Unmanned aerial vehicle patrol path optimization method and system based on mobile unmanned aerial vehicle nest

Abstract

The invention relates to the technical field of patrol path optimization, and particularly discloses an unmanned aerial vehicle patrol path optimization method and system based on a mobile unmanned aerial vehicle nest, and the method comprises the steps: carrying out the dynamic environment perception and multi-source data fusion, generating a global environment situation map, carrying out the task intelligent analysis and dynamic priority sorting, and obtaining a task distribution matrix; the method comprises the steps of performing multi-target hierarchical path planning, outputting a dynamic flight path set with a timestamp, constructing a real-time obstacle avoidance and emergency self-healing system, obtaining a safety correction path set, performing mobile nest collaborative scheduling and energy optimization, outputting a nest mobile instruction set, triggering global environment situation map updating, obtaining an unmanned aerial vehicle execution log, and performing knowledge evolution. And end-to-end technical iteration is completed. According to the method, the problems that the route planning of the traditional patrol path planning unmanned aerial vehicle cannot achieve refinement and is not high in applicability under complex terrain conditions are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of inspection path optimization, and specifically to an unmanned aerial vehicle (UAV) inspection path optimization method and system based on a mobile UAV nest. Background Technique

[0002] A mobile UAV nest can directly deploy UAVs to the operation site, solve the problem of manual commuting with UAVs, enhance the emergency operation ability of UAVs, improve the operation efficiency. At the same time, it can provide accommodation conditions and all-round protection for UAVs, such as charging, protection, etc.; moreover, the mobile nest makes the deployment of UAVs more flexible and can expand the inspection and patrol range. With the continuous development of sensor technology, communication technology, artificial intelligence technology, etc., it provides technical support for the optimization of UAV inspection paths. For example, high-precision lidar sensors can more accurately map the terrain and sense obstacles; advanced communication technology can achieve real-time data transmission between UAVs and ground control terminals; artificial intelligence algorithms can be used for path planning, obstacle avoidance decision-making, etc., to improve the autonomous flight ability and intelligent level of UAVs.

[0003] Existing UAV inspection routes are often planned manually according to the on-site terrain conditions and surrounding buildings, etc. This method fails to achieve fine-grained planning of UAV routes and is not very applicable to complex terrain conditions. During the flight of UAVs, a large number of flight obstacles will be encountered. Some are controlled manually to avoid, which is rather troublesome, and when meeting the collision requirements, UAVs cannot automatically avoid. In addition, in order to maintain a safe distance from obstacles, UAVs will make multiple turns, resulting in inflection points and overlapping lines in the flight distance, too long flight distance, high power consumption, and thus affecting the inspection efficiency and range. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides an unmanned aerial vehicle inspection path optimization method and system based on a mobile unmanned aerial vehicle nest, which solves the problems that the route planning of traditional inspection path planning UAVs fails to achieve refinement and is not very applicable to complex terrain conditions.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: An unmanned aerial vehicle inspection path optimization method based on a mobile unmanned aerial vehicle nest includes the following steps: Step 1: Dynamically sense the environment and fuse multi-source data of the UAV inspection area based on the mobile UAV nest to generate a global environment situation map; Step 2: Obtain inspection work orders and a historical database of equipment defects, and based on the global environment situation map, perform intelligent task analysis and dynamic priority sorting to obtain a task assignment matrix with time windows and priorities; Step 3: Collect the performance parameters of the drones in the mobile drone nest. Based on the global environmental situation map and the task assignment matrix, perform multi-objective hierarchical path planning and output a dynamic flight path set with timestamps; Step 4: Build a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe correction path set, and conduct drone inspections based on the safe correction path set; Step 5: Obtain the remaining power of the nest and highway network data. Combine with the safe correction path set to perform collaborative scheduling and energy optimization of the mobile nest, output a nest movement instruction set, and trigger the update of the global environmental situation map; Step 6: Obtain the drone execution logs, conduct knowledge evolution, extract the characteristics of successful obstacle avoidance cases executed by the drones, and complete end-to-end technology iteration.

[0006] Preferably, in the above Step 1, the process of generating the global environmental situation map is as follows: Perform dynamic environmental perception and multi-source data fusion based on the drone inspection area of the mobile drone nest, including: Satellite remote sensing image analysis: Use multi-spectral analysis technology to identify areas with excessive tree growth through the NDVI vegetation index. If the tree crown height exceeds the wire sag by 5 meters, it will be automatically marked. Synchronously access meteorological satellite data to obtain the cloud distribution and lightning warning information of meteorological satellite data in real time; LiDAR point cloud modeling: Use a ground filtering algorithm to separate the ground points and tower point clouds. Through a wire fitting algorithm, i.e., the least squares method, reconstruct the spatial trajectory of the wire and calculate the sag error of each span. If it exceeds the sag error threshold range of each span stored in the database, it will be marked as an abnormal section; Dynamic obstacle annotation: After accessing the civil aviation ADS-B signal, analyze the flight altitude layer information of flights, generate a three-dimensional cylindrical no-fly zone with a radius of 1 km and a height extending to the defined height of the flight route. The vehicle-mounted camera uses the YOLOv7 target detection model to identify construction vehicles and dynamically expand the avoidance range; Obtain the initial monitoring level of the drone inspection area based on the mobile drone nest; Wildfire risk prediction: Integrate historical wildfire data and real-time drought indices to construct a risk heat map. When the humidity in the area is less than 30% for 3 consecutive days and the wind speed is greater than level 4, automatically upgrade the area monitoring level by one level; The risk heat map is realized through Geographic Information System (GIS) technology. Superimpose historical wildfire data including the occurrence location, time, and area, and real-time drought indices including vegetation water content and soil humidity on the map, and use the Kriging interpolation algorithm to calculate the risk level of each area. Red represents high-risk areas, and green represents low-risk areas; Nest status synchronization: The blockchain technology is adopted to ensure the data consistency of multiple nests. After any node updates the location information, other nodes verify the validity of the updated data and add it to the local ledger to achieve status synchronization; Integrate satellite remote sensing image parsing, LiDAR point cloud modeling, dynamic obstacle annotation, wildfire risk prediction, and nest status synchronization to generate a global environmental situation map.

[0007] Preferably, in the second step, the process of obtaining the task assignment matrix with time windows and priorities is as follows: Obtain the inspection work order, and based on the global environmental situation map, perform task semantic parsing: Based on the natural language processing engine, disassemble the inspection work order text into: inspection object, target section, inspection requirements; Obtain the priorities corresponding to each inspection requirement stored in the database; Obtain the equipment defect history library, perform dynamic priority adjustment based on the global environmental situation map, and perform priority sorting based on the adjusted priorities; Resource matching degree calculation: Based on the resolution, battery life, and wind resistance level of the drone camera, construct a drone capability matrix, and use the Euclidean distance algorithm to calculate the matching degree with the task requirements. Combinations with a matching degree less than 60% are automatically eliminated; Obtain the task assignment matrix with time windows and priorities.

[0008] Preferably, the method for performing dynamic priority adjustment includes: Temperature trigger: When the monitored value of the conductor temperature in a certain section exceeds the ambient temperature by 30°C and does not drop for 5 consecutive minutes, trigger task escalation; Composite event processing: If a certain task involves both lightning strike warning and equipment defects at the same time, generate a double priority label and mark it as urgent + important.

[0009] Preferably, in the third step, the process of outputting the dynamic flight path set with timestamps is as follows: Collect the performance parameters of the drones in the mobile drone nest, including drone battery life and drone speed; Based on the global environmental situation map and the task assignment matrix, perform multi-objective hierarchical path planning, including: Global rough planning: Genetic algorithm improvement: Population initialization: The initial population includes the random inspection paths of drones, the historical optimal inspection paths of drones, and the manually preset inspection paths of drones. Except for the random paths, inject the mutated copies of the historical optimal paths, retain 80% of the core waypoints, and perform local perturbation on the historical paths; Crossover strategy: Use the order crossover method to ensure that the tower access order of the offspring paths conforms to the task priorities; Dynamic mutation rate: When the optimal solution has not been improved for 5 consecutive generations, the mutation rate is increased to the improved mutation rate stored in the database to jump out of the local optimum; Safety avoidance modeling: A three-dimensional avoidance sphere with a radius of 300 meters is set in the strong electromagnetic area, and the time cost increases when the path passes through; Local rolling optimization: Wind field compensation model: Establish an aerodynamic model for the UAV. When the crosswind speed is greater than the crosswind speed threshold stored in the database, the waypoint spacing is automatically increased by 10% to offset the drift error; Battery power adaptability: Calculate the remaining battery power coverage distance in real time. If it is unable to reach the next drone nest, insert a temporary charging task nearby; Airspace conflict resolution: Altitude layer allocation: The nearest ground obstacle altitude layer is used for emergency tasks, the default altitude layer is used for regular tasks, and the far ground altitude layer is used for meteorological monitoring tasks; Intersection negotiation: When paths intersect, the system compares the urgency of both tasks, and the low-priority UAV waits; Output a dynamic flight path set with timestamps, including UAV waypoint coordinates, UAV flight altitude, and UAV speed commands, to ensure that the obstacle avoidance system knows the planned arrival time of each UAV.

[0010] Preferably, in the fourth step, the process of UAV inspection based on the safe correction path set includes: Perception fusion: Millimeter-wave radar: Adopt the frequency-modulated continuous-wave mode to analyze the radial velocity of obstacles and predict the position in the next 3 seconds; Binocular vision: Calculate the motion vector of obstacles through the dense optical flow method to distinguish between static and dynamic obstacles; Obstacle avoidance decision: Establish a multi-level early warning mechanism: First-level early warning: When the distance to the obstacle > the farthest early warning distance stored in the database, adjust the path deflection angle to maintain the original task progress; Second-level early warning: When the nearest early warning distance stored in the database < the distance to the obstacle ≤ the farthest early warning distance stored in the database, start the Dubins path for rapid detour, with increased time consumption but ensuring a safe distance; Third-level emergency: When the distance to the obstacle ≤ the nearest early warning distance stored in the database, force the UAV to hover and start laser scanning to generate a local high-precision map for re-planning; Progressive offset strategy: In the first-level early warning, slowly deviate from the original path at the defined acceleration stored in the database to ensure the stability of the gimbal shooting; Dubins path generation: In the second-level early warning, calculate the curvature continuous path with the minimum turning radius stored in the database, and control the time consumption increment within the time consumption increment threshold stored in the database; Establish a group collaboration mechanism; Adopt the resume interrupted transfer technology: when the task is interrupted, record the completed detection data and the breakpoint coordinates, and automatically calibrate the shooting angle after restart; Obtain a safe correction path set, and acquire the energy consumption increment data of the detour section, and conduct UAV inspections based on the safe correction path set.

[0011] Preferably, the establishment of the group collaboration mechanism includes: Auction protocol: when there are multi-aircraft conflicts, the system opens a detour priority auction. For each level of increase in task urgency, the auction weight increases by 50%; Trajectory negotiation: exchange path modification plans through V2V communication, and select the group avoidance strategy with the shortest overall time consumption.

[0012] Preferably, in step five, the process of triggering the update of the global environmental situation map is as follows: Obtain the remaining power of the nest and highway network data; Charging demand prediction: Battery health model: track the capacity attenuation of each charge-discharge cycle of the nest battery, obtain the capacity attenuation rate of the nest battery for each charge-discharge cycle, obtain the remaining capacity of the nest battery, and based on the remaining capacity of the nest battery - nest charging threshold mapping set in the database, dynamically adjust the nest charging threshold according to the current remaining capacity of the nest battery; Dynamic path planning: Junction point calculation: according to the safe correction path set, obtain the estimated arrival time of the UAV, and reverse calculate the time for the nest to reach the target position; The nest movement route preferably selects paved roads and avoids height and width restricted sections through OpenStreetMap data; Multi-nest collaboration: Load balancing: when the task volume of a certain nest exceeds its carrying capacity, send a task subcontracting request to the 3 neighboring nests. Each neighboring nest calculates the feasibility and cost of the subcontracting task according to its current task volume and location, and selects the optimal subcontracting plan through the auction negotiation algorithm; Output the nest movement instruction set, including the nest target coordinates, the nest arrival time, and the nest path, triggering the update of the global environmental situation map.

[0013] Preferably, in step six, the process of completing the end-to-end technology iteration is as follows: Obtain the UAV execution log; Feature extraction of successful obstacle avoidance cases for UAV execution: abstract the successful obstacle avoidance trajectory into a three-dimensional feature vector of the average speed of the UAV - the maximum curvature of the obstacle avoidance path - the total length of the obstacle avoidance path, and store it in the graph database Neo4j; Generate a case feature set from the characteristics of successful obstacle avoidance cases executed by the drone; Encrypt the generated case feature set; Send the encrypted case feature set to all drone nests and drones via a wireless communication network; After receiving the case feature set, the drone nests and drones perform decryption processing; The drone nests and drones perform remote updates based on the decrypted case feature set and feedback the update results to the system to complete the end-to-end technology iteration.

[0014] A drone inspection path optimization system based on a mobile drone nest for implementing the above method, including: A global environment situation map generation module for dynamically perceiving the environment and fusing multi-source data of the drone inspection area based on the mobile drone nest to generate a global environment situation map; A task allocation matrix acquisition module for obtaining inspection work orders and a historical database of equipment defects, and performing intelligent task analysis and dynamic priority sorting based on the global environment situation map to obtain a task allocation matrix with time windows and priorities; A dynamic flight path set output module for collecting the performance parameters of the drones in the mobile drone nest, and performing multi-objective hierarchical path planning based on the global environment situation map and the task allocation matrix to output a dynamic flight path set with timestamps; An emergency self-healing system construction module for constructing a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe correction path set, and performing drone inspections based on the safe correction path set; A global environment situation map update trigger module for obtaining the remaining battery power of the drone nest and highway network data, and performing coordinated scheduling and energy optimization of the mobile drone nest in combination with the safe correction path set to output a drone nest movement instruction set and trigger the update of the global environment situation map; A knowledge evolution module for obtaining the drone execution logs, performing knowledge evolution, and extracting the characteristics of successful obstacle avoidance cases executed by the drone to complete the end-to-end technology iteration.

[0015] The present invention has the following beneficial effects: The present invention generates a global environment situation map through dynamic environment perception and multi-source data fusion of the inspection area, enabling the system to comprehensively understand information such as the terrain, obstacle distribution, and meteorological conditions of the inspection area, providing an accurate basis for subsequent task planning and path decision-making, and avoiding flight accidents or task failures caused by unclear environmental information. Combining inspection work orders and the historical database of equipment defects, performing intelligent task analysis and dynamic priority sorting helps to reasonably arrange the drone inspection tasks according to the actual situation, prioritize the handling of important and urgent tasks, improve resource utilization efficiency, and ensure that key areas and equipment can be inspected and maintained in a timely manner.

[0016] The present invention conducts multi-objective hierarchical path planning based on the global environmental situation map and the task assignment matrix, comprehensively considering task objectives, environmental factors, and UAV performance, to plan a more efficient and reasonable flight path for the UAV, reduce flight time and energy consumption, improve inspection efficiency and coverage. A real-time obstacle avoidance and emergency self-healing system is constructed based on the dynamic flight path set, enabling the UAV to timely sense and avoid obstacles during flight, automatically adjust the path in case of emergencies, ensure flight safety, and reduce the risk of UAV damage and mission interruption caused by collisions or other accidents. The reasonable movement and energy distribution of the mobile UAV nest are realized to ensure that the nest can provide support for the UAV at a suitable position, avoid energy waste, extend the overall operation time of the UAV, and conduct knowledge evolution to continuously improve the path optimization method and the overall system performance, better adapting to different inspection scenarios and task requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of the overall step flow of the method of the present invention; Figure 2 is a schematic diagram of the detailed step flow of the method of the present invention; Figure 3 is a schematic diagram of the module connection of the system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention.

[0019] Embodiment 1: As Figure 1 , Figure 2 shown, a UAV inspection path optimization method based on a mobile UAV nest includes the following steps: Step 1: Conduct dynamic environmental perception and multi-source data fusion on the UAV inspection area based on the mobile UAV nest to generate a global environmental situation map; Step 2: Obtain the inspection work order and the equipment defect history database, and conduct task intelligent analysis and dynamic priority sorting based on the global environmental situation map to obtain a task assignment matrix with time windows and priorities; Step 3: Collect the UAV performance parameters of the mobile UAV nest, and conduct multi-objective hierarchical path planning based on the global environmental situation map and the task assignment matrix to output a dynamic flight path set with time stamps; Step 4: Construct a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe corrected path set, and conduct UAV inspections based on the safe corrected path set; Step 5: Obtain the remaining power of the drone nest and road network data, combine with the safety-corrected path set, perform collaborative scheduling and energy optimization of the mobile drone nest, output the drone nest movement instruction set, and trigger the update of the global environmental situation map; Step 6: Obtain the UAV execution logs, perform knowledge evolution, extract the characteristics of successful obstacle avoidance cases executed by the UAV, and complete the end-to-end technology iteration.

[0020] In Step 1, the process of generating the global environmental situation map is as follows: Perform dynamic environmental perception and multi-source data fusion for the UAV inspection area based on the mobile UAV nest, including: Satellite remote sensing image analysis: Use multi-spectral analysis technology to identify areas with overgrown tree growth through the NDVI vegetation index. If the crown height exceeds the wire sag by 5 meters, it will be automatically marked. Synchronously access meteorological satellite data to obtain the cloud distribution and lightning warning information of meteorological satellite data in real time; LiDAR point cloud modeling: Use a ground filtering algorithm to separate the ground points and tower pole point clouds. Through a wire fitting algorithm, namely the least squares method, reconstruct the spatial trajectory of the wire, calculate the sag error of each span. If it exceeds the sag error threshold range of each span stored in the database, it will be marked as an abnormal section; Dynamic obstacle marking: After accessing the civil aviation ADS-B signal, analyze the flight altitude layer information, generate a three-dimensional cylindrical no-fly zone with a radius of 1 km and a height extending to the route-defined height. The vehicle-mounted camera uses the YOLOv7 target detection model to identify construction vehicles and dynamically expand the avoidance range; Obtain the initial monitoring level of the UAV inspection area based on the mobile UAV nest; Wildfire risk prediction: Integrate historical wildfire data and real-time drought indices to construct a risk heat map. When the humidity in the area is less than 30% and the wind speed is greater than level 4 for 3 consecutive days, automatically upgrade the area monitoring level by one level; The risk heat map is realized through geographic information system (GIS) technology. The historical wildfire data including the occurrence location, time, and area, as well as the real-time drought indices including vegetation water content and soil humidity, are overlaid on the map. The Kriging interpolation algorithm is used to calculate the risk level of each area. Red indicates high-risk areas and green indicates low-risk areas; Drone nest status synchronization: Use blockchain technology to ensure the consistency of multi-drone nest data. After any node updates the location information, other nodes verify the validity of the updated data and add it to the local ledger to achieve status synchronization; Integrate satellite remote sensing image analysis, LiDAR point cloud modeling, dynamic obstacle marking, wildfire risk prediction, and drone nest status synchronization to generate the global environmental situation map.

[0021] Using multispectral analysis technology and NDVI vegetation index to identify areas where tree growth exceeds the limit, automatically marking the tree crowns with a vertical distance of more than 5 meters from the overhead wire, can detect potential tree hazards that may affect line safety in advance, provide accurate information for line maintenance and UAV inspection planning, and avoid threats caused by trees to transmission lines and the like.

[0022] By accessing meteorological satellite data to obtain cloud distribution and lightning warning information, UAVs can avoid bad weather conditions in advance, ensure flight safety, and at the same time help to reasonably arrange the inspection task time and route according to the meteorological conditions. Through LiDAR point cloud modeling, using ground filtering algorithm and wire fitting algorithm (least squares method), calculate the sag error and mark the abnormal section, can accurately master the status of infrastructure such as transmission lines, detect line anomalies in time, and improve the accuracy and pertinence of inspection.

[0023] By accessing civil aviation ADS-B signals to generate no-fly zones, and combining vehicle-mounted cameras with YOLOv7 to identify construction vehicles and expand the avoidance range, it can effectively avoid conflicts between UAVs and flights and ground moving obstacles, and ensure the flight safety of UAVs and airspace order.

[0024] Fusing historical wildfire data and real-time drought index to construct a risk heat map, automatically raising the monitoring level according to conditions such as humidity and wind speed, can early warn of wildfire risks, enable UAVs to more specifically strengthen inspections in high-risk areas, provide strong support for preventing and responding to wildfires, and reduce losses caused by wildfires. Using blockchain technology to ensure the consistency of multi-nest data, enabling information such as the positions of each nest node to be synchronized in a timely and accurate manner, is conducive to multi-nest collaborative operations, improving the overall operation efficiency and reliability of the mobile UAV nest system, and ensuring the orderly development of UAV inspection tasks.

[0025] In step two, the process of obtaining the task assignment matrix with time window and priority is as follows: Obtain the inspection work order, and based on the global environment situation map, conduct task semantic parsing: Based on the natural language processing engine, disassemble the inspection work order text into: inspection object, target section, inspection requirements; Obtain the priorities corresponding to each inspection requirement stored in the database; Obtain the equipment defect history database, based on the global environment situation map, conduct dynamic priority adjustment, and based on the adjusted priorities, conduct priority sorting; Resource matching degree calculation: Based on the UAV camera resolution, battery life, and wind resistance level, construct a UAV capability matrix, and use the Euclidean distance algorithm to calculate the matching degree with the task requirements. Combinations with a matching degree less than 60% are automatically eliminated; Obtain the task assignment matrix with time window and priority.

[0026] The ways to conduct dynamic priority adjustment include: Temperature Trigger: When the monitored temperature of the wire in a certain section exceeds the ambient temperature by 30°C and remains unchanged for 5 minutes, the task is upgraded. Compound Event Handling: If a task involves both lightning warning and equipment defects, a dual-priority label is generated and marked as Urgent + Important.

[0027] Using a natural language processing engine to disassemble the inspection work order text into elements such as inspection objects, target sections, and inspection requirements can accurately understand the task requirements, avoid ambiguities that may arise from manual interpretation of work orders, and provide an accurate basis for subsequent task planning and resource allocation.

[0028] Combining the priorities corresponding to each inspection requirement in the database and dynamically adjusting the priorities based on the equipment defect history database and the global environmental situation map. For example, temperature trigger and compound event handling mechanisms can flexibly increase the task priority according to the actual situation, ensure that high-risk and urgent tasks are processed first, allocate resources reasonably, and improve the response speed of inspections to key issues.

[0029] Based on the resolution, battery life, wind resistance level, etc. of the drone camera, a drone capability matrix is constructed, and the matching degree with the task requirements is calculated through the Euclidean distance algorithm, and combinations with low matching degrees are eliminated. It can ensure that the drones involved in the task are adapted to the task requirements, avoid resource waste, and improve the execution efficiency and quality of inspection tasks.

[0030] In step three, the process of outputting a dynamic flight path set with timestamps is as follows: Collect the performance parameters of the drones in the mobile drone nest, including drone battery life and drone speed; Based on the global environmental situation map and the task assignment matrix, multi-objective hierarchical path planning is carried out, including: Global Rough Planning: Genetic Algorithm Improvement: Population Initialization: The initial population includes random inspection paths of drones, the historical optimal inspection paths of drones, and manually preset inspection paths of drones. Except for the random paths, mutated copies of the historical optimal paths are injected, 80% of the core waypoints are retained, and local perturbations are made to the historical paths; Crossover Strategy: The order crossover method is adopted to ensure that the tower access order of the offspring paths conforms to the task priority; Dynamic Mutation Rate: When the optimal solution has not been improved for 5 consecutive generations, the mutation rate is increased to the improved mutation rate stored in the database to jump out of the local optimum; Safety Avoidance Modeling: A three-dimensional avoidance sphere with a radius of 300 meters is set in the strong electromagnetic area, and the time cost increases when the path passes through; Local Rolling Optimization: Wind field compensation model: Establish an aerodynamic model of the UAV. When the crosswind speed is greater than the crosswind speed threshold stored in the database, automatically increase the waypoint spacing by 10% to offset the drift error; Battery power self - adaptation: Calculate the distance that the remaining battery power can cover in real time. If it cannot reach the next nest, insert a temporary charging task nearby; Airspace conflict resolution: Altitude layer allocation: Use the nearest ground obstacle altitude layer for emergency tasks, the default altitude layer for regular tasks, and the far - ground altitude layer for meteorological monitoring tasks; Intersection negotiation: When paths intersect, the system compares the urgency of both tasks, and the UAV with lower priority waits; Output a dynamic flight path set with timestamps, including UAV waypoint coordinates, UAV flight altitude, and UAV speed commands, ensuring that the obstacle avoidance system knows the planned arrival time of each UAV.

[0031] The population initialization integrates random, historical optimal, and manually preset paths, injects mutated copies of the historical optimal path and perturbs locally, enriches the search space, and increases the probability of finding the global optimal solution. The order crossover method ensures that the tower access order of the offspring path conforms to the task priority, and the dynamic mutation rate avoids the algorithm falling into local optimality, and can plan a more globally optimal inspection path for the UAV that better meets the task requirements.

[0032] The wind field compensation model considers the influence of crosswinds and automatically adjusts the waypoint spacing to offset the drift error, improving the accuracy of the UAV flight path under complex meteorological conditions. The battery power self - adaptation function plans the battery power in real time, ensuring that the UAV can complete the task smoothly and replenish energy, and improving the reliability of task execution.

[0033] Allocate altitude layers according to task types. For example, emergency tasks, regular tasks, and meteorological monitoring tasks belong to different altitude layers, reducing the risk of airspace conflicts. When paths intersect, the UAV with lower priority waits, ensuring flight order and guaranteeing the safety of multi - UAV simultaneous operations.

[0034] The output dynamic flight path set with timestamps contains information such as waypoint coordinates, flight altitude, and speed commands, providing precise guidance for UAV flight. At the same time, it enables the obstacle avoidance system to know the planned arrival time, facilitating real - time obstacle avoidance and collaborative operations, and improving flight safety and task execution efficiency.

[0035] In step four, the process of UAV inspection based on the safety - corrected path set includes: Perception fusion: Millimeter - wave radar: Adopt the frequency - modulated continuous - wave mode, analyze the radial velocity of obstacles, and predict the position in the next 3 seconds; Binocular vision: Calculate the motion vector of obstacles through the dense optical flow method to distinguish between stationary and dynamic obstacles; Obstacle avoidance decision: Establish a multi - level early warning mechanism: Level 1 early warning: When the distance to the obstacle > the farthest early warning distance stored in the database, adjust the path deflection angle and maintain the original task progress; Level 2 early warning: When the nearest early warning distance stored in the database < the distance to the obstacle ≤ the farthest early warning distance stored in the database, start the Dubins path for quick detour, with increased time consumption but ensuring a safe distance; Level 3 emergency: When the distance to the obstacle ≤ the nearest early warning distance stored in the database, force the drone to hover and start laser scanning to generate a local high - precision map for re - planning; Progressive offset strategy: In level 1 early warning, slowly deviate from the original path with the defined acceleration stored in the database to ensure the stability of the gimbal shooting; Dubins path generation: In level 2 early warning, calculate the curvature - continuous path with the minimum turning radius stored in the database, and control the time - consumption increment within the time - consumption increment threshold stored in the database; Establish a group cooperation mechanism; Adopt the breakpoint - resuming technology: When the task is interrupted, record the completed detection data and the breakpoint coordinates, and automatically calibrate the shooting angle after restart; Obtain a set of safe correction paths, and acquire the energy - consumption increment data for the detour section. Conduct drone inspections based on the set of safe correction paths.

[0036] Establish a group cooperation mechanism, including: Auction protocol: When there are multi - drone conflicts, the system opens a bid for the detour priority. For each level increase in the task urgency, the bid weight increases by 50%; Trajectory negotiation: Exchange path modification plans through V2V communication and select the group avoidance strategy with the shortest overall time consumption.

[0037] The millimeter - wave radar uses the frequency - modulated continuous - wave mode to analyze the radial velocity of the obstacle and predict its position in the next 3 seconds. The binocular vision calculates the motion vector of the obstacle through the dense optical flow method to distinguish between static and moving objects, can accurately obtain obstacle information, give early warnings, gain time for obstacle - avoidance decisions, and improve the accuracy and timeliness of obstacle avoidance.

[0038] Under the multi - level early warning mechanism, different distances trigger different levels of early warnings and corresponding countermeasures. Level 1 early warning fine - tunes the path to maintain the task progress, level 2 early warning uses the Dubins path for quick detour to ensure a safe distance, and level 3 emergency forces the drone to hover and re - plan, which can ensure the safety of the drone under different levels of danger, while taking into account task execution, and control the risk and time consumption within a reasonable range.

[0039] The progressive offset strategy ensures the stability of the gimbal shooting, and the Dubins path generation controls the time - consumption increment, balancing the requirements of obstacle avoidance and task execution.

[0040] In case of multi - aircraft conflict, the bidding protocol assigns detour priorities according to the task urgency. The trajectory negotiation selects the shortest - time - consuming avoidance strategy through V2V communication, avoiding the chaos of multiple aircraft avoiding obstacles simultaneously, improving the collaborative operation efficiency of multiple UAVs, and ensuring the smooth progress of group operations.

[0041] The breakpoint - resume technology records data and coordinates when the task is interrupted, and automatically calibrates the shooting angle after restart, ensuring that the task can continue to execute, reducing data loss, and enhancing the integrity and reliability of task execution.

[0042] In step five, the process of triggering the update of the global environmental situation map is as follows: Obtain the remaining power of the hangar and road network data; Charging demand prediction: Battery health model: Track the capacity attenuation of each charge - discharge cycle of the hangar battery, obtain the capacity attenuation rate of each charge - discharge cycle of the hangar battery, obtain the remaining capacity of the hangar battery, and based on the mapping set of the remaining capacity of the hangar battery - hangar charging threshold in the database, dynamically adjust the hangar charging threshold according to the current remaining capacity of the hangar battery; Dynamic path planning: Relay point calculation: According to the safe corrected path set, obtain the estimated arrival time of the UAV, and inversely calculate the time for the hangar to reach the target position; The hangar's movement route preferably selects paved roads and avoids height - limited and width - limited sections through OpenStreetMap data; Multi - hangar collaboration: Load balancing: When the task volume of a certain hangar exceeds its carrying capacity, it sends a task subcontracting request to the adjacent 3 hangars. Each adjacent hangar calculates the feasibility and cost of the subcontracting task according to its current task volume and location, and selects the optimal subcontracting plan through the auction negotiation algorithm; Output the hangar movement instruction set, including the hangar target coordinates, the hangar arrival time, and the hangar path, triggering the update of the global environmental situation map.

[0043] According to the result of load balancing, update the task allocation of each hangar. In the global environmental situation map, the task volume, task priority, and task coverage area of each hangar are displayed in real - time to ensure the transparency and real - time nature of task allocation. Combine the latest meteorological data, obstacle information (such as no - fly zones and construction vehicle information in dynamic obstacle markings), and wildfire risk prediction results to update the environmental information in the global environmental situation map. For example, if the wildfire risk level in a certain area changes, the situation map will timely adjust the color marking of that area (such as changing from green for low risk to red for high risk). After completing the above data synchronization and update, the global environmental situation map generation module triggers the global refresh mechanism and pushes the updated situation map information to each module of the system.

[0044] Tracking the capacity attenuation of the drone nest's charge and discharge cycles through a battery health model, dynamically adjusting the charging threshold, can more accurately predict the charging demand of the drone nest, avoid overcharging or insufficient power situations, extend the service life of the drone nest battery, ensure the stable energy supply of the drone nest, and provide reliable support for the continuous operation of the drone.

[0045] Based on the estimated arrival time of the drone, calculate the arrival time of the drone nest at the target location in reverse, perform dynamic path planning, and preferentially select paved roads and avoid sections with height and width restrictions, making the movement path of the drone nest more reasonable, reducing movement time and risk, ensuring that the drone nest can reach the designated location in time, and improving the collaborative operation efficiency of the drone nest and the drone.

[0046] When the task volume of the drone nest is overloaded, send a task subcontracting request to neighboring drone nests through a load balancing mechanism, and use the auction negotiation algorithm to select the optimal subcontracting plan to achieve a reasonable distribution of tasks among multiple drone nests, avoid overloading a single drone nest, and improve the task processing ability and resource utilization efficiency of the entire mobile drone nest system.

[0047] Output a drone nest movement instruction set containing information such as the target coordinates, arrival time, and path of the drone nest and trigger an update of the global environmental situation map, enabling the system to adjust in real time according to the movement of the drone nest, maintain an accurate perception of the environment and task status, and ensure the continuous, efficient, and stable progress of the drone inspection task.

[0048] In step six, the process of completing the end-to-end technology iteration is as follows: Obtain the execution log of the drone; Feature extraction of successful obstacle avoidance cases of the drone: Abstract the successful obstacle avoidance trajectory into a three-dimensional feature vector of the average speed of the drone - the maximum curvature of the obstacle avoidance path - the total length of the obstacle avoidance path, and store it in the graph database Neo4j; Generate a case feature set from the successful obstacle avoidance case features of the drone; Encrypt the generated case feature set; Send the encrypted case feature set to all drone nests and drones through a wireless communication network; After the drone nests and drones receive the case feature set, perform decryption processing; The drone nests and drones perform remote updates according to the decrypted case feature set and feedback the update results to the system to complete the end-to-end technology iteration.

[0049] Extract and store the successful obstacle avoidance case features, abstract them into a three-dimensional feature vector and store them in the graph database to form a case feature set. This enables the accumulation of the successful experience of the drone in obstacle avoidance, provides knowledge reserves for subsequent flights, and facilitates the inheritance and reuse of effective obstacle avoidance strategies.

[0050] By encrypting and transmitting the case feature set, the nest and the drone receive the decrypted data for remote update and then feedback the results. This enables the entire system to continuously absorb new successful experiences, continuously improve the obstacle avoidance algorithm and strategy based on actual execution cases, achieve end-to-end technology iteration, and enhance the overall obstacle avoidance ability and intelligence level of the system.

[0051] As the environment and tasks change, new successful obstacle avoidance case features are continuously incorporated into the system, enabling the drone and the nest to better adapt to complex and diverse scenarios, improve the ability to handle various potential obstacles and emergencies, and ensure flight safety and the stability of mission execution.

[0052] Example 2: Integrate satellite remote sensing image analysis, LiDAR point cloud modeling, dynamic obstacle annotation, wildfire risk prediction, and nest status synchronization to generate a global environmental situation map. The specific integration method is as follows: Perform spatio-temporal unified processing on the collected multi-source data and convert it to a unified spatial coordinate system and time coordinate system. Convert data in different formats to a unified format. For example, convert satellite remote sensing image data to the GeoTIFF format and convert LiDAR point cloud data to the LAS format.

[0053] For satellite remote sensing images and LiDAR point cloud data, they can be integrated according to their geographical spatial positions. Overlay the two sets of data in the same geographical coordinate system and use the spectral information of the image and the elevation information of the point cloud to generate a more comprehensive terrain and landform feature layer. For dynamic obstacle annotation information, update it to the corresponding spatial position in real time according to its dynamic change characteristics.

[0054] Higher-precision data is given higher weight during integration. For example, when integrating satellite remote sensing images and LiDAR point cloud data, if the LiDAR point cloud data has higher precision, then when generating terrain and landform features, the LiDAR data is mainly used and the satellite remote sensing image data is used as a supplement for integration.

[0055] For data with high real-time requirements, such as dynamic obstacle annotation and wildfire risk prediction results, the latest data is preferentially used for integration and updated to the global environmental situation map in a timely manner. For relatively stable data, such as satellite remote sensing images and LiDAR point cloud data, they can be updated and integrated regularly.

[0056] Use fusion algorithms, such as Kalman filtering, D-S evidence theory, or algorithms based on machine learning, to comprehensively process the information from different data sources and generate a unified environmental situation data set. Perform noise reduction and smoothing processing on the fused data to remove false targets and abnormal information, highlight key information, and form a data set that accurately reflects the environmental situation of the inspection area.

[0057] Prepare the geographical information data of the inspection area, including basic geographical elements such as topography, river systems, road networks, and populated areas, and construct a basic geographical framework. Classify the fused data according to different situation elements, and construct multiple layers, such as vegetation layer, power facility layer, obstacle layer, environmental risk layer, airframe layer, etc. Use GIS software to render and visualize each layer, distinguish them with different colors, symbols, and line types, intuitively display the environmental situation of the inspection area, and generate an overall environmental situation map.

[0058] Embodiment 3: As Figure 3 shown, a UAV inspection path optimization system based on a mobile UAV airframe is used to implement the method in Embodiment 1 or Embodiment 2, and includes: An overall environmental situation map generation module, which is used to perform dynamic environmental perception and multi-source data fusion on the UAV inspection area based on the mobile UAV airframe, and generate an overall environmental situation map; A task assignment matrix acquisition module, which is used to obtain inspection work orders and an equipment defect history database, and based on the overall environmental situation map, perform intelligent task analysis and dynamic priority sorting to obtain a task assignment matrix with time windows and priorities; A dynamic flight path set output module, which is used to collect the UAV performance parameters of the mobile UAV airframe, and based on the overall environmental situation map and the task assignment matrix, perform multi-objective hierarchical path planning and output a dynamic flight path set with timestamps; An emergency self-healing system construction module, which is used to construct a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe correction path set, and perform UAV inspections based on the safe correction path set; An overall environmental situation map update trigger module, which is used to obtain the remaining power of the airframe and highway network data, combine with the safe correction path set, perform coordinated scheduling and energy optimization of the mobile airframe, output an airframe movement instruction set, and trigger the update of the overall environmental situation map; A knowledge evolution module, which is used to obtain the UAV execution log, perform knowledge evolution, extract the characteristics of successful UAV obstacle avoidance cases, and complete end-to-end technology iteration.

Claims

1. An optimization method for the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest, characterized in that, It includes the following steps: Step 1: Perform dynamic environment perception and multi-source data fusion on the drone inspection area based on the mobile drone nest to generate a global environment situation map; Step 2: Obtain inspection work orders and the equipment defect history database, and based on the global environment situation map, perform intelligent task parsing and dynamic priority sorting to obtain a task allocation matrix with time windows and priorities; Step 3: Collect the drone performance parameters of the mobile drone nest, and based on the global environment situation map and the task allocation matrix, perform multi-objective hierarchical path planning to output a dynamic flight path set with timestamps; Step 4: Build a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe corrected path set, and conduct drone inspections based on the safe corrected path set; Step 5: Obtain the remaining battery power of the nest and highway network data, combine with the safe corrected path set, perform collaborative scheduling and energy optimization of the mobile nest, output a nest movement instruction set, and trigger the update of the global environment situation map; Step 6: Obtain the drone execution logs, conduct knowledge evolution, extract the characteristics of successful obstacle avoidance cases executed by the drone, and complete end-to-end technical iteration.

2. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In the said Step 1, the process of generating the global environment situation map is as follows: Performing dynamic environment perception and multi-source data fusion on the drone inspection area based on the mobile drone nest includes: Satellite remote sensing image analysis: Using multi-spectral analysis technology, identify areas with excessive tree growth through the NDVI vegetation index. If the crown height exceeds the conductor sag by 5 meters, it will be automatically marked, and at the same time, access meteorological satellite data to obtain the cloud distribution and lightning warning information of the meteorological satellite data in real time; LiDAR point cloud modeling: Adopt a ground filtering algorithm to separate the ground points and tower point clouds, and through the conductor fitting algorithm, that is, the least squares method, reconstruct the conductor spatial trajectory, calculate the sag error of each span. If it exceeds the sag error threshold range of each span stored in the database, it will be marked as an abnormal section; Dynamic obstacle annotation: After accessing the civil aviation ADS-B signal, parse the flight altitude layer information, generate a three-dimensional cylindrical no-fly zone with a radius of 1 km and a height extending to the route-defined height, and the vehicle-mounted camera identifies construction vehicles through the YOLOv7 target detection model to dynamically expand the avoidance range; Obtain the initial monitoring level of the drone inspection area based on the mobile drone nest; Wildfire risk prediction: Integrate historical wildfire data and real-time drought indices to construct a risk heat map. When the humidity in the area is less than 30% and the wind speed is greater than level 4 for 3 consecutive days, automatically upgrade the area monitoring level by one level; The risk heat map is realized through geographic information system (GIS) technology. The historical wildfire data including the occurrence location, time, and area, as well as the real-time drought indices including vegetation water content and soil humidity, are overlaid on the map, and the Kriging interpolation algorithm is used to calculate the risk level of each area. Red indicates high-risk areas, and green indicates low-risk areas; Nest status synchronization: Use blockchain technology to ensure the consistency of multi-nest data. After any node updates the location information, other nodes verify the validity of the updated data and then add it to the local ledger to achieve status synchronization; Integrate satellite remote sensing image analysis, LiDAR point cloud modeling, dynamic obstacle annotation, wildfire risk prediction, and nest status synchronization to generate a global environmental situation map.

3. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In the second step, the process of obtaining the task assignment matrix with time windows and priorities is as follows: Obtain inspection work orders and perform task semantic parsing based on the global environmental situation map: Based on the natural language processing engine, disassemble the inspection work order text into: inspection object, target section, and inspection requirements; Obtain the priorities corresponding to each inspection requirement stored in the database; Obtain the equipment defect history library, perform dynamic priority adjustment based on the global environmental situation map, and perform priority sorting based on the adjusted priorities; Resource matching degree calculation: Based on the resolution, battery life, and wind resistance level of the drone camera, construct a drone capability matrix, and use the Euclidean distance algorithm to calculate the matching degree with the task requirements. Combinations with a matching degree less than 60% are automatically eliminated; Obtain the task assignment matrix with time windows and priorities.

4. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 3, wherein: The methods for performing dynamic priority adjustment include: Temperature trigger: When the monitored value of the conductor temperature in a certain section exceeds the ambient temperature by 30°C and does not decrease for 5 consecutive minutes, trigger task escalation; Composite event processing: If a certain task involves both lightning warning and equipment defects, generate a dual priority label and mark it as urgent + important.

5. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In the third step, the process of outputting the dynamic flight path set with timestamps is as follows: Collect the drone performance parameters of the mobile drone nest, including drone battery life and drone speed; Based on the global environmental situation map and the task assignment matrix, perform multi-objective hierarchical path planning, including: Global rough planning: Genetic algorithm improvement: Population initialization: The initial population includes the random inspection paths of drones, the historical optimal inspection paths of drones, and the manually preset inspection paths of drones. Except for the random paths, inject the mutated copies of the historical optimal paths, retain 80% of the core waypoints, and perform local perturbations on the historical paths; Crossover strategy: Use the order crossover method to ensure that the tower access order of the offspring paths conforms to the task priorities; Dynamic mutation rate: When the optimal solution has not been improved for 5 consecutive generations, the mutation rate is increased to the improved mutation rate stored in the database to jump out of the local optimum; Safety avoidance modeling: Set a three-dimensional avoidance sphere with a radius of 300 meters in the strong electromagnetic area, and increase the time cost when the path passes through; Local rolling optimization: Wind field compensation model: Establish a drone aerodynamic model. When the crosswind speed is greater than the crosswind speed threshold stored in the database, automatically increase the waypoint spacing by 10% to offset the drift error; Battery power adaptability: Calculate the remaining battery power coverage distance in real time. If it is impossible to reach the next nest, insert a temporary charging task nearby; Airspace conflict resolution: Altitude layer allocation: Use the nearest ground obstacle altitude layer for urgent tasks, the default altitude layer for regular tasks, and the far ground altitude layer for meteorological monitoring tasks; Intersection negotiation: When paths intersect, the system compares the urgency of the two tasks, and the drone with a lower priority waits; Output the dynamic flight path set with timestamps, including drone waypoint coordinates, drone flight altitude, and drone speed commands, to ensure that the obstacle avoidance system knows the planned arrival time of each drone.

6. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In step 4, the process of conducting drone inspections based on the safe correction path set includes: Perception fusion: Millimeter-wave radar: Adopt the frequency-modulated continuous-wave mode to analyze the radial velocity of obstacles and predict their positions in the next 3 seconds; Binocular vision: Calculate the motion vectors of obstacles through the dense optical flow method to distinguish between static and dynamic obstacles; Obstacle avoidance decision-making: Establish a multi-level early warning mechanism: First-level early warning: When the distance to the obstacle > the farthest early warning distance stored in the database, adjust the path deflection angle and maintain the original task progress; Second-level early warning: When the nearest early warning distance stored in the database < the distance to the obstacle ≤ the farthest early warning distance stored in the database, start the Dubins path for rapid detouring, increasing the time consumption but ensuring a safe distance; Third-level emergency: When the distance to the obstacle ≤ the nearest early warning distance stored in the database, force the drone to hover and start laser scanning to generate a local high-precision map for re-planning; Progressive offset strategy: In the first-level early warning, slowly deviate from the original path at the defined acceleration stored in the database to ensure the stability of the gimbal shooting; Dubins path generation: In the second-level early warning, calculate the curvature continuous path with the minimum turning radius stored in the database, and control the time consumption increment within the time consumption increment threshold stored in the database; Establish a group collaboration mechanism; Adopt the breakpoint resumption technology: When the task is interrupted, record the completed detection data and breakpoint coordinates, and automatically calibrate the shooting angle after restart; Obtain the safe correction path set, and acquire the energy consumption increment data of the detour section. Conduct drone inspections based on the safe correction path set.

7. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 6, wherein: The establishment of the group collaboration mechanism includes: Auction protocol: When there are multi-drone conflicts, the system opens the auction for the detour priority. For each level increase in the task urgency, the auction weight increases by 50%; Trajectory negotiation: Exchange path modification plans through V2V communication and select the group avoidance strategy with the shortest overall time consumption.

8. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In step 5, the process of triggering the update of the global environmental situation map is as follows: Obtain the remaining power of the drone nest and highway network data; Charging demand prediction: Battery health model: Track the capacity attenuation of each charge-discharge cycle of the drone nest battery to obtain the capacity attenuation rate of each charge-discharge cycle of the drone nest battery, and obtain the remaining capacity of the drone nest battery. Based on the mapping set of the remaining capacity of the drone nest battery - the charging threshold of the drone nest in the database, dynamically adjust the charging threshold of the drone nest according to the current remaining capacity of the drone nest battery; Dynamic path planning: Connection point calculation: According to the safe correction path set, obtain the estimated arrival time of the drone, and reverse calculate the time for the drone nest to reach the target position; The drone nest's movement route preferably selects paved roads and avoids sections with height and width restrictions through OpenStreetMap data; Multi-drone nest collaboration: Load balancing: When the task volume of a certain drone nest exceeds its carrying capacity, send a task subcontracting request to the 3 adjacent drone nests. Each adjacent drone nest calculates the feasibility and cost of the subcontracting task based on its current task volume and position, and selects the optimal subcontracting plan through the auction negotiation algorithm; Output the drone nest movement instruction set, including the target coordinates of the drone nest, the arrival time of the drone nest, and the path of the drone nest, triggering the update of the global environmental situation map.

9. The method for optimizing the inspection path of an unmanned aerial vehicle based on a mobile unmanned aerial vehicle nest according to claim 1, wherein: In step 6, the process of completing the end-to-end technical iteration is as follows: Obtain the drone execution log; Feature extraction of successful obstacle avoidance cases executed by drones: Abstract the successful obstacle avoidance trajectory into a three-dimensional feature vector of the average speed of the drone - the maximum curvature of the obstacle avoidance path - the total length of the obstacle avoidance path, and store it in the graph database Neo4j; Generate a case feature set from the features of successful obstacle avoidance cases executed by drones; Perform encryption processing on the generated case feature set; Send the encrypted case feature set to all drone nests and drones via a wireless communication network; After receiving the case feature set, the drone nests and drones perform decryption processing; The drone nests and drones perform remote updates based on the decrypted case feature set and feedback the update results to the system to complete the end-to-end technology iteration.

10. An unmanned aerial vehicle inspection path optimization system based on a mobile unmanned aerial vehicle nest, for implementing the method according to any one of claims 1-9, characterized in that, Including: Global environment situation map generation module, used to perform dynamic environment perception and multi-source data fusion on the drone inspection area based on the mobile drone nest, and generate a global environment situation map; Task assignment matrix acquisition module, used to obtain inspection work orders and equipment defect history libraries, and perform intelligent task analysis and dynamic priority sorting based on the global environment situation map to obtain a task assignment matrix with time windows and priorities; Dynamic flight path set output module, used to collect the performance parameters of the drones in the mobile drone nest, and perform multi-objective hierarchical path planning based on the global environment situation map and the task assignment matrix, and output a dynamic flight path set with timestamps; Emergency self-healing system construction module, used to construct a real-time obstacle avoidance and emergency self-healing system based on the dynamic flight path set to obtain a safe correction path set, and perform drone inspections based on the safe correction path set; Global environment situation map update trigger module, used to obtain the remaining battery power of the drone nest and highway network data, combine with the safe correction path set, perform collaborative scheduling and energy optimization of the mobile drone nest, output a drone nest movement instruction set, and trigger the update of the global environment situation map; Knowledge evolution module, used to obtain drone execution logs, perform knowledge evolution, extract the features of successful obstacle avoidance cases executed by drones, and complete the end-to-end technology iteration.

Images