Unmanned aerial vehicle inspection method and system based on airport layout optimization

Abstract

The invention discloses an unmanned aerial vehicle inspection method and system based on airport layout optimization. The method comprises the following steps: firstly, receiving target point set data, an unmanned aerial vehicle and airport parameter data; and generating an optimal airport position set by adopting a dynamic candidate set strategy through iterative optimization under a plurality of preset service radiuses, balancing the coverage range and deployment cost, and calculating the total cost and the average inspection interval. Constructing a multi-objective function, screening a non-dominated solution set, forming a two-dimensional Pareto frontier, and achieving the balance of the cost and the inspection efficiency; and outputting a three-dimensional decision space diagram and a two-dimensional Pareto frontier diagram, and displaying the optimal airport site selection, coverage radius and inspection path on a geographic information system to assist a user in decision making. And finally, according to an optimization result, arranging airports and allocating unmanned aerial vehicle tasks to ensure efficient operation. According to the method, through fine optimization and visual visualization, an efficient, economical and practical solution is provided for unmanned aerial vehicle inspection. The invention is used in the field of low-altitude economy.

Description

Technical Field

[0001] The present invention relates to the field of low-altitude economy, and particularly to a method and system for UAV inspection based on airport layout optimization. Background Art

[0002] In the UAV inspection network, airport location selection and path planning are closely intertwined core issues. The rationality of airport layout directly affects the flight distance, energy consumption, and task execution efficiency of UAVs, while the quality of path planning in turn restricts the utilization rate of the airport service radius. However, existing technical solutions often treat the two separately: on the one hand, layout optimization only considers static coverage requirements and ignores the regulatory potential of different service radii on path dynamic decision-making; on the other hand, path planning mostly conducts local optimization under fixed airport positions, resulting in a lack of adjustability in global resource allocation. This fragmented technical path is difficult to meet the requirements of multi-objective collaborative optimization such as cost, efficiency, and emergency response in dynamic scenarios.

[0003] The existing limitations are mainly reflected in two major contradictions: one is the contradiction between the fixed service radius assumption and dynamic airspace restrictions. Traditional solutions allocate the airport coverage range based on a constant radius, which cannot adapt to the actual endurance fluctuations and sudden task scheduling of UAVs, resulting in redundant airport construction or coverage blind spots in key areas; the other is the index imbalance caused by isolated optimization, which single-mindedly focuses on layout coverage rate or local path optimality, sacrificing the overall operation efficiency of the network. For example, maximizing the coverage rate may lead to increased energy consumption due to circuitous inspection paths, while shortening the inspection distance may force airports to concentrate in resource-rich areas.

[0004] From a theoretical perspective, the coupling relationship between the airport service radius and path planning directly determines the dynamic adaptability of the network topology. The lack of joint modeling of the two will cause the optimization model to deviate from the actual constraint conditions; from the perspective of engineering practice, large-scale UAV inspection networks need to balance infrastructure investment costs and long-term operation efficiency. If the airport density cannot be reduced and the utilization rate of single-point resources cannot be improved through collaborative optimization, it will seriously restrict the commercialization process in complex scenarios. Especially in high-frequency and high-real-time applications such as power line inspection and disaster monitoring, this technical bottleneck has become a common challenge restricting the industry's efficiency. Summary of the Invention

[0005] The present invention provides a method and system for UAV inspection based on airport layout optimization, which is used to provide efficient and practical UAV inspection with comprehensive coverage of multiple target points and dynamic adaptability.

[0006] The technical solution adopted by the present invention is as follows: A method for UAV inspection based on airport layout optimization, comprising the following steps:

[0007] S1: Receive target point set data, UAV data, and airport parameter data;

[0008] S2: Iteratively optimize multiple preset service radii, call the optimization algorithm to generate the set of airport locations and the subset of covered target points. The algorithm adopts a dynamic candidate set strategy, retaining the top N optimal candidate airports in each iteration, calculating the total cost of each plan and the average inspection interval.

[0009] S3: Construct the objective function, screen the non-dominated solution set, and generate the two-dimensional Pareto front.

[0010] S4: Synchronously output the three-dimensional decision space diagram and the two-dimensional Pareto front diagram, and overlay and display the optimal airport location, coverage radius, and UAV inspection path on the geographic information system.

[0011] S5: Arrange the airports according to the optimal airport location. Each airport is equipped with an unmanned control system, and each unmanned control system assigns tasks and plans paths for the UAVs within the airport according to the coverage radius and target point allocation.

[0012] Optionally, in step S3, the non-dominated solution set is screened by the Pareto optimization algorithm. The specific implementation is as follows:

[0013] Establish a two-objective optimization model: min{f1 = C total , f2 = T avg};

[0014] Use the NSGA-II algorithm for population evolution, and screen the Pareto front solution set through fast non-dominated sorting and crowding degree calculation.

[0015] Optionally, the airport is deployed through a mobile UAV airport. The mobile UAV airport is a UAV vehicle. The mobile UAV airport constructs a three-dimensional environmental map around the airport through lidar and performs dynamic obstacle avoidance deployment.

[0016] Optionally, according to the change of the target point set data, update the airport location to obtain the updated new airport location, and optimize to obtain the new inspection route. The UAVs perform inspections along the new inspection route. Subsequently, after the mobile UAV airport is arranged according to the new airport location, the UAVs match the new airport location with the intersection points of the new inspection route. The intersection point matching of the new inspection route includes space-time corridor connection operations.

[0017] Optionally, it further includes a dynamic task integration step: In step S5, the unmanned control system dynamically receives external task requests, calculates the spatial proximity and time window coincidence degree between the task points and the existing inspection paths through the spatio-temporal similarity matching algorithm; selects the UAVs and the departure airports that can reuse the paths, inserts the task points and dynamically adjusts the flight speed or path, and ensures that the total inspection cycle remains unchanged through the air relay or speed compensation mechanism.

[0018] Optionally, in the spatio-temporal similarity matching algorithm, spatial proximity is determined by an Euclidean distance threshold, and the time window overlap degree is calculated by the intersection ratio of the task execution period and the UAV idle period.

[0019] Optionally, it further includes a dynamic task integration step: in step S2, reserve 5% - 10% of the transport capacity for each airport in the optimization algorithm to support emergency tasks.

[0020] Optionally, it further includes an emergency role dynamic switching step: in step S5, when the unmanned control system receives an emergency event instruction, assign a preset role module to the UAV according to the event type and update its task priority; based on the event geographical location, screen the nearest airport from the Pareto optimal solution set to generate an emergency path of "airport → target point → next target point"; highlight the emergency path in the visualization interface and dynamically adjust the cooperation strategy by real-time predicting multi-aircraft task conflicts through edge computing.

[0021] Optionally, the role module includes at least one of traffic command, target tracking, and material delivery, and the role switching time cost is predefined in the parameter initialization stage.

[0022] The present invention also provides a system for implementing a UAV inspection method based on airport layout optimization, which is characterized by including:

[0023] A data receiving module for efficiently integrating and verifying multi-source data;

[0024] An optimization processing module for improving the coverage rate and cost-effectiveness of airport site selection by screening dynamic candidate sets;

[0025] A decision-making optimization module for achieving a global optimal balance between cost and timeliness by using the multi-objective Pareto front;

[0026] A visualization module for realizing multi-dimensional dynamic deduction of the scheme by integrating three-dimensional efficiency space and geographical information;

[0027] An execution control module for automatically generating an airport deployment plan and a UAV cooperative inspection path based on the optimal solution.

[0028] Advantages of the present invention: The present invention provides a method and system for UAV inspection based on airport layout optimization. Through multi-stage system optimization and visual decision support, the efficiency and economy of UAV inspection are significantly improved. First, in step S1, the system receives the target point set, UAV, and airport parameter data, laying the foundation for subsequent optimization. Then, in step S2, through iterative optimization under multiple preset service radii, a dynamic candidate set strategy is used to generate an optimal airport location set, balancing the coverage range and deployment cost, and calculating the total cost and average inspection interval to evaluate the benefits of the plan. In step S3, a multi-objective function is constructed and the non-dominated solution set is screened to form a two-dimensional Pareto front, realizing the trade-off between cost and inspection efficiency. Next, in step S4, a three-dimensional decision space diagram and a two-dimensional Pareto front diagram are output, and the optimal airport location, coverage radius, and inspection path are intuitively displayed on the geographic information system to assist the user in decision-making. Finally, in step S5, the airport is arranged according to the optimization results and UAV tasks are assigned to ensure efficient operation. This method provides an efficient, economical, and practical solution for UAV inspection through fine optimization and intuitive visualization. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The present invention will be further described below with reference to the accompanying drawings:

[0030] Figure 1 is the multi-objective optimization flowchart of the UAV inspection method based on airport layout optimization;

[0031] Figure 2 is the system interface diagram of the UAV inspection method based on airport layout optimization;

[0032] Figure 3 is the three-dimensional decision space visualization schematic diagram of the UAV inspection method based on airport layout optimization;

[0033] Figure 4 is the Pareto front curve and plan selection interface diagram of the UAV inspection method based on airport layout optimization. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0034] Referring to Figures 1 to 4 a UAV inspection method based on airport layout optimization provided by the present invention:

[0035] In step S1, the system receives multi-source data sets such as the target point set (points to be inspected), UAVs, and airports. These data respectively cover the specific location points to be inspected, the performance parameters of the UAVs (such as endurance time, flight speed, etc.), and the cost information related to airport construction and UAV operation.

[0036] In step S2, the system iteratively optimizes according to multiple preset service radii, and determines the optimal set of airport locations and the subset of target points that each airport can cover by calling a specialized optimization algorithm. In this process, the algorithm adopts a dynamic candidate set strategy, that is, in each iteration, the top N airport locations with the best effects are retained from numerous candidate solutions, and the total cost and average inspection interval time of each solution are comprehensively calculated.

[0037] In step S3, the system constructs an objective function to screen out the non-dominated solution set from numerous optimization solutions, that is, the set of solutions that achieve the best balance between cost and inspection efficiency, and generates a two-dimensional Pareto front graph to intuitively display the trade-off relationships of different solutions.

[0038] In step S4, the system synchronously outputs a three-dimensional decision space graph and a two-dimensional Pareto front graph, and superimposes and displays the optimal airport location, the coverage radius of each airport, and the inspection path of the unmanned aerial vehicle (UAV) on the geographic information system (GIS) to help users clearly understand the optimization results.

[0039] In step S5, the actual airport layout is carried out according to the selected optimal airport location. Each airport is equipped with a set of unmanned control systems, and these systems will allocate specific inspection tasks for the UAVs within the airport and plan flight paths according to the coverage radius and the distribution of target points.

[0040] This method makes the optimization process of the airport layout for UAV inspection more efficient and scientific through a series of systematic steps. Receiving comprehensive data lays a foundation for subsequent optimization, and the design of multiple rounds of iterative optimization ensures that the airport locations can cover all target points while minimizing costs as much as possible. By generating intuitive two-dimensional and three-dimensional charts, users can easily compare the advantages and disadvantages of different solutions, and thus select the most suitable airport layout plan according to actual needs. The final unmanned control system further guarantees the efficient allocation and execution of UAV tasks, making the entire inspection process not only cost-effective but also able to meet the actual requirements in terms of time and coverage.

[0041] As a preferred implementation, in step S3, the method screens the non-dominated solution set through a specific optimization method. Specifically, the system first establishes a two-objective optimization model aiming to minimize both the total inspection cost and the average inspection interval time. To achieve this goal, the system sets some constraint conditions that are automatically extracted from historical optimization solutions to ensure that the screening process can refer to past empirical data. During the optimization process, the system adopts an algorithm called NSGA-II, which continuously improves candidate solutions by simulating the evolution process of the population. It sets certain crossover probabilities and mutation probabilities to combine different airport layout plans or randomly adjust them to explore more possibilities. After multiple rounds of calculations, the system sorts all plans into layers according to their superiority through fast non-dominated sorting, and combines crowding degree calculation to ensure that the finally selected solution set contains both the best-performing solutions and maintains the diversity between solutions, ultimately forming a clear Pareto front solution set.

[0042] This screening method significantly improves the efficiency and quality of multi-objective optimization. By considering both the cost and inspection interval, two key indicators, the system can find a balance in a complex decision-making environment and provide users with diverse choices. The application of the NSGA-II algorithm makes the optimization process more intelligent and comprehensive, while fast non-dominated sorting and crowding degree calculation ensure the diversity and practicality of the screening results. Therefore, users can select the most suitable solution from the solution set according to specific requirements (such as lower cost or higher inspection frequency).

[0043] As a preferred implementation, these movable airports are actually special vehicles capable of carrying drones (which can be large drones or land aircraft carriers) and can be flexibly adjusted in position according to needs. To ensure the safety and accuracy of deployment, each movable drone airport is equipped with lidar equipment. During deployment, the lidar scans the surrounding environment of the airport to generate a detailed three-dimensional map containing the location information of terrain, buildings, or other obstacles. Based on this map, the airport can automatically identify potential obstacles and adjust its position or path to achieve dynamic obstacle avoidance deployment, ensuring smooth layout in complex environments.

[0044] The design of the movable drone airport greatly enhances the flexibility and adaptability of the system. Traditional fixed airports may be limited by terrain or environmental changes, while this movable airport can be adjusted in position according to actual needs, especially performing well in complex scenarios such as cities or remote areas. The application of lidar further improves the safety and accuracy of deployment, enabling the airport to quickly take its position while avoiding collisions and providing reliable support for drone inspection tasks.

[0045] As a preferred embodiment, this method can dynamically adjust the airport layout and inspection path according to the changes in the target point set data. Specifically, when the position or quantity of the target points changes, the system will re-analyze and update the optimal airport location, generating a new airport layout plan. At the same time, the system will re-optimize the inspection path of the UAV according to the new airport location to ensure that all target points can be effectively covered. The UAV will then execute the task according to the new inspection path. After the airport location update is completed, the movable UAV airport will be re-arranged according to the new location. Once the arrangement is completed, the UAV will match with the latest inspection route and airport location. This matching process includes a special spatio-temporal corridor connection operation, that is, by coordinating the intersection points in time and space, ensuring that the UAV can seamlessly connect the airport location and inspection task, avoiding resource waste or task interruption.

[0046] This dynamic adjustment ability enables the system to show extremely strong adaptability in the face of changes. Whether the number of target points increases, decreases or their positions move, the system can quickly update the airport layout and inspection path, maintaining the continuity and efficiency of the inspection task. The spatio-temporal corridor connection operation further optimizes the cooperation between the UAV and the airport, ensuring the smoothness of task execution, and is particularly suitable for scenarios with strong dynamics or frequent changes in requirements.

[0047] As a preferred embodiment, this method adds a function of dynamic task integration in step S5 to handle external task requests. Specifically, the unmanned control system can receive real-time temporary task requirements from the outside, such as emergency material delivery or inspection of emergencies. The system will analyze the relationship between the new task points and the existing inspection path through a spatio-temporal similarity matching algorithm, including their proximity in space and the possibility of matching in time. According to the analysis results, the system will select a reusable inspection path and determine the most suitable UAV and departure airport for executing the task. Subsequently, the system will insert the new task points into the existing path and adapt to the new task by adjusting the flight speed of the UAV or optimizing the path order. To ensure that the overall inspection cycle is not disrupted, the system will also adopt mechanisms such as aerial relay (i.e., multiple UAVs cooperate to complete the task) or speed compensation (i.e., making up for the time difference by increasing the flight speed).

[0048] The addition of dynamic task integration significantly improves the multi-task processing ability of the system. Through intelligent matching and path adjustment, the system can quickly respond to external demands without disturbing the original inspection plan. This flexibility is particularly suitable for scenarios where tasks need to be added temporarily, such as emergency rescue or handling of emergencies. At the same time, the aerial relay and speed compensation mechanisms ensure the stability of the overall inspection efficiency.

[0049] As a preferred implementation, during the process of dynamic task integration, the spatio-temporal similarity matching algorithm is one of the core technologies. Specifically, this algorithm evaluates from two dimensions: space and time. In terms of space, the system calculates the distance between the new task point and the existing inspection path and sets a distance threshold. Only when the distance is close enough is the path considered suitable for reuse. In terms of time, the algorithm compares the execution time period of the new task with the idle time period of the drone and calculates their coincidence ratio. Only when the coincidence degree is relatively high will the drone and path be selected. Through this dual evaluation, the system can accurately find the combination of the drone and path most suitable for executing the new task.

[0050] This algorithm ensures the efficient integration of new tasks into the existing inspection plan through precise matching in space and time. The evaluation of spatial proximity reduces the flight distance of the drone, while the calculation of time coincidence avoids problems such as task conflicts or overly long waiting times. This high-precision matching method improves the efficiency and accuracy of task allocation, providing solid technical support for dynamic task integration.

[0051] As a preferred implementation, additional considerations are added to the design of the optimization algorithm, that is, during the iterative optimization process of step S2, 5% to 10% of the transport capacity is reserved for each airport. This reserved capacity is specifically used to support the execution of emergency tasks. Specifically, when calculating the location and coverage of the airport, the system deliberately limits the maximum task load of each drone to ensure that there is still a certain margin to cope with temporary needs, such as emergency inspections or coverage of additional target points, outside the normal inspection tasks. This reservation strategy is embedded in the optimization algorithm to ensure that all plans have the ability to handle emergencies during planning.

[0052] The design of reserving transport capacity provides a buffer space for the system to handle emergency tasks, enabling the system to remain stable even under high load or emergency conditions. This forward-looking planning improves the reliability and robustness of the system. Especially in an environment with uncertain task requirements, it can effectively avoid task failures caused by insufficient resources.

[0053] As a preferred embodiment, the method adds a function of dynamic switching of emergency roles in step S5 to cope with emergencies. When the unmanned control system receives an emergency instruction, such as fire monitoring or traffic accident response, the system will assign a preset role module to the drone according to the specific type of the event (such as rescue, monitoring, etc.), and raise the priority of this task to the highest. Then, based on the geographical location of the event, the system will screen out the nearest airport from the previously optimized Pareto optimal solution set, and quickly generate an emergency path that starts from the departure airport, passes through the event point, and then connects to the next inspection point. On the visualization interface, this emergency path will be highlighted for the user to monitor in real time. At the same time, through edge computing technology, the system analyzes the task arrangements among multiple drones in real time, predicts possible conflicts, and dynamically adjusts the flight strategy to ensure smooth cooperation among all drones.

[0054] The function of dynamic switching of emergency roles enables the system to quickly respond and execute tasks in case of emergencies. By quickly assigning roles and generating emergency paths, drones can reach the event site in the shortest time, and the application of edge computing avoids multi-drone task conflicts, ensuring the safety and efficiency of task execution. This mechanism is particularly suitable for scenarios that require immediate response, significantly enhancing the system's emergency handling ability.

[0055] As a preferred embodiment, in the dynamic switching of emergency roles, the system designs multiple preset role modules for the drone, including at least three functions: traffic command, target tracking, and material delivery. For example, the traffic command role allows the drone to monitor road conditions and provide real-time guidance, the target tracking role is suitable for tracking moving targets, and the material delivery role is used for emergency transportation of small materials. To ensure the efficiency of role switching, the system pre-defines the time cost required for each role switch during the parameter initialization phase. For example, it may take a few seconds of preparation time to switch from the inspection mode to the material delivery mode. These time costs will be entered into the system in advance so that rapid execution can be achieved during actual switching.

[0056] The pre-defined role modules and switching time costs enable the system to quickly adjust the functions of the drone in case of emergencies, ensuring that it enters the task state in the shortest time. This modular design not only improves the system's response speed and flexibility but also ensures the predictability and stability of task execution by pre-planning the switching costs, providing a guarantee for the efficient completion of emergency tasks.

[0057] Example 1: Drone Airport Layout Planning

[0058] 1) Preparation of target point data

[0059] 1.1) Input target point data: The target points are the locations that the UAV needs to inspect or serve, and each target point includes longitude and latitude information.

[0060] 1.2) Verification of target point data: If the input target point data is incorrect, the system will automatically prompt the user to modify the data.

[0061] 2) Parameter initialization

[0062] 2.1) Initialize algorithm parameters, including: UAV speed (unit: m / s); UAV endurance time (unit: minutes); UAV inspection time (unit: minutes); UAV charging time (unit: minutes); airport construction cost (unit: ten thousand yuan); UAV unit price (unit: ten thousand yuan)

[0063] 2.2) Set the service radius range: Define the minimum radius (e.g., 1000 m), maximum radius (e.g., 30000 m), and service radius adjustment step (e.g., 1000 m).

[0064] 3) Loop through different service radii

[0065] 3.1) Initialize the Pareto solution and index dataset arrays to store the solution sets and plotting data for inspection plans with different coverage radii.

[0066] 3.2) Enter the loop: Start from the minimum service radius and gradually increase the radius in steps until the maximum radius is reached.

[0067] 3.3) For each service radius value, perform the following operations:

[0068] 3.3.1) Call the enhanced greedy algorithm to generate the airport locations and covered target points.

[0069] 3.3.2) Check if a valid solution is found: If a valid solution is found, proceed to the next step; otherwise, skip the current service radius and continue the loop.

[0070] 3.3.3) Calculate the total cost of the plan and the average inspection interval time:

[0071] Total cost C total = Number of airports × Airport construction cost + Number of UAVs × UAV unit price

[0072] Average interval time T avg = Total inspection cycle time of all airports / Total number of target points

[0073] 3.3.4) The system will store each solution in the Pareto solution set and index dataset arrays.

[0074] 4) Pareto front screening

[0075] 4.1) Check whether the Pareto solution set is empty: If it is empty, prompt the user that no valid solution is found; otherwise, proceed to the next step.

[0076] 4.2) Call the Pareto front algorithm to screen for the optimal solution: Screening criteria: Solutions that are optimal in both the total cost and average interval time objectives.

[0077] 4.3) Check whether a Pareto solution is found: If found, proceed to the next step; otherwise, prompt the user that no Pareto solution is found.

[0078] As a preferred implementation, the non-dominated solution set can be screened by the Pareto optimization algorithm, and the specific implementation is as follows:

[0079] Establish a bi-objective optimization model: min{f1 = C total , f2 = T avg}, that is, optimize the total cost C total and the average interval time T avg . Use the NSGA-II algorithm for population evolution, and adopt adaptive crossover and mutation rules.

[0080] The crossover probability and mutation probability are dynamically adjusted according to the population evolution state: When the population individuals are densely distributed or the fitness converges, the mutation probability is automatically increased to enhance the global search ability, and at the same time, the crossover probability is decreased to avoid gene homogenization; when the population diversity is sufficient, the crossover probability is increased to promote the combination of excellent genes, and the mutation probability is decreased to maintain the local optimization stability. This rule autonomously balances the algorithm's exploration and exploitation capabilities by real-time feedback of the population diversity level, without the need for artificial preset fixed parameters.

[0081] In addition, the Pareto front solution set is screened by fast non-dominated sorting and crowding degree calculation.

[0082] Non-dominated sorting: Stratify the solution set according to the dominance relationship of the objective function values. The first layer is the Pareto front solutions that are not simultaneously surpassed by other solutions in terms of C total and T avg .

[0083] Crowding degree calculation: First, normalize the two objective functions of cost and average interval time to eliminate the influence of dimension differences on distance calculation. Second, calculate the distance between adjacent solutions. For the individuals within the same non-dominated layer, sort them in ascending order according to each objective function value; for each individual, calculate the Euclidean distance between its two adjacent individuals in the normalized objective space as the initial crowding degree of this individual. For the individuals corresponding to the minimum and maximum values of each objective function (i.e., the boundary solutions in the objective space), assign an infinite large crowding degree to ensure their absolute retention. Finally, perform crowding degree sorting: Sort according to the individual crowding degree values from large to small, and preferentially retain the individuals with high crowding degrees to maintain the uniform distribution of the solution set in the objective space.

[0084] Since the optimal solutions in the Pareto solution set are all feasible, we use the following calculation method to select the final solution from this set:

[0085] minG n =(C n / C * )*(T n / T * ).

[0086] where n is the index of the nth solution on the Pareto front, C * , T * are the single-objective optimal solutions for cost and inspection interval time. The solution with the minimum G n value can be regarded as the final solution to be used. In practical applications, different objective weights can be set according to emergency situations. For example, select the solution with the least inspection interval time because ensuring the inspection frequency is the most important in emergency situations.

[0087] 5) Result visualization and output

[0088] 5.1) Plot the 3D graph of index data: Taking the service radius as the X-axis, the total cost as the Y-axis, and the average interval time as the Z-axis, plot the three-dimensional distribution map of all solutions.

[0089] 5.2) Plot the Pareto chart: Taking the total cost as the X-axis and the average interval time as the Y-axis, plot the Pareto front curve.

[0090] 5.3) Select the optimal solution: Select the solution with the minimum G n value from the Pareto front.

[0091] 5.4) Display the optimization results on the map: Mark the airport location, draw the airport coverage area, and display the flight path and target point coverage.

[0092] 6) End

[0093] 6.1) Output the final results, including information such as the number of airports, total cost, average inspection interval time, etc.

[0094] 6.2) End the algorithm process.

[0095] Taking the UAV inspection in Wuhan City as an example, verify the steps of the above Embodiment 1:

[0096] UAV parameter settings: Endurance 50 min, speed 10 m / s, charging time 32 min, airport cost 70,000 yuan, UAV unit price 18,000 yuan, stay and inspection time 2 minutes. Traverse the service radius from 5 - 30 km with a step of 1 km. The parameters of the target points are as follows:

[0097]

[0098]

[0099] Finally, the Pareto optimal solution is obtained, and the solution under the above parameter settings is obtained: 30 unmanned airports should be built, the total cost of purchasing unmanned airports and drones is 2.64 million yuan, the target point coverage rate is 100%, and the average inspection interval is 23.3 minutes. The specific three-dimensional decision space visualization diagram and Pareto solution are as Figure 3 、 4 shown, and the specific inspection plan and airport layout are as Figure 2 shown.

[0100] Example 2: Dynamic Task Integration and Collaborative Delivery

[0101] After the unmanned aerial vehicle (UAV) inspection path planning is completed (Step 5.4 of Example 1), the system supports dynamically receiving external tasks (such as urgent express delivery or emergency material delivery). When the demand arrives, through the spatio-temporal similarity matching algorithm, the spatio-temporal relevance analysis of the new task and the existing inspection path is carried out:

[0102] 1) Dynamic task matching mechanism:

[0103] After the inspection path planning is completed (Step 5.4 of Example 1), the system receives and processes external tasks (such as urgent express delivery, emergency material delivery) through the spatio-temporal similarity matching algorithm.

[0104] Spatial matching: Calculate the Euclidean distance between the task point and the inspection path. If it is less than 20% of the current service radius (for example, when the radius is 3000 meters, the task points within 600 meters are matched), it is determined that it can be covered by the way.

[0105] Time matching: Analyze the overlap degree between the time window for the UAV to reach the task point and the demand timeliness. If the coincidence rate ≥ 70%, the task insertion is triggered.

[0106] 2) Path dynamic optimization:

[0107] Adopt the dynamic time warping (DTW) algorithm to insert task points into the original inspection path, and compensate for the time delay in the following ways:

[0108] Speed adjustment: Increase the UAV cruise speed to 110% of the nominal value (for example, from 10 m / s to 11 m / s) to ensure that the total inspection cycle remains unchanged.

[0109] Air relay: If a single UAV cannot meet the timeliness, an idle UAV is dispatched from an adjacent airport to perform the task by relay, and returns to the original path after the task is completed.

[0110] 3) Resource reservation strategy:

[0111] When generating the airport layout using the enhanced greedy algorithm (Step 3.3.1 of Example 1), 10% of the drone capacity is forcibly reserved for each airport (for example, if a certain airport plans 10 drones, then an additional 1 drone is reserved) for emergency task response.

[0112] The applicable scenarios of Example 2 are as follows:

[0113] Logistics distribution: In areas with inconvenient transportation such as mountains and islands, when drones are patrolling power grids or base stations, they can deliver small express packages or maintenance tools and other supplies on the way.

[0114] Delivery of emergency supplies: When an emergency occurs, use the existing patrol routes to transport supplies such as medicines and medical kits to the accident site to avoid opening a separate flight path.

[0115] Example 3: Dynamic switching of emergency roles and multi-functional scheduling

[0116] After the airport and routes are determined (Step 5.4 of Example 1), drones can switch to specific roles according to real-time emergency needs (such as traffic control, target tracking):

[0117] 1) Modular task switching:

[0118] Drones are equipped with pluggable task modules (such as megaphones, searchlights, infrared cameras), and the module switching time (such as 5 minutes) and energy consumption parameters are preset during the parameter initialization stage (Step 2.1 of Example 1).

[0119] When receiving an emergency instruction (such as traffic control, fire monitoring), the system automatically matches the nearest airport, schedules the drone to replace the module and updates the task priority.

[0120] 2) Dynamic priority scheduling:

[0121] During the Pareto front screening stage (Step 4.2 of Example 1), an event impact factor (such as event level, covered population) is introduced to adjust the solution set weight. For example, the priority of the traffic control task is higher than that of regular patrols, and the system preferentially selects a solution with a total cost increase ≤ 15% but the shortest response time.

[0122] 3) Multi-drone collaborative obstacle avoidance:

[0123] In the visualization interface (Step 5.4 of Example 1), the emergency task paths are highlighted in red, and the multi-drone path conflicts are predicted in real time through edge computing. If the conflict probability > 30%, the path replanning is automatically triggered, and the distributed negotiation algorithm is used to adjust the flight altitude or speed of the drones.

[0124] The applicable scenarios of Example 2 are as follows:

[0125] Traffic control: In case of highway accidents, the drone switches to the loudspeaker module and takes off from the nearest airport to direct traffic flow.

[0126] Target tracking: When the police are chasing fugitives, drones equipped with thermal imaging cameras are dispatched to lock the target positions.

[0127] Disaster inspection: At the fire scene, the drone switches to the gas sensor module to monitor the spreading direction of the fire in real time.

[0128] Example 4: Dynamic obstacle avoidance deployment and path coordination optimization of mobile airports

[0129] Based on the static layout of Example 1, this example realizes the elastic expansion and seamless scheduling of the drone network in complex urban environments through dynamic obstacle avoidance deployment of mobile airports, airport migration triggering and path coordination mechanisms, local path replanning and spatio-temporal corridor connection.

[0130] 1) Dynamic obstacle avoidance deployment of mobile airports

[0131] This step is used to solve the problem of invalid site selection caused by environmental obstacles during airport deployment, and ensure the safe deployment of mobile drone airports in complex terrains.

[0132] 1.1) Environmental perception and obstacle avoidance strategy

[0133] Based on the airport layout generated in Example 1, the mobile drone airport performs the following operations:

[0134] 1.1.1) LiDAR modeling:

[0135] Based on the drone speed v nominal (unit: m / s) and service radius R cover (unit: meters) defined in step 2.1 of Example 1, the airport is equipped with a LiDAR with a scanning radius of 0.2R cover range (e.g., when R cover = 3000 meters, it scans 600 meters) to generate a three-dimensional environmental map. Identify static obstacles (such as buildings) and dynamic obstacles (such as vehicles) around the deployment point to provide data support for obstacle avoidance decisions.

[0136] 1.1.2) Obstacle avoidance decision:

[0137] For static obstacles (buildings, trees): Trigger the airport to translate to the pre-equipped backup point (distance from the original coordinates ≤ 0.15R cover ). Avoid site selection conflicts. For dynamic obstacles (vehicles, pedestrians): Adjust the takeoff and landing direction (deflection angle ≤ 20°) through real-time point cloud updates to ensure the safety of takeoff and landing paths.

[0138] 1.2) Adaptive deployment verification

[0139] 1.2.1) Terrain Adaptation: According to the flight endurance time T of the drone in step 2.1 of Embodiment 1 endurance (unit: minutes), detect the slope α of the deployment point (formula: α = arctan(Δz / Δx)). If α > 5°, automatically level the drone to ensure the stability of takeoff and landing.

[0140] 1.2.2) Electromagnetic Calibration: To prevent communication interference from affecting the inspection cycle generated in Embodiment 1, detect communication frequency band interference. If the signal-to-noise ratio is lower than the inspection cycle threshold calculated in step 3.3.3 of Embodiment 1, switch to an anti-interference protocol.

[0141] 2) Airport Migration Trigger and Path Coordination Mechanism

[0142] When the target point set or the environment changes, dynamically optimize and adjust the airport location and optimize the associated path of the drone to maintain the integrity and efficiency of the inspection network.

[0143] 2.1) Migration Trigger Conditions

[0144] 2.1.1) Update of Target Point Set: When the number of newly added / deleted target points ΔN satisfies ΔN / Nt otal ≥ 5% (N total is the total number of target points input in step 1.1 of Embodiment 1), or the coverage gap rate η gap = N uncovered / N total > 10%, trigger migration to re-meet the coverage requirements.

[0145] 2.1.2) Environmental Constraint: When an obstacle invades the safety zone (within <0.1R from the airport cover ) and cannot be avoided, force migration.

[0146] 2.2) Path Coordination Optimization Process

[0147] 2.2.1) Path Inheritance: Retain the unaffected flight segments in the visualization result of step 5.4 of Embodiment 1 to reduce the cost of repeated calculation.

[0148] 2.2.2) Local Re-planning: For the flight segments affected by airport migration, use the bidirectional A* algorithm to generate a new path.

[0149] 3) Local Path Re-planning and Connection with the Space-Time Corridor

[0150] After the airport relocation is triggered by changes in the target point set or the environment, the system will perform local replanning on the affected UAV paths according to the new layout requirements. Since the airport deployment takes a certain amount of time, the UAVs will first execute the inspection tasks according to the optimized new paths, while the airport is deployed at the new location. This deployment lag may lead to inconsistent scheduling of UAVs before and after the completion of the airport relocation. Therefore, this operation ensures that the UAVs seamlessly switch to the new paths during the airport relocation process and dynamically match the newly deployed airport location through local path replanning and spatio-temporal corridor connection.

[0151] 3.1) Local path replanning

[0152] For the flight segments affected by the airport relocation, generate new UAV inspection paths to ensure that all target points are effectively covered under the new layout, while minimizing the impact of path adjustment on the overall inspection cycle.

[0153] Analyze the impact of airport relocation on the existing inspection paths and identify the flight segments that need to be adjusted due to the change in airport location. Specifically, the flight segments directly associated with the relocated airport (i.e., the segments taking off or landing at this airport) will be marked as affected segments. For each affected segment, use the bidirectional A* algorithm to generate a new inspection path.

[0154] The constraints for replanning using the bidirectional A* algorithm include:

[0155] Single-segment path length limit: The single-segment length L of the new path j satisfies L i ≤v nominal ×T endurance ×60, where v nominal is the nominal speed of the UAV (m / s) and T endurance is the endurance time (minutes).

[0156] Turning angle limit: The turning angle θ in the new path ≤ 30° to avoid overly circuitous paths.

[0157] The bidirectional A* algorithm searches from both the starting point and the ending point simultaneously to improve the planning efficiency. Here, a speed adjustment strategy is used to minimize the impact of path adjustment on the overall inspection cycle. Dynamically adjust the cruising speed v of the UAV according to the relocation distance ΔD adjusted , compensate for the change in the flight range, and ensure that the total inspection cycle remains unchanged. Speed adjustment rule:

[0158]

[0159] where, R cover is the service radius (meters).

[0160] 3.2) Spatio-temporal corridor connection operation

[0161] In the case where the airport deployment lags behind the new path inspection of the drone, through the connection operation of the spatio-temporal corridor, ensure that the drone seamlessly switches between the old and new airport positions and dynamically matches the newly deployed airport position.

[0162] Define the spatio-temporal corridor: The spatio-temporal corridor is a flight path restricted in both time and space, used to coordinate the flight path of the drone during airport relocation. Specifically, the spatio-temporal corridor defines the geographical area that the drone must pass through within a specific time period to ensure that it can reach the new airport position in time after the airport deployment is completed.

[0163] 3.2.1) Calculate the intersection point of the optimal inspection route: On the new inspection path, identify the inspection point (i.e., the intersection point) closest to the new airport position as the key node for the drone to match with the new airport. The selection criteria for the optimal intersection point are as follows:

[0164] Time window: The time t when the drone reaches the intersection point intersect Must be after the airport deployment completion time t deploy And the delay time Δt = t intersect -t deploy ≤0.1T cycle where t deploy is the inspection cycle.

[0165] Spatial distance: The distance d from the intersection point to the new airport intersect ≤0.2R cover .

[0166] 3.2.2) Dynamic path adjustment:

[0167] If an intersection point that meets the conditions cannot be directly found, the system creates a new intersection point by adjusting the flight speed or path:

[0168] Speed fine-tuning: Without violating the energy consumption limit, fine-tune the flight speed so that the drone reaches the inspection point near the new airport after t deploy .

[0169] Path insertion: Insert temporary stop points or detour points in the inspection path to delay the arrival time to match the airport deployment progress.

[0170] 3.2.3) Spatio-temporal corridor generation:

[0171] Generate a spatio-temporal corridor for each affected drone to ensure that it continues to take off and land at the old airport before t deploy and switches to the new airport after t deploy . The definition of the spatio-temporal corridor includes:

[0172] Spatial constraint: When t < t deploy , the path connects to the old airport; when t ≥ t deployWhen, the path connects to the new airport.

[0173] Time constraint: The UAV must be within the service range of the new airport at time t deploy or be able to return to the new airport in time during subsequent flights.

[0174] 3.2.4) Multi-UAV cooperation and conflict avoidance:

[0175] In the spatio-temporal corridor, the flight paths of multiple UAVs are monitored in real time. If a conflict is predicted (two UAVs enter the same airspace at the same time), obstacle avoidance is carried out by adjusting the flight altitude or speed. If the conflict probability P conflict > 20%, trigger path replanning and use a distributed negotiation algorithm to adjust the flight parameters.

[0176] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0177] Each embodiment in this specification is described in a progressive manner. The key point of each embodiment is to illustrate the differences from other embodiments. The same or similar parts among the embodiments can be referred to each other.

[0178] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a device, or a computer program product. Therefore, the embodiments of the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0179] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal devices generate a device for realizing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0180] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0181] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0182] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0183] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or terminal device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of additional identical elements in the process, method, article, or terminal device that includes the element.

[0184] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments can still be modified, or some of the technical features thereof can be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for drone inspection based on airport layout optimization, characterized in that, It includes the following steps: S1: Receive target point set data, UAV and airport parameter data; S2: Iteratively optimize multiple preset service radii, call the optimization algorithm to generate an airport location set and a covered target point subset. The algorithm adopts a dynamic candidate set strategy, retains the top N optimal candidate airports in each iteration, and calculates the total cost and average inspection interval of each solution; S3: Construct an objective function, screen the non-dominated solution set, and generate a two-dimensional Pareto front; S4: Synchronously output a three-dimensional decision space diagram and a two-dimensional Pareto front diagram, and overlay and display the optimal airport location, coverage radius, and UAV inspection path on the geographic information system; S5: Arrange the airport according to the optimal airport location. Each of the airports is provided with an unmanned control system, and each of the unmanned control systems performs task allocation and path planning for the UAVs within the airport according to the coverage radius and target point distribution.

2. The method for unmanned aerial vehicle inspection based on airport layout optimization according to claim 1, wherein, In step S3, the screening of the non-dominated solution set through the Pareto optimization algorithm is specifically implemented as follows: Establish a bi-objective optimization model: min{f1 = C total , f2 = T avg}; Use the NSGA-II algorithm for population evolution, and screen the Pareto front solution set through fast non-dominated sorting and crowding degree calculation.

3. The method for drone inspection based on airport layout optimization according to claim 1, wherein The airport is deployed through a mobile UAV airport. The mobile UAV airport is a UAV vehicle. The mobile UAV airport constructs a three-dimensional environmental map around the airport through lidar and performs dynamic obstacle avoidance deployment.

4. The drone inspection method based on airport layout optimization according to claim 1, wherein According to the change of the target point set data, update the airport location to obtain the updated new airport location, and optimize to obtain a new inspection route. The UAV performs inspections along the new inspection route. Subsequently, after the mobile UAV airport is arranged according to the new airport location, the UAV matches the new airport location with the intersection point of the new inspection route. The intersection point matching of the new inspection route includes space-time corridor connection operations.

5. The drone inspection method based on airport layout optimization according to any one of claims 1 to 4, characterized in that, It also includes a dynamic task integration step: in step S5, the unmanned control system dynamically receives external task requests, calculates the spatial proximity and time window coincidence degree between the task point and the existing inspection path through the spatio-temporal similarity matching algorithm; selects the UAVs and departure airports with reusable paths, inserts the task points and dynamically adjusts the flight speed or path, and ensures that the total inspection cycle remains unchanged through an air relay or speed compensation mechanism.

6. The method for drone inspection based on airport layout optimization according to claim 5, wherein, In the spatio-temporal similarity matching algorithm, the spatial proximity is determined by the Euclidean distance threshold, and the time window coincidence degree is calculated by the intersection ratio of the task execution period and the UAV idle period.

7. The method for UAV inspection based on airport layout optimization according to claim 6, wherein It also includes a dynamic task integration step: in step S2, reserve 5% - 10% of the transport capacity for each airport in the optimization algorithm to support emergency tasks.

8. The method for unmanned aerial vehicle inspection based on airport layout optimization according to any one of claims 1 to 4, characterized in that It also includes an emergency role dynamic switching step: in step S5, when the unmanned control system receives an emergency event instruction, assign a preset role module to the UAV according to the event type, and update its task priority; Based on the event geographical location, screen the nearest airport from the Pareto optimal solution set to generate an emergency path of "airport → target point → next target point"; highlight the emergency path in the visualization interface, and dynamically adjust the cooperation strategy by real-time predicting multi-UAV task conflicts through edge computing.

9. The method for unmanned aerial vehicle inspection based on airport layout optimization according to claim 8, wherein, The role module includes at least one of traffic command, target tracking, and material delivery, and the role switching time cost is predefined in the parameter initialization stage.

10. A system for implementing the drone inspection method based on airport layout optimization according to any one of claims 1 to 9, characterized in that It includes: A data receiving module for efficient integration and verification of multi-source data; An optimization processing module to improve the coverage rate and cost-effectiveness of airport site selection through dynamic candidate set screening; A decision-making optimization module to achieve the global optimal balance between cost and timeliness using the multi-objective Pareto front; A visualization module to realize multi-dimensional dynamic deduction of the solution by integrating the three-dimensional efficiency space and geographical information; An execution control module to automatically generate an airport deployment plan and a collaborative UAV inspection path based on the optimal solution.

Images