Unmanned aerial vehicle airport site selection method suitable for large-range power line inspection

Abstract

The invention discloses an unmanned aerial vehicle airport site selection method suitable for large-scale electric power line inspection, and the method comprises the steps: S1, carrying out the primary selection of a geographic space, constructing a geographic constraint evaluation model, carrying out the processing of elements and historical disaster data through discrete coding, generating a safe and suitable region through the multi-factor weighted stacking space analysis, and carrying out the calculation of an evaluation result; and a K-Means clustering algorithm based on Euclidean distance is combined to screen a primary selection region. S2, risk assessment preselection: constructing an entropy weight-TOPSIS risk assessment model, determining a risk factor weight by using an entropy evaluation method, calculating a scheme closeness degree through a TOPSIS algorithm, and screening a low-risk preselection region; and S3, constructing a binary relationship between a demand center set and a pre-selected site set by combining an electric power inspection target and taking an electric power inspection task target as a demand node, measuring weighted traffic cost through Euclidean distance, designing a double-layer nested 0-1 decision variable, and realizing optimization of an airport deployment area by taking minimization of a total weighted distance as a target function.

Description

Technical Field

[0001] The present invention relates to technical fields such as power inspection drones, intelligent optimization algorithms, and geographic information technologies, and particularly relates to a method for selecting the location of a drone airport suitable for large-scale power line inspection. Background Art

[0002] I. Technical Challenges in Drone Airport Location Selection

[0003] Restriction of Terrain Diversity

[0004] Existing models (such as set covering, p-median) rely on Euclidean distance or static path planning and do not integrate high-precision geographic data (such as DEM, feature layers), resulting in difficulties in quantifying problems such as construction cost differences (such as increased construction costs in mountainous terrains) and communication blockages (such as signal interruption by mountains). Typical cases show that the construction cost prediction error of traditional models in mountainous scenarios is as high as 30%, and the airport damage rate in strong coastal wind environments exceeds 35%.

[0005] Frequent Disasters Aggravate Communication and Disaster Resistance Shortcomings

[0006] Existing solutions rely on fixed communication base stations, but typhoons and floods are likely to damage ground facilities (see Patent 202411002238.8), and there is a lack of a dynamic risk assessment mechanism. For example, when the post-disaster inspection demand surges by 80%, the traditional model lacks a pre-set dynamic weight adjustment logic, resulting in low response efficiency.

[0007] Insufficient Multi-Objective Collaborative Optimization Ability

[0008] Traditional models one-sidedly pursue a single indicator (such as coverage radius or path length) and do not overall consider construction costs (such as terrain construction coefficients), maintenance costs (disaster resistance facility requirements), and inspection efficiency (surge in post-disaster tasks). For example, the set covering model of Patent 202310389714.5 ignores the differences in terrain construction costs, and the p-median model of Patent 202311302316.1 does not dynamically integrate disaster historical data.

[0009] II. Existing Technical Solutions and Their Core Defects

[0010] Solution of the static location selection model based on GIS (Patent 202310389714.5): Discretize transmission, transformation, and distribution facilities into geographic center points through GIS, construct a demand-preselected facility network, and optimize the deployment with the goal of minimizing the total cost.

[0011] Defect: Assuming constant facility costs, it does not consider terrain complexity and dynamic disaster risks (such as a sharp increase in maintenance costs in flood-prone areas), resulting in poor applicability in scenarios such as mountains and coastal areas.

[0012] Two-stage Stochastic Programming Model (Patent 202311302316.1) Solution: Generate a random task scenario library based on Monte Carlo simulation, quantify the impact of dynamic changes in task priorities on site selection costs, and reduce the model dimension through the Sample Average Approximation (SAA) algorithm.

[0013] Deficiency: Monte Carlo simulation takes too long in complex terrains (such as mountainous and disaster-prone environments), and high-precision geographical data (such as no-fly zone buffer analysis) is not integrated, resulting in insufficient coverage of candidate areas.

[0014] Multi-source Data-driven Logistics Site Selection Method (Patent 202411002238.8) Solution: Integrate pedestrian flow, obstacle distribution, and climate data, screen candidate solutions through EWM-TOPSIS, and optimize the global solution in combination with the Particle Swarm Optimization (PSO) algorithm.

[0015] Deficiency: There is no dynamic optimization mechanism for disaster prevention facilities designed for disasters such as typhoons and floods, and the spatial constraints of airspace restricted areas and obstacle buffers (such as mountain communication blind spots) are not deeply processed.

[0016] III. Common Shortcomings of Existing Technologies Insufficient integration of geographical data:

[0017] Existing models do not fully integrate high-precision GIS data (such as DEM, feature layers), resulting in the failure to cover construction cost differences and communication occlusion risks.

[0018] Weak dynamic response to disasters: The design of disaster prevention facilities relies on static parameters (Patent 202411002238.8 does not dynamically adjust in combination with historical disaster data), and the failure risk is high under extreme weather conditions.

[0019] Fragmented multi-objective optimization: The construction, maintenance, and inspection costs are not optimized synergistically (not solved in both Patent 202310389714.5 and 202311302316.1), resulting in poor long-term economic efficiency. Summary of the Invention

[0020] In view of the limitations of existing UAV airport site selection technologies in complex disaster scenarios, the present invention proposes a UAV airport site selection method suitable for large-scale power line inspection, targeting the multi-topography and multi-disaster environmental characteristics of the areas through which the power grid passes. First, in response to the challenges of terrain diversity, frequent disasters, and communication reliability, a multi-constraint spatial database is constructed through a Geographic Information System (GIS), integrating airspace restricted areas, obstacle distributions, and historical disaster data to complete the geographical space screening and clustering analysis of the preliminary selection areas, systematically excluding high-risk zones of debris flows and landslides and airspace no-fly zones, and generating macroscopically suitable areas. On this basis, combined with the entropy weight method and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), a dynamic weight evaluation system for risk factors is established to quantify key indicators such as terrain complexity, disaster prevention facility requirements, and communication stability, and the candidate areas with the best risk resistance ability are selected through proximity ranking, significantly reducing the selection probability of satellite communication blind spots and high-maintenance-cost points. Further, in combination with the power inspection objectives, based on the weighted distance analysis method, with the minimization of the total weighted distance as the objective function, the optimal site layout is achieved through linear programming solution.

[0021] The implementation of the solution includes: S1 Geospatial preliminary selection, constructing a geographical constraint evaluation model, processing elements such as airspace restricted areas, slopes, land use, and historical disaster data using discretized coding, generating a safe and suitable area through multi-factor weighted overlay spatial analysis, and screening the preliminary selection areas in combination with the K-Means clustering algorithm based on Euclidean distance. S2 Risk assessment and preliminary selection, constructing an entropy weight-TOPSIS risk assessment model, using the entropy value method to determine the weights of risk factors such as terrain complexity and disaster prevention requirements, calculating the proximity of the solution through the TOPSIS algorithm, and screening low-risk preliminary selection areas; S3 Combining the power inspection objectives, and taking the power inspection task objectives as demand nodes, constructing a binary relationship between the demand center set and the preliminary selected site set, measuring the weighted traffic cost through Euclidean distance, designing a two-layer nested 0-1 decision variable, and taking the minimization of the total weighted distance as the objective function to achieve the optimal selection of the airport deployment area. This technical system provides a theoretical support and technical innovation path for the layout of intelligent power grid operation and maintenance facilities.

[0022] This method innovatively realizes multi-objective dynamic trade-off through a "three-stage progressive site selection framework (preliminary selection - preliminary selection - optimal selection)":

[0023] Preliminary selection stage: Based on the K-Means clustering based on Euclidean distance and GIS overlay analysis, restricted areas such as mountains and no-fly zones are excluded to generate a macroscopically suitable area, breaking through the neglect of terrain construction costs and disaster risks in traditional site selection models;

[0024] Preliminary selection stage: Dynamically quantify the weights of factors such as typhoon frequency and communication interruption risk through the EWM-TOPSIS model, and select highly robust preliminary selection areas from the preliminary selection areas to achieve the deep integration of risk quantification and objective decision-making;

[0025] Optimal selection stage: Combining the power inspection objectives, based on the weighted distance analysis method, with the minimization of the total weighted distance as the objective function, the optimal site layout is achieved through linear programming solution.

[0026] Compared with the traditional single-objective site selection method, the present invention solves the three core problems of UAV airport site selection in complex disaster scenarios through the collaborative integration of geospatial analysis, multi-attribute decision-making, and operations research optimization techniques:

[0027] Coverage blind area: Dynamically screening obstacle areas based on GIS and airspace restriction data to improve the integrity of geospatial coverage;

[0028] Insufficient disaster resistance ability: Quantifying the requirements for disaster prevention facilities and communication stability indicators through EWM-TOPSIS, and screening preselected areas with high robustness;

[0029] Cost redundancy: Combining the power inspection objectives, based on the weighted distance analysis method of the P-median model, to achieve an optimal layout that meets the actual effectiveness of inspection requirements.

[0030] The present invention specifically adopts the following technical solutions:

[0031] A method for UAV airport site selection suitable for large-scale power line inspection, comprising the following steps:

[0032] S1: Geospatial primary selection, based on the geographic information system GIS, constructing a multi-constraint spatial database integrating airspace restriction areas, obstacle distributions, and disaster history data, and screening out unsuitable areas through spatial overlay analysis (including spatial analysis and data fusion, specifically involving logical intersection operations on airspace restriction areas, slope grading, and historical disaster layers, and screening out grid areas that are simultaneously located in no-fly zones, with slopes > 25°, or disaster-prone areas), buffer generation, and terrain interpolation calculation, and using the K-means clustering algorithm based on Euclidean distance to generate a set of primary selection areas;

[0033] S2: Risk assessment preselection, establishing a risk assessment system including terrain complexity, disaster prevention facility requirements, and communication stability, calculating the weights of risk factors by the entropy weight method, sorting and eliminating high-risk areas by combining the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), and generating a set of preselected areas;

[0034] S3: Collaborative comparison considering comprehensive inspection requirements, taking the inspection requirement center as a constraint, constructing an improved p-median model, with the minimization of the weighted total distance as the objective function, and solving it by combining the Lagrangian relaxation method and the greedy strategy to generate an optimal site layout plan.

[0035] The main framework of the solution of the present invention adopts a three-stage progressive technical framework of "geospatial primary selection - dynamic risk assessment - coordinated comparison of comprehensive inspection requirements". Through in-depth integration of geographical data, quantification of dynamic risks, and coordinated optimization of comprehensive inspection requirements, it systematically solves the three major problems of incomplete geographical coverage, insufficient disaster resistance, and long-term cost redundancy in traditional models.

[0036] Further, S1 includes the following steps:

[0037] Extract airspace restricted areas, slopes (for example, define slopes > 25° as high-risk areas), and historical disaster information (landslide and debris flow locations) through a GIS platform (such as ArcGIS 10.2), and generate a binary evaluation matrix through discretized coding (0 / 1 represents suitability) to construct a binary evaluation matrix and generate a thematic layer;

[0038] Implement multi-factor weighted overlay spatial analysis, generate a site selection suitability layer through weighted overlay, and extract grid areas and centroid coordinate points above the area threshold;

[0039] Use the K-means clustering algorithm to perform spatial self-organization clustering on the grid areas, screen candidate centers from the disaster buffer zones, and achieve the screening of the primary selection areas.

[0040] Further, the risk factors include:

[0041] The frequency of typhoon and flood disasters;

[0042] The probability of communication interruption based on the ground communication coverage range (calculated by the geometric overlap of the UAV flight radius and the base station coverage);

[0043] The complexity of terrain construction (which can be calculated based on DEM interpolation);

[0044] Among them, the probability of communication interruption is dynamically calculated through the geometric relationship between the UAV flight radius and the ground base station signal coverage range.

[0045] Further, the sorting logic of the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) includes:

[0046] Perform range standardization processing on the original evaluation matrix, calculate the entropy value and weight;

[0047] Construct a weighted normalized matrix;

[0048] Calculate the Euclidean distances from the candidate areas to the positive ideal solution and the negative ideal solution;

[0049] Calculate the closeness index and sort.

[0050] Thus, the recognition accuracy of high-risk areas is improved, and the selection rate of communication blind spots is reduced.

[0051] Further, the collaborative comparison and selection stage for comprehensive inspection requirements includes: an objective function, the construction of a model for fusing weights of terrain construction coefficients and post-disaster inspection frequencies, and the setting of constraint conditions, where:

[0052] The objective function of the improved p-median model is:

[0053]

[0054] where ω u is the weight of the inspection requirement center, l uv is the Euclidean distance from the pre-selected site area to the requirement center, and f uv is a 0-1 decision variable.

[0055] Further, the constraint conditions of the objective function include:

[0056]

[0057] It means that each requirement center is served by only one airport, and the number of UAV airports is constrained, where p is the number of UAV airports;

[0058]

[0059] It means that only when airport v is selected can it provide inspection services for requirement center u, and at the same time, f v and f uv belong to the range of {0, 1}.

[0060] Further, the constraint conditions of the objective function include: The Lagrangian relaxation method generates a dual problem by relaxing the single-point service constraint, iteratively updates the Lagrange multipliers, and corrects the feasible solutions.

[0061] Further, the greedy strategy optimizes the initial solution selection through local search and gradually approaches the global optimal solution.

[0062] Further, the basis for dividing the airspace restricted area includes:

[0063] The signal coverage radius of the ground communication base station;

[0064] The buffer range of obstacles.

[0065] In addition, an electronic device includes a memory, a processor, and a computer program stored on the memory. It is characterized in that when the processor executes the program, the steps of the above-mentioned method are implemented.

[0066] A non-transitory computer-readable storage medium stores a computer program. It is characterized in that when the computer program is executed by a processor, the steps of the above-mentioned method are implemented.

[0067] Compared with the prior art, the main design points and beneficial effects of the present invention and its preferred solutions at least include:

[0068] (1) Innovation in the intelligent site selection method of UAV airports for multi - terrain and multi - disaster environments

[0069] Through a multi - stage optimization framework based on GIS, integrating airspace restricted areas and disaster historical data, systematically screening out high - risk zones of debris flows and landslides and airspace no - fly zones, breaking through the limitations of traditional models in evaluating terrain construction costs and disaster adaptability.

[0070] (2) Upgrade of the dynamic risk quantification assessment method

[0071] Aiming at the frequent occurrence characteristics of disasters such as typhoons and floods, a multi - attribute decision - making model based on the entropy weight method (EWM) and the technique for order preference by similarity to an ideal solution (TOPSIS) is constructed to dynamically assign weights to terrain complexity, disaster prevention facility requirements, and communication stability indicators, realizing the objective quantification and hierarchical ranking of risk factors.

[0072] (3) Optimization of the three - stage progressive site selection framework

[0073] A dynamic verification process of primary selection (geographical screening) - preselection (risk assessment) - optimization (distance optimization) is established. Relying on the synergistic effect of GIS and K - Means clustering based on Euclidean distance, a macro - suitable area is generated in the primary selection stage; in the preselection stage, a highly robust preselected area is quantitatively screened through the EWM - TOPSIS model; in the optimization stage, with the goal of minimizing the weighted total distance, avoiding the local optimal trap of traditional single - objective models, and finally outputting the optimal airport layout plan to achieve the balanced optimization of disaster resistance ability and cost - effectiveness.

[0074] Comprehensive technical breakthrough and engineering value

[0075] Through the deep integration of geospatial analysis, multi - attribute decision - making, and operations research optimization technologies, the present invention realizes three major core improvements in complex disaster scenarios:

[0076] Improved coverage efficiency: Based on GIS and real obstacle data, communication occlusion areas (such as mountain signal blockages) are dynamically excluded, and the accessibility of inspection paths is improved;

[0077] Enhanced disaster resistance ability: Through dynamic risk assessment and quantification of disaster prevention facility requirements, the typhoon damage rate is reduced;

[0078] Reduced cost redundancy: Based on the optimization model of inspection objectives, the site selection areas with deviations from actual inspection requirements are further reduced, reducing redundant costs;

[0079] Final effect verification: Combining airspace division and ground communication coverage optimization, a high - reliability and low - cost UAV airport deployment plan is provided for large - scale power inspection scenarios. Brief Description of the Drawings

[0080] The present invention will be further described in detail below in conjunction with the drawings and specific embodiments:

[0081] Figure 1 It is the overall flowchart of the technical solution for the location selection of the UAV airport in the embodiment of the present invention;

[0082] Figure 2 It is the flowchart of the coverage algorithm in the embodiment of the present invention;

[0083] Figure 3 It is the flowchart of the preliminary location selection of the UAV airport in the embodiment of the present invention;

[0084] Figure 4 It is the schematic diagram of obstacle analysis in the embodiment of the present invention;

[0085] Figure 5 It is the schematic diagram of dividing the UAV flight area in the embodiment of the present invention;

[0086] Figure 6 It is the schematic diagram of initially selecting 12 UAV airport areas based on GIS in the embodiment of the present invention;

[0087] Figure 7 It is the schematic diagram of re-screening 10 preselected UAV airport areas based on EWM-TOPSIS in the embodiment of the present invention;

[0088] Figure 8 It is the schematic diagram of selecting 7 areas from 10 preselected areas in the embodiment of the present invention. Specific Embodiments

[0089] In the following, specific embodiments of the present application will be described in detail with reference to the drawings. According to these detailed descriptions, those skilled in the art can clearly understand the present application and can implement the present application. Without departing from the principle of the present application, the features in different embodiments can be combined to obtain new embodiments, or some features in certain embodiments can be replaced to obtain other preferred embodiments.

[0090] To make the features and advantages of the present invention more obvious and understandable, specific embodiments are given below and described in detail in conjunction with the drawings as follows:

[0091] As Figure 1 shown, it is the implementation process of the complete technical solution provided by the embodiment of the present invention, including:

[0092] Step S1: To exclude restrictive factors such as airspace restricted areas, residential areas, rivers, mountains, and historical disaster information of debris flows and landslides within the research area, functions such as screening, buffer analysis, interpolation analysis, and overlay analysis in ArcGIS 10.2 software are used to conduct a detailed analysis of the research area; exclude restricted areas and determine the preliminary site selection area.

[0093] Step S2: In the preliminary site selection stage, some potential operation risks may exist in some of the obtained preliminary site selection areas. Therefore, in the pre-selection stage, fully consider the risk factors in the preliminary selection areas, use the entropy weight method to determine the weight values of the risk factors, and use the weighted TOPSIS method to rank the advantages and disadvantages of the preliminary site selection plans to screen out pre-site selection areas with better safe operation conditions.

[0094] Step S3: Determine the optimal area in the pre-selected areas to minimize the total inspection distance. Take the power inspection task points as the demand centers and establish an inspection demand set. Based on the P-median model, with the minimum weighted total distance as the objective function, solve through linear programming to achieve the optimal site layout and realize the optimization of the airport site selection area.

[0095] The following introduces the specific technical content and the construction process of the solution of the embodiments of the present invention:

[0096] (1) GIS-based geographical space preliminary selection of UAV airports

[0097] 1) Draw thematic layers of each restrictive influencing factor. Divide the research area into suitable areas and unsuitable areas according to the restrictive factors, and use "reclassification" in the ArcToolbox toolbox to assign them "0" and "1" respectively, and draw thematic layers such as airspace restriction grading maps, slope grading maps, land use grading maps, water system grading maps, residential area grading maps, and nature reserve grading maps.

[0098] 2) To ensure the influence of each layer on the weights of the influencing factors during overlay, adopt the steps of pairwise weighted overlay and then reclassification to obtain the preliminary large area that meets the macro safety site selection conditions.

[0099] 3) To further accurately locate the scope of the site selection area, use the ArcMap raster calculator to extract the raster map with a polygon area greater than 3 km 2 and the centroid coordinate points of each raster, and determine K clustering centers according to the actual situation based on the Euclidean distance-based K-Means clustering algorithm.

[0100] 4) Obtain the preliminary selection area. Comprehensively analyze the K clustering centers, and combine the actual situation of the inspection area to screen out several preliminary selection area sets. The process is shown in Figure 3 .

[0101] (2) Pre-selection of drone airport risk assessment based on EWM-TOPSIS

[0102] 1. Standardization. Suppose there are m potential operational risk factors, n proposed site selection options, and x ij For the value of the jth site selection scheme under the i-th influencing factor, the original evaluation matrix X = (x ij ) m×n In order to eliminate the order of magnitude differences between different indicators, each evaluation indicator is dimensionless, and the evaluation matrix Y = (y ij ) m×n , where i = 1, 2,…, m; j = 1, 2,…, n.

[0103]

[0104] 2. Calculate the proportion of the jth solution under the i-th impact factor.

[0105]

[0106] 3. Calculate the entropy value of the i-th influencing factor.

[0107]

[0108] Among them, it is defined that when R ij =0, R ij ln ij =0.

[0109] 4. Calculate the weight of the i-th impact factor.

[0110]

[0111] 5. Use vector normalization to calculate the original evaluation matrix. X=(x ij ) m×n The standard matrix X'=(x' ij ) m×n for:

[0112]

[0113] 6. Determine the weighted norm matrix Z.

[0114] Z ij =w i x′ ij (7)

[0115] 7. Calculate the positive and negative ideal solutions for each impact factor.

[0116]

[0117] 8. Calculate the distance from the site selection plan to the positive ideal solution and the negative ideal solution.

[0118]

[0119] 9. Calculate the closeness of the evaluation object to the positive ideal solution.

[0120]

[0121] (3) Collaborative comparison of comprehensive inspection requirements based on the p-median model

[0122] 1) After analyzing the above steps, N pre-selected areas with good overall conditions can be screened. Next, considering the economic status of the pre-selected areas and the transportation distances between each pre-selected area and each inspection demand center, P areas are selected from the N pre-selected areas to minimize the weighted total distance from each of the P pre-selected areas to each demand center.

[0123] 2) Model assumptions: 1) Different demand centers are represented by power inspection task objectives; 2) The distance from each pre-selected site to each demand center is represented by Euclidean distance; 3) Each demand center should be served by a proposed airport.

[0124] 3) U and V represent the set of demand centers and pre-selected sites, respectively, u∈U, v∈V; ω u l uv represents the weighted distance from the pre-selected site v to each demand center u; f v is a variable of 0 or 1. If you choose to build an airport at location v, the value is 1, otherwise it is 0. uv It is a 0, 1 variable. If airport v provides services to demand center u, the value is 1, otherwise it is 0.

[0125] 4) Based on the analysis and assumptions in the previous article, the p-median model for drone airport site selection is determined as follows.

[0126]

[0127] St1:

[0128] St2:

[0129] The objective function Q represents the minimum total distance from the pre-selected location to each demand center, ω u is the weight, l uv is the distance; st1 indicates that each demand center is served by only one airport, and the number of drone airports is constrained, and p is the number of drone airports; st2 indicates that only when airport v is selected can it provide inspection services for demand center u, and at the same time constrain f v and f uvBelongs to the range of {0, 1}.

[0130] For the above optimization function, further combining the Lagrangian relaxation method and the greedy strategy for solution, a new set of site selection areas can be obtained.

[0131] Technical verification

[0132] Airspace obstacle calibration

[0133] In the Fujian region, the geographical information system (GIS) is used to calibrate the obstacles in the UAV flight environment. The main purpose is to identify potential obstacles in the flight area through spatial data analysis, so as to optimize the flight path and ensure flight safety.

[0134] Terrain data: Obtain the digital elevation model (DEM) data of the Fujian region, which are used to obtain information such as ground elevation and building height;

[0135] Feature data: Collect the feature data of the Fujian region, including the spatial positions and attributes of obstacles such as urban buildings, bridges, towers, wind turbines, transmission lines, and trees;

[0136] Aviation obstacle data: Specific obstacle information in the UAV flight environment can be obtained through government departments, pilot reports, or remote sensing equipment of UAVs; Aviation restricted areas: Obtain flight restricted areas (such as no-fly zones, temporary no-fly zones, controlled airspaces, etc.) and airway data.

[0137] Obstacle analysis: Analyze the height information of obstacles such as buildings, mountains, and towers to determine which positions will affect UAV flight. By comparing the heights of obstacles and the UAV flight height, potential collision risk areas are calibrated, such as Figure 4 shown by the rectangle in.

[0138] No-fly zone analysis: Set buffers for points or lines (such as buildings, roads, transmission lines, etc.) that may become obstacles to generate flight safety areas. For example, if the vertical distance between a building and the UAV flight path is less than a certain threshold, this area can be marked as a flight risk area, such as Figure 4 shown by the circle in.

[0139] Table 1 Obstacle information table

[0140] Obstacle ID Type Location (Longitude) Location (Latitude) Height Radius (m) 1 Building A 117.982053 25.741278 50m 10m 2 Building B 118.637479 26.503429 80m 15m 3 Mountain A 116.904633 27.144930 60m 18m 4 Mountain B 119.198025 24.927890 70m 20m 5 Tower 120.037395 27.309845 40m 10m 6 No-Fly Zone A 117.241724 25.503682 0 10m 7 No-Fly Zone B 119.686407 28.084106 0 20m 8 No-Fly Zone C 118.371530 27.779907 0 30m 9 Threat Zone A 116.592210 25.975681 0 15m 10 Threat Zone B 120.180510 28.206741 0 20m

[0141] Such as Figure 4As shown in the figure, based on the real geographical data of Fujian Province, a total of 10 obstacles and no-fly zones were demarcated, and the influence ranges of the obstacles and no-fly zones were demarcated according to their sizes and positions, and were shown in the figure with red rectangles and red circles. Users can modify the positions and sizes of the no-fly zones according to the changes in the airspace environment. The position information and influence range information of the obstacles are stored in the obstacle information database in the form of the data in Table 1.

[0142] UAV Airspace Division

[0143] UAV airspace division is the planning and management of the UAV flight area to ensure that different types of UAVs can fly safely and efficiently and avoid conflicts. The UAV airspace division of this project is considered according to the following factors:

[0144] Restricted airspace: including obstacles (mountains, buildings), no-fly zones, special use airspaces, etc. These areas usually have relatively high restrictions on all aircraft.

[0145] Open airspace: The flight area applicable to UAVs, usually referring to the area where UAVs can fly without special approval.

[0146] Communication range: The flyable area is divided according to the communication range of the UAV to ensure that the UAV flight does not exceed the ground communication range.

[0147] As Figure 5 shown, the UAV flight areas in Fujian Province were divided, comprehensively considering restricted areas such as obstacles and no-fly zones, and the flight areas applicable to UAVs were divided.

[0148] Table 2 Airspace Information Table

[0149]

[0150]

[0151] As Figure 5 shown, according to the obstacle and no-fly zone information demarcated in the previous step, and the functions of each airspace, the entire UAV flight airspace was divided, and a total of 26 effective airspaces where UAVs can fly were divided. During the division process, it was ensured that the airspace boundaries would not pass through the demarcated obstacles and no-fly zones, that is, the formed airspace network would envelope the obstacles and no-fly zones outside the safety threshold. The airspace information data is shown in Table 2.

[0152] UAV Airport Site Selection

[0153] (1) 12 preliminary selection areas of UAV airports based on GIS

[0154] The process of preliminary selection of UAV airport areas based on GIS includes four steps: First, draw thematic layers of restrictive factors within the study area, such as airspace restrictions, slopes, land use, water systems, residential areas, and nature reserves, and divide the area into suitable and unsuitable areas. Second, use the weighted overlay analysis method to overlay each layer to screen out the large preliminary selection areas that meet the macro safety siting conditions. Then, apply the K-Means clustering algorithm based on Euclidean distance to conduct clustering analysis on the preliminary selection areas to determine multiple potential airport siting centers. Finally, combined with the actual situation in Fujian Province, comprehensively evaluate the clustering results and screen out 12 eligible preliminary selection areas, such as Figure 6 shown. Twelve UAV airport areas were preliminarily selected based on GIS, mainly considering the impacts of obstacles and no-fly zones in the airspace environment, as well as the actual geographical environment information in Fujian Province. According to the airspace environment information, use the K-Means algorithm to conduct clustering analysis on the preliminary selection areas to determine 12 potential airport siting centers. The location information of the 12 preliminary selected airports is shown in Table 3.

[0155] Table 3 Airport Location Information Table

[0156] Airport ID Location (Longitude) Location (Latitude) Airport 1 119.257139 25.213488 Airport 2 119.261731 25.217293 Airport 3 119.265000 25.220000 Airport 4 119.268512 25.215817 Airport 5 119.252290 25.218380 Airport 6 119.250616 25.213721 Airport 7 119.271044 25.213100 Airport 8 119.227742 25.236743 Airport 9 119.248041 25.238994 Airport 10 119.258727 25.237869 Airport 11 119.237827 25.223582 Airport 12 119.241776 25.234297

[0157] (2) Ten preliminary selected areas for UAV airports based on EWM-TOPSIS

[0158] In the preliminary selection stage, fully consider the risk factors in the preliminary selection areas, use the entropy weight method to determine the weight values of the risk factors, and use the weighted TOPSIS method to rank the pros and cons of the preliminary siting schemes. Through the EWM-TOPSIS algorithm, eliminate the areas with greater risks to screen out 10 preliminary selected areas with better safe operation conditions, such as Figure 7 shown. The 10 selected airport candidate areas are mainly distributed in areas far from obstacles and no-fly zones. The location information of the airport candidate areas is shown in Table 4.

[0159] Table 4 Airport Location Information Table

[0160] Airport ID Location (Longitude) Location (Latitude) Airport 1 119.257139 25.213488 Airport 2 119.261731 25.217293 Airport 3 119.265000 25.220000 Airport 4 119.268512 25.215817 Airport 5 119.252290 25.218380 Airport 6 119.250616 25.213721 Airport 7 119.271044 25.213100 Airport 8 119.227742 25.236743 Airport 9 119.248041 25.238994 Airport 10 119.258727 25.237869

[0161] (3) Seven alternative areas based on the p-median model

[0162] After the above steps, 10 preliminary selected areas with better conditions can be screened out. Then, consider the transportation distance between the preliminary selected areas and the power grid line inspection target demand center. Such as Figure 8As shown, the blue triangles represent the inspection target demand centers of the power grid lines. Considering the transportation distance between the pre-selected areas and the inspection demand centers, 7 areas are selected from 10 pre-selected areas to minimize the weighted total distance from the 7 selected areas to each demand center and reduce the costs caused by the distance. The location information of the airport alternative areas is shown in Table 5

[0163] Table 5 Airport Location Information Table

[0164]

[0165]

[0166] In the description of this specification, the descriptions referring to terms such as "one embodiment", "example", "specific example", etc. mean that the specific features, structures, materials or characteristics described in connection with that embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0167] The above shows and describes the basic principles, main features and advantages of the present disclosure. Those skilled in the art of this industry should understand that the present disclosure is not limited by the above embodiments. The above embodiments and the descriptions in the specification only illustrate the principles of the present disclosure. Without departing from the spirit and scope of the present disclosure, the present disclosure will have various changes and improvements, and these changes and improvements all fall within the scope of the present disclosure claimed.

[0168] The present invention is not limited to the above best implementation manner. Anyone can derive other various forms of a method for selecting the location of an unmanned aerial vehicle airport suitable for large-scale power line inspection under the inspiration of the present invention. All equal changes and modifications made according to the scope of the patent application of the present invention shall fall within the scope covered by the present invention.

Claims

1. A method for selecting a location of a drone airport suitable for large - range power line inspection, characterized in that, It includes the following steps: S1: Geospatial primary selection. Based on the Geographic Information System (GIS), a multi-constraint spatial database integrating airspace restricted areas, obstacle distributions, and historical disaster data is constructed. Inappropriate areas are screened out through spatial overlay analysis, buffer zone generation, and terrain interpolation calculations, and a set of primary selection areas is generated using the K-means clustering algorithm based on Euclidean distance; S2: Risk assessment preselection. A risk evaluation system including terrain complexity, disaster prevention facility requirements, and communication stability is established. The weights of risk factors are calculated by the entropy weight method, and the high-risk areas are sorted and excluded by combining the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), generating a set of preselected areas; S3: Collaborative comparison considering comprehensive inspection requirements. With the inspection requirement center as a constraint, an improved p-median model is constructed, and the weighted total distance is minimized as the objective function. It is solved by combining the Lagrangian relaxation method and the greedy strategy to generate the optimal site layout plan.

2. The method for selecting a UAV airport location suitable for large - range power line inspection according to claim 1, wherein, S1 includes the following steps: Extract airspace restricted areas, slopes, and historical disaster information through the GIS platform, construct a binary evaluation matrix, and generate a thematic layer; Implement multi-factor weighted overlay spatial analysis, and extract raster areas and centroid coordinate points above the area threshold; Use the K-means clustering algorithm to perform spatial self-organization clustering on the raster areas and screen the primary selection areas.

3. The method for selecting a location of a UAV airport suitable for large-range power line inspection according to claim 1, characterized in that, The risk factors include: The frequency of typhoon and flood disasters; The probability of communication interruption based on the ground communication coverage range; The complexity of terrain construction; Among them, the probability of communication interruption is dynamically calculated through the geometric relationship between the flight radius of the unmanned aerial vehicle and the signal coverage range of the ground base station.

4. A method for selecting a location of a drone airport suitable for large - range power line inspection according to claim 1, characterized in that, The sorting logic of the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) includes: Perform range standardization processing on the original evaluation matrix; Construct a weighted normalized matrix; Calculate the Euclidean distances from the candidate areas to the positive ideal solution and the negative ideal solution; Calculate the closeness index and sort.

5. A method for selecting the location of a drone airport suitable for large-range power line inspection according to claim 1, characterized in that: The objective function of the improved p-median model is: Among them, ω u is the weight of the inspection demand center, l uv is the Euclidean distance from the pre-selected site area to the demand center, f uv is a 0-1 decision variable.

6. The method according to claim 5, wherein The constraint conditions of the objective function include: It means that each demand center is served by only one airport, and the number of unmanned aerial vehicle airports is constrained, where p is the number of unmanned aerial vehicle airports; It means that only when airport v is selected can the inspection service be provided for demand center u, while constraining f v and f uv belongs to the range of {0, 1}.

7. The method according to claim 6, characterized in that, The constraint conditions of the objective function include: The Lagrangian relaxation method generates a dual problem by relaxing the single-point service constraint, iteratively updates the Lagrangian multiplier, and corrects the feasible solution.

8. A method for selecting the location of a drone airport suitable for large-range power line inspection according to claim 1, characterized in that: The greedy strategy optimizes the initial solution selection through local search and gradually approaches the global optimal solution.

9. A method for selecting the location of a drone airport suitable for large - range power line inspection according to claim 1, characterized in that: The basis for the division of the airspace restricted area includes: The signal coverage radius of the ground communication base station; The buffer zone range of obstacles.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, When the processor executes the program, it implements the steps of the method described in any one of claims 1-9.

Images