Transmission, transformation and distribution unmanned aerial vehicle gridding deployment method based on GIS

Abstract

The invention discloses a transmission, transformation and distribution unmanned aerial vehicle gridding deployment method based on a geographic information system (GIS), and belongs to the technical field of power inspection. In order to solve the problems of low unmanned aerial vehicle utilization rate, low inspection efficiency and the like in the existing electric power inspection, the invention provides a new solution which comprises the following steps: firstly, collecting distribution data of electric power facilities and inputting the data into a GIS (Geographic Information System) platform; then, according to the distribution of electric power facilities and the cruising ability of the unmanned aerial vehicle, carrying out nest gridding site selection; thirdly, grid division is carried out on the electric power facilities by utilizing a GIS spatial analysis function; afterwards, an efficient unmanned aerial vehicle path planning algorithm is provided, and an optimal flight path is planned for the unmanned aerial vehicle; and finally, deploying and scheduling the unmanned aerial vehicle according to the planned path and the nest position. Through combination of the GIS technology and the unmanned aerial vehicle technology, efficient and accurate inspection of electric power facilities is realized, the inspection efficiency and quality are improved, and safe and stable operation of an electric power system is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of electric power inspection, and in particular to a grid deployment method of unmanned aerial vehicles (UAVs) for power transmission and distribution based on GIS. Background Art

[0002] In the field of power inspection, with the rapid development of drone technology, drones have gradually become an important tool for power transmission, transformation and distribution facility inspection. Traditional power inspection methods mainly rely on manual walking or ground vehicles, which is not only inefficient but also difficult to cover all power facilities, especially those located in remote areas or difficult to reach places. The introduction of drones has greatly improved the efficiency and coverage of inspections, making the inspection of power facilities more comprehensive and timely.

[0003] At present, the application of drones in power inspections is mainly concentrated in the following aspects: first, using drones equipped with high-definition cameras to conduct aerial photography and visual inspections of power facilities; second, using infrared thermal imagers and other equipment carried by drones to monitor the temperature of power facilities and promptly detect potential overheating problems; third, using the flying capabilities of drones to inspect power lines across obstacles, such as across rivers, mountains and other complex terrains.

[0004] However, although drones have shown great potential in power inspections, existing drone inspection technologies still have some obvious problems and shortcomings. First, drones have limited endurance and short single flight times, making it difficult to complete large-scale inspection tasks. This results in the need to frequently replace or recharge batteries in practical applications, which seriously affects inspection efficiency. Secondly, the path planning problem of drones is also a major challenge. In a complex power facility environment, how to plan the optimal flight path to ensure that drones can complete inspection tasks efficiently and safely is an urgent problem to be solved. In addition, existing drone inspection technologies lack intelligent nest site selection and grid management, resulting in inflexible and inefficient deployment and scheduling of drones.

[0005] In view of the above problems, the present invention proposes a grid deployment method for transmission, transformation and distribution UAVs based on Geographic Information System (GIS). This method aims to solve the problems of insufficient endurance, complex path planning and inflexible deployment and scheduling in existing UAV inspection technologies through accurate machine nest site selection, optimized grid division and efficient UAV path planning, thereby improving the efficiency and quality of power inspection. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a GIS-based grid deployment method for transmission and distribution UAVs, so as to solve the technical problems of insufficient endurance, complex path planning and inflexible deployment and scheduling existing in the UAV inspection technology in the field of power inspection.

[0007] In order to solve the above technical problems, the technical solution adopted by the present invention is: a grid deployment method of transmission and distribution UAV based on GIS, comprising the following steps: Step 1: Calculate the coverage radius based on the maximum flight time and geographical factors of the transmission, transformation and distribution joint inspection drone; Step 2: Use the coverage radius and polygon intersection set method PIPS to obtain the candidate area of ​​the machine nest; Step 3: Consider the deployment environment restrictions of the machine nest to determine the candidate points of the machine nest in each candidate area and obtain the candidate point set; Step 4: Using GIS and convex hull, calculate the Euclidean shortest path ESP between the candidate points of the machine nest in Step 3; Step 5: Combine ESP and weighted set coverage model to establish a UAV grid site selection model; Step 6: Use Python to call Gurobi to solve the global optimal solution of the model and obtain the location of the machine nest; Step 7: Use ArcGIS spatial analysis functions to perform coverage visualization and grid division, and build a UAV leapfrog flight network based on the Euclidean shortest path between the site selection points obtained in Step 4.

[0008] In the preferred solution, in Step 1, an impact factor is introduced to calculate the nest coverage radius , the specific expression is: (1) In the formula, In order to consider the margin left by the drone due to incomplete charging, battery aging, low temperature, and high-frequency operation, the value is [0, 1]; and are terrain influence factor and wind speed influence factor, respectively, with values ​​ranging from [0, 1]; In a windless environment, the test speed is The longest flight time measured when flying at a constant speed until the remaining battery power is 0%; It is the inspection time of power equipment; is the maximum flight time of the drone; the calculated distance is divided by 2 to ensure that the drone can return to the starting nest after the inspection.

[0009] In the preferred solution, the specific method and steps of step 2 are: using the ArcGIS fishing net tool to divide the study area into regular grids, and mapping all transmission and distribution transformer demands to the corresponding grids to obtain a demand grid; taking the vertex of the demand grid as the center, The intersection of multiple circular regions with a radius of is the coverage boundary of the required grid, and the intersection area of ​​multiple coverage boundaries is the candidate machine nest area.

[0010] In the preferred solution, in the step 4, the steps of calculating the Euclidean shortest path ESP between candidate points are as follows: Step 4.1: Generate a straight line segment between two candidate points and check whether the straight line intersects with the obstacle. If not, the Euclidean distance between the two points is the shortest distance. If yes, continue to Step 4.2. Step 4.2: Obstacles that intersect the straight line are considered direct obstacles, and other obstacles are indirect obstacles; Step 4.3: Generate the convex hull of the starting point, the end point and each direct obstacle; Step 4.4: If any arc in the convex hull intersects with another obstacle, replace the arc until there is no arc intersecting with the obstacle; Step 4.5: Use Dijkstra's shortest path algorithm to calculate the shortest distance; Step 4.6: Repeat the above steps to obtain the distance matrix of the candidate points of the machine nest.

[0011] In the preferred solution, in the step 5, the objective function of the UAV grid site selection model is: (2) In the formula, Indicates the center point of the demand grid A collection of; represents the total cost of machine nest deployment; The cost of building a machine nest is an abstract construction difficulty determined by the actual situation. The influencing factors include topography, land cost, power supply network cost, and distance cost. A binary variable indicating whether the facility is chosen.

[0012] In the preferred solution, in step 5, the constraints of the UAV grid site selection model are: (3) (4) (5) (6) (7) In the formula, Represents the set of candidate points of the machine nest; Candidate points for the machine nest The construction and development suitability score; To develop suitability for the theoretically highest construction; Representation and candidate points Distance less than Candidate points Collection of; A variable that represents whether the distance between the facility point and the demand point is less than the coverage radius; Indicates candidate points A binary variable of whether or not it was selected; In the preferred solution, the constraints include: each demand point is covered by at least one facility, the value of the decision variable, the distance limit of the frog jump inspection, and the calculation of the nest construction cost, and constraint (2) indicates that the economic cost of covering all demands is the minimum; constraint (3) stipulates that each demand point is covered by at least one facility; constraint (4) indicates the value of the decision variable; constraint (5) indicates that the binary variable The value of; constraint (6) represents the distance limit of frog jumping inspection, and constraint (7) is used to calculate the cost of computer nest construction.

[0013] In the preferred scheme, in Step 7, the site selection data obtained in Step 6 is imported into ArcGIS, and the buffer tool is established using GIS for coverage visualization, and the inspection area is gridded according to the coverage range of the machine nest. At the same time, a UAV leapfrog flight network is constructed based on the Euclidean shortest path between the site selection points obtained in Step 4.

[0014] In the preferred solution, in the process of dividing the inspection area into grids according to the coverage of the machine nest, the importance of the power facilities and the inspection capabilities of the drones are also considered to ensure that the power facilities in each grid can be inspected in a timely and effective manner.

[0015] In a preferred solution, the UAV is a combination of one or more multi-rotor UAVs, fixed-wing UAVs or unmanned helicopters to meet the inspection needs of different power facilities.

[0016] The GIS-based transmission and distribution UAV grid deployment method provided by the present invention has the following beneficial effects: 1. The present invention solves the technical problems of insufficient endurance, complex path planning and inflexible deployment and scheduling of UAV inspection technology in the field of power inspection; 2. The present invention calculates the coverage radius by introducing influencing factors, taking into account the margin left by factors such as incomplete charging, battery aging, low temperature, high-frequency operation, and the influence of terrain and wind speed, so that the coverage radius calculation is more accurate; 3. The present invention combines GIS and convex hull technology to optimize ESP calculations, uses GIS and convex hull technology to efficiently construct the most effective graph, optimizes the solution of the Euclidean shortest path, and improves calculation efficiency and accuracy; 4. The present invention establishes a site selection model that comprehensively considers multiple factors. The site selection model not only considers the minimization of economic costs, but also combines multiple factors such as the suitability of machine nest construction and frog leap inspection distance restrictions, making the site selection results more reasonable and practical; 5. Through grid deployment and leapfrog flight network planning, this technology improves the efficiency and coverage of drone inspections, reduces inspection costs, and ensures the safe operation of power facilities; 6. The present invention solves the problem of insufficient endurance of UAVs through grid deployment and machine nest construction, and realizes autonomous charging and continuous operation of UAVs; 7. Compared with traditional inspection work, the grid deployment and autonomous flight operation of drones in this invention reduce the skill requirements for inspection staff and improve the automation level of inspection operations; 8. The present invention solves the problem of redundant investment and waste of resources caused by the independent operation of power transmission, transformation and distribution businesses of power grid companies through joint inspection of transmission, transformation and distribution and integrated linkage inspection; 9. The present invention reduces inspection costs and improves economic benefits by optimizing the location of the machine nest and reducing redundant investment; 10. The implementation of the present invention promotes the construction of smart grids and improves the digital management level and operation and maintenance efficiency of power systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present invention will be further described below in conjunction with the accompanying drawings and implementation examples: Figure 1 It is a technical flow chart of the present invention; Figure 2 A fishnet map created for the study area of ​​this invention; Figure 3 It is a grid-based machine nest candidate area judgment diagram of the present invention; Figure 4 For Example 2 of the present invention and There is an obstacle between 's picture; Figure 5 This is a diagram of the machine nest site selection and grid division results of an embodiment of the present invention. DETAILED DESCRIPTION

[0018] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments: Example 1 The grid deployment method of UAVs for transmission, transformation and distribution based on GIS includes the following steps: Step 1: Calculate the coverage radius based on the maximum flight time and geographical factors of the transmission, transformation and distribution joint inspection drone; Step 2: Use the coverage radius and polygon intersection point set method (PIPS, Polygon Intersection PointSet) to obtain the candidate area of ​​the machine nest; Step 3: Consider the deployment environment restrictions of the machine nest to determine the candidate points of the machine nest in each candidate area and obtain the candidate point set; Step 4: Using GIS and convex hull, calculate the Euclidean shortest path ESP between the candidate points of the machine nest in Step 3; Step 5: Combine ESP (Euclidean Shortest Path) and weighted set coverage model to establish a UAV grid site selection model; Step 6: Use Python to call Gurobi to solve the global optimal solution of the model and obtain the location of the machine nest; Step 7: Use ArcGIS spatial analysis functions to perform coverage visualization and grid division, and build a UAV leapfrog flight network based on the Euclidean shortest path between the site selection points obtained in Step 4.

[0019] In this embodiment, in Step 1, an impact factor is introduced to calculate the nest coverage radius. , the specific expression is: (1) In the formula, In order to consider the margin left by the drone due to incomplete charging, battery aging, low temperature, and high-frequency operation, the value is [0, 1]; and are terrain influence factor and wind speed influence factor, respectively, with values ​​ranging from [0, 1]; In a windless environment, the test speed is The longest flight time measured when flying at a constant speed until the remaining battery power is 0%; It is the inspection time of power equipment; is the maximum flight time of the drone; the calculated distance is divided by 2 to ensure that the drone can return to the starting nest after the inspection.

[0020] Furthermore, the specific method and steps of Step 2 are to use the ArcGIS fishing net tool to divide the study area into regular grids, and map all transmission and distribution transformer demands to the corresponding grids to obtain a demand grid; with the vertex of the demand grid as the center, The intersection of multiple circular regions with a radius of is the coverage boundary of the required grid, and the intersection area of ​​multiple coverage boundaries is the candidate machine nest area.

[0021] Furthermore, in the step 4, the steps of calculating the Euclidean shortest path ESP between candidate points are as follows: Step 4.1: Generate a straight line segment between two candidate points and check whether the straight line intersects with the obstacle. If not, the Euclidean distance between the two points is the shortest distance. If yes, continue to Step 4.2. Step 4.2: Obstacles that intersect the straight line are considered direct obstacles, and other obstacles are indirect obstacles; Step 4.3: Generate the convex hull of the starting point, the end point and each direct obstacle; Step 4.4: If any arc in the convex hull intersects with another obstacle, replace the arc until there is no arc intersecting with the obstacle; Step 4.5: Use Dijkstra's shortest path algorithm to calculate the shortest distance; Step 4.6: Repeat the above steps to obtain the distance matrix of the candidate points of the machine nest.

[0022] Furthermore, in the step 5, the objective function of the UAV grid site selection model is: (2) In the formula, Indicates the center point of the demand grid A collection of; represents the total cost of machine nest deployment; The cost of building a machine nest is an abstract construction difficulty determined by the actual situation. The influencing factors include topography, land cost, power supply network cost, and distance cost. A binary variable indicating whether the facility is chosen.

[0023] Furthermore, in the step 5, the constraints of the UAV grid site selection model are: (3) (4) (5) (6) (7) In the formula, Represents the set of candidate points of the machine nest; Candidate points for the machine nest The construction and development suitability score; The theoretical maximum construction and development suitability is 100; Representation and candidate points Distance less than Candidate points Collection of; A variable that represents whether the distance between the facility point and the demand point is less than the coverage radius; Indicates candidate points A binary variable of whether or not it was selected; Furthermore, the constraints include: each demand point is covered by at least one facility, the value of the decision variable, the distance limit of the frog jump inspection, and the calculation of the construction cost of the machine nest, and constraint (2) indicates that the economic cost of covering all demands is the minimum; constraint (3) stipulates that each demand point is covered by at least one facility; constraint (4) indicates the value of the decision variable; constraint (5) indicates that the binary variable The value of; constraint (6) represents the distance limit of frog jumping inspection, and constraint (7) is used to calculate the cost of computer nest construction.

[0024] Furthermore, in Step 7, the site selection data obtained in Step 6 is imported into ArcGIS, and a buffer zone tool is established using GIS for coverage visualization. The inspection area is gridded according to the coverage of the machine nest, and a UAV leapfrog flight network is constructed based on the Euclidean shortest path between the site selection points obtained in Step 4.

[0025] Furthermore, in the process of dividing the inspection area into grids according to the coverage of the machine nest, the importance of the power facilities and the inspection capabilities of the drones are also considered to ensure that the power facilities in each grid can be inspected in a timely and effective manner.

[0026] Furthermore, the UAV is a combination of one or more multi-rotor UAVs, fixed-wing UAVs or unmanned helicopters to meet the inspection needs of different power facilities.

[0027] Example 2 In another preferred embodiment, based on the above embodiment 1, Figures 1 to 5 As shown in the figure, the grid deployment method of UAV for transmission, transformation and distribution based on GIS includes the following steps: Step S1: Calculate the coverage radius based on the maximum flight time of the transmission, transformation and distribution joint inspection drone and the geographical factors of the study area ; Step S2: Using coverage radius The candidate area of ​​the machine nest is obtained by using the polygon intersection set method (PIPS); Step S3: Consider the restrictions of the machine nest deployment environment to determine the machine nest candidate points in each candidate area, and obtain the candidate point set ; Step S4: UAV flight does not need to follow the established traffic network, but obstacles such as high-rise buildings and flight restricted areas may affect its flight path; using GIS and convex hull, calculate the Euclidean shortest path (ESP) between the candidate points of the drone nest in step S3, providing the necessary basis for the frog leap distance constraint in the subsequent site selection model; Step S5: Combine ESP and weighted set coverage model to establish a UAV grid site selection model; Step S6, using Python to call Gurobi to solve the global optimal solution of the model and obtain the location of the machine nest; Step S7, covering visualization and grid division, and obtaining a feasible and efficient UAV leapfrog flight network based on the Euclidean shortest path between the site selection points obtained in step 4; In the present embodiment, in step S1, the inspection targets of the grid inspection mode of the distribution transformer are numerous and relatively complex, and the drones need to operate at a high frequency to ensure that all inspection tasks are completed as required. Such scenarios are usually limited by the maximum flight time of the drones. Therefore, the present invention introduces an influencing factor to calculate the coverage radius of the computer nest, which is specifically expressed as follows: (1) To ensure that the drone can return to the departure nest after the inspection is completed, this embodiment divides the calculated distance by 2.

[0028] Furthermore, in step S2, the research area is divided into regular grids using the ArcGIS fishnet tool, and all transmission and distribution transformer demands are mapped to the corresponding grids to obtain a demand grid; with the vertex of the demand grid as the center, The intersection of multiple circular domains with a radius of is the coverage boundary of the demand grid. Facilities at any position on or within this boundary can cover the demand object. The intersection area of ​​multiple coverage boundaries is the candidate machine nest area.

[0029] Furthermore, in step S3, when selecting the site for the fixed hangar of the UAV, it is necessary to consider the restrictions on the hangar deployment environment. The ArcGIS clipping tool is used to delete the areas that are not suitable for station construction (lakes and rivers, traffic roads, no-fly zones, and densely populated areas) on the candidate areas for the machine nest determined in step S2, and obtain the candidate areas for the machine nest to be built. Then, in the candidate areas for the machine nest to be built, the expert's experience and knowledge are first used to roughly select the candidate points for the machine nest in each area, and then the structural entropy weight method is used for fine selection, and a certain number of candidate points for the machine nest are selected in proportion, and the construction suitability scores of the candidate points are obtained. .

[0030] Further, step S4 includes the following sub-steps: Step S4.1, Euclidean shortest path (ESP) problem analysis; The Euclidean shortest path problem between two locations is to find the middle point The problem of the number and location of is expressed as: (8) In the formula, Starting point The coordinates of For the end The coordinates of The middle point The coordinates of is the number of intermediate points; and the line between two consecutive points in the path is required not to intersect with any obstacles.

[0031] Step S4.2: Efficiently construct the most effective graph through the convex hull , optimize the solution of ESP; In the GIS environment, obstacles in the study area are represented as polygons; the expression of obstacles is as follows: (9) In the formula, is the index of the obstacle, there are obstacles; For obstacles The vertex coordinates of Another factor to be considered is the regional boundary. Generally, travel outside the regional boundary is prohibited. The expression of the regional boundary is: (10) In the formula, are the vertex coordinates of the study area boundary, indivual.

[0032] Other symbols are defined as follows: is a set of obstacles, for The set of vertices in , (11) is the set of all obstacle vertices and the start and end points, For the arc segment and the vertex Connected The set of vertices in , and for The index of the vertex in , From the vertex arrive distance.

[0033] The steps of computing the Euclidean shortest path (ESP) between candidate points of a nest are as follows, where: , ,and , The candidate point set of the machine nest obtained in step S3: 1) Generate a straight line segment between two candidate points , check whether the straight line intersects with the obstacle. If there is no intersection, the Euclidean distance between the two points is the shortest distance; if there is any obstacle If they intersect, go to (12); 2) Any obstacle that intersects the straight line is considered a direct obstacle; the set Other obstacles are considered indirect obstacles; 3) Generate starting point ,end and the convex hull of each direct obstacle. The convex hull is an important concept in computational geometry. The convex hull is a The intersection of all convex sets of , or the smallest and unique convex polygon that contains all the points, the convex hull boundary contains the shortest path around the obstacle; 4) If any arc in the convex hull intersects with another obstacle, the arc will be replaced by another convex hull until no arc intersects with the obstacle. ; 5) Use Dijkstra's shortest path algorithm to calculate the graph The shortest distance in The objective function is: (12) Constraints: (13) (14) (15) (16) The objective function (12) is to connect the start and end points and The total length of the line segment is minimized; constraints (13) and (14) stipulate that To the end The flow direction; Constraint (15) ensures the flow conservation at each intermediate vertex except the starting point and the end point; Constraint (16) restricts the decision variables to binary; 6) Repeat the above steps to calculate the Euclidean shortest path (ESP) between the candidate points of the nest and obtain the distance matrix of the candidate points of the nest .

[0034] Furthermore, in step S5, the UAV grid site selection model is as follows: Objective function: (2) Constraints: (3) (4) (5) (6) (7) Furthermore, in step S7, the site selection data obtained in step S6 is imported into ArcGIS, and the buffer tool of GIS is used to visualize the coverage effect of the machine nest site selection results on the inspection needs; at the same time, according to the Euclidean shortest path between the site selection points obtained in step S4, a UAV leapfrog flight network for the study area is constructed.

[0035] Example 3 In another preferred embodiment, based on the above embodiments 1 and 2, a grid deployment method of transmission, transformation and distribution UAV based on GIS is provided, such as Figure 1 As shown, the following steps are included: Step S1: Calculate the coverage radius based on the maximum flight time of the transmission, transformation and distribution joint inspection drone and the geographical factors of the study area ; The grid inspection mode of power transmission and distribution has many inspection targets and relatively complex types, which requires high-frequency operation of drones to ensure that all inspection tasks are completed as required; this type of scenario is usually limited by the maximum flight time of the drone, rather than the flight distance; in the actual inspection process, the flight of the drone is affected by the terrain and wind speed. A large altitude difference or high wind speed in the inspection area will shorten the flight distance of the drone in the same time, thereby reducing the coverage radius; among them, the terrain and wind speed of the inspection area can be obtained through the map service platform and actual investigation. Therefore, this paper introduces the impact factor to calculate the nest coverage radius, which is specifically expressed as: (1) The coverage radius of the machine nest in this example is 4km.

[0036] Step S2: Using the coverage radius The candidate area of ​​the machine nest is obtained by using the polygon intersection set method (PIPS).

[0037] Further preferably, step S2 includes the following sub-steps: Step S2.1: The drone nest site selection problem belongs to the continuous space site selection problem. There are infinite candidate facility locations in the study area, and the demand objects are continuously distributed transmission and distribution lines and substation areas. The PIPS method requires the coverage radius Significantly larger than the size of the demand object to achieve full coverage of the demand. Therefore, the ArcGIS fishnet tool is used to divide the study area into regular grids. The length of the fishnet is 1 km. The center of the fishnet is marked with a green dot, and the pink is the study area. All transmission and distribution transformer demands are mapped to the corresponding grids to obtain the demand grid, such as Figure 2 As shown in Table 1, the attribute values ​​of some demand grids include the demand grid number, the corresponding fishing net number, the total length of the transmission and distribution lines in the grid, and the XY coordinates of the center point of the demand grid:

[0038] Step S2.2: The required object is the intersection area of ​​the polygon; take the polygon vertex as the center, The intersection of multiple circular domains with a radius of is the coverage boundary of the demand object. Facilities at any position on or within this boundary can cover the demand object. The intersection area of ​​multiple coverage boundaries is the candidate area of ​​the machine nest. Figure 3 As shown, there are two demand grids and The process of generating candidate regions. and To cover the boundary, cover the boundary intersection area is the equivalent coverage area (candidate area).

[0039] Step S3: Consider the hangar deployment environment restrictions to determine the candidate points of the hangar in each candidate area and obtain the candidate point set .

[0040] Step S3.1: When selecting a site for a fixed hangar for UAVs, it is necessary to consider the restrictions on the hangar deployment environment, such as the hangar's power supply network, operation and maintenance management, image transmission occlusion, land cost, safety and other key factors in site selection, and avoid lakes and rivers, traffic roads, no-fly zones, densely populated areas, no-network areas, flammable and explosive dangerous areas and other areas that are not suitable for station construction; use ArcGIS's clipping tool to delete the areas that are not suitable for station construction on the candidate area for the machine nest determined in step S3 to obtain the candidate area for the machine nest to be built.

[0041] Step S3.2: The site selection of the machine nest is flexible, and fixed machine nests can be arranged in substations, power supply stations, and outdoor self-owned assets; therefore, in the candidate areas where the machine nest can be built, the expert's experience and knowledge are first used to roughly select the candidate machine nest points in each area, and then the structural entropy weight method is used for fine selection. The natural terrain conditions, distance conditions, communication conditions, and land costs are selected to establish a suitability evaluation index system, and the construction suitability score of each roughly selected candidate point is calculated. , select a certain number of candidate machine nest points in proportion. In practical applications, 1~2 are usually selected for each machine nest area to obtain a candidate point set .

[0042] Step 4: Using GIS and convex hull, calculate the Euclidean shortest path (ESP) between the candidate points of the nest in step S3; UAV flight does not need to follow the established transportation network, but obstacles such as high-rise buildings, mountains and flight restricted areas may affect its flight path; therefore, calculating the Euclidean shortest path between the candidate points of the nest provides the necessary basis for the leapfrog distance constraint in the subsequent nest site selection model.

[0043] Step S4.1, Euclidean shortest path (ESP) problem analysis; The Euclidean shortest path problem between two locations is to find the middle point The problem of the number and location of is expressed as: (8).

[0044] Step S4.2: Efficiently construct the most effective graph through the convex hull , optimize the solution of ESP; In the GIS environment, obstacles in the study area are represented as polygons; the expression of obstacles is as follows: (9) Another factor to consider is the regional boundary. Generally, travel outside the regional boundary is prohibited. The expression of the research regional boundary is: (10).

[0045] Other symbols are defined as follows: is a set of obstacles, for The set of vertices in , (11).

[0046] The steps of computing the Euclidean shortest path (ESP) between candidate points of a nest are as follows, where: , ,and , It is the candidate point set of the machine nest obtained in step S3.

[0047] 1) Generate straight line segments , check whether the straight line intersects with the obstacle. If there is no intersection, the Euclidean distance between the two points is the shortest distance; if there is any obstacle If they intersect, go to (12); 2) Any obstacle that intersects the straight line is considered a direct obstacle; the set Other obstacles are considered indirect obstacles; 3) Generate starting point ,end and the convex hull of each direct obstacle. The convex hull is an important concept in computational geometry. The convex hull is a The intersection of all convex sets of , or the smallest and unique convex polygon that contains all the points, the convex hull boundary contains the shortest path around the obstacle; 4) If any arc in the convex hull intersects with another obstacle, the arc will be replaced by another convex hull until no arc intersects with the obstacle. ; 5) Use Dijkstra's shortest path algorithm to calculate the graph The shortest distance in The objective function is: (12) Constraints: (13) (14) (15) (16) 6) Repeat the above steps to calculate the Euclidean shortest path (ESP) between the candidate points of the nest and obtain the distance matrix of the candidate points of the nest ;like Figure 4 As shown, there is an obstacle k between A and B. The most effective graph is constructed using the convex hull. .

[0048] Step S5: Combine ESP and set coverage model to establish a UAV grid site selection model; The grid-based UAVs require that the UAVs can perform “leapfrog” inspections between the nests, i.e., “one machine, multiple sites, multiple grids”, to improve the inspection efficiency of the UAVs. Therefore, the deployment of the UAV nests should consider the distance limit of the leapfrog inspection, i.e., the maximum flight distance of the UAVs. At this time, the UAVs do not need to return, and the maximum flight distance of the UAVs is At the same time, when achieving full coverage of inspection targets, the economic cost is minimized, and the drone grid site selection model is as follows: Objective function: (2) Constraints: (3) (4) (5) (6) (7).

[0049] Step S6: Solve the machine nest site selection model, and convert the above obtained data: the machine nest candidate point distance matrix Suitability of candidate sites for construction of machine nests , substitute it into the weighted set coverage site selection model, use Python to call Gurobi to solve the global optimal solution of the model, and obtain the location of the machine nest site.

[0050] Step S7: Overlay visualization and grid division, and obtain a feasible and efficient UAV leapfrog flight network based on the Euclidean shortest path between the site selection points obtained in step 4.

[0051] Import the site selection data obtained in step 6 into ArcGIS, and use GIS to create a buffer tool to visualize the coverage effect of the machine nest site selection results on the inspection needs, such as Figure 5 As shown in the figure, the yellow circular area is the coverage of the machine nest, and the gray area is the obstacle. Finally, according to the coverage of the machine nest, the inspection area is divided into grids based on the coverage of the machine nest to form a transmission and distribution inspection grid area with solid coverage of the machine nest. At the same time, according to the diagram between the site selection points obtained in step 5, , construct the UAV frog-leaping flight network in the study area, such as Figure 5 Shown by the orange line in the middle.

[0052] Example 4 In another preferred embodiment, based on the above embodiments 1 and 2, as Figures 1 to 5 As shown, this embodiment takes a certain power inspection area as the research object, and specifically demonstrates the implementation process of the GIS-based transmission and distribution UAV grid deployment method.

[0053] Step S1: Calculate the coverage radius First, the coverage radius is calculated based on the maximum flight time of the transmission, transformation and distribution joint inspection drone and the geographical factors of the study area; it is assumed that the maximum flight time of the drone in a windless environment is 60 minutes, the flight speed is 30 kilometers per hour, and the average inspection time of power equipment is 10 minutes; considering the impact of incomplete charging of the drone, battery aging, low temperature and high-frequency operation, the margin coefficient is set to 0.8; at the same time, the terrain of the study area is complex, with mountainous areas and hills, the terrain influence factor is 0.7, and the wind speed influence factor is 0.9 on average. According to the formula: (17) The calculated coverage radius R of the machine nest is about 4 kilometers.

[0054] Step S2: Obtaining the candidate area of ​​the machine nest Using the ArcGIS Fishnet tool, the study area was divided into a 1 km × 1 km regular grid; all transmission and distribution transformer demands were mapped to the corresponding grid to obtain the demand grid; a circle was drawn with a radius of 4 km and the intersection of multiple circular areas was the candidate machine nest area; Figure 2 As shown, the generation process of some demand grids and their coverage boundaries is demonstrated.

[0055] Step S3: Determine the candidate points of the machine nest Taking into account the environmental restrictions on hangar deployment, the ArcGIS clipping tool is used to delete areas that are not suitable for hangar construction, such as lakes and rivers, traffic roads, no-fly zones, and densely populated areas, to obtain candidate areas for hangar nests. Within the candidate areas for hangar nests, experts first make a rough selection of each area based on their experience, and then use the structural entropy weight method to make a fine selection of each roughly selected point, taking into account factors such as natural terrain conditions, distance conditions, communication conditions, and land costs, and calculating the construction suitability score of each point. Finally, candidate points for hangar nests are selected in proportion to obtain a candidate point set.

[0056] Step S4: Calculate the Euclidean Shortest Path (ESP) For the path planning between candidate points of the machine nest, the influence of obstacles such as high-rise buildings and mountains is considered; using GIS and convex hull technology, the ESP between candidate points of the machine nest is calculated according to the following steps: Generate a straight line segment between two candidate points and check whether it intersects with obstacles; If they intersect, generate the convex hull of the starting point, end point, and direct obstacles; Check whether the arc in the convex hull intersects with other obstacles. If so, replace it with a new convex hull until there are no intersecting arc segments. Use Dijkstra algorithm to calculate the shortest path, repeat the above steps, and obtain the distance matrix of the candidate points of the machine nest.

[0057] Step S5: Establishing a UAV grid site selection model Combining ESP and weighted set coverage model, a grid-based site selection model for UAVs is established. The objective function is to minimize the total cost of nest deployment, while satisfying constraints such as each demand point being covered by at least one nest, reasonable values ​​of decision variables, and frog leapfrog inspection distance restrictions; the nest construction cost takes into account factors such as topography, land use cost, power supply network cost, and distance cost.

[0058] Step S6: Solving the model Use Python to call the Gurobi solver, input the distance matrix of the candidate points of the machine nest and the construction suitability score, solve the global optimal solution of the model, and obtain the location of the machine nest.

[0059] Step S7: Coverage visualization and meshing The site selection results were imported into ArcGIS, and the buffer tool was used to visualize the coverage effect of the machine nest site selection on the inspection needs. According to the coverage of the machine nest, the inspection area was gridded to form a transmission and distribution inspection grid area with solid coverage of the machine nest; at the same time, based on the Euclidean shortest path between the site selection points, a UAV leapfrog flight network was constructed.

[0060] In the preferred solution, the specific method and steps of step 2 are: using the ArcGIS fishing net tool to divide the study area into regular grids, and mapping all transmission and distribution transformer demands to the corresponding grids to obtain a demand grid; taking the vertex of the demand grid as the center, The intersection of multiple circular areas with a radius of is the coverage boundary of the required grid, and the intersection area of ​​multiple coverage boundaries is the candidate area of ​​the machine nest; the above settings can automatically calculate and optimize the location and number of candidate areas of the machine nest through the algorithm to ensure comprehensive coverage and reasonable resource allocation. Subsequently, further screening is carried out in combination with factors such as terrain and weather to obtain the final recommended plan for the construction of the machine nest.

[0061] In the preferred solution, the constraints include: each demand point is covered by at least one facility, the value of the decision variable, the distance limit of the frog leap inspection, and the calculation of the nest construction cost, and constraint (2) indicates the minimum economic cost of covering all demands; constraint (3) stipulates that each demand point is covered by at least one facility; constraint (4) indicates the value of the decision variable; constraint (6) indicates the distance limit of the frog leap inspection, and constraint (7) is used to calculate the nest construction cost; the above settings are intended to ensure that the solution is both efficient and economical. Although constraint (5) is not explicitly mentioned, it is implied as a restriction on the number or type of facilities to balance the coverage breadth and cost input, and work together to build the optimization model to promote the optimization of resource allocation.

[0062] In the preferred scheme, in Step 7, the site selection data obtained in Step 6 is imported into ArcGIS, and the buffer tool is established by GIS for coverage visualization, and the inspection area is gridded according to the coverage range of the machine nest. At the same time, a UAV leapfrog flight network is constructed according to the Euclidean shortest path between the site selection points obtained in Step 4; the above settings can ensure that the UAV can efficiently jump between the site selection points when performing inspection tasks, reduce flight path overlap, and improve inspection efficiency; finally, combined with the environmental assessment results in Step 4, the flight path within the grid is further optimized.

[0063] In the preferred scheme, in the process of gridding the inspection area according to the coverage of the machine nest, the importance of the power facilities and the inspection capabilities of the drones are also taken into consideration to ensure that the power facilities in each grid can be inspected in a timely and effective manner; the above settings not only improve the inspection efficiency, but also enhance the safety of power grid operation; at the same time, the scheme ensures the continuity and accuracy of the inspection work by dynamically adjusting the grid division and flexibly responding to changes in the layout of power facilities.

[0064] In the preferred scheme, the UAV is a combination of one or more multi-rotor UAVs, fixed-wing UAVs or unmanned helicopters to meet the inspection needs of different power facilities; the above settings can ensure that the UAVs can operate flexibly in complex environments. For example, multi-rotor UAVs are suitable for narrow spaces, fixed-wing UAVs are suitable for long-distance and rapid inspections, and unmanned helicopters can fly stably at high altitudes or in complex terrains, thereby comprehensively improving the efficiency and quality of power inspections.

[0065] In summary, the grid deployment method of UAVs for power transmission and distribution proposed in the present invention solves the technical problems of insufficient endurance, complex path planning and inflexible deployment and scheduling of UAV inspection technology in the field of power inspection; the present invention combines geographic information system (GIS) with UAV inspection technology, and uses the powerful spatial analysis and data processing capabilities of GIS to provide a new perspective and method for the site selection and deployment of UAV nests; when calculating the coverage radius of UAV nests, not only natural factors such as the maximum flight time of the UAV and terrain wind speed are considered, but also human factors such as incomplete charging, battery aging, low temperature, and high-frequency operations are introduced as influencing factors, so that the calculation of the coverage radius is more accurate and closer to reality; the study area is divided into regular grids by the polygon intersection set method (PIPS), and the demand for power transmission and distribution is mapped to the grid, so as to determine the candidate area of ​​the nest. This method is not widely used in the field of UAV inspection; the present invention calculates the Euclidean shortest path (ES) between the candidate nest points in the computer P), the convex hull technology is used to efficiently construct the most effective graph, and the ESP solution process is optimized. This method has a high novelty in the field of UAV path planning; the UAV grid site selection model established by the present invention not only takes into account the minimization of economic costs, but also combines multiple factors such as the suitability of machine nest construction and frog-leaping inspection distance restrictions. This site selection model that comprehensively considers multiple factors is highly creative in the field of UAV inspection; through grid deployment and frog-leaping flight network planning, flexible jumping and sharing of UAVs between multiple machine nests are realized, and the efficiency and coverage of UAV inspections are improved; the implementation of the technical solution of the present invention not only solves the problems of insufficient endurance and high skill requirements of inspection staff in the process of UAV inspections, but also reduces inspection costs by optimizing machine nest site selection and reducing redundant investment, promotes the construction and development of smart grids, and provides intelligent and efficient inspection solutions for the power industry, promotes the deep integration of UAV technology and smart grids, and injects new vitality into the digital transformation of the power industry.

Claims

1. A grid deployment method for UAVs in power transmission, transformation and distribution based on GIS, characterized in that: The method comprises the following steps: Step 1: Calculate the coverage radius based on the maximum flight time and geographical factors of the transmission, transformation and distribution joint inspection drone; Step 2: Use the coverage radius and polygon intersection set method PIPS to obtain the candidate area of ​​the machine nest; Step 3: Consider the deployment environment restrictions of the machine nest to determine the candidate points of the machine nest in each candidate area and obtain the candidate point set; Step 4: Using GIS and convex hull, calculate the Euclidean shortest path ESP between the candidate points of the machine nest in Step 3; Step 5: Combine ESP and weighted set coverage model to establish a UAV grid site selection model; Step 6: Use Python to call Gurobi to solve the global optimal solution of the model and obtain the location of the machine nest; Step 7: Use ArcGIS spatial analysis functions to perform coverage visualization and grid division, and build a UAV leapfrog flight network based on the Euclidean shortest path between the site selection points obtained in Step 4.

2. The GIS-based transmission and distribution UAV grid deployment method according to claim 1 is characterized in that: In Step 1, the impact factor is introduced to calculate the nest coverage radius , the specific expression is: （1）； In the formula, In order to consider the margin left by the drone due to incomplete charging, battery aging, low temperature, and high-frequency operation, the value is [0, 1]; and are terrain influence factor and wind speed influence factor, respectively, with values ​​ranging from [0, 1]; In a windless environment, the test speed is The longest flight time measured when flying at a constant speed until the remaining battery power is 0%; It is the inspection time of power equipment; is the maximum flight time of the drone; the calculated distance is divided by 2 to ensure that the drone can return to the starting nest after the inspection.

3. The GIS-based transmission and distribution UAV grid deployment method according to claim 2 is characterized in that: The specific method and steps of Step 2 are: using the ArcGIS fishnet tool to divide the study area into regular grids, and mapping all transmission and distribution transformer demands to corresponding grids to obtain a demand grid; Centered on the vertex of the demand grid, The intersection of multiple circular regions with a radius of is the coverage boundary of the required grid, and the intersection area of ​​multiple coverage boundaries is the candidate machine nest area.

4. The GIS-based transmission and distribution UAV grid deployment method according to claim 3 is characterized in that: In the step 4, the steps of calculating the Euclidean shortest path ESP between the candidate points are as follows: Step 4.1: Generate a straight line segment between two candidate points and check whether the straight line intersects with the obstacle. If not, the Euclidean distance between the two points is the shortest distance. If yes, continue to Step 4.

2. Step 4.2: Obstacles that intersect the straight line are considered direct obstacles, and other obstacles are indirect obstacles; Step 4.3: Generate the convex hull of the starting point, the end point and each direct obstacle; Step 4.4: If any arc in the convex hull intersects with another obstacle, replace the arc until there is no arc intersecting with the obstacle; Step 4.5: Use Dijkstra's shortest path algorithm to calculate the shortest distance; Step 4.6: Repeat the above steps to obtain the distance matrix of the candidate points of the machine nest.

5. The GIS-based transmission, transformation and distribution UAV grid deployment method according to claim 4 is characterized in that: In the step 5, the objective function of the UAV grid site selection model is: （2）； In the formula, Indicates the center point of the demand grid A collection of; represents the total cost of machine nest deployment; The cost of building a machine nest is an abstract construction difficulty determined by the actual situation. The influencing factors include topography, land cost, power supply network cost, and distance cost. A binary variable indicating whether the facility is chosen.

6. The GIS-based transmission, transformation and distribution UAV grid deployment method according to claim 5 is characterized in that: In the step 5, the constraints of the UAV grid site selection model are: （3）； （4）； （5）； （6）； （7）； In the formula, Represents the set of candidate points of the machine nest; Candidate points for the machine nest The construction and development suitability score; To develop suitability for the theoretically highest construction; Representation and candidate points Distance less than Candidate points Collection of; A variable that represents whether the distance between the facility point and the demand point is less than the coverage radius; Indicates candidate points Binary variable whether or not it is selected.

7. The GIS-based transmission, transformation and distribution UAV grid deployment method according to claim 6 is characterized in that: The constraints include: each demand point is covered by at least one facility, the value of the decision variable, the distance limit of the frog jump inspection, and the calculation of the construction cost of the machine nest. Constraint (2) indicates that the economic cost of covering all demands is the minimum; constraint (3) stipulates that each demand point is covered by at least one facility; constraint (4) indicates the value of the decision variable; constraint (5) indicates that the binary variable The value of; constraint (6) represents the distance limit of frog jumping inspection, and constraint (7) is used to calculate the cost of computer nest construction.

8. A GIS-based transmission and distribution UAV grid deployment method according to claim 7, characterized in that: In Step 7, the site selection data obtained in Step 6 is imported into ArcGIS, and a buffer zone tool is used to create coverage visualization, and the inspection area is gridded according to the coverage of the machine nest. At the same time, a UAV leapfrog flight network is constructed based on the Euclidean shortest path between the site selection points obtained in Step 4.

9. The GIS-based transmission, transformation and distribution UAV grid deployment method according to claim 8 is characterized by: In the process of dividing the inspection area into grids according to the coverage of the machine nest, the importance of the power facilities and the inspection capabilities of the drones are also considered to ensure that the power facilities in each grid can be inspected in a timely and effective manner.

10. The GIS-based transmission and distribution UAV grid deployment method according to claim 9 is characterized by: The UAV is a combination of one or more of a multi-rotor UAV, a fixed-wing UAV or an unmanned helicopter to meet the inspection needs of different power facilities.

Images