A method for continuous unmanned aerial vehicle (UAV) airport inspection and dispatch

Abstract

This invention provides a sequential UAV airport inspection scheduling method. The inspection platform creates a flight route task, searches for the most suitable airport A for flight, sets 80% of the flight points of the route within the airport's operational range, ensures that the airport's operational power meets flight requirements, detects the airport status, searches for the most suitable airport B for landing, and determines that the last point of the route is the nearest airport. After confirming that the airport status is normal, a leapfrog task is issued, taking off from airport A and executing until landing at airport B, starting from airport A again. Applying this technical solution can improve the efficiency of UAV inspection and save time and energy.

Description

Technical Field

[0001] This invention relates to the field of power drone inspection technology, and in particular to a continuous drone airport inspection and scheduling method. Background Technology

[0002] Unmanned aerial vehicle (UAV) inspection operations have been widely adopted in the power sector, gradually shifting from manual inspections to fully autonomous, unmanned inspections based on UAV airports. UAV airports (also known as hangars or drone shelters) provide storage space, environmental monitoring, data transmission, and power replenishment for UAVs. They consist of a main control module, electrical module, communication module, monitoring module, positioning module, and mechanical structure. Some customized UAV airports also feature artificial intelligence edge computing, battery swapping, and load replacement capabilities. UAV airports enable remote inspection operations by UAVs, effectively improving the efficiency and potential of power sector inspections.

[0003] However, existing drone airports are mostly deployed at single points, requiring drones to travel back and forth for operations. There is insufficient communication between airports, and they lack the capability for continuous or leapfrog inspection operations. Given the numerous and complex routes of power distribution network lines, existing drone airports require the deployment of a large number of airports and drones, resulting in low construction efficiency.

[0004] Existing drone inspection methods require the deployment of a large number of airports and drones. In addition, the relatively short distances of power distribution lines, dense grids, and complex branch paths in practical applications greatly increase construction costs. Furthermore, there is no completely safe and reliable solution for drones that cannot return to their designated routes when their power is low.

[0005] The main challenges currently facing drone inspection path planning include insufficient adaptability and robustness to complex environments and uncertainties, as well as inadequate effectiveness and real-time performance in data processing and analysis. When conducting inspection missions, drones must contend with complex geographical environments, obstacles, weather changes, and other uncertainties, all of which can impact their flight and inspection tasks. Furthermore, efficient and accurate methods are needed to process and analyze large amounts of inspection data to obtain useful information and conclusions. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a sequential UAV airport inspection scheduling method. Through the sequential inspection method, this invention can improve the efficiency of UAV inspection and save time and energy.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a continuous UAV airport inspection and scheduling method, wherein the inspection platform creates a flight route task, searches for the most suitable airport A for flight, 80% of the flight points of the route are set within the airport's operating range, the airport's operating power needs to meet the flight requirements, detects the airport status, searches for the most suitable airport B for landing, the last point of the route is the nearest airport, after determining that the airport status is normal, issues a leapfrog task to take off from airport A and execute to land at airport B, starting from airport A; detects whether there is a UAV at airport A, if there is a UAV, detects whether airport B meets the landing conditions, if airport B meets the landing conditions, detects whether the airspace is safe, if the airspace is safe, the aircraft takes off from airport A, executes the task, and lands at airport B; if the airspace is not safe, the trip ends; if airport B does not meet the landing conditions, the trip also ends.

[0008] The system connects to the airport, which is located within the power supply station's jurisdiction, via a wired connection. A convergence switch and firewall are used to transmit the communication signals to the drone command center. The airport acts as a front-end communication node, communicating with the drone via a wireless encryption protocol and connecting to the drone command center via a wired network. Simultaneously, in airport locations where wired network deployment is inconvenient, 4G can be used to resolve communication connectivity issues. The airport solution utilizes mesh equipment, 5G, and 4G / converged terminals for flexible networking.

[0009] An improved Bald Eagle optimization algorithm is used to optimize the UAV inspection path planning method, enabling the UAV to select the shortest inspection path with the fewest obstacles from the starting point to the starting point when performing a mission. This includes the following steps:

[0010] Step 1: Construct the drone flight environment and set the starting and target points for the drone's mission;

[0011] Step 2: Based on the set starting and target points of the UAV's mission flight, the UAV path planning problem is transformed into an optimization mathematical model. The UAV path planning problem is a shortest distance problem and a minimum obstacle problem. The optimization mathematical model is used to find the optimal path point for the UAV using the improved Bald Eagle optimization algorithm.

[0012] Step 3: Improve the Bald Eagle optimization algorithm, including the following two steps:

[0013] D1. Introduce an adaptive nonlinear step size factor to optimize the population update strategy of the Bald Eagle optimization algorithm;

[0014] D2. Introduce a probability factor mechanism. When the probability factor is greater than the set threshold, execute the population position update strategy of the vulture algorithm in the exploration phase; otherwise, execute the population position update strategy in the search phase.

[0015] Step 4: Optimize the UAV path planning mathematical model using the improved Bald Eagle optimization algorithm to obtain the shortest path with the fewest obstacles for the UAV when performing the mission, i.e., the optimal path.

[0016] In a preferred embodiment, when a drone takes off from airport A, other drones are allowed to land at airport A, while airport B enters a mission state, meaning that other drones are not allowed to land.

[0017] In a preferred embodiment, if an unexpected situation occurs during the flight of the drone, Airport A reports its current location to the inspection platform and requests the most suitable return point. The inspection platform searches for the most suitable airport for return and replies to Airport A. Airport A refreshes the return point and notifies Airport B to prepare to recover the drone.

[0018] In a preferred embodiment, the bald eagle optimization algorithm is improved by introducing an adaptive nonlinear step size s, the mathematical formula of which is:

[0019]

[0020] In the formula, s is the adaptive nonlinear step size; min The minimum step size is set to 0; s max The maximum step size is set to 2; t is the current iteration number; t_mid is half the current iteration number; η is the step size change rate, which is 1.

[0021] In a preferred embodiment, the bald eagle optimization algorithm is improved by introducing a probability factor mechanism, the specific process of which is as follows:

[0022] Step 1: Define the upper bound U and the lower bound L of the fitness value of the algorithm; if the current fitness value is greater than the upper bound U, execute the exploration phase of the vulture optimization algorithm; if the current fitness value is less than the lower bound L, execute the search phase of the vulture optimization algorithm.

[0023] Step 2: If the current fitness value is between the upper and lower bounds, then according to the probability factor P, if P is greater than 0.5, execute the search phase; otherwise, execute the vulture optimization algorithm exploration phase.

[0024] Step 3, the mathematical formula for probability factors is:

[0025]

[0026] In the formula, F(t) is the fitness value of the current solution, and λ is the threshold factor, which takes a value of 2 to control the rate of change of P.

[0027] In a preferred embodiment, the improved Bald Eagle optimization algorithm is used to optimize the mathematical model of UAV path planning. The specific process is as follows:

[0028] S1. Based on the optimization problem of UAV path planning, design an improved fitness function for the Bald Eagle optimization algorithm. The algorithm determines the optimal path point for the UAV through the fitness function; the fitness function formula is:

[0029]

[0030] In the formula, t1 is the total time of the entire drone path, length is the length of the drone path, w2 is the weight coefficient with a value of 1.5, and count is the number of obstacles in the identified drone path.

[0031] S2. Initialize the improved vulture optimization algorithm, including initializing the initial position of the vultures, the population size N, the search space range of the vultures, the upper bound ub and the lower bound lb, the dimension of the optimization problem dim, and the maximum number of iterations T. max ;

[0032] The initial position of the bald eagle is the initial path point in the UAV path selection, and the search space range is the range of the UAV path selection.

[0033] S3. In the improved vulture optimization algorithm, the position code of each vulture is used as a path for UAV inspection.

[0034] S4. Calculate the fitness value of each bald eagle in the current iteration of the improved bald eagle optimization algorithm, and record the minimum fitness value to determine the position of the bald eagle corresponding to the minimum fitness value, and then determine the current best solution, which is the best path of the current UAV.

[0035] S5. Compare the minimum fitness value of the current iteration with the minimum fitness value of the previous iteration, determine the minimum fitness value between the two, update it to the best fitness value, and then execute the population position update formula of the improved vulture optimization algorithm.

[0036] S6. Activate the probability factor mechanism; S7. Recalculate the fitness value of the bald eagle population, introduce a dynamic reverse refraction strategy, and dynamically adjust the search direction and search range according to the current optimal solution and the change of fitness value, so as to facilitate the optimization of the UAV inspection path;

[0037] S8. Repeat the iteration process from S4 to S7 until the maximum number of iterations is met. Then, output the optimal solution and decode to obtain the best path point for the UAV inspection.

[0038] In a preferred embodiment, step 6 specifically includes:

[0039] S6.1 When the fitness value is less than the lower bound L, it means that the algorithm search has not yet converged. In this case, the search phase continues. An adaptive nonlinear step size factor s is introduced to optimize the position update strategy. The improved search position update formulas are Equations (3) and (4):

[0040]

[0041] In the formula, X best This represents the optimal path for the drone, which is also the optimal position for the current vulture. X represents the latest position of the i-th individual bald eagle. i Let X be the position of the i-th vulture in the previous iteration, a be a linear control factor, r be a random number between (0, 1), and X be a random number between (0, 1). mean is the average position of the bald eagle population, and S is the adaptive nonlinear step size;

[0042]

[0043] In the formula, X i Let X be the position of the i-th individual bald eagle in the previous iteration, which is the path of the i-th drone in the previous iteration of the algorithm. i+1 Let x(i) be the position of the (i+1)th vulture in the last iteration, and y(i) be the vulture position in polar coordinates, and S be the adaptive nonlinear step size.

[0044] S6.2 When the fitness value is greater than the upper bound U, it indicates that a relatively good solution has been found, and the exploration stage begins to try to find a better solution. The exploration location update formula is Equation (5):

[0045]

[0046] In the formula, rand is a uniformly random number in [0,1], and c1 and c2 are constants with values ​​of 1.2;

[0047] S6.3 When the fitness value is between the upper and lower bounds, calculate the probability P according to formula (2). If P is greater than the preset threshold of 0.5, continue to execute the search phase; otherwise, enter the exploration phase.

[0048] Compared with the prior art, the present invention has the following advantages: the present invention supports multiple communication methods, and improves the security of data transmission while ensuring the stability of data transmission; through the automatic return-to-home function, the present invention can ensure the safety of the drone in case of unexpected situations and reduce the risks in the inspection process.

[0049] By introducing an adaptive nonlinear step size factor, the population update strategy of the Bald Eagle optimization algorithm is optimized. A probability factor mechanism is added during the algorithm iteration process. Through this mechanism, we can determine whether the current algorithm is executing the search phase or the exploration phase based on the statistical information of the fitness value and the termination condition of the search phase, thereby achieving better search results and improving the algorithm's optimization speed. This provides better paths for UAV inspection mission path planning. Through the leapfrog inspection method and path planning optimization algorithm, this invention can improve the efficiency of UAV inspection, save time, energy and construction costs; at the same time, it can improve the adaptability and robustness of UAVs to complex environments and uncertainties, as well as improve the effectiveness and real-time performance of data processing and analysis. Attached Figure Description

[0050] Figure 1 This is a flowchart of the scheduling algorithm of a preferred embodiment of the present invention;

[0051] Figure 2 This is a schematic diagram of the main tasks of each part in a preferred embodiment of the present invention;

[0052] Figure 3 A simplified flowchart (I) of the scheduling algorithm of a preferred embodiment of the present invention;

[0053] Figure 4 A simplified flowchart (II) of the scheduling algorithm of a preferred embodiment of the present invention;

[0054] Figure 5 The flowchart illustrates the scheduling process for a drone encountering unexpected situations, as per a preferred embodiment of the present invention. Detailed Implementation

[0055] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0056] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0057] It should be noted that the terminology used herein is for the purpose of describing particular implementations only and is not intended to limit the exemplary implementations according to this application; as used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise; furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0058] A method for continuous unmanned aerial vehicle (UAV) airport inspection and dispatching, with reference to Figure 1-5The inspection platform creates a flight route task, uses a scheduling algorithm to find the most suitable airport A for flight, and sets 80% of the flight points of the route within the airport's operational range. The airport's operational power must meet flight requirements. It checks the airport status and finds the most suitable airport B for landing, which is the airport from which the last point of the route is located. After determining that the airport status is normal, it issues a leapfrog mission to take off from airport A and land at airport B. The mission starts from airport A. It checks if there are drones at airport A. If there are drones, it checks if airport B meets the landing conditions. If airport B meets the landing conditions, it checks if the airspace is safe. If the airspace is safe, the aircraft takes off from airport A, performs the mission, and lands at airport B. If the airspace is unsafe, the mission ends. If airport B does not meet the landing conditions, the mission ends.

[0059] The continuous drone airport is connected via wired connection to the airport installed at the location under the jurisdiction of the power supply station, and uses an aggregation switch and firewall to connect the communication signal to the drone command center. The airport, as the front-end communication node, communicates with the drones through a wireless encryption protocol and connects to the drone command center through a wired network. In airport locations where it is inconvenient to deploy wired networks, 4G / 5G and Mesh self-organizing networks can be used to solve the communication connectivity problem. The airport solution uses the addition of mesh equipment, 5G, 4G / converged terminals for flexible networking, allowing airport operations in areas with weak signals to be carried out without restriction. Different communication methods are selected for different areas, and there are no specific restrictions.

[0060] When a drone takes off from airport A, other drones are allowed to land at airport A, while airport B enters mission mode, meaning that other drones are not allowed to land.

[0061] If an unexpected situation occurs during the drone's flight, Airport A reports its current location to the inspection platform and requests the most suitable return point B. The inspection platform searches for the most suitable return point and replies to Airport A. Airport A then refreshes the return point and notifies Airport B to prepare to recover the drone.

[0062] This invention employs multiple communication methods, including wired networks, 4G / 5G, and Mesh device networking, to achieve flexibility and reliability in data transmission.

[0063] The power supply station, as a key node in the communication link, connects to the airport located within its jurisdiction via wired connections and utilizes aggregation switches and firewalls to transmit communication signals to the drone command center. The airport, as a front-end communication node, communicates with the drones not only via wireless encryption protocols but also via wired network connections to the drone command center. Furthermore, in airport locations where wired network deployment is inconvenient, 4G can be used to resolve communication connectivity issues. Since weak or even nonexistent signal areas are unavoidable on the distribution network lines, this airport solution allows for flexible networking using mesh devices, 5G, and 4G / converged terminals, enabling unrestricted airport operations in areas with weak signals. Different communication methods can be selected for different regions, with no specific restrictions.

[0064] If the drone encounters an unexpected situation during flight, it will automatically return to the nearest airport where it can land.

[0065] Path planning algorithm: Generally, UAV route planning involves flying to the specified number of waypoints, starting from the takeoff point and returning to it. This route planning is often limited by the UAV's return time, leading to a sacrifice of effective operational time due to the need to consider the return journey. This proposed solution optimizes path planning by: if the return time to the final waypoint is insufficient, but there is an adjacent airport with sufficient time, the UAV will be instructed to land at the adjacent airport.

[0066] Supplement: Assuming the drone's battery consumption rate (v1) is 3% / minute at a flight speed of 2 m / s, then 3% of the battery allows for a flight distance of 120 m. (Note: This is an empirical value, pre-set in the algorithm; the specific numerical value is not directly related to the core of this invention.)

[0067] Assuming the low battery emergency landing protection power (s2) of the drone is 8% (empirical value, same as above), the relationship between the drone's return distance and the battery level is: Drone's return distance (s1) = [Drone's current battery level (B1) - Drone's low battery emergency landing protection power (s2)] / battery depletion rate (v1) * 60; The unit of drone's return distance (s1) is meters.

[0068] Let the latitude and longitude of the UAV's final destination be (lat1, longt1) (obtained from the flight path). Divide this into intervals of 100m to obtain the interval [p1, d1], where (lat1, longt1) lies within the interval [p1, p2]. Similarly, obtain the location interval [px, dx] of the airports entered in the system (all with known latitude and longitude). By adding the UAV's return distance (s1), obtain the interval range [tp1, td1] centered at [p1, d1] and extending outwards by s1 distances. By comparing the interval range [px, dx], the matching landing airports are obtained.

[0069] An improved Bald Eagle optimization algorithm is used to optimize the UAV inspection path planning method, enabling the UAV to select the shortest inspection path with the fewest obstacles from the starting point to the starting point when performing a mission. This includes the following steps:

[0070] Step 1: Construct the drone flight environment and set the starting point and target point for the drone's mission.

[0071] Step 2: Based on the starting point and target point of the drone's mission flight, the drone path planning problem is transformed into an optimization mathematical model. The drone path planning problem is a shortest distance problem and a minimum obstacle problem. The optimization mathematical model is used to find the optimal path point of the drone using the improved Bald Eagle optimization algorithm.

[0072] Step 3: Improve the Bald Eagle optimization algorithm, including the following two steps:

[0073] D1. Introduce an adaptive nonlinear step size factor to optimize the population update strategy of the Bald Eagle optimization algorithm;

[0074] D2. Introduce a probability factor mechanism. When the probability factor is greater than the set threshold, execute the population position update strategy of the vulture algorithm in the exploration phase; otherwise, execute the population position update strategy in the search phase.

[0075] Step 4: Optimize the UAV path planning mathematical model using the improved Bald Eagle optimization algorithm to obtain the shortest path with the fewest obstacles for the UAV when performing the mission, i.e., the optimal path.

[0076] Furthermore, in step one, constructing the UAV flight environment, it is necessary to determine the UAV's environmental parameters and performance parameters. The UAV performance parameters include maximum airspeed, maximum climb rate, maximum descent rate, maximum turning radius, and minimum turning radius. The UAV's environmental parameters include terrain elevation data, obstacle location data, and meteorological data.

[0077] Furthermore, in step two, the UAV path planning problem is a shortest distance problem and a minimum obstacle problem. The UAV starts from the starting point (x) start ,y start ,z start ) to the destination (x goal ,y goal ,z goal The path length of the drone can be represented as a series of coordinate points (x, y). i ,y i ,z i ), where i = 1, 2, ..., N, and these coordinate points constitute the flight path of the UAV.

[0078] Furthermore, in this invention, during UAV path planning, obstacles are objects or terrain features that affect the UAV's flight, such as mountains or buildings. A Gaussian function is used to represent the influence of obstacles, and the formula is:

[0079]

[0080] In the formula, A takes a value of 2, representing the intensity of the obstacle's influence, and σ x ,σ y ,σ z The standard deviations of the Gaussian function in the X, Y, and Z axes of the UAV in a three-dimensional environment determine the range of influence of obstacles.

[0081] Furthermore, in step three, an adaptive nonlinear step size factor is introduced to optimize the population update strategy of the Bald Eagle optimization algorithm. In the initial search phase, the step size gradually increases from s_min, providing the algorithm with sufficient motivation to explore new and potentially more promising regions. As the search progresses, the step size gradually decreases, giving the algorithm more opportunities to conduct a more refined search within known regions that may contain the optimal solution.

[0082] This ensures that the algorithm has sufficient exploration capability in the early stages of the search to avoid getting trapped in local optima; in the later stages of the search, the algorithm will gradually reduce the step size to increase the accuracy of the search.

[0083] Furthermore, in step three, the bald eagle optimization algorithm is improved by introducing a probability factor mechanism. The specific process is as follows:

[0084] Step 1: Define the upper bound U and the lower bound L of the fitness value of the algorithm; if the current fitness value is greater than the upper bound U, execute the exploration phase of the vulture optimization algorithm; if the current fitness value is less than the lower bound L, execute the search phase of the vulture optimization algorithm.

[0085] Step 2: If the current fitness value is between the upper and lower bounds, then according to the probability factor P, if P is greater than 0.5, execute the search phase; otherwise, execute the vulture optimization algorithm exploration phase.

[0086] Step 3, the mathematical formula for probability factors is:

[0087]

[0088] In the formula, F(t) is the fitness value of the current solution, and λ is the threshold factor, which takes a value of 2 to control the rate of change of P.

[0089] Furthermore, in Step 1, the upper and lower bounds of the fitness value play a crucial role. The upper bound U indicates that we have found a fairly good solution, while the lower bound L indicates that the search has not yet converged and we need to continue exploring. In Step 3, the threshold factor controls the rate of change of the probability P. As the search progresses, the algorithm's desire to explore new solutions will gradually decrease. In other words, the algorithm will increasingly tend to find better solutions in the known search space.

[0090] Furthermore, in step four, the improved Bald Eagle optimization algorithm is used to optimize the UAV path planning mathematical model. The specific process is as follows:

[0091] S1. Based on the optimization problem of UAV path planning, design an improved fitness function for the Bald Eagle optimization algorithm. The algorithm determines the optimal path point for the UAV through the fitness function; the fitness function formula is:

[0092] In the formula, t1 is the total time of the entire drone path, length is the drone path length, w2 is the weight coefficient with a value of 1.5, and count is the number of obstacles identified in the drone path. S2, Initialize the improved vulture optimization algorithm, including initializing the initial position of the vultures, the population size N of the vultures, the search space range of the vultures, the upper bound ub and the lower bound lb, the optimization problem dimension dim, and the maximum number of iterations T. max ;

[0093] The initial position of the bald eagle is the initial path point in the UAV path selection, and the search space range is the range of the UAV path selection.

[0094] S3. In the improved vulture optimization algorithm, the position code of each vulture is used as a path for UAV inspection.

[0095] S4. Calculate the fitness value of each bald eagle in the current iteration of the improved bald eagle optimization algorithm, and record the minimum fitness value to determine the position of the bald eagle corresponding to the minimum fitness value, and then determine the current best solution, which is the best path of the current UAV.

[0096] S5. Compare the minimum fitness value of the current iteration with the minimum fitness value of the previous iteration, determine the minimum fitness value between the two, update it to the best fitness value, and then execute the population position update formula of the improved vulture optimization algorithm.

[0097] S6. Activate the probability factor mechanism, the steps are as follows:

[0098] S6.1 When the fitness value is less than the lower bound L, it means that the algorithm search has not yet converged. In this case, the search phase continues. An adaptive nonlinear step size factor s is introduced to optimize the position update strategy. The improved search position update formulas are Equations (3) and (4):

[0099]

[0100] In the formula, X best This represents the optimal path for the drone, which is also the optimal position for the current vulture. X represents the latest position of the i-th individual bald eagle. i Let X be the position of the i-th vulture in the previous iteration, a be a linear control factor, r be a random number between (0, 1), and X be a random number between (0, 1). mean is the average position of the bald eagle population, and S is the adaptive nonlinear step size;

[0101]

[0102] In the formula, X i Let X be the position of the i-th individual bald eagle in the previous iteration, which is the path of the i-th drone in the previous iteration of the algorithm. i+1 Let x(i) be the position of the (i+1)th vulture in the last iteration, and y(i) be the vulture position in polar coordinates, and S be the adaptive nonlinear step size.

[0103] S6.2 When the fitness value is greater than the upper bound U, it indicates that a relatively good solution has been found, and the exploration stage begins to try to find a better solution. The exploration location update formula is Equation (5):

[0104]

[0105] In the formula, rand is a uniformly random number in [0,1], and c1 and c2 are constants with values ​​of 1.2;

[0106] S7. Calculate the fitness value of the bald eagle population again, and introduce a dynamic reverse refraction strategy to dynamically adjust the search direction and search range based on the current optimal solution and the change in fitness value, thereby facilitating the optimization of the UAV inspection path.

[0107] S8. Repeat the iteration process from S4 to S7 until the maximum number of iterations is met. Then, output the optimal solution and decode to obtain the best path point for the UAV inspection.

[0108] Furthermore, in S6.1, the adaptive nonlinear step size factor operation mechanism is that the step size gradually increases from α_min to α_max, and then gradually decreases. This ensures that the algorithm has sufficient exploration ability in the early stage of the search and avoids getting trapped in local optima. In the later stage of the search, the algorithm gradually decreases the step size to increase the search accuracy.

[0109] This invention improves the efficiency of drone inspections by employing a sequential inspection method, saving time and energy. It utilizes multiple communication methods, including wired networks, 4G / 5G, and mesh device networking. Through an automatic return-to-home function, this invention ensures the safety of the drone in unexpected situations, reducing risks during the inspection process.

Claims

1. A method for continuous unmanned aerial vehicle (UAV) airport inspection and scheduling, characterized in that, The inspection platform creates a flight route task, uses a scheduling algorithm to retrieve the most suitable airport A for flight, and sets 80% of the flight points of the route within the airport's operational range. The airport's operational power must meet flight requirements. It checks the airport status and retrieves the most suitable airport B for landing, which is the airport from which the last point of the route is located. After confirming that the airport status is normal, it issues a leapfrog mission, taking off from airport A and executing to airport B, starting from airport A. It checks if there are drones at airport A; if so, it checks if airport B meets the landing conditions. If airport B meets the landing conditions, it checks if the airspace is safe. If the airspace is safe, the aircraft takes off from airport A, executes the mission, and lands at airport B. If the airspace is unsafe, the journey ends. If airport B does not meet the landing requirements, the trip ends; The continuous drone airport is connected to the airport installed at the jurisdiction location via a wired connection, and uses a convergence switch and firewall to connect the communication signal to the drone command center; The airport serves as a front-end communication node, communicating with the drone via a wireless encryption protocol and connecting to the drone command center via a wired network. In airport locations where it is inconvenient to deploy wired networks, 4G / 5G and Mesh self-organizing networks are used for communication. The airport solution uses the addition of mesh devices, 5G, and 4G / converged terminals for flexible networking. The scheduling algorithm employs an improved Bald Eagle optimization algorithm to optimize the UAV inspection path planning method. Its goal is to enable the UAV to select the inspection path with the shortest distance from the starting point to the starting point and the fewest obstacles when performing a task. The algorithm includes the following steps: Step 1: Construct the drone flight environment and set the starting and target points for the drone's mission; Step 2: Based on the set starting and target points of the UAV's mission flight, the UAV path planning problem is transformed into an optimization mathematical model. The UAV path planning problem is a shortest distance problem and a minimum obstacle problem. The optimization mathematical model is used to find the optimal path point for the UAV using the improved Bald Eagle optimization algorithm. Step 3: Improve the Bald Eagle optimization algorithm, including the following two steps: D1. Introduce an adaptive nonlinear step size factor to optimize the population update strategy of the Bald Eagle optimization algorithm; D2. Introduce a probability factor mechanism. When the probability factor is greater than the set threshold, execute the population position update strategy of the vulture algorithm in the exploration phase; otherwise, execute the population position update strategy in the search phase. Step 4: Optimize the UAV path planning mathematical model using the improved Bald Eagle optimization algorithm to obtain the shortest path with the fewest obstacles for the UAV when performing the mission, i.e., the optimal path.

2. The continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 1, characterized in that... When a drone takes off from airport A, other drones are allowed to land at airport A, while airport B enters mission mode, meaning that other drones are not allowed to land.

3. The continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 1, characterized in that, If an unexpected situation occurs during the drone's flight, Airport A reports its current location to the inspection platform and requests the most suitable return airport B. The inspection platform searches for the most suitable return airport and replies to Airport A. Airport A then refreshes the return point and notifies Airport B to prepare to recover the drone.

4. The continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 1, characterized in that, The improved Bald Eagle optimization algorithm introduces an adaptive nonlinear step size s, the formula of which is: In the formula: s is the adaptive nonlinear step size; s min This is the minimum step size set, typically 0; s max The maximum step size is set, typically 2; t is the current iteration number; t_mid is half the current iteration number. η is the step size change rate, with a typical value of 1.

5. The continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 1, characterized in that, The improved bald eagle optimization algorithm introduces a probability factor mechanism. The specific process is as follows: Step 1: Define the upper bound U and the lower bound L of the fitness value of the algorithm; if the current fitness value is greater than the upper bound U, execute the exploration phase of the vulture optimization algorithm; if the current fitness value is less than the lower bound L, execute the search phase of the vulture optimization algorithm. Step 2: If the current fitness value is between the upper and lower bounds, then according to the probability factor P, if P is greater than 0.5, execute the search phase; otherwise, execute the vulture optimization algorithm exploration phase. Step 3, the mathematical formula for probability factors is: In the formula, F(t) is the fitness value of the current solution, and λ is the threshold factor, typically 2, which controls the rate of change of P.

6. The continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 1, characterized in that, The improved Bald Eagle optimization algorithm is used to optimize the mathematical model of UAV path planning. The specific process is as follows: S1. Based on the optimization problem of UAV path planning, design an improved fitness function for the Bald Eagle optimization algorithm. The algorithm determines the optimal path point for the UAV through the fitness function; the fitness function formula is: In the formula, t1 is the total time of the entire drone path, length is the length of the drone path, w2 is the weight coefficient, with a typical value of 1.5, and count is the number of obstacles identified in the drone path. S2. Initialize the improved vulture optimization algorithm, including initializing the initial position of the vultures, the population size N, the search space range of the vultures, the upper bound U and the lower bound L, the dimension of the optimization problem dim, and the maximum number of iterations T. max ; The initial position of the bald eagle is the initial path point in the UAV path selection, and the search space range is the range of the UAV path selection. S3. In the improved vulture optimization algorithm, the position code of each vulture is used as a path for UAV inspection. S4. Calculate the fitness value of each bald eagle in the current iteration of the improved bald eagle optimization algorithm, and record the minimum fitness value to determine the position of the bald eagle corresponding to the minimum fitness value, and then determine the current best solution, which is the best path of the current UAV. S5. Compare the minimum fitness value of the current iteration with the minimum fitness value of the previous iteration, determine the minimum fitness value between the two, update it to the best fitness value, and then execute the population position update formula of the improved vulture optimization algorithm. S6. Activate the probability factor mechanism; S7. Calculate the fitness value of the bald eagle population again, and introduce a dynamic reverse refraction strategy to dynamically adjust the search direction and search range based on the current optimal solution and the change in fitness value, thereby facilitating the optimization of the UAV inspection path. S8. Repeat the iteration process from S4 to S7 until the maximum number of iterations is met. Then, output the optimal solution and decode to obtain the best path point for the UAV inspection.

7. A continuous unmanned aerial vehicle (UAV) airport inspection and scheduling method according to claim 6, characterized in that, Step 6 specifically includes: S6.1 When the fitness value is less than the lower bound L, it means that the algorithm search has not yet converged. In this case, the search phase continues. An adaptive nonlinear step size factor s is introduced to optimize the position update strategy. The improved search position update formulas are Equations (3) and (4): In the formula, X best This represents the optimal path for the drone, which is also the optimal position for the current vulture. X represents the latest position of the i-th individual bald eagle. i Let X be the position of the i-th vulture in the previous iteration, a be a linear control factor, r be a random number between (0, 1), and X be a random number between (0, 1). mean is the average position of the bald eagle population, and S is the adaptive nonlinear step size; In the formula, X i Let X be the position of the i-th individual bald eagle in the previous iteration, which is the path of the i-th drone in the previous iteration of the algorithm. i+1 Let x(i) be the position of the (i+1)th vulture in the last iteration, and y(i) be the vulture position in polar coordinates, and S be the adaptive nonlinear step size. S6.2 When the fitness value is greater than the upper bound U, it indicates that a relatively good solution has been found, and the exploration stage begins to try to find a better solution. The exploration location update formula is Equation (5): In the formula, rand is a uniformly random number in [0,1], and c1 and c2 are constants with values ​​of 1.2; S6.3 When the fitness value is between the upper and lower bounds, calculate the probability P according to formula (2). If P is greater than the preset threshold of 0.5, continue to execute the search phase; otherwise, enter the exploration phase.

Images