Unmanned aerial vehicle electrified water washing remote cooperative control platform and method

Abstract

The invention relates to the technical field of water washing control, and particularly discloses an unmanned aerial vehicle electrified water washing remote cooperative control platform and method, and the platform comprises a thermodynamic diagram generation module which is used for generating a three-dimensional pollution thermodynamic diagram containing the pollution conditions of the surfaces of all power transmission tower insulators in a target area; the thermodynamic diagram equipartition module is used for carrying out multidirectional pollution uniform clustering division on the three-dimensional pollution thermodynamic diagram to obtain a pollution equipartition space set under various preset division dimensions; the task distribution module is used for distributing a subarea inclusion task for the unmanned aerial vehicle group based on all the pollution equipartition space sets to obtain a task distribution result; according to the method, the scientificity and reasonability of task allocation are realized, so that the unmanned aerial vehicle group can efficiently cooperate and pointedly wash the insulator of the power transmission tower, the working efficiency of electrified water washing is improved, safe and stable operation of a power transmission line is guaranteed, and the intelligent level of whole remote cooperative control is improved.

Description

Technical Field

[0001] This invention relates to the field of water flushing control technology, and in particular to a remote collaborative control platform and method for electrified water flushing of unmanned aerial vehicles (UAVs). Background Technology

[0002] In the operation and maintenance of power systems, the cleaning and maintenance of transmission tower insulators is a crucial link in ensuring a stable power supply. Due to the wide distribution of transmission lines and the complex terrain and harsh environments in some areas, traditional manual cleaning methods are not only inefficient but also pose significant safety risks. With the continuous development of drone technology, live-line water washing technology using drones is gradually being applied to the cleaning of transmission tower insulators. It can efficiently clean insulators while they are energized, greatly improving cleaning efficiency and safety. However, when dealing with large-scale transmission networks, relying solely on a single drone is far from sufficient; multiple drones need to work collaboratively. This places higher demands on the collaborative control of live-line water washing using drones. A remote collaborative control platform and method for live-line water washing using drones is of great significance for improving power operation and maintenance efficiency and ensuring the safe and stable operation of the power grid. With the increasing demand for intelligent and efficient operation and maintenance in the power industry, such a remote collaborative control platform has broad application prospects and is expected to drive the development of power operation and maintenance models towards greater intelligence and automation.

[0003] However, existing methods for controlling live-line water flushing by drones have many shortcomings. In terms of presenting the contamination status of transmission tower insulators, it is difficult to intuitively and comprehensively understand the distribution of contamination. This lack of a holistic and intuitive presentation of the contamination status leads to an unreasonable and unscientific division of task areas, and makes it impossible to scientifically allocate tasks to drone swarms. This results in low efficiency for multi-drone collaborative operations, failing to fully leverage the advantages of live-line water flushing technology and impacting the overall effectiveness of power operation and maintenance.

[0004] Therefore, this invention proposes a remote collaborative control platform and method for electric water flushing of unmanned aerial vehicles (UAVs). Summary of the Invention

[0005] This invention provides a remote collaborative control platform and method for live water flushing of unmanned aerial vehicles (UAVs). The platform can generate a three-dimensional pollution heat map to intuitively display the pollution status of the surface of the insulator of the transmission tower. Then, the heat map can be reasonably divided, and tasks can be scientifically allocated to the UAV swarm according to the characteristics of different division spaces and the performance of the UAVs, so as to realize the efficient collaborative operation of multiple UAVs.

[0006] This invention provides a remote collaborative control platform for electro-hydraulic water flushing of unmanned aerial vehicles (UAVs), comprising: The heat map generation module is used to generate a three-dimensional pollution heat map containing the pollution status of the surfaces of all transmission tower insulators within the target area; The heatmap uniform division module is used to perform multi-directional uniform clustering of the three-dimensional pollution heatmap based on all the pollution excess space in the three-dimensional pollution heatmap to obtain pollution uniform division space sets under multiple preset division dimensions. The task allocation module is used to allocate partitioned tasks to the UAV swarm based on the comprehensive cost matrix under each preset partitioning dimension, and obtain the task allocation results.

[0007] Optionally, it also includes: The task update module is used to update the task allocation results in real time based on the current three-dimensional contamination heat map when a task interruption is detected during the execution of an electrified water flushing task by the drone swarm based on the task allocation results.

[0008] Optionally, the heatmap generation module includes: The pollution status acquisition submodule is used to collect pollution status data of the surface of all transmission tower insulators using an infrared thermal imager and an ultraviolet detector mounted on a drone. The heat map generation submodule is used to generate a three-dimensional pollution heat map based on the pollution status data of all transmission tower insulator surfaces and combined with GIS geographic information.

[0009] Optionally, the heatmap distribution module includes: The first division submodule is used to divide the three-dimensional contamination heat map into multiple division spatial scales of no less than the preset division spatial scale, obtain multiple first division spaces under each division spatial scale, and regard all first division spaces under all division spatial scales as contamination excess spaces with a contamination ratio of no less than the preset contamination ratio threshold. The second partitioning submodule is used to cluster and partition all the excessive filth space to obtain multiple filth concentration areas in the three-dimensional filth heat map. The third sub-module is used to determine the number of divisions based on the number of drones in the drone swarm, and to further divide the three-dimensional pollution heat map based on the number of divisions, multiple preset division dimensions, and all pollution concentration areas to obtain the pollution distribution space set under each preset division dimension.

[0010] Optionally, the second sub-module includes: The first division unit is used to determine the physical center location of each filth concentration area, and arbitrarily divide all filth concentration areas to obtain multiple area clusters. It calculates the distance between the physical center locations of any two filth concentration areas in each area cluster, and calculates the ratio of each distance to a preset distance as the relative deviation of the corresponding two filth concentration areas. Based on the relative deviation of the two filth concentration areas, it calculates the concentration degree of the two filth concentration areas. The product of the average concentration degree of all two filth concentration areas in each area cluster and the total number of all filth concentration areas in the corresponding area cluster is taken as the comprehensive concentration degree of each area cluster. The first judgment unit is used to determine whether the average of the comprehensive concentration of all currently obtained regional clusters is not less than the preset concentration threshold. If so, all the polluted concentrated areas in each currently obtained regional cluster are merged to obtain a polluted concentrated area. Otherwise, all the polluted concentrated areas are arbitrarily re-divided to obtain multiple new regional clusters until the average of the comprehensive concentration of all the latest obtained regional clusters is not less than the preset concentration threshold. Then, all the polluted concentrated areas in each latest obtained regional cluster are merged to obtain a polluted concentrated area.

[0011] Optionally, the third sub-module includes: The second division unit is used to divide the three-dimensional pollution heat map based on the number of divisions and multiple preset division dimensions, to obtain multiple second division spaces under each preset division dimension, and there are no two or more second division spaces that simultaneously intersect with the pollution concentration area in all the second division spaces under each preset division dimension. The spatial analysis unit is used to analyze the contamination percentage, spatial volume, and obstacle coefficient of each second partitioned space; The second judgment unit is used to determine whether the average degree of each dimension of the filth ratio, space volume, and obstacle coefficient of all second division spaces under each preset division dimension is less than the preset average degree threshold. If so, the filth heat map is re-divided based on the corresponding preset division dimension until the average degree of the filth ratio, space volume, obstacle coefficient, and total flushing cost of the latest obtained second division spaces is not less than the preset average degree threshold. Then, all the latest obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension. Otherwise, all the currently obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension.

[0012] Optionally, the task assignment module includes: The comprehensive cost analysis unit is used to analyze the comprehensive cost of allocating each second partition space of each type of contamination space to each UAV based on the contamination ratio, space volume, obstacle coefficient, and flushing path of each second partition space in the contamination space set. The cost matrix construction unit is used to construct a comprehensive cost matrix for each preset partition dimension based on the comprehensive cost of allocating each type of pollution space to each second partition space and then to each UAV. The task optimal allocation unit is used to generate the optimal task allocation result for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The final task allocation unit is used to allocate partitioned tasks to the UAV swarm based on the optimal task allocation result under each preset partition dimension, and obtain the task allocation result.

[0013] Optionally, the comprehensive cost analysis unit includes: The flushing cost analysis subunit is used to calculate the flushing cost of the corresponding second partition space based on the proportion of dirt, space volume, and obstacle coefficient of each second partition space in each type of dirt-sharing space set. The distance cost analysis subunit is used to plan the flushing path for each second partition space in the space for each type of filth distribution, and to analyze the distance cost of the corresponding second partition space based on the flushing path. The matching cost analysis subunit is used to analyze the equipment matching cost between each drone and each second partition space of each type of pollution equal distribution space based on the working parameter range and equipment status of each drone in the drone swarm, as well as the pollution ratio, space volume, and obstacle coefficient of each second partition space in each pollution equal distribution space set. The comprehensive cost analysis subunit is used to calculate the comprehensive cost of allocating each second partition space of each type of filth distribution space to each drone based on the equipment matching cost of each drone with each second partition space of each type of filth distribution space set, the flushing cost of the corresponding second partition space, and the distance cost.

[0014] Optionally, the optimal task allocation unit includes: The task allocation subunit is used to generate multiple hypothetical task allocation schemes for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The multidimensional evaluation subunit is used to generate a comprehensive evaluation index value for each hypothetical task allocation scheme based on the total allocation cost and imbalance coefficient of each hypothetical task allocation scheme. The result filtering subunit is used to filter out the optimal task allocation result under each preset partition dimension from all the hypothetical allocation schemes under each preset partition dimension based on the comprehensive evaluation index value of all hypothetical allocation schemes under each preset partition dimension.

[0015] This invention provides a remote collaborative control method for electro-hydraulic water flushing of unmanned aerial vehicles (UAVs), comprising: Generate a three-dimensional pollution heat map containing the pollution status of all transmission tower insulator surfaces within the target area; Based on all the excess space of pollution in the 3D pollution heat map, the 3D pollution heat map is divided into pollution uniform clusters in multiple directions to obtain pollution uniform space sets under multiple preset division dimensions; Based on the comprehensive cost matrix under each preset partitioning dimension, the drone swarm is assigned partitioned tasks to obtain the task allocation results.

[0016] The beneficial effects of this invention compared to existing technologies are as follows: It generates a three-dimensional pollution heat map containing the surface contamination status of transmission tower insulators, comprehensively presenting the pollution distribution within the target area in an intuitive and visual manner, providing clear and accurate basic information for subsequent analysis. The heat map is divided into multi-directional uniform pollution clusters, resulting in pollution distribution spatial sets under various preset dimensions. This detailed division allows for analysis of pollution distribution from different angles, facilitating precise location of pollution characteristics in different areas and laying the foundation for rational task allocation. Based on all pollution distribution spatial sets, the UAV swarm is assigned zoned tasks, and the task allocation results are obtained, achieving scientific and rational task allocation. This enables the UAV swarm to collaborate efficiently, specifically cleaning transmission tower insulators, improving the efficiency of live-line water cleaning, ensuring the safe and stable operation of transmission lines, and enhancing the overall intelligence level of remote collaborative control.

[0017] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0018] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0019] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic diagram of a remote collaborative control platform for electric water flushing of unmanned aerial vehicles (UAVs) in an embodiment of the present invention. Figure 2 This is a schematic diagram of the heat map generation module in an embodiment of the present invention. Detailed Implementation

[0020] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0021] like Figure 1As shown, the present invention provides an implementation of a remote collaborative control platform for electro-hydraulic water flushing of unmanned aerial vehicles (UAVs), comprising: The heat map generation module is used to generate a three-dimensional pollution heat map containing the pollution status of the surfaces of all transmission tower insulators within the target area; The heatmap uniform division module is used to perform multi-directional uniform clustering of the three-dimensional pollution heatmap based on all the pollution excess space in the three-dimensional pollution heatmap to obtain pollution uniform division space sets under multiple preset division dimensions. The task allocation module is used to allocate partitioned tasks to the UAV swarm based on the comprehensive cost matrix under each preset partitioning dimension, and obtain the task allocation results.

[0022] In this embodiment, the target area refers to the specific geographical area where live-line water flushing of transmission tower insulators is required. This area includes numerous transmission towers and is the scope of the entire cleaning task; its boundaries and specific range are determined by actual power operation and maintenance needs. For example, it could be the coverage area of ​​all transmission lines within a specific area of ​​a city, or the area involved in a specific transmission network in a mountainous region.

[0023] In this embodiment, the contamination condition of the transmission tower insulator surface refers to the amount of dirt, impurities, etc., adhering to the surface of the insulator. It encompasses information such as the type of contamination (e.g., dust, oil, salt, etc.), the degree of contamination (light, moderate, heavy, etc.), and the distribution of the contamination. These conditions affect the electrical performance and operational safety of the insulator and are crucial for determining whether and how to perform cleaning. For example, if the insulator surface is heavily contaminated with salt and alkali, reaching a heavy level, cleaning is urgently needed to ensure power transmission safety.

[0024] In this embodiment, the three-dimensional pollution heat map is a visualization generated by combining data on the pollution status of transmission tower insulator surfaces collected by an infrared thermal imager and an ultraviolet detector mounted on a UAV, along with GIS geographic information. It visually displays the pollution status of all transmission tower insulator surfaces within a target area in a three-dimensional space. Different colors or brightness levels may represent different degrees of pollution. This heat map provides a clear and comprehensive understanding of the pollution distribution within the target area, offering intuitive and accurate basic information for subsequent task allocation and analysis. For example, darker areas on the heat map may indicate areas with more severe pollution.

[0025] In this embodiment, the excessive contamination space refers to the first partitioned spaces in which the contamination ratio is not less than a preset contamination ratio threshold among all the first partitioned spaces under all partitioned spatial scales after the three-dimensional contamination heat map is divided into multiple partitioned spatial scales of not less than a preset partitioned spatial scale. These spaces contain relatively large amounts of contamination and are the key areas to focus on during subsequent clustering and task allocation because they have a significant potential impact on power transmission safety and require more reasonable planning of flushing tasks.

[0026] In this embodiment, multiple preset division dimensions refer to different pre-defined division methods or standards for dividing the three-dimensional pollution heat map from different perspectives. For example, the three-dimensional pollution heat map can be divided according to multiple dimensions such as spatial location dimension (e.g., horizontal direction, vertical direction, etc.) and geographical region dimension (e.g., division according to topography, etc.), thereby obtaining a pollution evenly distributed spatial set under multiple preset division dimensions, so as to more accurately locate the pollution characteristics of different areas and provide more reference for the reasonable allocation of tasks.

[0027] In this embodiment, the zoned task allocation is a specific cleaning task mode assigned to the drone swarm based on a pollution distribution space set under multiple preset division dimensions. That is, the target area is divided into multiple sub-regions according to certain rules, and each drone is responsible for the insulator washing task of one or more sub-regions. Each drone clearly defines its assigned area and completes the live water washing work in that area. This task allocation method helps improve the collaborative operation efficiency of the drone swarm, enabling them to wash the transmission tower insulators more effectively.

[0028] In this embodiment, the task allocation result is based on the analysis of the comprehensive cost of allocating each second partition space in all pollution-divided space sets to each UAV, and is the final task arrangement determined for the UAV swarm. This result clarifies the specific washing area that each UAV is responsible for. It is derived by comprehensively considering various factors such as the pollution ratio, space volume, obstacle coefficient, washing path, and the UAV's own operating parameters and equipment status in each partition space. The aim is to achieve scientific and rational task allocation, enabling the UAV swarm to cooperate efficiently and complete the washing task of the transmission tower insulators.

[0029] To ensure the continuous and efficient execution of tasks and allow for real-time adjustment of task allocation when a drone swarm encounters an interruption during a live-water flushing mission, the following additional measures are proposed: The task update module is used to update the task allocation results in real time based on the current three-dimensional contamination heat map when a task interruption is detected during the execution of an electrified water flushing task by the drone swarm based on the task allocation results.

[0030] In this embodiment, the live-line water washing task refers to the task of using a drone equipped with relevant equipment to spray water onto the insulators of the transmission tower to remove surface contaminants while the transmission line is energized. This task requires the drone to accurately wash the insulators while ensuring its own safety and the normal operation of the transmission line, in order to restore or maintain the good electrical performance of the insulators and ensure the safe and stable operation of the transmission line.

[0031] In this embodiment, mission interruption refers to various situations that occur during the execution of an electrified water rinsing task by a drone swarm based on the task allocation results, preventing the rinsing task from continuing as originally planned. For example, drone equipment malfunctions, such as power system failures or communication failures, may prevent the drones from flying normally or performing rinsing operations; external environmental factors, such as sudden severe weather (heavy rain, strong winds, etc.), may also interfere, affecting drone flight safety and rinsing effectiveness; and problems directly related to the rinsing operation, such as water flow interruptions, may also occur. All of these situations fall under the category of mission interruption.

[0032] In this embodiment, the task allocation results are updated in real time based on the current 3D pollution heat map. This means that when a task interruption is detected, the system re-analyzes the pollution status of the transmission tower insulators in the target area based on the latest generated 3D pollution heat map. Since the pollution status may change over time or due to other factors after a task interruption, the current 3D pollution heat map reflects the latest pollution distribution and other information. Based on this latest information, the system performs multi-directional uniform clustering of pollution in the 3D pollution heat map again, re-analyzes the comprehensive cost of allocating each second partition space of each type of pollution to each UAV, and then re-allocates the partitioned tasks to the UAV swarm, generating new task allocation results. This ensures that the flushing task can continue to be carried out efficiently and scientifically, minimizing the impact of task interruption on the overall live-line flushing operation and ensuring the safe and stable operation of the transmission line.

[0033] like Figure 2 As shown, in order to generate a three-dimensional pollution heat map that intuitively reflects the pollution status of transmission tower insulators by collecting pollution status data and combining it with geographic information, a heat map generation module is proposed, including: The pollution status acquisition submodule is used to collect pollution status data of the surface of all transmission tower insulators using an infrared thermal imager and an ultraviolet detector mounted on a drone. The heat map generation submodule is used to generate a three-dimensional pollution heat map based on the pollution status data of all transmission tower insulator surfaces and combined with GIS geographic information.

[0034] In this embodiment, infrared thermal imagers and ultraviolet detectors mounted on a drone collect data on the contamination status of all transmission tower insulators. This utilizes the characteristics of these two instruments to obtain relevant information about the contamination on the insulator surface. Infrared thermal imagers can detect differences in thermal radiation on an object's surface. Since the presence of contamination alters the thermal characteristics of the insulator surface, the distribution and extent of contamination can be inferred by analyzing the thermal images. For example, areas with more contamination exhibit different thermal radiation than clean areas, appearing as different colors or shades of gray on the thermal image. Ultraviolet detectors are used to detect corona discharge phenomena on the insulator surface caused by contamination and other factors. The intensity and distribution of corona discharge are closely related to the contamination status, thus providing information about the contamination. The drone, with its mobility, can inspect all transmission tower insulators at close range and from multiple angles, comprehensively collecting this contamination status data.

[0035] In this embodiment, GIS geographic information refers to the information contained in a Geographic Information System (GIS). It integrates geospatial data of the target area, such as topography, geographic coordinates, and the spatial distribution of transmission towers. This information is stored and managed digitally, and can intuitively display the geographical features of the target area and the positional relationships of the transmission towers within it. For example, GIS geographic information can clearly indicate whether a transmission tower is located in a mountainous area or a plain, as well as its precise latitude and longitude coordinates, providing a geospatial positioning basis for subsequently generating a three-dimensional pollution heat map by combining pollution status data.

[0036] In this embodiment, a three-dimensional pollution heat map is generated based on the pollution status data of all transmission tower insulator surfaces and combined with GIS geographic information. This involves fusing pollution status data collected by infrared thermal imagers and ultraviolet detectors with GIS geographic information. Using the spatial positioning information provided by GIS, the pollution status data is accurately mapped to the corresponding transmission tower insulator locations. Then, it is displayed in a three-dimensional visualization, using different visual elements such as color, brightness, or height to represent the degree and distribution of pollution. For example, in the three-dimensional heat map, red represents severely polluted areas, and green represents lightly polluted areas. The actual spatial location of these areas is determined based on GIS geographic information, thereby generating a clear and comprehensive three-dimensional pollution heat map that displays the pollution status of all transmission tower insulator surfaces within the target area, providing clear and accurate basic data for subsequent task planning and analysis.

[0037] To obtain reasonable waste distribution space sets under various preset partitioning dimensions by dividing, clustering, and further dividing a 3D waste heatmap at different scales, a heatmap partitioning module is proposed, including: The first division submodule is used to divide the three-dimensional contamination heat map into multiple division spatial scales of no less than the preset division spatial scale, obtain multiple first division spaces under each division spatial scale, and regard all first division spaces under all division spatial scales as contamination excess spaces with a contamination ratio of no less than the preset contamination ratio threshold. The second partitioning submodule is used to cluster and partition all the excessive filth space to obtain multiple filth concentration areas in the three-dimensional filth heat map. The third sub-module is used to determine the number of divisions based on the number of drones in the drone swarm, and to further divide the three-dimensional pollution heat map based on the number of divisions, multiple preset division dimensions, and all pollution concentration areas to obtain the pollution distribution space set under each preset division dimension.

[0038] In this embodiment, the preset spatial scale is a pre-defined standard for dividing the three-dimensional pollution heat map. This scale determines the minimum size of each spatial unit when initially dividing the three-dimensional pollution heat map. It is the basic parameter for subsequent spatial division and analysis, and its value is determined based on factors such as the actual distribution of transmission towers, the required pollution detection accuracy, and computing resources. For example, if the transmission towers are densely distributed, a smaller preset spatial scale may be set to more accurately analyze the pollution situation in each area.

[0039] In this embodiment, multiple spatial scales refer to a series of spatial division standards of different sizes set based on a preset spatial scale. By employing multiple spatial scales, the three-dimensional pollution heat map can be analyzed at different granularities to obtain more comprehensive pollution distribution information. For example, in addition to the preset spatial scale, scales one and two levels larger may also be set, which allows for both a macroscopic understanding of the general pollution distribution and a microscopic analysis of local areas.

[0040] In this embodiment, the three-dimensional pollution heat map is spatially divided into multiple scales, each no smaller than a preset scale, to obtain multiple first-division spaces at each scale. This involves dividing the three-dimensional pollution heat map into spatial units of different sizes according to pre-defined standards of no smaller than the preset scale. Based on three-dimensional spatial coordinates, the entire target area is divided into small spatial blocks according to different scale requirements; each such block is a first-division space. For example, assuming the preset scale is a cube with a side length of 1 meter, the entire three-dimensional pollution heat map will be divided into many cube spaces with a side length of 1 meter; these are the first-division spaces at that scale. If there is a larger scale, such as a cube with a side length of 5 meters, another set of first-division spaces of different sizes will be obtained.

[0041] In this embodiment, the pollution percentage within each first-divided space is determined as follows: First, the number of transmission tower insulators included in each first-divided space and the total amount of pollution on these insulators are determined (measured by the product of the polluted area and the degree of pollution). This can be achieved by combining previously collected pollution data with spatial division information to determine the relevant data for the insulators in each space. Then, the ratio of the total pollution in that space to the total pollution in the entire target area is calculated. This ratio is the pollution percentage within each first-divided space. For example, if the total pollution of the insulators in a certain first-divided space is 10, and the total pollution in the entire target area is 1000, then the pollution percentage in that first-divided space is 10 ÷ 1000 = 1%.

[0042] In this embodiment, the preset contamination percentage threshold is a pre-defined proportion used to filter out areas with relatively high levels of contamination. After calculating the contamination percentage within each first-divided space, it is compared with the preset contamination percentage threshold. If the contamination percentage of a certain first-divided space is not less than this threshold, it is considered that the amount of contamination in that space is relatively high, belonging to an area that requires special attention, and will be further processed as a space with excessive contamination. For example, if the preset contamination percentage threshold is set to 5%, then a first-divided space with a contamination percentage of 5% or higher will be identified as a space with excessive contamination.

[0043] To achieve effective clustering of areas with concentrated pollution by rationally dividing the regions into clusters and determining their overall concentration, a second partitioning submodule is proposed, including: The first division unit is used to determine the physical center location of each filth concentration area, and arbitrarily divide all filth concentration areas to obtain multiple area clusters. It calculates the distance between the physical center locations of any two filth concentration areas in each area cluster, and calculates the ratio of each distance to a preset distance as the relative deviation of the corresponding two filth concentration areas. Based on the relative deviation of the two filth concentration areas, it calculates the concentration degree of the two filth concentration areas. The product of the average concentration degree of all two filth concentration areas in each area cluster and the total number of all filth concentration areas in the corresponding area cluster is taken as the comprehensive concentration degree of each area cluster. The first judgment unit is used to determine whether the average of the comprehensive concentration of all currently obtained regional clusters is not less than the preset concentration threshold. If so, all the polluted concentrated areas in each currently obtained regional cluster are merged to obtain a polluted concentrated area. Otherwise, all the polluted concentrated areas are arbitrarily re-divided to obtain multiple new regional clusters until the average of the comprehensive concentration of all the latest obtained regional clusters is not less than the preset concentration threshold. Then, all the polluted concentrated areas in each latest obtained regional cluster are merged to obtain a polluted concentrated area.

[0044] In this embodiment, the physical center location of each contamination area refers to the point in three-dimensional space that represents the geometric center of that contamination area. For a specific contamination area, the physical center location can be determined by mathematically calculating all its contained spatial locations (such as three-dimensional coordinates), for example, by averaging the coordinate values. This location point helps to measure the area's positional distribution in the entire three-dimensional space, as well as its spatial relationship with other contamination areas. For example, when analyzing a three-dimensional contamination heatmap, clearly defining the physical center location of each contamination area provides a visual understanding of the distances and distribution patterns between different areas, offering spatial location references for subsequent clustering and task allocation.

[0045] In this embodiment, the concentration degree of each pair of contaminated areas is calculated based on the relative deviation between them. The specific process is as follows: First, the relative deviation between each pair of contaminated areas has been obtained (i.e., the ratio of the distance between the physical centers of each pair of contaminated areas in each cluster to a preset distance). Then, the concentration degree is determined by taking the difference between 1 and the relative deviation.

[0046] In this embodiment, the preset concentration threshold is a pre-set standard value used to judge whether the division of region clusters is reasonable. After dividing all the polluted areas and calculating the comprehensive concentration of each region cluster (the product of the average concentration of all pairs of polluted areas in each region cluster and the total number of polluted areas in the corresponding region cluster), the average comprehensive concentration of all the currently obtained region clusters is compared with the preset concentration threshold. If the average is not less than the threshold, it means that the polluted areas in the currently divided region clusters are spatially clustered and meet the requirements. At this time, all the polluted areas in the current region cluster can be merged to obtain the final polluted area. If the average is less than the threshold, it means that the current region cluster division is not reasonable and the areas are not clustered tightly enough. All polluted areas need to be re-divided arbitrarily until the average comprehensive concentration of all the newly obtained region clusters is not less than the preset concentration threshold before merging. For example, the preset concentration threshold is set to 40. When the calculated average comprehensive concentration of all region clusters is greater than or equal to 40, the region cluster division is considered reasonable and subsequent operations can be performed.

[0047] To obtain a set of equally distributed pollution spaces that meets the requirements of average level by dividing a three-dimensional pollution heatmap and analyzing the characteristics of each division space, a third division submodule is proposed, including: The second division unit is used to divide the three-dimensional pollution heat map based on the number of divisions and multiple preset division dimensions, to obtain multiple second division spaces under each preset division dimension, and there are no two or more second division spaces that simultaneously intersect with the pollution concentration area in all the second division spaces under each preset division dimension. The spatial analysis unit is used to analyze the contamination percentage, spatial volume, and obstacle coefficient of each second partitioned space; The second judgment unit is used to determine whether the average degree of each dimension of the filth ratio, space volume, and obstacle coefficient of all second division spaces under each preset division dimension is less than the preset average degree threshold. If so (i.e., at least one dimension has an average degree less than the preset average degree threshold), the filth heat map is re-divided based on the corresponding preset division dimension until the average degree of the filth ratio, space volume, obstacle coefficient, and total flushing cost of all the latest obtained second division spaces is not less than the preset average degree threshold. Then, all the latest obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension. Otherwise, all the currently obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension.

[0048] In this embodiment, the number of divisions is a value determined based on the number of drones in the drone swarm. It determines the number of parts the 3D contamination heatmap will be divided into during subsequent partitioning, aiming to rationally allocate the entire target area to different drones to achieve a balanced task distribution. For example, if there are 5 drones, the number of divisions might be set to 5, thus roughly dividing the 3D contamination heatmap into 5 parts, each corresponding to the working area of ​​one drone.

[0049] In this embodiment, the three-dimensional pollution heat map is divided based on the number of divisions and multiple preset division dimensions to obtain multiple second division spaces under each preset division dimension. This means that the three-dimensional pollution heat map is further divided according to a pre-set division dimension (such as vertical direction, horizontal direction, topography, etc.) and combined with the previously determined number of divisions. For example, when dividing by spatial location dimension, the three-dimensional space is evenly divided into 5 parts along a certain direction according to the number of divisions of 5. Each part is a second division space under the preset division dimension; this allows for the subdivision of the target area from different angles, providing a more detailed regional division basis for subsequent task allocation.

[0050] In this embodiment, analyzing the contamination percentage, space volume, and obstacle coefficient of each second-divided space is an evaluation of multiple characteristics of each second-divided space. The contamination percentage measures the proportion of contamination in that space relative to the total contamination in the entire target area, reflecting the severity of contamination in that area. The space volume refers to the size of the second-divided space in three-dimensional space, which affects the workload and difficulty of the rinsing operation. The obstacle coefficient indicates the degree to which obstacles (such as other electrical equipment, natural obstacles, etc.) within the space hinder the drone rinsing operation; a higher coefficient indicates a greater impact of the obstacles on the rinsing operation. The obstacle coefficient is determined by: First, identify the types of obstacles that may exist in the area, such as other electrical equipment (e.g., adjacent transmission lines, transformers, etc.) and natural obstacles (e.g., trees, mountains, etc.). For each second partitioned space, we determine the obstacle coefficient as follows: Suppose that within a second partitioned space, there are 3 adjacent power lines and 2 trees affecting the drone's washing operation. We assign base impact values ​​to different types of obstacles. For example, each adjacent power line has a significant impact on the drone's flight, so we assign a base impact value of 3; each tree has a certain impact on the drone's flight, so we assign a base impact value of 1. Therefore, based solely on the number of obstacles and their base impact values, the initial impact value of obstacles in this space is 3×3 + 2×1 = 11.

[0051] If these obstacles are concentrated on one side of the space, they will have a greater impact on the drone's flight path planning and operational difficulty compared to a uniform distribution. We can set a distribution adjustment coefficient based on the dispersion of the obstacle distribution. Assuming that the evaluation shows that the distribution adjustment coefficient is 1.5 for this concentrated distribution, then the impact value after considering the distribution becomes 11 × 1.5 = 16.5.

[0052] Different drones have different obstacle avoidance capabilities. If the drone participating in the mission is equipped with an advanced obstacle avoidance system, it will have a stronger ability to cope with such obstacles, and we can set a response capability adjustment coefficient. Assuming that this coefficient is 0.8 based on the drone's performance, then the final obstacle coefficient for this second partitioned space is 16.5 × 0.8 = 13.2.

[0053] By comprehensively considering factors such as the type, quantity, and distribution of obstacles, as well as the drone's response capabilities, a reasonable obstacle coefficient can be determined for each second partition space, providing an accurate basis for subsequent calculations of flushing costs, overall costs, and task allocation.

[0054] If a second partition space has a high percentage of contamination, it indicates that the area is severely contaminated. A large space volume means a large amount of cleaning work, and a high obstacle coefficient means that the drone will encounter more obstacles during cleaning. All of this information together provides a basis for assessing the cleaning difficulty and cost of each second partition space.

[0055] In this embodiment, the average degree of each dimension of the dirt percentage, space volume, and obstacle coefficient of all second-divided spaces under each preset division dimension (i.e., the average degree of dirt percentage, space volume, and obstacle coefficient of all second-divided spaces) is a measure of the uniformity of the characteristics of all second-divided spaces in these three aspects under each preset division dimension. The average degree is determined by calculating the average values ​​of dirt percentage, space volume, and obstacle coefficient for all second-divided spaces and analyzing the dispersion of these values ​​(e.g., calculating variance and other statistics). For example, if the dirt percentage values ​​of all second-divided spaces are relatively close under a certain preset division dimension, it indicates a high degree of average, meaning that these spaces are relatively evenly distributed in terms of dirt percentage characteristics; conversely, if the values ​​differ significantly, the average degree is low. A high degree of average helps to achieve a balanced distribution of drone washing tasks, avoiding some drones being overloaded or underloaded.

[0056] In this embodiment, the preset average degree threshold is a pre-set standard used to determine whether the total average degree of the pollution ratio, space volume, and obstacle coefficient of all second partition spaces under each preset partition dimension meets the requirements. After calculating the average degree of these features under each preset partition dimension, it is compared with the preset average degree threshold. If the average degree is less than the threshold, it indicates that the distribution of each second partition space on this feature is not uniform enough under the current partition, and the pollution heatmap needs to be further partitioned based on the corresponding preset partition dimension; the partition is considered reasonable only when the average degree of the three features of all the latest obtained second partition spaces is not less than the preset average degree threshold, and all the latest obtained second partition spaces are regarded as the pollution evenly distributed space set under the corresponding preset partition dimension. For example, the preset average degree threshold is set to 0.8 (assuming that the average degree quantification value obtained by calculating variance, etc., is in the range of 0-1). When the calculated average degree is less than 0.8, it needs to be re-partitioned; when it reaches or exceeds 0.8, the partitioning result meets the requirements.

[0057] To comprehensively consider multiple factors in analyzing the overall cost of allocating each second partition space to each UAV, and based on this, construct a matrix, generate the optimal task allocation result, and thus rationally allocate tasks to the UAV swarm, a task allocation module is proposed, including: The comprehensive cost analysis unit is used to analyze the comprehensive cost of allocating each second partition space of each type of contamination space to each UAV based on the contamination ratio, space volume, obstacle coefficient, and flushing path of each second partition space in the contamination space set. The cost matrix construction unit is used to construct a comprehensive cost matrix for each preset partition dimension based on the comprehensive cost of allocating each type of pollution space to each second partition space and then to each UAV. The task optimal allocation unit is used to generate the optimal task allocation result for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The final task allocation unit is used to allocate partitioned tasks to the UAV swarm based on the optimal task allocation result under each preset partition dimension, and obtain the task allocation result.

[0058] In this embodiment, the comprehensive cost of allocating each type of contamination space to each drone within a second partitioned space is a value derived from a comprehensive consideration of multiple factors. This value is used to assess the overall cost or difficulty of assigning a specific second partitioned space to a particular drone for cleaning. It combines factors such as the contamination percentage, space volume, obstacle coefficient, and cleaning path of the second partitioned space. For example, a high contamination percentage means potentially greater cleaning difficulty and required more resources; a large space volume results in a larger cleaning workload; a high obstacle coefficient increases the difficulty and risk of drone operation; and a long cleaning path consumes more energy. By comprehensively analyzing these factors, the comprehensive cost of allocating the space to different drones is derived, providing a quantitative basis for task allocation.

[0059] In this embodiment, a comprehensive cost matrix is ​​constructed based on the comprehensive cost of allocating each second partition space to each drone when allocating each type of contamination space to each preset partition dimension. This matrix presents the comprehensive cost of allocating each second partition space to each drone in matrix form. Rows in the matrix may represent different drones, columns may represent different second partition spaces, and matrix elements represent the corresponding comprehensive costs. For example, assuming there are 5 drones and 5 second partition spaces, the matrix size might be 5×5, and the element (i,j) in the matrix represents the comprehensive cost of allocating the j-th second partition space to the i-th drone. Such a matrix clearly shows the cost of different drones performing different area washing tasks under each preset partition dimension, facilitating subsequent optimization calculations for task allocation.

[0060] In this embodiment, the optimal task allocation result under each preset partition dimension is obtained by analyzing and calculating the comprehensive cost matrix under each preset partition dimension, resulting in a task allocation scheme that minimizes the overall task execution cost or maximizes efficiency. For example, by using optimization algorithms (such as the Hungarian algorithm) to process the comprehensive cost matrix, considering the capabilities of each UAV and the characteristics of each second partition space, an allocation method is found that minimizes the total cost of all UAVs performing tasks or minimizes the task completion time. This allocation method is the optimal task allocation result under that preset partition dimension. It ensures that the task allocation of the UAV swarm reaches a relatively optimal state under a specific partition dimension.

[0061] In this embodiment, the drone swarm is assigned zoned tasks based on the optimal task allocation results under each preset division dimension. Obtaining the task allocation result means that after obtaining the optimal task allocation results under each preset division dimension, the situation of each dimension is comprehensively considered, and the most suitable allocation method is selected to finally determine the specific zoned tasks for the drone swarm. For example, an optimal allocation may be obtained under the spatial location dimension, and another optimal allocation may be obtained under the degree of contamination dimension. By comparing the feasibility, efficiency, and other factors of the allocation schemes under different dimensions in practical applications, one of them is selected as the final task allocation result, clarifying the specific area that each drone is responsible for washing, realizing the scientific and rational nature of the drone swarm task allocation, and improving the efficiency of electrified water washing. For example: The first allocation scheme is to divide the entire target area into five secondary partition spaces according to their spatial location. Each space has a relatively regular shape, roughly in the form of a cube.

[0062] Drone 1 is responsible for the top left corner area, Drone 2 is responsible for the top right corner area, Drone 3 is responsible for the bottom left corner area, Drone 4 is responsible for the bottom right corner area, and Drone 5 is responsible for the middle area.

[0063] Feasibility analysis: From a spatial perspective, each area is relatively independent, with minimal interference between drones, and the flight path planning for each drone is relatively simple, making it highly feasible.

[0064] Efficiency Analysis: However, after analyzing the dirt conditions of each area, it was found that the central area was extremely dirty and required a long rinsing time, while the dirt in the other four corner areas was relatively light. This would result in the drone 5 having an excessive workload and reduced overall efficiency.

[0065] The second allocation scheme: The target area is divided into lightly, moderately, and heavily polluted areas based on the degree of contamination, and further subdivided according to the number of drones. Two high-performance drones (Drone 1 and Drone 2) are responsible for the two heavily polluted areas; two medium-performance drones (Drone 3 and Drone 4) are responsible for the moderately polluted area, each also responsible for one sub-area; and Drone 5, with slightly lower performance, is responsible for the lightly polluted area.

[0066] Feasibility Analysis: This allocation considers the matching of drone performance with the degree of contamination. Using high-performance drones to handle heavily contaminated areas is theoretically highly feasible. However, heavily contaminated areas may be spatially dispersed, requiring drones 1 and 2 to frequently switch flight areas, increasing flight distance and time costs.

[0067] Efficiency Analysis: From the perspective of overall flushing efficiency, theoretically, by rationally allocating drones with different performance levels to areas with corresponding levels of dirt, the advantages of each drone can be fully utilized to improve flushing efficiency. However, as mentioned above, increased flight distance may offset some of the efficiency advantages.

[0068] By comparing these two allocation schemes under different dimensions: In terms of feasibility: the spatial location-based approach has a simple flight path and minimal mutual interference, but it does not adequately consider the degree of contamination; the contamination-level approach considers the matching of UAV performance with contamination levels, but the flight distance may increase. Overall, although the contamination-level approach has a more complex flight path, it is still achievable through proper planning, and it has better adaptability to the mission itself, making it more feasible.

[0069] In terms of efficiency: the spatial location dimension solution has low overall efficiency due to uneven workload; although the pollution level dimension solution has flight distance issues, it can be optimized through reasonable planning and can give full play to the performance of drones, making it more efficient overall.

[0070] Taking into account factors such as feasibility and efficiency, an allocation scheme based on the degree of contamination was ultimately selected as the final task allocation result. It was determined that drone 1 and drone 2 would each be responsible for two sub-areas of the heavily contaminated area, drone 3 and drone 4 would each be responsible for a sub-area of ​​the moderately contaminated area, and drone 5 would be responsible for the lightly contaminated area, thus defining the specific area each drone would be responsible for washing.

[0071] To analyze the costs from aspects such as washing, distance, and equipment matching, and to comprehensively calculate the overall cost of allocating the second partition space to the UAV, a comprehensive cost analysis unit is proposed, including: The flushing cost analysis subunit is used to calculate the flushing cost of the corresponding second partition space based on the proportion of dirt, space volume, and obstacle coefficient of each second partition space in each type of dirt-sharing space set. The distance cost analysis subunit is used to plan the flushing path for each second partition space in the space for each type of filth distribution, and to analyze the distance cost of the corresponding second partition space based on the flushing path. The matching cost analysis subunit is used to analyze the equipment matching cost between each drone and each second partition space of each type of pollution equal distribution space based on the working parameter range and equipment status of each drone in the drone swarm, as well as the pollution ratio, space volume, and obstacle coefficient of each second partition space in each pollution equal distribution space set. The comprehensive cost analysis subunit is used to calculate the comprehensive cost of allocating each second partition space of each type of filth distribution space to each drone based on the equipment matching cost of each drone with each second partition space of each type of filth distribution space set, the flushing cost of the corresponding second partition space, and the distance cost.

[0072] In this embodiment, the flushing cost for each second partitioned space is calculated based on the percentage of contamination, the space volume, and the obstacle coefficient of each type of contamination-divided space. A higher percentage of contamination means more water, longer flushing time, and more complex operations are required to remove the contamination, thus increasing the flushing cost. A larger space volume means a wider coverage area for the drone, increasing the flushing workload and also raising the flushing cost. A higher obstacle coefficient means more obstacles the drone encounters during flushing, potentially requiring more complex flight paths or additional measures to avoid obstacles, which also increases the flushing cost. For example, a calculation formula incorporating these factors can be established, such as Flushing Cost = Contamination Percentage × Flushing Difficulty Coefficient + Space Volume × Flushing Cost per Unit Volume + Obstacle Coefficient × Obstacle Treatment Cost, to quantify the flushing cost for each second partitioned space.

[0073] In this embodiment, a planned flushing path is planned for each second partitioned space in the space where each type of pollution is evenly distributed, and the distance cost of the corresponding second partitioned space is analyzed based on the flushing path. Planning the flushing path needs to consider factors such as the location of the transmission tower insulators, the shape of the second partitioned space, and the distribution of obstacles to ensure that the UAV can complete the flushing task safely and efficiently. The distance cost is directly related to the length of the flushing path; the longer the path, the more energy the UAV requires to fly, and the greater the equipment wear, thus resulting in a higher distance cost. For example, using a path planning algorithm, combined with spatial information and obstacle information from a 3D pollution heat map, the optimal path from the UAV's starting position to each flushing point within the second partitioned space is planned. Then, based on parameters such as path length and UAV energy consumption per unit distance, the distance cost of the corresponding second partitioned space is calculated.

[0074] In this embodiment, the equipment matching cost between each drone and each second partition space in each pollution-saturated space is analyzed based on the operating parameter range and equipment status of each drone in the drone swarm, as well as the pollution percentage, space volume, and obstacle coefficient of each second partition space in each pollution-saturated space set. Different drones have different operating parameter ranges, such as flight speed, payload capacity, and water jet pressure, and their equipment status also varies, such as age and performance stability. The pollution percentage, space volume, and obstacle coefficient of each second partition space impose different requirements on the drone's operating parameters. If the drone's operating parameters do not match the requirements of the second partition space, additional adjustments may be required, or the task may not be completed efficiently, resulting in an equipment matching cost. For example, if a second partition space is heavily polluted and has a large space volume, requiring drones with high water jet pressure and large payload capacity, but a drone has insufficient water jet pressure or limited payload capacity, then assigning that drone to this space will result in a high equipment matching cost. This cost can be quantified by analyzing the degree of matching between the drone and the characteristics of the second partition space. For example: Suppose we have a drone swarm consisting of three drones (drone A, drone B, and drone C) that needs to perform live water flushing on the insulators of a certain power transmission tower area. This area has been divided into multiple secondary partition spaces. We will take three of these secondary partition spaces (space 1, space 2, and space 3) as examples to explain the analysis process of equipment matching costs.

[0075] Drone A: Maximum flight speed is 5m / s, maximum payload is 10kg, water spray pressure range is 3-6MPa, the equipment is relatively new and has stable performance.

[0076] Drone B: Maximum flight speed is 4m / s, maximum payload is 8kg, water spray pressure range is 2-5MPa, the equipment has a long service life, and its stability is average.

[0077] Drone C: Maximum flight speed is 6m / s, maximum payload is 12kg, water spray pressure range is 4-8MPa, the equipment is brand new and in good condition.

[0078] Characteristics of the second partitioned space: Space 1: 30% contamination, 20m³ volume, obstacle coefficient 1.2. This space is moderately contaminated, relatively large, and contains a number of obstacles, which somewhat hinder drone flight and washing operations.

[0079] Space 2: 50% filth, 15m³ volume, obstacle coefficient 1.5. This space is highly filthy and of moderate size, but has many obstacles, increasing the difficulty of cleaning.

[0080] Space 3: 10% contamination, 30m³ volume, obstacle coefficient 1.0. This space has a low level of contamination but a large volume, although it has few obstacles.

[0081] Cost of matching equipment between drone A and space 1: Regarding payload: The rinsing task in Space 1 is expected to require carrying 5kg of cleaning water and related equipment. Drone A has a maximum payload of 10kg, which can easily meet the requirements. The payload is well matched, and the cost of this item is low, so we assume it is assigned a value of 1.

[0082] Regarding flight speed: Since there are certain obstacles in space 1, the drone needs to fly flexibly. The maximum flight speed of drone A is 5m / s, which can handle the situation well. The speed matching cost is low, so it is assigned a value of 2.

[0083] Regarding water spray pressure: Space 1 has a moderate level of contamination. Based on experience, a water spray pressure of around 4 MPa is suitable. The water spray pressure range of drone A is 3-6 MPa, which meets the requirements. The cost of matching the water spray pressure is 1.

[0084] Regarding equipment status: Drone A is relatively new and has stable performance. It can operate stably in the relatively complex environment of Space 1. The equipment status matching cost is 1.

[0085] Overall equipment matching cost: Considering the importance of each factor, assuming the weights of load, flight speed, water jet pressure, and equipment status are 0.3, 0.2, 0.3, and 0.2 respectively, the equipment matching cost between UAV A and space 1 is calculated by weighting: 1×0.3+2×0.2+1×0.3+1×0.2=1.2.

[0086] Cost of matching equipment between drone B and space 2: Regarding load capacity: Space 2 is expected to carry 6kg of supplies for rinsing, while Drone B has a maximum load capacity of 8kg. Although this meets the requirement, it is close to the load capacity limit. The load capacity matching cost is relatively high, so a value of 3 is assigned.

[0087] Regarding flight speed: Space 2 has many obstacles, requiring fast and flexible flight. Drone B's maximum flight speed is 4m / s, which is relatively slow, and the speed matching cost is 3.

[0088] Regarding water spray pressure: Space 2 is highly polluted and requires a water spray pressure of at least 5MPa. Drone B's water spray pressure limit is 5MPa, which just meets the requirement. The cost of matching the water spray pressure is 2.

[0089] Regarding equipment status: UAV B has average equipment stability and may experience malfunctions when operating in the complex environment of Space 2. The equipment status matching cost is 3.

[0090] Overall equipment matching cost: Calculated according to the same weights as above, the equipment matching cost between UAV B and space 2 is 3×0.3+3×0.2+2×0.3+3×0.2=2.7.

[0091] The above examples demonstrate that by comprehensively analyzing various factors, including the operating parameter range of the drone, its equipment status, and the characteristics of the second partitioned space, the equipment matching cost between each drone and each second partitioned space can be determined, providing an important reference for task allocation.

[0092] In this embodiment, the comprehensive cost of allocating each second partitioned space of each type of contamination distribution space to each drone is calculated based on the equipment matching cost of each drone with each type of contamination distribution space in each second partitioned space, the rinsing cost of the corresponding second partitioned space, and the distance cost. This comprehensively considers the above three cost factors to fully evaluate the total cost of allocating a specific second partitioned space to a drone. The calculation of the comprehensive cost can provide a more accurate and comprehensive basis for task allocation, ensuring that task allocation not only takes into account the difficulty and cost of the rinsing work itself, but also the adaptability of drone equipment to the task and the cost of flight distance. For example, by setting weighting coefficients (e.g., the weight of the equipment matching cost between each drone and each second partition space in each type of contamination space set is 0.4, the weight of the rinsing cost in the corresponding second partition space is 0.3, and the weight of the distance cost is 0.3), the equipment matching cost, rinsing cost, and distance cost are weighted and summed to obtain the formula for calculating the comprehensive cost: Comprehensive cost = Equipment matching cost × Matching weight + Rinsing cost × Rinsing weight + Distance cost × Distance weight. The weighting coefficients can be adjusted according to the actual situation and importance, thereby obtaining the comprehensive cost when each second partition space in each type of contamination space set is allocated to each drone.

[0093] To generate multiple hypothetical task allocation schemes based on the comprehensive cost matrix, the optimal task allocation result under each preset partition dimension is selected through multi-dimensional evaluation indicators, and an optimal task allocation unit is proposed, including: The task allocation subunit is used to generate multiple hypothetical task allocation schemes for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The multidimensional evaluation subunit is used to generate a comprehensive evaluation index value for each hypothetical task allocation scheme based on the total allocation cost and imbalance coefficient of each hypothetical task allocation scheme. The result filtering subunit is used to filter out the optimal task allocation result under each preset partition dimension from all the hypothetical allocation schemes under each preset partition dimension based on the comprehensive evaluation index value of all hypothetical allocation schemes under each preset partition dimension.

[0094] In this embodiment, generating multiple hypothetical task allocation schemes for each preset partition dimension based on the comprehensive cost matrix is ​​achieved by utilizing the information in the comprehensive cost matrix to construct different task allocation possibilities. Since the comprehensive cost matrix displays the cost of each drone performing a flushing task in each second partition space, multiple hypothetical task allocation schemes can be generated through different combinations. For example, for a comprehensive cost matrix containing 4 drones and 4 second partition spaces, one hypothetical task allocation scheme might be: drone 1 is responsible for space 1, drone 2 for space 3, drone 3 for space 2, and drone 4 for space 4. Multiple such schemes can be generated through exhaustive search or specific algorithms, providing a basis for subsequent selection of the optimal scheme.

[0095] In this embodiment, the total allocation cost and imbalance coefficient for each hypothetical task allocation scheme are two important indicators for evaluating the scheme. The total allocation cost refers to the sum of the combined costs incurred by all UAVs performing their respective tasks under the hypothetical allocation scheme. The imbalance coefficient measures the degree of unevenness in task allocation among the UAVs. For example, if the task costs borne by each UAV are similar, it indicates a relatively balanced task allocation and a low imbalance coefficient; conversely, if the task cost borne by one UAV is much higher than that of other UAVs, it indicates an unbalanced task allocation and a high imbalance coefficient. The calculation of the imbalance coefficient may involve the calculation of statistical measures such as the variance and standard deviation of the task costs of each UAV to quantify the degree of balance in task allocation.

[0096] In this embodiment, a comprehensive evaluation index value for each hypothetical task allocation scheme is generated based on the total allocation cost and imbalance coefficient of each scheme. This involves combining the total allocation cost and imbalance coefficient to obtain a single value that more comprehensively reflects the merits of each hypothetical task allocation scheme. For example: ; In the formula, This is a comprehensive evaluation index value for a single hypothetical allocation scheme. , , These are weighting coefficients (e.g., values ​​of 0.4, 0.5, and 0.1 respectively), determined through training with historical data. The average of the total costs allocated to history. The total cost of allocation for a single hypothetical allocation scheme is also the sum of the combined costs of all drones performing their missions. It is an exponential function with the natural constant e as its base, and e has a value of 2.72. The imbalance coefficient for a single hypothetical allocation scheme. For the number of drones, For the first The deviation between the total cost of allocating each drone and the average total cost of allocating them. The mean of the total cost deviation is used to allocate the total cost, and all of the above parameters are dimensionless.

[0097] In this embodiment, based on the comprehensive evaluation index value of all hypothetical task allocation schemes under each preset partition dimension, the optimal task allocation result under each preset partition dimension is selected from all hypothetical task allocation schemes under each preset partition dimension. Under each preset partition dimension, multiple hypothetical task allocation schemes have been generated, and their respective comprehensive evaluation index values ​​have been calculated. By comparing these index values, the hypothetical task allocation scheme with the smallest index value (or the optimal one selected according to the actual evaluation criteria) is selected as the optimal task allocation result under that preset partition dimension. For example, if there are 10 hypothetical task allocation schemes under a certain preset partition dimension, with comprehensive evaluation index values ​​of 10, 8, 12, 9, 7, 11, 6, 13, 9, and 8 respectively, then the hypothetical task allocation scheme with an index value of 6 is the optimal task allocation result under that preset partition dimension. In this way, relatively optimal task allocation results can be obtained under different preset partition dimensions, providing multiple candidate schemes for finally determining the task allocation of the UAV swarm, so as to further comprehensively consider and select the most suitable task allocation method.

[0098] This invention provides an implementation method for a remote collaborative control method of electro-hydraulic water flushing of unmanned aerial vehicles (UAVs), comprising: Generate a three-dimensional pollution heat map containing the pollution status of all transmission tower insulator surfaces within the target area; Based on all the excess space of pollution in the 3D pollution heat map, the 3D pollution heat map is divided into pollution uniform clusters in multiple directions to obtain pollution uniform space sets under multiple preset division dimensions; Based on the comprehensive cost matrix under each preset partitioning dimension, the drone swarm is assigned partitioned tasks to obtain the task allocation results.

[0099] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A remote collaborative control platform for electro-hydraulic water flushing of unmanned aerial vehicles (UAVs), characterized in that, include: The heat map generation module is used to generate a three-dimensional pollution heat map containing the pollution status of the surfaces of all transmission tower insulators within the target area; The heatmap uniform division module is used to perform multi-directional uniform clustering of the three-dimensional pollution heatmap based on all the pollution excess space in the three-dimensional pollution heatmap to obtain pollution uniform division space sets under multiple preset division dimensions. The task allocation module is used to allocate partitioned tasks to the UAV swarm based on the comprehensive cost matrix under each preset partitioning dimension, and obtain the task allocation results.

2. The remote collaborative control platform for electrified water flushing of unmanned aerial vehicles according to claim 1, characterized in that, Also includes: The task update module is used to update the task allocation results in real time based on the current three-dimensional contamination heat map when a task interruption is detected during the execution of an electrified water flushing task by the drone swarm based on the task allocation results.

3. The remote collaborative control platform for electrified water flushing of unmanned aerial vehicles according to claim 1, characterized in that, The heatmap generation module includes: The pollution status acquisition submodule is used to collect pollution status data of the surface of all transmission tower insulators using an infrared thermal imager and an ultraviolet detector mounted on a drone. The heat map generation submodule is used to generate a three-dimensional pollution heat map based on the pollution status data of all transmission tower insulator surfaces and combined with GIS geographic information.

4. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 1, characterized in that, The heatmap distribution module includes: The first division submodule is used to divide the three-dimensional contamination heat map into multiple division spatial scales of no less than the preset division spatial scale, obtain multiple first division spaces under each division spatial scale, and regard all first division spaces under all division spatial scales as contamination excess spaces with a contamination ratio of no less than the preset contamination ratio threshold. The second partitioning submodule is used to cluster and partition all the excessive filth space to obtain multiple filth concentration areas in the three-dimensional filth heat map. The third sub-module is used to determine the number of divisions based on the number of drones in the drone swarm, and to further divide the three-dimensional pollution heat map based on the number of divisions, multiple preset division dimensions, and all pollution concentration areas to obtain the pollution distribution space set under each preset division dimension.

5. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 4, characterized in that, The second sub-module includes: The first division unit is used to determine the physical center location of each filth concentration area, and arbitrarily divide all filth concentration areas to obtain multiple area clusters. It calculates the distance between the physical center locations of any two filth concentration areas in each area cluster, and calculates the ratio of each distance to a preset distance as the relative deviation of the corresponding two filth concentration areas. Based on the relative deviation of the two filth concentration areas, it calculates the concentration degree of the two filth concentration areas. The product of the average concentration degree of all two filth concentration areas in each area cluster and the total number of all filth concentration areas in the corresponding area cluster is taken as the comprehensive concentration degree of each area cluster. The first judgment unit is used to determine whether the average of the comprehensive concentration of all currently obtained regional clusters is not less than the preset concentration threshold. If so, all the polluted concentrated areas in each currently obtained regional cluster are merged to obtain a polluted concentrated area. Otherwise, all the polluted concentrated areas are arbitrarily re-divided to obtain multiple new regional clusters until the average of the comprehensive concentration of all the latest obtained regional clusters is not less than the preset concentration threshold. Then, all the polluted concentrated areas in each latest obtained regional cluster are merged to obtain a polluted concentrated area.

6. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 4, characterized in that, The third sub-module includes: The second division unit is used to divide the three-dimensional pollution heat map based on the number of divisions and multiple preset division dimensions, to obtain multiple second division spaces under each preset division dimension, and there are no two or more second division spaces that simultaneously intersect with the pollution concentration area in all the second division spaces under each preset division dimension. The spatial analysis unit is used to analyze the contamination percentage, spatial volume, and obstacle coefficient of each second partitioned space; The second judgment unit is used to determine whether the average degree of each dimension of the filth ratio, space volume, and obstacle coefficient of all second division spaces under each preset division dimension is less than the preset average degree threshold. If so, the filth heat map is re-divided based on the corresponding preset division dimension until the average degree of the filth ratio, space volume, obstacle coefficient, and total flushing cost of the latest obtained second division spaces is not less than the preset average degree threshold. Then, all the latest obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension. Otherwise, all the currently obtained second division spaces are regarded as the filth evenly distributed space set under the corresponding preset division dimension.

7. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 1, characterized in that, The task allocation module includes: The comprehensive cost analysis unit is used to analyze the comprehensive cost of allocating each second partition space of each type of contamination space to each UAV based on the contamination ratio, space volume, obstacle coefficient, and flushing path of each second partition space in the contamination space set. The cost matrix construction unit is used to construct a comprehensive cost matrix for each preset partition dimension based on the comprehensive cost of allocating each type of pollution space to each second partition space and then to each UAV. The task optimal allocation unit is used to generate the optimal task allocation result for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The final task allocation unit is used to allocate partitioned tasks to the UAV swarm based on the optimal task allocation result under each preset partition dimension, and obtain the task allocation result.

8. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 7, characterized in that, The comprehensive cost analysis unit includes: The flushing cost analysis subunit is used to calculate the flushing cost of the corresponding second partition space based on the proportion of dirt, space volume, and obstacle coefficient of each second partition space in each type of dirt-sharing space set. The distance cost analysis subunit is used to plan the flushing path for each second partition space in the space for each type of filth distribution, and to analyze the distance cost of the corresponding second partition space based on the flushing path. The matching cost analysis subunit is used to analyze the equipment matching cost between each drone and each second partition space of each type of pollution equal distribution space based on the working parameter range and equipment status of each drone in the drone swarm, as well as the pollution ratio, space volume, and obstacle coefficient of each second partition space in each pollution equal distribution space set. The comprehensive cost analysis subunit is used to calculate the comprehensive cost of allocating each second partition space of each type of filth distribution space to each drone based on the equipment matching cost of each drone with each second partition space of each type of filth distribution space set, the flushing cost of the corresponding second partition space, and the distance cost.

9. The remote collaborative control platform for live water flushing of unmanned aerial vehicles according to claim 1, characterized in that, The optimal task allocation unit includes: The task allocation subunit is used to generate multiple hypothetical task allocation schemes for each preset partition dimension based on the comprehensive cost matrix under each preset partition dimension. The multidimensional evaluation subunit is used to generate a comprehensive evaluation index value for each hypothetical task allocation scheme based on the total allocation cost and imbalance coefficient of each hypothetical task allocation scheme. The result filtering subunit is used to filter out the optimal task allocation result under each preset partition dimension from all the hypothetical allocation schemes under each preset partition dimension based on the comprehensive evaluation index value of all hypothetical allocation schemes under each preset partition dimension.

10. A method for remote collaborative control of an unmanned aerial vehicle (UAV) with electrified water flushing, characterized in that, include: Generate a three-dimensional pollution heat map containing the pollution status of all transmission tower insulator surfaces within the target area; Based on all the excess space of pollution in the 3D pollution heat map, the 3D pollution heat map is divided into pollution uniform clusters in multiple directions to obtain pollution uniform space sets under multiple preset division dimensions; Based on the comprehensive cost matrix under each preset partitioning dimension, the drone swarm is assigned partitioned tasks to obtain the task allocation results.

Images