Unmanned aerial vehicle autonomous flight path planning method and system based on transformer substation high-density point cloud

By building the occupancy grid map of the substation and using the improved A-star algorithm and Floyd algorithm, the safety and efficiency problems of drone track planning in complex environments are solved, and a safe and efficient flight path is generated, suitable for substations Automated patrols and emergency responses.

CN119044999APending Publication Date: 2024-11-29JIANGSU FRONTIER ELECTRIC TECH
View PDF 0 Cites 9 Cited by

Patent Information

Application Number
CN202411356531.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-27
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

The existing technology is difficult to generate safe and efficient drone track planning schemes in complex substation environments, especially when considering high voltage equipment and safety, existing path planning algorithms cannot effectively balance safety, efficiency and environment. Factors that may cause safety hazards or inefficiency in the generated path.

Method used

By obtaining the substation cloud data scanned by three-dimensional lidar, an occupation grid map containing obstacle information is built, and the path planning is carried out using improved A-star algorithm, taking into account the safety evaluation value and obstacle distance, dynamically adjusting the safety distance, combining the combination of The Floyd algorithm removes redundant points and turning points to optimize and simplify the path.

Benefits of technology

It significantly improves the safety and adaptability of path planning, maximizes flight efficiency, reduces unnecessary turning points and energy consumption, and generates safer and more efficient flight paths suitable for routine inspections and emergency responses. The automation inspection capabilities and overall operation level of the substation have been improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119044999A_ABST
    Figure CN119044999A_ABST
Patent Text Reader

Abstract

The invention discloses an unmanned aerial vehicle autonomous flight path planning method and system based on transformer substation high-density point cloud, and belongs to the technical field of unmanned aerial vehicle flight path planning, and the method comprises the steps: obtaining transformer substation point cloud data scanned by a three-dimensional laser radar, converting the transformer substation point cloud data into a three-dimensional grid space, constructing an occupied grid map containing obstacle information, and carrying out the three-dimensional grid space; carrying out expansion processing on obstacles occupying the grid map to obtain an unmanned aerial vehicle navigation map; performing path planning on the unmanned aerial vehicle navigation map by adopting an improved A star algorithm to obtain an initial path; and applying a Floyd algorithm to the initial path, and deleting redundant points and turning points in the initial path to obtain a simplified unmanned aerial vehicle route planning scheme. The method is not only suitable for conventional inspection tasks, but also can provide quick response support in emergency, so that the automatic inspection capability, the equipment maintenance efficiency and the overall operation level of the transformer substation are greatly improved, and powerful support is provided for intelligence and safety of a power system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) trajectory planning, and in particular relates to an autonomous UAV trajectory planning method and system based on a high-density point cloud of a substation. Background Art

[0002] The intelligent and unmanned substations have become the main development trend at present. Conventional manual inspections have gradually been replaced by drones, ground robots, cameras, etc. Among them, the application of drone inspections can make up for the shortcomings of other inspection methods and has broad application prospects.

[0003] Manual inspection methods are labor-intensive and inefficient, and the inspection data is not well-informed, which is greatly affected by the quality of personnel. Although ground robots and cameras can reduce the intensity of manual inspections and improve the informatization of inspection operations to a certain extent, ground robots are affected by the road environment, are inflexible, and have low inspection efficiency. Fixed cameras cannot achieve a large number of multi-point arrangements, and both have limited viewing angles, which can only achieve monitoring of low-altitude areas and limited angles, and high-altitude multi-angle inspections are still impossible. With the advantages of high maneuverability, compact and flexible fuselage, precise hovering, vertical take-off and landing, drones can observe equipment at close range and cover all aspects, eliminating the visual blind spots of other inspection methods, and realizing multi-angle, refined, and efficient inspections of substations, timely discovering equipment defects, and reducing the occurrence of accidents.

[0004] At present, there are various ways to supervise the safety of the work site in the substation, including on-site manual monitoring, remote video monitoring, safety inspections, etc. However, these supervision methods require the participation of a large number of safety management personnel. With the increasing size of the power grid and the increasing workload, it is becoming increasingly difficult to adapt to the safety management needs of the site. How to build an accurate and suitable three-dimensional grid map for drone navigation based on high-density point cloud data in a complex substation environment, how to consider safety, efficiency and special substation environments (such as high-voltage equipment) when planning the path, and how to optimize and simplify the path while ensuring safety and considering the performance limitations of the drone to obtain a more efficient and feasible drone trajectory planning solution, have all become problems that need to be solved. Summary of the invention

[0005] In view of at least one of the above technical problems, the present application provides a method and system for autonomous trajectory planning of unmanned aerial vehicles based on high-density point cloud of substation.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] The method for autonomous trajectory planning of unmanned aerial vehicles based on high-density point cloud of substations comprises: obtaining substation site cloud data scanned by three-dimensional laser radar, converting the substation site cloud data into three-dimensional grid space, constructing an occupied grid map containing obstacle information, and performing expansion processing on obstacles in the occupied grid map to obtain a navigation map of unmanned aerial vehicles; on the navigation map of unmanned aerial vehicles, adopting an improved A-star algorithm to perform path planning to obtain an initial path; wherein the improved A-star algorithm comprises incorporating a safety assessment value into node cost calculation; setting a safety distance as a grid unit, and when it is detected that there are obstacle nodes or high-voltage equipment on the left, right, above or below the current node, prohibiting the generation of path nodes passing through the diagonal of the current node, and increasing the actual cost of adjacent nodes in the direction where there are obstacles or high-voltage equipment; applying the Floyd algorithm to the initial path, deleting redundant points and turning points in the initial path, and obtaining a simplified trajectory planning scheme of the unmanned aerial vehicle.

[0008] As a preferred solution of the method for autonomous trajectory planning of unmanned aerial vehicles based on high-density point cloud of substations described in the present invention, wherein: substation cloud data scanned by three-dimensional laser radar is obtained, the substation cloud data is converted into three-dimensional grid space, an occupation grid map containing obstacle information is constructed, and obstacles in the occupation grid map are expanded to obtain a navigation map of the unmanned aerial vehicle, including: obtaining substation cloud data scanned by three-dimensional laser radar, performing noise reduction processing on the substation cloud data to obtain the noise-reduced substation cloud data; the noise-reduced substation cloud data is processed Filter the ground point cloud to obtain the substation cloud data after filtering the ground point cloud; create a three-dimensional grid space according to the spatial distribution range of the substation cloud data after filtering the ground point cloud, and map the substation cloud data after filtering the ground point cloud to the three-dimensional grid space; calculate the point cloud density of each grid in the three-dimensional grid space, mark the grids with point cloud density higher than a preset threshold as obstacle grids, and construct an occupied grid map containing obstacle information; according to the size parameters of the drone, expand the obstacle grids in the occupied grid map to obtain the drone navigation map.

[0009] As a preferred solution of the autonomous trajectory planning method of unmanned aerial vehicles based on high-density point cloud of substations described in the present invention, wherein: the substation site cloud data after filtering out the ground point cloud is mapped to the three-dimensional grid space, including: calculating the x-axis index of a point in the substation site cloud data after filtering out the ground point cloud in the grid space: calculating the difference between the x-coordinate of the point and the minimum x-value, dividing it by the grid size and rounding it down to obtain the grid index to which the point belongs in the x direction; calculating the y-axis index of a point in the substation site cloud data after filtering out the ground point cloud in the grid space. : Calculate the difference between the y coordinate of the point and the minimum y value, divide it by the grid size and round it down to get the grid index to which the point belongs in the y direction; calculate the z-axis index of a point in the substation cloud data in the grid space after filtering out the ground point cloud : Calculate the difference between the point's z coordinate and the minimum z value, divide it by the grid size, and round it down to get the grid index to which the point belongs in the z direction.

[0010] As a preferred solution of the autonomous trajectory planning method of UAV based on substation high-density point cloud described in the present invention, wherein: on the navigation map of the UAV, the improved A-star algorithm is used for path planning to obtain the initial path, including: setting the starting point and the end point on the navigation map of the UAV, and adding the starting point to the open list. The cost function value of each node in the open list is calculated according to the actual cost, the estimated cost and the safety evaluation function. The node with the smallest cost function value is selected as the current node, the current node is moved from the open list to the closed list, and the adjacent nodes of the current node are checked; if the adjacent node is the end point, the path search is completed; otherwise, check whether there are obstacle nodes or high-voltage equipment on the left, right, above and below each adjacent node. If so, it is prohibited to generate path nodes passing through the diagonal of the current node, and the actual cost of the adjacent nodes in the direction where there are obstacles or high-voltage equipment is increased.

[0011] Perform node evaluation, wherein the node evaluation includes: calculating the security evaluation value of the non-prohibited adjacent node based on the security evaluation function, and when the security evaluation value is less than or equal to the preset security threshold, performing the following operations: if the non-prohibited adjacent node is not in the open list and the closed list, adding the non-prohibited adjacent node to the open list, calculating the cost function value, and setting the current node as the parent node of the non-prohibited adjacent node; if the non-prohibited adjacent node is already in the open list, calculating the new actual cost of reaching the non-prohibited adjacent node via the current node, and calculating the new cost function value according to the new actual cost, the estimated cost and the security evaluation function, comparing the new cost function value with the cost function value currently recorded for the non-prohibited adjacent node, and if the new cost function value is less than the cost function value currently recorded, updating the information of the non-prohibited adjacent node: setting the current node as the new parent node of the non-prohibited adjacent node, updating the actual cost currently recorded to the new actual cost, and updating the cost function value currently recorded to the new cost function value.

[0012] When the security assessment value is greater than a preset security threshold, the non-prohibited adjacent node is considered unsafe, and the following operations are performed: if the non-prohibited adjacent node is not in the open list, the non-prohibited adjacent node is not added to the open list; if the non-prohibited adjacent node is already in the open list, the non-prohibited adjacent node is removed from the open list.

[0013] Repeat the cost function value of each node in the open list calculated according to the actual cost, estimated cost and security evaluation function, and evaluate the node until the end node is found or the path is determined to be non-existent. If the end node is found, start from the end node, trace back through the parent node information of each node to construct the initial path; if it is determined that the path does not exist, output the path search failure information.

[0014] As a preferred solution of the method for autonomous trajectory planning of unmanned aerial vehicles based on high-density point cloud of substation described in the present invention, the safety assessment value of adjacent nodes that are not prohibited is calculated based on the safety assessment function, specifically including: determining an adjustment constant; scanning and calculating the distance from the current node to the nearest obstacle in the unmanned aerial vehicle navigation map; measuring the electromagnetic field strength at the current position using an electromagnetic field sensor and normalizing it; determining the safety assessment value of the current node based on the adjustment constant, the distance from the current node to the nearest obstacle and the normalized electromagnetic field strength.

[0015] As a preferred solution of the autonomous trajectory planning method for unmanned aerial vehicles based on high-density point cloud of substation described in the present invention, wherein: the Floyd algorithm is applied to the initial path, redundant points and turning points in the initial path are deleted, and a simplified trajectory planning scheme for the unmanned aerial vehicle is obtained, including: taking all nodes in the initial path as the vertex set of the Floyd algorithm; for every three consecutive nodes A, B, and C in the initial path, the following operations are performed: calculating the length of path ABC; calculating the length of direct path AC; checking whether the direct path AC intersects with any obstacle; calculating the minimum distance between direct path AC and the obstacle; if direct path AC satisfies: not intersecting with any obstacle, the length of direct path AC is less than the length of path ABC, the minimum distance between the obstacle is not less than the preset safety distance, and the inclination angle of path AC does not exceed the maximum climb / descent angle of the unmanned aerial vehicle, then node B is deleted, the path is updated to AC, and an intermediate checkpoint is inserted between A and C; and the above steps are repeated until the path cannot be further simplified.

[0016] As a preferred solution of the autonomous trajectory planning method for unmanned aerial vehicles based on high-density point cloud of substation described in the present invention, it also includes: performing a safety requirement check on the simplified path; the safety requirement check includes recalculating the safety assessment value of the node, dynamically adjusting the preset safety distance according to the result of the node assessment, checking the node safety distance requirement, and the path turning angle limit; if a node that does not meet the safety requirements is found, it is marked and the surrounding area is locally replanned, and the replanned local path is integrated into the simplified overall path; the path parameters before and after simplification are calculated and compared; the path parameters include the total path length, the number of turns and the energy consumption estimation based on the path characteristics; if the path parameters meet the requirements of reducing the total length, reducing the number of turns, reducing the estimated energy consumption and meeting all safety requirements, then it is adopted as the final unmanned aerial vehicle trajectory planning scheme, otherwise the original path is retained or locally adjusted.

[0017] To further solve the above technical problems, the present invention also provides the following technical solutions: an autonomous UAV trajectory planning system based on a high-density point cloud of a substation, comprising: a map construction module, used to obtain substation site cloud data scanned by a three-dimensional laser radar, convert the substation site cloud data into a three-dimensional grid space, construct an occupied grid map containing obstacle information, and expand the obstacles in the occupied grid map to obtain a UAV navigation map; a path planning module, used to use an improved A-star algorithm to perform path planning on the UAV navigation map to obtain an initial path; a path optimization module, used to apply the Floyd algorithm to the initial path, delete redundant points and turning points in the initial path, and obtain a simplified UAV trajectory planning scheme.

[0018] A computer device includes 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 method for autonomous trajectory planning of an unmanned aerial vehicle based on a high-density point cloud of a substation as described above is implemented.

[0019] A computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the method for autonomous trajectory planning of an unmanned aerial vehicle based on a high-density point cloud of a substation as described above is implemented.

[0020] Beneficial effects of the present invention: The present invention can not only accurately construct a three-dimensional grid map suitable for drone navigation, but also generate a safe and efficient flight path in a complex substation environment. By introducing a safety assessment function that comprehensively considers obstacle distance and electromagnetic field strength, as well as a mechanism for dynamically adjusting the safety distance, the present invention significantly improves the safety and adaptability of path planning. At the same time, through the intelligent path simplification and optimization process, flight efficiency is maximized while ensuring safety, and unnecessary turns and energy consumption are reduced. The present invention is not only suitable for routine inspection tasks, but also can provide rapid response support in emergency situations, thereby greatly improving the substation's automated inspection capabilities, equipment maintenance efficiency and overall operational level, providing strong support for the intelligence and safety of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0022] Figure 1 A schematic diagram of the overall process of a method for autonomous trajectory planning of an unmanned aerial vehicle based on a high-density point cloud of a substation provided by an embodiment of the present invention.

[0023] Figure 2 A schematic diagram of a computer device provided for one embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0025] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0026] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides an autonomous trajectory planning method for UAV based on high-density point cloud of substation.

[0027] The existing methods for autonomous trajectory planning of substation drones have the following main problems: First, traditional path planning algorithms are not flexible and safe enough when dealing with highly complex environments such as substations with special safety requirements. They often cannot effectively balance safety, efficiency and environmental factors unique to substations (such as the presence of high-voltage equipment), resulting in the generated paths that may have safety hazards or low efficiency. Secondly, in the path optimization stage, it is difficult for existing methods to consider the performance limitations of drones while ensuring safety, such as turning radius, climbing angle, etc., which may make theoretically feasible paths difficult to implement in actual operations. In addition, existing path simplification methods are often oversimplified and may ignore potential risks in the substation environment in the pursuit of efficiency, and fail to achieve the best balance between safety and efficiency. These problems seriously restrict the degree of autonomy and operating efficiency of drones in substation environments.

[0028] The present application provides an effective solution to the above-mentioned problems. Next, multiple embodiments will be combined to explain in detail how to implement the autonomous trajectory planning method of UAV based on substation high-density point cloud.

[0029] Figure 1 The overall flow chart of the autonomous trajectory planning method of UAV based on high-density point cloud of substation is shown, including:

[0030] S1: Obtain the substation cloud data scanned by the three-dimensional lidar, convert the substation cloud data into a three-dimensional grid space, construct an occupancy grid map containing obstacle information, and expand the obstacles in the occupancy grid map to obtain a UAV navigation map.

[0031] S1.1: Obtain substation cloud data scanned by a three-dimensional laser radar, perform noise reduction processing on the substation cloud data, and obtain noise-reduced substation cloud data.

[0032] For the noise reduction of substation cloud data, the statistical outlier filtering algorithm can be used, and its formula is as follows:

[0033] ;

[0034] in, is the local density of the kth point; is the number of points in the neighborhood; is the coordinate of the i-th point in the neighborhood; is the average coordinate of all points in the neighborhood. If it is greater than the set threshold, the point is considered a noise point and is removed.

[0035] S1.2: Filter the ground point cloud from the noise-reduced substation cloud data to obtain the substation cloud data after filtering the ground point cloud.

[0036] Specifically, the ground point cloud can be filtered out using the height threshold method:

[0037] ;

[0038] in, is the height value of point p; is the preset height threshold; if the height of a point is less than the threshold, it is considered to be a ground point and is filtered out.

[0039] S1.3: Create a three-dimensional grid space according to the spatial distribution range of the substation cloud data after filtering out the ground point cloud, and map the substation cloud data after filtering out the ground point cloud into the three-dimensional grid space.

[0040] Specifically, the grid size calculation formula is:

[0041] ;

[0042] is the grid size; , , It is the maximum value of the substation cloud data on the x, y, and z axes after filtering out the ground point cloud; , , It is the minimum value of the substation cloud data on the x, y, and z axes after filtering out the ground point cloud; The preset number of grids.

[0043] Furthermore, the mapping calculation from point cloud to grid includes: 1. Calculating the x-axis index of the substation cloud data in the grid space after filtering out the ground point cloud : Calculate the difference between the x-coordinate of the point and the minimum x-value, divide it by the grid size and round it down to get the grid index of the point in the x-direction. The formula is:

[0044] ;

[0045] in, It is a point in the substation cloud data after filtering out the ground point cloud. coordinate; It is the minimum value of the substation cloud data on the x-axis after filtering out the ground point cloud.

[0046] 2. Calculate the y-axis index of the substation cloud data in the grid space after filtering out the ground point cloud : Calculate the difference between the y coordinate of the point and the minimum y value, divide it by the grid size and round it down to get the grid index of the point in the y direction. The formula is:

[0047] ;

[0048] in, It is a point in the substation cloud data after filtering out the ground point cloud. coordinate.

[0049] 3. Calculate the z-axis index of the substation cloud data in the grid space after filtering out the ground point cloud : Calculate the difference between the z coordinate of the point and the minimum z value, divide it by the grid size and round it down to get the grid index of the point in the z direction. The formula is:

[0050] ;

[0051] in, It is a point in the substation cloud data after filtering out the ground point cloud. coordinate.

[0052] S1.4: Calculate the point cloud density of each grid in the three-dimensional grid space, mark the grids with point cloud density higher than a preset point cloud density threshold as obstacle grids, and construct an occupancy grid map containing obstacle information.

[0053] Specifically, the formula for calculating the grid point cloud density is:

[0054] ;

[0055] in, For Grid Point cloud density; For Grid The number of point clouds; is the grid volume, equal to .

[0056] The obstacle grid marking conditions are:

[0057] ;

[0058] in, It is the preset point cloud density threshold.

[0059] S1.5: According to the size parameters of the UAV, the obstacle grid in the occupied grid map is expanded to obtain the UAV navigation map.

[0060] The expansion processing formula is:

[0061] ;

[0062] Among them, E is the expansion parameter, which indicates the number of grids expanded by the obstacle; r is the sum of the UAV radius and the minimum safety distance; is the grid size.

[0063] S2: On the UAV navigation map, the improved A-star algorithm is used for path planning to obtain the initial path.

[0064] Among them, the improved A-star algorithm includes setting the safety distance as a grid unit. When an obstacle node is detected on the left, right, above or below the current node, it is prohibited to generate a path node passing through the diagonal of the current node.

[0065] S2.1: On the drone navigation map, set the start point and the end point, and add the start point to the open list.

[0066] S2.2: Calculate the cost function value of each node in the open list based on the actual cost, estimated cost and security evaluation function:

[0067] ;

[0068] Among them, G(n) is the actual cost from the starting point to the current node; H(n) is the estimated cost from the current node to the end point; S(n) is the safety evaluation function; n represents the node.

[0069] ;

[0070] Among them, K is the adjustment constant, which is used to adjust the weight of S(n) in the total cost; is the distance from the current node to the nearest obstacle; is a small positive number used to avoid division by zero; E(n) is the normalized electromagnetic field strength at the current node position (the value range is 0 to 1).

[0071] This safety assessment function provides an important safety indicator for path planning by comprehensively considering obstacle distance and electromagnetic field strength.

[0072] Furthermore, G(n) is the actual cost from the starting point to the current node n. In the grid map, G(n) represents the number of grids passed from the starting point to the current node. If terrain or other factors are considered, different cost values ​​may be assigned to different grids. In this embodiment, when obstacles or high-voltage equipment are detected in a certain direction, the G(n) value of the adjacent node in that direction is increased, which can reduce the possibility of the algorithm selecting these potentially dangerous paths.

[0073] The traditional A-star algorithm usually uses a simple Euclidean distance or Manhattan distance as a heuristic function, while the heuristic function of this embodiment takes into account the obstacle density and voltage level. H(n) is the estimated cost from the current node n to the end point, also known as the heuristic function. In this embodiment, H(n) takes into account the Manhattan distance from the node to the end point, the surrounding obstacle density, and the voltage level of the area. In this way, this embodiment can guide the algorithm to prioritize safer and more suitable paths for drone flights.

[0074] The calculation formula for the estimated cost H(n) from the current node to the end point is:

[0075] ;

[0076] Where D is the distance between adjacent grids; (x, y, z) is the coordinate of the current node; (x t , y t , z t ) is the end point coordinate; is the obstacle density weight coefficient; Q is the obstacle density within a certain range around the current node; m is the voltage level weight coefficient; V is the nominal voltage level of the area where the current node is located.

[0077] S2.3: Select the node with the smallest cost function value F(n) as the current node, move it from the open list to the closed list, and check its adjacent nodes; if the adjacent node is the end point, the path search is completed; otherwise, for each adjacent node, check whether there are obstacle nodes or high-voltage equipment on its left, right, above and below. If so, it is prohibited to generate path nodes passing through the diagonal of the current node, and increase the G(n) value of the adjacent node in the direction where the obstacle or high-voltage equipment exists.

[0078] The present invention further enhances the algorithm's ability to avoid dangerous areas by increasing the G(n) value of adjacent nodes in the direction of obstacles or high-voltage equipment. This method not only improves the safety of path planning, but also optimizes the energy efficiency of drones, because it can help drones choose safer and less energy-consuming paths. Combining these improvements, the trajectory planning method of the present invention can significantly improve the operating efficiency of drones in substation environments while ensuring safety. It can support drones to complete inspection, monitoring and other tasks more accurately, reduce manual intervention, reduce personnel safety risks, and improve the maintenance efficiency and reliability of substation equipment. This improvement is not only applicable to conventional inspection tasks, but also can provide reliable path support for rapid response and disposal in emergency situations, which is of great significance to improving the overall operation level of the power system.

[0079] S2.4: Perform node evaluation, wherein the node evaluation specifically includes: calculating the security evaluation value of the adjacent nodes that are not prohibited based on the security evaluation function, and when the security evaluation value is less than or equal to the preset security threshold, performing the following operations:

[0080] If the non-banned adjacent node is not in the open list and the closed list, then add the non-banned adjacent node to the open list, calculate the cost function value, and set the current node as the parent node of the non-banned adjacent node;

[0081] If the non-prohibited adjacent node is already in the open list, calculate the new actual cost of reaching the non-prohibited adjacent node via the current node, and calculate the new cost function value based on the new actual cost, the estimated cost and the security evaluation function, compare the new cost function value with the cost function value currently recorded for the non-prohibited adjacent node, if the new cost function value is less than the cost function value currently recorded, update the information of the non-prohibited adjacent node: set the current node as the new parent node of the non-prohibited adjacent node, update the actual cost currently recorded to the new actual cost, and update the cost function value currently recorded to the new cost function value;

[0082] When the security assessment value is greater than the preset security threshold, the adjacent node that has not been banned is considered unsafe, and the following operations are performed:

[0083] If the neighboring node that is not banned is not in the open list, it is not added to the open list;

[0084] If the non-banned neighboring node is already in the open list, it is removed from the open list.

[0085] Repeat S2.2 to calculate the cost function value of each node in the open list, and then go to S2.4 to evaluate the nodes until the end node is found or it is determined that the path does not exist.

[0086] S2.5: If the end node is found, start from the end node and trace back through the parent node information of each node to build the initial path; if it is determined that the path does not exist, output the path search failure information.

[0087] Preferably, S2 uses an improved A-star algorithm for path planning, and its beneficial effects are reflected in the following aspects: First, by introducing a safety assessment function S(n) that comprehensively considers obstacle distance and electromagnetic field strength, the algorithm's adaptability to the complex environment of the substation and path safety are significantly improved. Secondly, the heuristic function H(n) not only considers the traditional distance factor, but also introduces the weights of obstacle density and voltage level, so that the algorithm can more intelligently avoid high-risk areas. Furthermore, by dynamically adjusting the actual cost G(n), especially increasing the G(n) value for adjacent nodes in the direction of obstacles or high-voltage equipment, the algorithm's ability to avoid dangerous areas is further enhanced. At the same time, setting a safe distance and prohibiting the generation of path nodes passing through the diagonal of the current node effectively reduces the sharp turn operation of the drone in a small space and improves the stability and controllability of the flight. Finally, the nodes in the open list are dynamically managed by presetting a safety threshold to ensure the overall safety of the selected path. These improvements not only optimize the energy efficiency of drones, but also significantly improve the operating efficiency and safety in substation environments, providing reliable path support for drones to perform tasks such as precise inspections and emergency responses, thereby greatly improving the automation level and overall safety of substation operations.

[0088] S3: Apply the Floyd algorithm to the initial path, delete the redundant points and turning points in the initial path, and obtain a simplified UAV trajectory planning scheme.

[0089] S3.1: Take all nodes in the initial path as the vertex set of Floyd's algorithm.

[0090] S3.2: For every three consecutive nodes A, B, C in the initial path, perform the following operations:

[0091] a) Calculate the length of path ABC;

[0092] b) Calculate the length of the direct path AC;

[0093] c) Check whether the direct path AC intersects any obstacle;

[0094] d) Calculate the security assessment value on the direct path AC.

[0095] S3.3: If the direct path AC satisfies all of the following conditions:

[0096] a) does not intersect with any obstacles;

[0097] b) its length is less than the length of path ABC;

[0098] c) The safety assessment value of each point on path AC is not greater than the preset safety threshold;

[0099] d) The inclination angle of path AC does not exceed the maximum climb / descent angle of the UAV;

[0100] Then delete node B, update the path to AC, and insert an intermediate checkpoint between A and C for subsequent path adjustment.

[0101] S3.4: Repeat steps S3.2 to S3.3 until the path cannot be further simplified.

[0102] S3.5: Perform safety requirement checks (i.e. safety and feasibility checks) on the simplified path:

[0103] a) Recalculate the safety evaluation value of each node on the simplified path through the safety evaluation function.

[0104] b) According to the node evaluation results in S2, dynamically adjust the preset safety distance of each node: if the safety evaluation value of the node is high (indicating a higher risk), increase the preset safety distance of the node; if the safety evaluation value of the node is low (indicating a lower risk), the preset safety distance can be appropriately reduced.

[0105] c) Check whether each node on the simplified path meets the dynamically adjusted preset safety distance requirements.

[0106] d) Check whether the turning angle of the path meets the steering capability limit of the drone.

[0107] S3.6: If nodes that do not meet the safety requirements are found in S3.5, these unsafe nodes are marked, and the areas around these unsafe nodes are locally replanned through S2, and the replanned local paths are integrated into the simplified overall path.

[0108] S3.7: Calculate and compare path parameters before and after simplification:

[0109] a) Calculate the total path length;

[0110] b) Count the number of turning points;

[0111] c) Estimation of energy consumption (estimate based on a comprehensive estimate of path length, number of turns and safety assessment value).

[0112] S3.8: If the simplified path meets the following conditions, it is adopted as the final UAV trajectory planning solution:

[0113] a) Overall length is reduced.

[0114] b) The number of turns is reduced.

[0115] c) Estimated reduction in energy consumption.

[0116] d) meet all safety and feasibility requirements.

[0117] Otherwise, keep the original path or make further local adjustments.

[0118] Preferably, step S3 optimizes the initial path by applying the improved Floyd algorithm, which significantly improves the efficiency and safety of the UAV's trajectory planning in the substation environment. Compared with the traditional path simplification method, the present invention has the following significant advantages:

[0119] First, the present invention introduces a dynamic safety assessment mechanism in the path simplification process, which not only considers the geometric distance, but also takes into account environmental factors unique to substations such as electromagnetic field strength. This comprehensive assessment method ensures that the simplified path not only maintains the optimal length and turning characteristics, but also effectively avoids potential safety hazards, which is a feature that traditional path planning algorithms do not have.

[0120] Secondly, the present invention realizes the dynamic adjustment of the preset safety distance. By analyzing the safety assessment value of the node in real time, the safety distance requirement can be flexibly adjusted according to the change of environmental risks. This adaptive mechanism enables the drone to optimize the flight path to the maximum extent while ensuring safety, which is particularly suitable for the complex and changeable environment inside the substation.

[0121] In addition, the present invention takes into account the physical limitations of the UAV, such as the maximum climb / descent angle and steering capability, during the path simplification process. This ensures that the simplified path is not only theoretically feasible, but also fully considers the actual flight characteristics of the UAV, greatly improving the practicality and feasibility of path planning.

[0122] Finally, the invention can accurately handle local safety issues that arise during the simplification process through an iterative local replanning mechanism without giving up the overall optimization results. This flexible adjustment capability enables the system to effectively deal with local complex situations while ensuring global optimization.

[0123] In summary, the present invention significantly improves the efficiency and quality of path planning while ensuring safety. Compared with the prior art, the present invention is more adaptable to the special requirements of the substation environment, and can generate a safer, more efficient and more feasible UAV trajectory planning scheme, providing strong technical support for the automated inspection and emergency response of substations.

[0124] Embodiment 2 is an embodiment of the present invention, which provides an autonomous trajectory planning system for unmanned aerial vehicles based on high-density point clouds of substations, including: a map construction module, used to obtain substation cloud data scanned by three-dimensional laser radar, convert the substation cloud data into a three-dimensional grid space, construct an occupation grid map containing obstacle information, and expand the obstacles in the occupation grid map to obtain a unmanned aerial vehicle navigation map; a path planning module, used to use an improved A-star algorithm to perform path planning on the unmanned aerial vehicle navigation map to obtain an initial path; a path optimization module, used to apply the Floyd algorithm to the initial path, delete redundant points and turning points in the initial path, and obtain a simplified unmanned aerial vehicle trajectory planning scheme.

[0125] Example 3, reference Figure 2 , is an embodiment of the present invention, which is different from the previous embodiment in that: if the function is implemented in the form of a software functional unit and sold or used as an independent product, it 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, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0126] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0127] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0128] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0129] Example 4 is an embodiment of the present invention, which provides an autonomous trajectory planning method for UAV based on high-density point cloud of substation. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0130] In order to comprehensively evaluate the performance of the present invention in the autonomous trajectory planning of substation drones, this embodiment designs a series of simulation experiments covering different environmental complexities, electromagnetic intensities and mission scenarios. The experimental environment is constructed based on a three-dimensional model of a real substation, including various typical equipment such as transformers, switch cabinets, cable trays, etc., and simulates the distribution of electromagnetic fields of different intensities. This embodiment selects three typical environmental complexities (low, medium, and high), three electromagnetic intensity levels (low, medium, and high), and three common task types (routine inspections, emergency response, and detailed inspections) for testing. In each case, the experiment was repeated at least 50 times to ensure the statistical significance of the data.

[0131] In the experiment, this embodiment compares the performance of the present invention with the traditional A-star algorithm and the RRT-based method under various conditions. The evaluation indicators include basic indicators such as path length, calculation time, successful planning rate, safety distance, number of turns, and specific task-related indicators such as electromagnetic interference avoidance success rate, energy consumption, task completion time, and detection accuracy. In order to simulate the actual situation, this embodiment also takes into account the physical limitations of the drone, such as maximum flight speed, turning radius, climb angle, etc. Through these comprehensive and detailed simulation experiments, this embodiment can objectively evaluate the actual performance and advantages of the present invention in a complex substation environment, and provide a reliable theoretical basis for its effect in practical applications.

[0132] Table 1: Performance comparison under different environment complexity

[0133]

[0134] As shown in Table 1, the performance of the present invention under different environmental complexities is better than that of the traditional A algorithm and the RRT-based method. Especially in high-complexity environments, the advantages of the present invention are more obvious. Its average path length is 197.5m, which is 14.8% shorter than the traditional A algorithm and 9.8% shorter than the RRT-based method. Although the calculation time is slightly longer (5.12s), the successful planning rate reaches 94.7%, which is much higher than the other two methods. In addition, the present invention also performs well in terms of safety distance and number of turns. The average safety distance is 1.76m, which is 43.1% higher than the traditional A algorithm, and the average number of turns is only 17.5 times, which is 32.2% less than the traditional A algorithm. These data fully demonstrate the superiority of the present invention in complex substation environments and can generate safer and more efficient flight paths.

[0135] Table 2: Performance comparison under different electromagnetic environments

[0136]

[0137] As shown in Table 2, the present invention exhibits excellent anti-interference ability and path optimization ability under different electromagnetic intensity environments. In a high-intensity electromagnetic environment, the electromagnetic interference avoidance success rate of the present invention reaches 94.2%, which is much higher than the 73.5% of the traditional A algorithm and the 84.6% based on the RRT method. At the same time, the average path deviation of the present invention is only 1.54m, which is 47.8% less than the traditional A algorithm and 29.4% less than the RRT method. In terms of energy consumption, the energy consumption increase ratio of the present invention in a high-intensity electromagnetic environment is only 8.3%, which is significantly lower than the other two methods. These data show that the present invention can effectively deal with electromagnetic interference in substations, maintain the stability and energy efficiency of the flight path, and provide a strong guarantee for the safe operation of drones in complex electromagnetic environments.

[0138] Table 3: Performance comparison under different task scenarios

[0139]

[0140] As shown in Table 3, the present invention shows significant advantages in different task scenarios. In conventional inspection tasks, the average task completion time of the present invention is 18.3 minutes, which is 19.4% faster than the traditional A algorithm, and the detection point coverage rate reaches 99.2%, while maintaining a high battery remaining capacity (62.7%). In the emergency response scenario, the average arrival time of the present invention is only 45.6 seconds, which is 21.6% faster than the traditional A algorithm, and the number of path conflicts is the least (0.3 times). In the fine inspection task, the average hovering stability of the present invention reaches 3.2cm, the image clarity score is 8.7 points (out of 10 points), and the detection accuracy reaches 97.8%, which are significantly better than the other two methods. These data fully illustrate the adaptability and efficiency of the present invention in various practical task scenarios, and can significantly improve the overall performance and reliability of substation drone inspections.

[0141] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. The autonomous trajectory planning method of UAV based on high-density point cloud of substation is characterized by: include: Obtaining substation cloud data scanned by a three-dimensional laser radar, converting the substation cloud data into a three-dimensional grid space, constructing an occupation grid map containing obstacle information, and performing expansion processing on the obstacles in the occupation grid map to obtain a drone navigation map; On the UAV navigation map, an improved A-star algorithm is used for path planning to obtain an initial path; The improved A-star algorithm includes incorporating the safety assessment value into the node cost calculation; setting the safety distance as a grid unit, and when it is detected that there is an obstacle node or high-voltage equipment on the left, right, above or below the current node, prohibiting the generation of a path node passing through the diagonal of the current node, and increasing the actual cost of the adjacent node in the direction where the obstacle or high-voltage equipment exists; The Floyd algorithm is applied to the initial path, redundant points and turning points in the initial path are deleted, and a simplified UAV trajectory planning scheme is obtained.

2. The autonomous trajectory planning method for UAV based on substation high-density point cloud according to claim 1 is characterized in that: Obtaining substation cloud data scanned by a three-dimensional laser radar, converting the substation cloud data into a three-dimensional grid space, constructing an occupied grid map containing obstacle information, and performing expansion processing on the obstacles in the occupied grid map to obtain a drone navigation map, including: Acquire substation cloud data scanned by a three-dimensional laser radar, and perform noise reduction processing on the substation cloud data to obtain the substation cloud data after noise reduction; Filtering the ground point cloud of the substation cloud data after noise reduction to obtain the substation cloud data after filtering the ground point cloud; According to the spatial distribution range of the substation cloud data after filtering out the ground point cloud, a three-dimensional grid space is created, and the substation cloud data after filtering out the ground point cloud is mapped to the three-dimensional grid space; Calculating the point cloud density of each grid in the three-dimensional grid space, marking grids with point cloud density higher than a preset threshold as obstacle grids, and constructing an occupied grid map containing obstacle information; According to the size parameters of the UAV, the obstacle grid in the occupied grid map is expanded to obtain the UAV navigation map.

3. The autonomous trajectory planning method for UAV based on substation high-density point cloud as claimed in claim 2 is characterized in that: Mapping the substation cloud data after filtering out the ground point cloud to the three-dimensional grid space includes: Calculate the x-axis index of a point in the grid space in the substation cloud data after filtering out the ground point cloud: calculate the difference between the x-coordinate of the point and the minimum x-value, divide it by the grid size and round it down to get the grid index of the point in the x direction; Calculate the y-axis index of a point in the substation cloud data after filtering out the ground point cloud in the grid space : Calculate the difference between the y coordinate of the point and the minimum y value, divide it by the grid size and round it down to get the grid index of the point in the y direction; Calculate the z-axis index of a point in the substation cloud data in the grid space after filtering out the ground point cloud : Calculate the difference between the point's z coordinate and the minimum z value, divide it by the grid size, and round it down to get the grid index to which the point belongs in the z direction.

4. The autonomous trajectory planning method for unmanned aerial vehicles based on high-density point cloud of substation according to claim 3 is characterized in that: On the UAV navigation map, the improved A-star algorithm is used for path planning to obtain the initial path, including: On the drone navigation map, set the starting point and the end point, and add the starting point to the open list; Calculate the cost function value of each node in the open list according to the actual cost, the estimated cost and the security evaluation function; Select the node with the smallest cost function value as the current node, move the current node from the open list to the closed list, and check the adjacent nodes of the current node; if the adjacent node is the end point, the path search is completed; otherwise, check whether there are obstacle nodes or high-voltage devices on the left, right, above and below each adjacent node. If so, it is forbidden to generate a path node passing through the diagonal of the current node, and increase the actual cost of the adjacent nodes in the direction where the obstacles or high-voltage devices exist; Performing node evaluation, wherein the node evaluation includes: calculating the security evaluation value of the adjacent nodes that are not prohibited based on the security evaluation function, and when the security evaluation value is less than or equal to a preset security threshold, performing the following operations: If the non-prohibited adjacent node is not in the open list and the closed list, the non-prohibited adjacent node is added to the open list, the cost function value is calculated, and the current node is set as the parent node of the non-prohibited adjacent node; If the non-prohibited adjacent node is already in the open list, then calculate the new actual cost of reaching the non-prohibited adjacent node via the current node, and calculate a new cost function value based on the new actual cost, the estimated cost and the security assessment function, compare the new cost function value with the cost function value currently recorded for the non-prohibited adjacent node, and if the new cost function value is less than the cost function value currently recorded, update the information of the non-prohibited adjacent node: set the current node as the new parent node of the non-prohibited adjacent node, update the actual cost currently recorded to the new actual cost, and update the cost function value currently recorded to the new cost function value; When the security assessment value is greater than a preset security threshold, the adjacent node that is not prohibited is considered unsafe, and the following operations are performed: If the non-prohibited neighboring node is not in the open list, then the non-prohibited neighboring node is not added to the open list; If the non-prohibited neighboring node is already in the open list, removing the non-prohibited neighboring node from the open list; Repeatedly calculating the cost function value of each node in the open list according to the actual cost, the estimated cost and the security evaluation function, and evaluating the node until the end node is found or it is determined that the path does not exist; If the end node is found, start from the end node and trace back through the parent node information of each node to build the initial path; if it is determined that the path does not exist, the path search failure information is output.

5. The autonomous trajectory planning method for unmanned aerial vehicles based on high-density point cloud of substation according to claim 4 is characterized in that: Calculating the security evaluation value of the adjacent node that is not prohibited based on the security evaluation function specifically includes: Determine the regulation constant; Scan and calculate the distance from the current node to the nearest obstacle in the drone navigation map; The electromagnetic field strength at the current position is measured and normalized using an electromagnetic field sensor; Based on the adjustment constant, the distance from the current node to the nearest obstacle and the normalized electromagnetic field strength, a safety assessment value of the current node is determined.

6. The autonomous trajectory planning method for UAV based on substation high-density point cloud according to claim 1 is characterized in that: Applying the Floyd algorithm to the initial path, deleting redundant points and turning points in the initial path, and obtaining a simplified UAV trajectory planning scheme, including: All nodes in the initial path are used as the vertex set of the Floyd algorithm; For every three consecutive nodes A, B, C in the initial path, the following operations are performed: calculating the length of the path ABC; calculating the length of the direct path AC; checking whether the direct path AC intersects any obstacle; calculating the minimum distance between the direct path AC and an obstacle; If the direct path AC satisfies: it does not intersect with any obstacle, the length of the direct path AC is less than the length of the path ABC, the minimum distance from the obstacle is not less than the preset safety distance, and the inclination angle of the path AC does not exceed the maximum climb / descent angle of the drone, then delete node B, update the path to AC, and insert an intermediate checkpoint between A and C; Repeat the above steps until the path cannot be simplified any further.

7. The autonomous trajectory planning method for unmanned aerial vehicles based on high-density point cloud of substation according to claim 6 is characterized in that: Also includes: Performing a safety requirement check on the simplified path; the safety requirement check includes recalculating the safety evaluation value of the node, dynamically adjusting the preset safety distance according to the result of the node evaluation, checking the node safety distance requirement, and the path turning angle limit; If a node that does not meet the safety requirements is found, it will be marked and the surrounding area will be partially replanned, and the replanned local path will be integrated into the simplified overall path; Calculating and comparing path parameters before and after simplification; the path parameters include total path length, number of turns, and energy consumption estimation based on path characteristics; If the path parameters satisfy the requirements of reducing the total length, reducing the number of turns, reducing the estimated energy consumption and meeting all safety requirements, the final UAV trajectory planning scheme is adopted, otherwise the original path is retained or partially adjusted.

8. The autonomous trajectory planning system for UAV based on high-density point cloud of substation is characterized by: include: A map construction module is used to obtain the substation cloud data scanned by the three-dimensional laser radar, convert the substation cloud data into a three-dimensional grid space, construct an occupation grid map containing obstacle information, and expand the obstacles in the occupation grid map to obtain a UAV navigation map; A path planning module is used to perform path planning on the UAV navigation map using an improved A-star algorithm to obtain an initial path; The path optimization module is used to apply the Floyd algorithm to the initial path, delete the redundant points and turning points in the initial path, and obtain a simplified UAV trajectory planning scheme.

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 method for autonomous trajectory planning of a UAV based on a high-density point cloud of a substation according to any one of claims 1 to 7 is 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 method for autonomous trajectory planning of a UAV based on a high-density point cloud of a substation as described in any one of claims 1 to 7 is implemented.

Citation Information

Cited By

  • Transformer substation inspection unmanned aerial vehicle local collision avoidance path planning method based on independent steering

    CN118500405A

  • A Local Collision Avoidance Path Planning Method for Substation Inspection UAVs Based on Individual Turning

    CN118500405B

  • Unmanned aerial vehicle system for transformer substation inspection

    CN119596993A

  • An unmanned aerial vehicle system for substation inspection

    CN119596993B

  • Three-dimensional dynamic path planning method based on high-density radar point cloud

    CN120370956A