the path generation algorithm to convert the matching results into a flight path with trajectory constraints, and recording the flight process data to continuously optimize the load parameters, thereby realizing cluster task scheduling, maximizing the use of flight time and rapid response to high-priority tasks in large-scale construction scenarios, and improving the accuracy and safety of the system.

[0050] In one embodiment, the method further includes, based on artificial intelligence visual recognition and multi-source data fusion, automatically detecting equipment defects and generating reports by drones during power inspections, determining the equipment defects, and triggering a repair verification process if the determination result is a directly repairable defect; and generating a repair work order and uploading it to the power management system if the determination result is a defect requiring manual intervention, including: During power construction, drones are equipped with cameras and infrared thermal imagers to obtain visible light imaging data and infrared imaging data on the surface of power facilities; Identify surface damage of the power facility using visible light imaging data, and locate temperature anomalies of the power facility using infrared imaging data; The intelligent recognition system carried by the drone is used to judge equipment defects based on the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area. When the probability of a defect exceeds a preset threshold, the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area are uploaded to the server; In some embodiments, this embodiment uses a high-definition visible light imaging system carried by a drone to detect surface damage to power facilities (such as broken insulators and broken conductors), and combines it with an infrared thermal imaging system to identify temperature anomalies (such as localized overheating of equipment and discharge traces). Simultaneously, it integrates meteorological data from the target area (such as wind speed, humidity, and precipitation) to perform multi-source data collaborative analysis. The onboard intelligent recognition system calculates the probability of equipment defects in real time. When the probability exceeds a preset threshold, the surface damage image, temperature heat map, and meteorological parameters are automatically uploaded to a cloud server to trigger fault warnings, generate operation and maintenance reports, and optimize subsequent inspection strategies. Identify equipment defects and generate a report using an AI model using surface damage of the power facility, temperature anomalies of the power facility, and meteorological data of the target area; The AI ​​makes an autonomous decision to determine the equipment defect. If a defect can be repaired by a drone equipped with a robotic arm, it will be considered a directly repairable defect. Otherwise, it will be considered a defect requiring manual intervention. For defects that can be repaired directly, the drone's robotic arm is triggered to perform tightening and cleaning operations based on the defect type, and the laser calibration equipment is used to fine-tune the component position. The drone then immediately conducts a re-inspection and verifies the repair results using an image comparison algorithm. In some embodiments, this embodiment automatically triggers an onboard robotic arm to perform refined repair operations based on the defect type identified by drone inspections. Once the repair is complete, a 3D scanning re-inspection is immediately initiated. A high-frame-rate visible light / infrared dual-mode imaging system is used to collect data on the repair area, quantifying and comparing feature differences before and after the repair. When the repair similarity reaches a threshold, the repair is deemed qualified. Defects requiring manual intervention automatically generate maintenance work orders containing location coordinates, defect images, and risk levels, and push them to the power management system; After manual maintenance is completed, the drone is called for a second inspection, and the acceptance report is automatically generated after AI review.

[0051] In some embodiments, this embodiment is based on a closed-loop management mechanism after the completion of manual maintenance work, and the system automatically triggers the drone's secondary refined inspection process: the drone equipped with high-definition visible light and infrared dual-mode cameras accurately re-inspects the maintenance points according to preset coordinates, and obtains complete data of the maintenance area through multi-angle image acquisition and three-dimensional point cloud scanning. The AI ​​review engine is started simultaneously, and the texture comparison algorithm is used to analyze the structural consistency of the images before and after maintenance. The infrared thermal map is combined to verify the elimination status of the temperature anomaly points, and the defect repair confidence score is output.

[0052] The beneficial effects of the above technical solution are: integrating visible light, infrared imaging and meteorological data to build a multi-dimensional defect perception system, realizing accurate correlation diagnosis of damage and temperature anomalies and generating structured reports through AI models; classifying defects into two categories based on the intelligent decision-making mechanism of repair feasibility: self-repairable and requiring manual intervention, triggering differentiated handling processes, for self-repairable defects, the onboard robotic arm is mobilized in real time to perform tightening, cleaning or laser calibration operations, and the repair effect is immediately verified through the image comparison algorithm; for those requiring manual intervention, a maintenance work order with spatial coordinates and risk maps is automatically generated and pushed to the management system, and a digital acceptance report is generated through drone re-inspection and AI review after manual handling, significantly reducing operational risks and maintenance costs, and taking into account both efficiency and safety.

[0053] A path planning system for UAVs in power construction, such as Figure 3 As shown, it is characterized by comprising: Determination module 101, for determining the target area where power construction needs to be carried out and the coordinates of the preset construction points; A map construction module 102 is configured to obtain a topographic map of the target area based on a geographic information database, construct a three-dimensional map model of the target area, and mark the coordinates of the preset construction points on the three-dimensional map model; The path generation module 103 is used to generate the initial construction path of the UAV based on the annotated three-dimensional map model; The weather adjustment module 104 is used to obtain real-time wind speed, precipitation, and temperature data from the target area weather station, establish a three-dimensional model of weather obstacles, and divide the area into a no-fly zone based on the severity level of the weather obstacle and the preset radius threshold with the weather obstacle as the center; A first adjustment module 105 is configured to adjust the initial construction path based on the no-fly zone using an improved artificial potential field method; The second adjustment module 106 is used to dynamically adjust the cruising altitude according to meteorological data and terrain fluctuations, and to generate a fluctuation tracking path using an ant colony algorithm to maintain a constant relative altitude; The allocation module 107 is used to introduce a load balancing algorithm, combine battery capacity, flight speed and task priority, dynamically allocate inspection sections, and generate corresponding flight paths for power construction drones.

[0054] Those skilled in the art should understand that the first and second in the present invention simply refer to different application stages.

[0055] Other embodiments of the present disclosure will readily occur to those skilled in the art after considering the specification and practicing the disclosure herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow from the general principles of the present disclosure and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the present disclosure being indicated by the following claims.

[0056] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for planning a path for a UAV in electric power construction, characterized in that: The following steps are involved: Determine the target area where electrical construction needs to be carried out and the coordinates of the preset construction points; Acquiring a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model; Generate the initial construction path of the UAV based on the annotated 3D map model; Real-time wind speed, precipitation, and temperature data are obtained from weather stations in the target area. A three-dimensional model of meteorological obstacles is established. Based on the severity of the meteorological obstacle, a preset radius threshold with the meteorological obstacle as the center is divided into a no-fly zone. Adjusting the initial construction path based on the no-flyable area using an improved artificial potential field method; Dynamically adjust the cruising altitude based on meteorological data and terrain undulations, using the artificial hummingbird algorithm to generate an undulating tracking path to maintain a constant relative altitude; A load balancing algorithm is introduced, combining battery capacity, flight speed and task priority to dynamically allocate inspection sections and generate corresponding flight paths for power construction drones.

2. A method for planning a path for a UAV for electric power construction according to claim 1, characterized in that: Determining the target area where power construction needs to be carried out and the coordinates of the preset construction points includes: Obtaining the coordinates of specific points where power construction needs to be carried out, and determining the power construction area based on the specific coordinates of the points; Gridding the power construction area, extracting the specific point coordinates in each grid within the power construction area, and taking the area where the average number of points exceeds a preset threshold as the core area; Divide the area where the vertical distance between the specific point coordinates and the core area is less than a preset distance threshold into an available area to form an available area combination; Based on the core area, secondary power construction areas are set according to the available area combination of each core area to form a secondary control network covering the power construction area, and the coordinates of each specific point are located in the secondary control network; Based on the topological structure of the power construction points in the secondary control network and the construction coverage range of each drone, the secondary control network is divided into regions through spatial calculation to form a third-level construction area network that can be fully covered by each drone.

3. A method for planning a path for a power construction drone according to claim 1, characterized in that: The step of acquiring a topographic map of the target area based on a geographic information database, constructing a three-dimensional map model of the target area, and marking the coordinates of the preset construction points on the three-dimensional map model includes: Obtain terrain data through satellite remote sensing data and generate point cloud data by combining it with lidar scanning; Obtaining the target area elevation data through a standard map library and superimposing the terrain data to generate a textured three-dimensional terrain surface; rasterizing the elevation data of the target area to generate a triangulated network model, combining the point cloud data to generate a refined three-dimensional structure, and rendering the three-dimensional map model using the textured three-dimensional terrain surface; Mapping the two-dimensional map coordinates of the preset construction point to the three-dimensional map model, assigning an elevation value to the preset construction point, and generating three-dimensional space coordinates; The three-dimensional map model is rendered according to the three-level construction area network to form a three-dimensional map of the construction area division of each drone.

4. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: Generating the initial construction path of the UAV based on the annotated three-dimensional map model includes: Divide the construction area of ​​each drone in the 3D map model into a uniform grid, and mark the flyable space and obstacles using grid attributes; Taking the drone launch point as the coordinate origin, the drone flight constraint parameters are set based on the drone working parameters, and the RRT algorithm is used to traverse all construction points to generate the initial path; A pruning algorithm is used to remove continuous collinear nodes in the initial path, and the initial path is smoothed based on a curve smoothing technology to generate an initial construction path for the UAV.

5. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: The method involves acquiring wind speed, precipitation, and temperature data in real time from the target area's weather stations, establishing a three-dimensional model of meteorological obstacles, and dividing a no-fly zone into a predetermined radius threshold with the meteorological obstacle as the center according to the severity of the meteorological obstacle. The method includes: Determine a target weather station site according to the target area, and obtain real-time weather radar data including wind speed, precipitation, and temperature of the target weather station site; Determine the polar coordinate format data of the weather radar, convert the polar coordinates into a rectangular coordinate system, and establish a three-dimensional weather data field; The interference intensity of UAV flight is graded according to meteorological targets. Meteorological targets without interference are classified as safe level. Meteorological targets with interference intensity that makes the UAV work efficiency lower than the rated efficiency are classified as low risk level. Meteorological targets with interference intensity that makes the UAV unable to work are classified as high risk level. The high risk level meteorological targets are defined as meteorological obstacles. Reconstructing the meteorological obstacle in three dimensions, synthesizing two-dimensional data at different heights of the meteorological obstacle to obtain a three-dimensional signal, and combining the spatial coordinates of the three-dimensional signal with the three-dimensional map model to generate a textured three-dimensional environment; The safety radius threshold is adjusted based on the severity level, and a spherical no-fly zone is generated using the center of the meteorological obstacle circle and the dynamic safety radius. The no-fly zone is then polygonized using a contour simplification algorithm. The no-fly zone is marked as a no-fly area in the three-dimensional map model.

6. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: The adjusting the initial construction path based on the no-flyable area using the improved artificial potential field method includes: The center of the meteorological obstacle is set as a repulsive force source, and a spherical repulsive force field is generated with the dynamic safety radius; Obtaining a no-fly zone in the three-dimensional map model, and marking the no-fly zone with a red potential energy field, wherein a potential energy value of the potential energy field decreases as the distance from the no-fly zone increases; Calculate the resultant force of the drone in the potential energy field. When the direction of the resultant force deviates from the original path by more than a threshold, path replanning is triggered. The collinear waypoints are deleted to obtain an adjusted UAV flight path, and the adjusted UAV flight path is processed so that a curvature radius of the UAV flight path is greater than a preset value.

7. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: The method of dynamically adjusting the cruising altitude based on meteorological data and terrain fluctuations, using an artificial hummingbird algorithm to generate an undulating tracking path, and maintaining a constant relative altitude includes: Calculate the constant relative altitude that the drone needs to maintain based on the relative height difference of the terrain and the meteorological safety altitude; The relative height difference between the path point and the terrain is used as a heuristic value to calculate the terrain fit weight. When the path point falls into the no-fly zone, a weather risk penalty strategy is implemented to forcibly avoid the no-fly zone. Obtain the adjusted 3D UAV flight path and perform reinforcement learning on the path that successfully avoids obstacles and has minimal altitude fluctuations; The objective function of minimizing the path length and height variance is used as a constraint to optimize the parameters of the artificial hummingbird algorithm results.

8. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: The aforementioned load balancing algorithm combines battery capacity, flight speed, and task priority to dynamically allocate inspection sections and generate corresponding flight paths for power construction drones, including: Importing the three-dimensional map model data and the data of the power facilities that need to be constructed, and marking the no-fly areas; Split the inspection section into independent task segments according to the preset construction point coordinates. Each task segment contains the start and end coordinates and length attributes. Obtaining the battery capacity and flight speed of multiple drones to be launched, and calculating the load rate of each of the drones to be launched based on the real-time data; Using a minimum load priority algorithm, the independent mission segment is matched with the load rate of each of the UAVs to be launched; Dynamically assign independent mission segments to each of the unmanned aerial vehicles to be launched based on the matching results, and generate corresponding flight paths for the power construction unmanned aerial vehicles; The trajectory deviation and energy consumption data of each flight are recorded, and the load balancing parameters of the minimum load priority algorithm are adjusted through reinforcement learning.

9. A method for planning a path for a UAV in electric power construction according to claim 1, characterized in that: The method further includes, based on artificial intelligence visual recognition and multi-source data fusion, automatically detecting equipment defects and generating reports by drones during power inspections, determining the equipment defects, and triggering a repair verification process if the determination result is a directly repairable defect; and generating a repair work order and uploading it to the power management system if the determination result is a defect requiring manual intervention, including: During power construction, drones are equipped with cameras and infrared thermal imagers to obtain visible light imaging data and infrared imaging data on the surface of power facilities; Identify surface damage of the power facility using visible light imaging data, and locate temperature anomalies of the power facility using infrared imaging data; The intelligent recognition system carried by the drone is used to judge equipment defects based on the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area. When the probability of a defect exceeds a preset threshold, the surface damage of the power facility, the abnormal temperature points of the power facility, and the meteorological data of the target area are uploaded to the server; Identify equipment defects and generate a report using an AI model using surface damage of the power facility, temperature anomalies of the power facility, and meteorological data of the target area; The AI ​​makes an autonomous decision to determine the equipment defect. If a defect can be repaired by a drone equipped with a robotic arm, it will be considered a directly repairable defect. Otherwise, it will be considered a defect requiring manual intervention. For defects that can be repaired directly, the drone's robotic arm is triggered to perform tightening and cleaning operations based on the defect type, and the laser calibration equipment is used to fine-tune the component position. The drone then immediately conducts a re-inspection and verifies the repair results using an image comparison algorithm. Defects requiring manual intervention automatically generate maintenance work orders containing location coordinates, defect images, and risk levels, and push them to the power management system; After manual maintenance is completed, the drone is called for a second inspection, and the acceptance report is automatically generated after AI review.

10. A power construction drone path planning system, characterized by: include: A determination module is used to determine the target area where power construction needs to be carried out and the coordinates of the preset construction points; A map construction module is used to obtain a topographic map of the target area based on a geographic information database, construct a three-dimensional map model of the target area, and mark the coordinates of the preset construction points on the three-dimensional map model; The path generation module is used to generate the initial construction path of the UAV based on the annotated 3D map model; The weather adjustment module is used to obtain real-time wind speed, precipitation, and temperature data from the target area weather station, establish a three-dimensional model of weather obstacles, and divide the preset radius threshold with the weather obstacle as the center into a no-fly zone based on the severity level of the weather obstacle; A first adjustment module is configured to adjust the initial construction path based on the no-flyable area using an improved artificial potential field method; The second adjustment module is used to dynamically adjust the cruising altitude according to meteorological data and terrain undulations, using an ant colony algorithm to generate an undulation tracking path to maintain a constant relative altitude; The allocation module is used to introduce a load balancing algorithm, combine battery capacity, flight speed and task priority, dynamically allocate inspection sections, and generate corresponding flight paths for power construction drones.

Images