Unmanned aerial vehicle task scheduling method and device for intelligent road network inspection

Abstract

The embodiment of the application provides a kind of unmanned aerial vehicle task scheduling method and device of intelligent road network patrol, the precise planning of unmanned aerial vehicle patrol route is realized by constructing three-dimensional road network topology model and obstacle distribution map.Unmanned aerial vehicle scheduling strategy based on task fitness is designed, and real-time position, endurance mileage and load state are combined to extract patrol rules from task template library for intelligent scheduling.A hierarchical early warning mechanism is introduced, and the influence range of the event is analyzed through target detection and scene recognition, to realize the differentiated early warning of roadside equipment and vehicle terminal.This method effectively solves the deficiencies of traditional technology in task planning, unmanned aerial vehicle scheduling and early warning linkage, and significantly improves the intelligent level and emergency disposal effect of road network patrol.

Description

Technical Field

[0001] This application relates to the field of data processing, specifically to a method and apparatus for scheduling unmanned aerial vehicle (UAV) tasks for intelligent road network inspection. Background Technology

[0002] Existing methods for dispatching drones for road network patrol have significant shortcomings. Traditional systems lack a systematic approach to task planning, making it difficult to effectively integrate 3D road network topology and obstacle distribution information, thus affecting the rationality of patrol routes.

[0003] Furthermore, existing technologies face bottlenecks in drone scheduling. Most systems fail to comprehensively consider drone location, flight range, and payload status, and lack task-adaptive scheduling mechanisms, resulting in suboptimal resource allocation.

[0004] The existing system has technical shortcomings in early warning and linkage. It lacks the ability to dynamically analyze the scope of an event's impact, making it difficult to achieve tiered early warning systems for roadside equipment and vehicle-mounted terminals, thus affecting the effectiveness of emergency response. Solving these problems is crucial for improving the level of road network patrol. Summary of the Invention

[0005] To address the problems in the existing technology, this application provides a method and device for scheduling unmanned aerial vehicle (UAV) missions for intelligent road network patrol, which can effectively solve the shortcomings of traditional technologies in mission planning, UAV scheduling and early warning linkage, and significantly improve the intelligence level and emergency response effect of road network patrol.

[0006] To solve at least one of the above problems, this application provides the following technical solution: Firstly, this application provides a method for scheduling drone missions for intelligent road network inspection, including: The system receives traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system, extracts the event type, location, and level attributes from the event information, constructs a three-dimensional road network topology model based on the road network electronic map, marks the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map, calculates the available emergency return points for UAVs around the event location, and plans UAV patrol routes and obstacle avoidance paths based on the obstacle distribution map. The real-time location information, remaining range, and payload status information of each UAV are obtained from the UAV management platform. The task fitness score of each UAV reaching the event location via the patrol route is calculated. The UAV with the highest task fitness score is selected as the target UAV. Patrol rules are extracted from the task template library based on the event type. The patrol rules and the obstacle avoidance path are sent to the target UAV to execute the patrol task. Image data acquired by the target drone during the patrol is collected, target detection and scene recognition are performed on the image data, information on the scope of event impact is extracted, the distribution area of ​​roadside equipment that needs to be coordinated is determined based on the information on the scope of event impact, graded early warning information is sent to electronic information boards and LED displays in the distribution area, different levels of early warning information are pushed to vehicle terminals based on the distance of the vehicle from the location of the event, and the content of the early warning information is updated in real time according to the progress of event handling.

[0007] Furthermore, it also includes: reading event information data streams from the road network monitoring system, establishing an event information cache queue to store the data streams, parsing the data packet headers of the data streams to extract event codes and timestamps, matching the event codes with a preset event type mapping table to obtain event type identifiers, extracting the latitude and longitude coordinates of the event occurrence location and the event level value from the data stream message body, and writing the event type identifier, the latitude and longitude coordinates, and the event level value into an event attribute table; The system reads vector data from the electronic road network map, constructs a road network skeleton line matrix based on the vector data, calculates the three-dimensional spatial coordinates of road network nodes to generate a road network node set, and constructs a three-dimensional road network topology model using the road network skeleton line matrix and the road network node set. It retrieves building outline data, overpass elevation data, and tunnel boundary data from a geographic information database, converts the outline data, elevation data, and boundary data into a set of obstacle boundary points, and marks the set of obstacle boundary points in the three-dimensional road network topology model to generate an obstacle distribution map.

[0008] Furthermore, it also includes: obtaining the center coordinates of the location where the event occurred, constructing a circular search space region with the center coordinates as the center, dividing the circular region into grid cells, calculating the center point coordinates of each grid cell, determining whether the center point of each grid cell overlaps with the boundary point of the obstacle based on the obstacle distribution map, filtering the center points of grid cells that do not overlap with the obstacle as a set of candidate return points, calculating the distance and height difference from each return point in the set of candidate return points to the location where the event occurred, sorting the set of candidate return points from near to far distance, and selecting the return points ranked higher as emergency return points; A three-dimensional coordinate system is constructed for the road network topology map. The coordinates of the emergency return point and the location of the incident are projected onto the three-dimensional coordinate system. The initial flight path from the emergency return point to the location of the incident is calculated based on the A-star search algorithm. The boundary point data of the obstacle distribution map is read, and the collision detection distance between the initial flight path and the obstacle is calculated. When a collision risk is detected, the flight path nodes are adjusted to generate an obstacle avoidance segment. The obstacle avoidance segment is combined with the flight path segment that has not collided to generate a patrol flight path. The patrol flight path is smoothed according to the turning radius and climb angle constraints of the UAV.

[0009] Furthermore, it also includes: receiving telemetry data packets from each drone from the drone management platform, parsing the telemetry data packets to obtain the drone identification code, reading the corresponding drone's position coordinates, remaining battery power, flight speed, and mission payload type based on the identification code, calculating the remaining range based on the remaining battery power, determining the camera specifications and image transmission equipment performance of the drone based on the mission payload type, and writing the position coordinates, remaining range, flight speed, camera specifications, and image transmission equipment performance into the drone status information table; A task fitness score calculation model is established, using the straight-line distance between the UAV's location coordinates and the location of the event, the ratio of the UAV's remaining flight range to the length of the patrol route, the matching degree between the UAV's flight speed and the expected patrol duration, and the adaptability of the camera specifications to the patrol task requirements as scoring indicators. The weight coefficients of each scoring indicator are determined based on the analytic hierarchy process (AHP). The task fitness score is obtained by weighting the scoring indicators with the weight coefficients, and the UAV with the highest task fitness score is selected as the target UAV.

[0010] Furthermore, it also includes: reading the preset mapping relationship between event types and patrol rules in the task template library, matching the corresponding patrol rule identifier based on the event type, extracting the patrol start point coordinates, patrol end point coordinates, flight path sampling interval, flight altitude parameters, flight speed parameters, camera pitch angle parameters, and image acquisition frequency parameters according to the patrol rule identifier, and combining the patrol rule parameters with the event occurrence location information and the target UAV identification code to generate a task configuration file; A drone mission control instruction set is constructed. The parameters in the mission configuration file are encoded according to the communication protocol specification to generate mission initialization instructions, flight path setting instructions, obstacle avoidance path instructions, and image acquisition instructions. A data communication link with the target drone is established. The control instruction set is sent to the target drone through the data communication link. The instruction verification code returned by the target drone is received, and the verification code is verified to confirm the reliability of the mission instructions.

[0011] Furthermore, it also includes: receiving image data streams transmitted back by the target drone, parsing image frames and shooting timestamps from the image data stream, storing the image frames and shooting timestamps in an image cache queue, reading the target drone position data to record the shooting position coordinates of the image frames, performing geometric correction and image enhancement processing on the image frames, using a target detection model to identify vehicle distribution density, road traffic status, and traffic control facilities in the image, extracting the accident scene range, vehicle backlog length, and road blockage location based on a scene recognition model, and generating a target detection result table and a scene recognition result table; The spatial location information in the target detection result table and the scene recognition result table is fused to construct the boundary contour of the event's influence range. The geometric center and coverage radius of the boundary contour are calculated. A local coordinate system is established with the geometric center as the origin. The locations of information boards, display screens, and broadcasting equipment along the road are marked in the local coordinate system. The distance from each device location to the geometric center is calculated. Devices with a distance smaller than the coverage radius are selected to generate the roadside device distribution area.

[0012] Furthermore, it also includes: constructing a graded template for early warning information, dividing the early warning level threshold based on the straight-line distance between the vehicle and the location of the incident, reading the communication address and device type identifier of each device in the distribution area of ​​roadside devices, matching the corresponding information release format according to the device type identifier, generating device control instructions for the early warning information content according to the information release format, establishing communication connections with each device, sending the device control instructions to the corresponding electronic information board and LED display screen, and receiving instruction response confirmation information returned by the device; The system obtains real-time location data of vehicles from the vehicle-to-everything (V2X) platform, calculates the straight-line distance between each vehicle and the location of the event, compares the straight-line distance with a warning level threshold to determine the warning information level, selects the corresponding information push template based on the warning information level, writes the event handling status information into the information push template to generate a vehicle terminal prompt message, pushes the prompt message to the vehicle terminal of the target vehicle through the V2X communication gateway, monitors the progress of event handling, and regenerates the warning information content when a progress update is received.

[0013] Secondly, this application provides an intelligent road network patrol drone task scheduling device, comprising: The path planning module is used to receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system, extract the event type, location, and event level attributes from the event information, construct a three-dimensional road network topology model based on the road network electronic map, mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map, calculate the available emergency return points for UAVs around the event location, and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. The task inspection module is used to obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location through the inspection route, select the UAV with the highest task fitness score as the target UAV, extract inspection rules from the task template library based on the event type, and send the inspection rules and the obstacle avoidance path to the target UAV to execute the inspection task. The anomaly alarm module is used to collect image data acquired by the target drone during the patrol process, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded warning information to electronic information boards and LED displays in the distribution area, push different levels of warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the warning information in real time according to the progress of event handling.

[0014] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the intelligent road network patrol drone task scheduling method.

[0015] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the aforementioned intelligent road network patrol drone task scheduling method.

[0016] Fifthly, this application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implements the steps of the aforementioned intelligent road network patrol drone task scheduling method.

[0017] As described above, this application provides an intelligent UAV task scheduling method and device for road network patrol. By constructing a three-dimensional road network topology model and obstacle distribution map, it achieves precise planning of UAV patrol routes. A task-adaptability-based UAV scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events, enabling differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively addresses the shortcomings of traditional technologies in task planning, UAV scheduling, and early warning linkage, significantly improving the intelligence level of road network patrol and emergency response effectiveness. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the UAV task scheduling method for intelligent road network patrol in this application embodiment; Figure 2 This is a structural diagram of the UAV task scheduling device for intelligent road network patrol in the embodiments of this application; Figure 3 This is a schematic diagram of the structure of the electronic device in the embodiments of this application.

[0020] Figure label: Electronic device 9600, central processing unit 9100, memory 9140, communication module 9110, input unit 9120, audio processor 9130, display 9160, power supply 9170, buffer memory 9141, application / function storage unit 9142, data storage unit 9143, driver storage unit 9144, antenna 9111, speaker 9131, microphone 9132. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0022] The acquisition, storage, use, and processing of data in this application all comply with the relevant provisions of national laws and regulations.

[0023] To address the shortcomings of existing technologies, this application provides an intelligent UAV mission scheduling method and apparatus for road network patrol. By constructing a 3D road network topology model and obstacle distribution map, it achieves precise planning of UAV patrol routes. A mission-adaptability-based UAV scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a mission template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively solves the deficiencies of traditional technologies in mission planning, UAV scheduling, and early warning linkage, significantly improving the intelligence level of road network patrol and emergency response effectiveness.

[0024] To effectively address the shortcomings of traditional technologies in task planning, drone scheduling, and early warning linkage, and to significantly improve the intelligence level and emergency response effectiveness of road network patrols, this application provides an embodiment of an intelligent road network patrol drone task scheduling method, see [link to embodiment]. Figure 1 The intelligent road network patrol drone task scheduling method specifically includes the following: Step S101: Receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system; extract the event type, location, and event level attributes from the event information; construct a three-dimensional road network topology model based on the road network electronic map; mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map; calculate the available emergency return points for UAVs around the event location; and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. Optionally, this embodiment addresses the problems of untimely event information processing, inaccurate road network modeling, and unreasonable UAV flight path planning in traditional road network inspections by innovatively designing an intelligent inspection scheme based on a 3D road network model. This embodiment first establishes a data interface with the road network monitoring system and uses distributed message queue technology to receive event information in real time. Key attributes are extracted through a message parsing engine: Event = Parse(Type, Location, Level), where Type is the event type code, Location is the latitude and longitude coordinates, and Level is the event level.

[0025] This embodiment deeply optimizes the road network modeling mechanism. A 3D road network topology model is constructed based on high-precision electronic map data, and a multi-level grid partitioning strategy is adopted to improve modeling accuracy. The system maps road network nodes to a 3D coordinate system: Node(x, y, z) = Transform(Lat, Lon, Alt), where Lat and Lon are latitude and longitude coordinates, and Alt is the elevation value. The road network skeleton is constructed through node connection relationships, with particular consideration given to geometric features such as road slope and curvature.

[0026] This embodiment innovatively implements an obstacle modeling strategy. Geometric data of buildings, overpasses, tunnels, and other facilities are retrieved from a geographic information database, and obstacle contour models are constructed using boundary extraction algorithms. The system employs an octree structure for spatial partitioning and management, improving collision detection efficiency. For large buildings, convex hull decomposition simplifies the geometric model, maintaining computational efficiency while ensuring safety margins.

[0027] This embodiment optimizes the emergency return-to-home point calculation mechanism. A search space is constructed centered on the event location, and candidate return-to-home points are determined through grid partitioning and accessibility analysis. The system considers terrain limitations and obstacle avoidance requirements, and evaluates the usability of return-to-home points based on UAV performance parameters. Special attention is paid to elevated road sections and complex interchange areas, and multi-level search is used to ensure the safety of return-to-home points.

[0028] This embodiment innovatively designs a route planning scheme. An improved A* algorithm is used for path search, with obstacle safety distance as a key factor in the cost function. The system optimizes route smoothness using Bézier curves to ensure that the UAV's motion constraints are met. For segments requiring obstacle avoidance, an optimal obstacle avoidance path is generated using a dynamic programming algorithm.

[0029] This embodiment features a deeply optimized route verification mechanism. A virtual flight environment is constructed to simulate and verify the planned route. The system evaluates the safety of the route using the Monte Carlo method, taking into account uncertainties such as wind field effects and positional errors. Particularly for segments traversing densely built-up areas, probabilistic collision detection ensures the reliability of the planned path.

[0030] This embodiment innovatively implements a task scheduling strategy. The optimization objectives of route planning are dynamically adjusted based on event type and urgency. The system balances flight time, energy consumption, and safety through a multi-objective optimization method to generate patrol plans that meet actual needs. For complex road conditions, segmented planning reduces computational complexity.

[0031] This embodiment achieves a precise representation of the road network environment through 3D modeling technology. It demonstrates powerful modeling capabilities and planning efficiency, particularly when handling complex urban scenarios. Through accurate environmental perception and intelligent path planning, the system can quickly respond to various emergencies, providing reliable mission support for drone patrols.

[0032] This embodiment's innovative design not only solves the modeling accuracy and planning efficiency problems of traditional methods but also establishes a continuously optimized task planning framework. Through continuous updates to the environmental model and dynamic adjustments to the planning strategy, the system can continuously improve its adaptability to various scenarios, providing strong support for intelligent road network inspection. This intelligent planning mechanism ensures that the system maintains high processing efficiency and reliable planning results when facing complex and ever-changing road conditions. In particular, through multi-dimensional safety assessments, the reliability of drone inspections is significantly improved, enabling rapid response to emergencies.

[0033] Step S102: Obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location via the patrol route, select the UAV with the highest task fitness score as the target UAV, extract patrol rules from the task template library based on the event type, and send the patrol rules and the obstacle avoidance path to the target UAV to execute the patrol task. Optionally, this embodiment addresses the problems of unreasonable drone scheduling, low task allocation efficiency, and unclear patrol rules in traditional road network patrols by innovatively designing a drone scheduling scheme based on task adaptability. This embodiment first acquires drone status information in real time through the management platform's data interface using a distributed acquisition architecture. Each drone is equipped with a high-precision positioning module and an intelligent power management system to ensure the accuracy of location information and battery life data.

[0034] This embodiment deeply optimizes the UAV state assessment mechanism. Task fitness is calculated using a data fusion algorithm: Score = w1 Distance+w2 Battery + w3 Speed+w4 The payload consists of Distance (distance score from the drone to the target location), Battery (endurance score), Speed ​​(speed matching score), and Payload (payload adaptability score), with w1 to w4 representing the corresponding weight coefficients. The system employs an adaptive weight adjustment strategy, dynamically optimizing the weight configuration based on the urgency of the task.

[0035] This embodiment innovatively implements a range assessment strategy. It uses real-time battery power data acquired by the battery management system and combines this data with weather conditions and flight load to calculate the remaining range. The system considers the impact of environmental factors such as wind speed and temperature on range and predicts the actual usable range using an energy consumption model. Particularly for long-distance patrol missions, multi-point range planning ensures reliable mission execution.

[0036] This embodiment optimizes the payload status analysis mechanism. For different types of patrol tasks, the system evaluates the adaptability of the equipment carried by the UAV. Through matching analysis of sensor performance parameters and task requirements, it ensures the selection of the most suitable equipment combination. For example, for traffic accident patrols, the resolution and zoom capability of high-definition cameras are considered; for natural disaster patrols, the detection performance of thermal imaging equipment is emphasized.

[0037] This embodiment innovatively designs a task rule extraction scheme. A structured task template library is established, with standardized inspection rules preset for different types of events. The system quickly matches applicable inspection templates through a rule mapping engine, extracting key control parameters and execution flows. For complex scenarios, a complete task execution plan is generated through rule combinations.

[0038] This embodiment features a deeply optimized command issuance mechanism. A reliable communication protocol is used to construct the command transmission channel, and encryption algorithms ensure the security of command transmission. The system monitors communication quality in real time and automatically switches to a backup channel when signal attenuation is detected. Especially for critical commands, an acknowledgment mechanism guarantees reliable delivery.

[0039] This embodiment innovatively implements a task execution monitoring strategy. A real-time feedback mechanism is established to monitor the drone's execution status and task progress. The system assesses task execution risks through a status prediction algorithm and adjusts the patrol strategy promptly upon detecting anomalies. For multi-drone collaborative scenarios, a task coordination mechanism ensures the continuity and integrity of patrols.

[0040] This embodiment achieves precise allocation of drone patrol tasks through intelligent scheduling technology. It demonstrates strong decision-making capabilities and response efficiency, particularly in handling emergencies. Through a scientific evaluation mechanism and reliable execution plans, the system can quickly select the most suitable drone to perform patrol tasks, providing strong support for road network monitoring.

[0041] This embodiment's innovative design not only solves the scheduling efficiency and task execution problems of traditional methods but also establishes a continuously optimizing task management framework. Through continuous optimization of the evaluation model and improvement of execution strategies, the system can continuously enhance its adaptability to various patrol scenarios, providing reliable assurance for intelligent road network patrol. This intelligent scheduling mechanism ensures that the system maintains efficient decision-making capabilities and reliable execution results when facing complex and ever-changing road conditions. In particular, standardized task templates and flexible rule management significantly improve the standardization and execution efficiency of patrol tasks.

[0042] Step S103: Collect image data acquired by the target drone during the patrol process, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded early warning information to electronic information boards and LED displays in the distribution area, push different levels of early warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the early warning information in real time according to the progress of event handling.

[0043] Optionally, this embodiment addresses the problems of inaccurate image analysis, imprecise impact range assessment, and untimely dissemination of early warning information in traditional road network inspections by innovatively designing an intelligent early warning scheme based on deep learning. This embodiment first receives image data streams collected by drones in real time via a high-speed image transmission link, employing hardware acceleration technology to ensure real-time image processing. The system preprocesses the received images, including geometric correction, illumination equalization, and noise suppression, improving the accuracy of subsequent analysis.

[0044] This embodiment deeply optimizes the target detection mechanism. An improved YOLOv5 model is used for multi-target detection: Detection = Model(Image, Confidence), where Image is the input image and Confidence is the detection confidence threshold. The model is optimized for traffic scenarios through transfer learning, accurately identifying key targets such as vehicles, pedestrians, and road facilities. Especially for images under complex weather conditions, an attention mechanism enhances the robustness of detection.

[0045] This embodiment innovatively implements a scene understanding strategy. A scene recognition model is constructed based on a semantic segmentation network, and semantic information of the scene is extracted through multi-scale feature fusion. The system accurately segments key areas such as accident scenes and disaster zones, and calculates the impact range by combining spatial location information. For example, in traffic accident scenarios, the impact is assessed based on the length of vehicle backlog and the degree of road blockage; for natural disasters, the threat level is determined by the area of ​​damaged areas and the spread trend.

[0046] This embodiment optimizes the impact range analysis mechanism. A spatial analysis model based on geographic information is constructed, mapping the detection results to the actual road network coordinate system. The system determines the boundaries of the impact area through spatial clustering algorithms and calculates the impact propagation range considering the road topology. For complex road network structures, graph theory algorithms are used to analyze the diffusion path of the impact, achieving precise range delineation.

[0047] This embodiment innovatively designs a device linkage scheme. A device selection strategy is constructed based on the impact range information, and roadside devices requiring linkage are quickly located using spatial indexing. The system considers device coverage and information display effects, optimizing the device combination scheme. Especially for critical road sections, multi-device collaboration ensures comprehensive coverage of early warning information.

[0048] This embodiment deeply optimizes the early warning grading mechanism. Based on the distance between the vehicle and the event location, a multi-level early warning strategy is established: Warning_Level = f(Distance, Severity), where Distance is the vehicle distance and Severity is the severity of the event. The system adjusts the early warning level through dynamic thresholds to ensure the targeted nature and effectiveness of the early warning information.

[0049] This embodiment innovatively implements an information push strategy. Different information delivery schemes are adopted for different types of terminal devices. The system ensures the timely delivery of critical information through message priority management. For vehicle-mounted terminals, the system dynamically adjusts the push strategy based on the vehicle's direction and speed, predicting the potential impact time.

[0050] This embodiment utilizes deep learning technology to achieve intelligent analysis and early warning of road network events. It demonstrates particularly strong perception capabilities and decision-making efficiency when handling emergencies. Through precise scene understanding and rapid information dissemination, the system can effectively reduce the impact of events and provide strong support for road safety management.

[0051] This embodiment's innovative design not only solves the problems of analytical accuracy and response speed in traditional methods, but also establishes a continuously optimizing early warning framework. Through continuous optimization of the analytical model and improvement of early warning strategies, the system can continuously enhance its ability to handle various events, providing reliable assurance for intelligent road network management. This intelligent early warning mechanism ensures that the system maintains efficient analytical capabilities and reliable early warning effects when facing complex and ever-changing road conditions. In particular, through a multi-level information dissemination strategy, the coverage and timeliness of early warning information are significantly improved, achieving proactive prevention and control of road safety.

[0052] As described above, the intelligent road network patrol drone task scheduling method provided in this application can achieve precise planning of drone patrol routes by constructing a three-dimensional road network topology model and obstacle distribution map. A task-adaptability-based drone scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively solves the shortcomings of traditional technologies in task planning, drone scheduling, and early warning linkage, significantly improving the intelligence level of road network patrol and emergency response effectiveness.

[0053] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S201: Read the event information data stream from the road network monitoring system, establish an event information cache queue to store the data stream, parse the data packet header of the data stream to extract the event code and timestamp, obtain the event type identifier by matching the event code with a preset event type mapping table, extract the latitude and longitude coordinates of the event occurrence location and the event level value from the data stream message body, and write the event type identifier, the latitude and longitude coordinates, and the event level value into the event attribute table; Step S202: Read the vector data of the road network electronic map, construct a road network skeleton line matrix based on the vector data, calculate the three-dimensional spatial coordinates of the road network nodes to generate a road network node set, construct a three-dimensional road network topology model using the road network skeleton line matrix and the road network node set, retrieve building outline data, overpass elevation data, and tunnel boundary data from the geographic information database, convert the outline data, elevation data, and boundary data into a set of obstacle boundary points, and mark the set of obstacle boundary points in the three-dimensional road network topology model to generate an obstacle distribution map.

[0054] Optionally, this embodiment addresses the problems of delayed event information processing, insufficient road network modeling accuracy, and inaccurate obstacle marking in traditional road network monitoring by innovatively designing a road network modeling scheme based on real-time data streams. This embodiment first establishes a high-speed data channel to access the road network monitoring system, employing zero-copy technology to achieve efficient data stream reading. An event information cache queue is constructed using memory mapping technology: Queue_Size = Buffer_Length Event_Size, where Buffer_Length is the buffer length and Event_Size is the size of a single event data.

[0055] This embodiment features a deeply optimized data parsing mechanism. A streaming parser is used to process data packets, parsing the packet header information through a custom protocol. The system employs an efficient event coding system, including event categories, subcategories, and attribute identifiers, quickly extracting coded information through bitwise operations. For timestamp processing, a unified time base is used to ensure the timing accuracy of the event sequence. Especially in high-concurrency scenarios, multi-level queue management guarantees real-time data processing.

[0056] This embodiment innovatively implements an event mapping strategy. A structured event type mapping table is constructed, and fast lookup is achieved through hash indexes. The system establishes standardized description templates for different types of traffic events, including scenarios such as accidents, disasters, and road damage. For example, traffic accidents are subdivided into specific types such as rear-end collisions, rollovers, and multi-vehicle collisions; natural disasters include subdivisions such as flooding, landslides, and icing.

[0057] This embodiment optimizes the location information processing mechanism. Latitude and longitude coordinates are extracted from the data stream message body and mapped to a unified spatial reference system using a coordinate transformation algorithm. The system employs a high-precision positioning model to correct coordinate errors, specifically considering the impact of GPS signal drift and multipath effects. For event severity assessment, a multi-dimensional grading standard is established, comprehensively considering the event's impact range, duration, and severity.

[0058] This embodiment innovatively designs a road network modeling scheme. The road network skeleton is constructed based on vector data: Network = Matrix(Nodes, Edges), where Nodes is the set of road network nodes and Edges represents the road segment connections. The system calculates the spatial coordinates of the nodes through 3D projection transformation, considering geometric features such as road longitudinal slope and superelevation to ensure the geometric accuracy of the model.

[0059] This embodiment deeply optimizes the topology construction mechanism. A road network topology model is generated using a skeleton line matrix and node sets, and road network connectivity is analyzed using graph theory algorithms. The system considers special structures such as complex intersections and overpasses, ensuring the accuracy of topological relationships through multi-level modeling. In particular, for complex road sections such as roundabouts and interchanges, a layered modeling strategy ensures a clear representation of spatial relationships.

[0060] This embodiment innovatively implements an obstacle modeling strategy. It retrieves data on buildings, overpasses, tunnels, and other facilities from a geographic information database and constructs obstacle outlines using boundary extraction algorithms. The system employs 3D spatial indexing technology to manage obstacle data, improving spatial query efficiency. For tall buildings, accurate outline models are generated through 3D scanning; for overpasses, the locations of piers and clearance clearances are highlighted.

[0061] This embodiment achieves accurate modeling of the road network environment through data stream processing technology. It demonstrates particularly strong modeling capabilities and processing efficiency when dealing with complex urban road networks. Through precise event parsing and efficient spatial modeling, the system can quickly construct a realistic road network environment model, providing a reliable environmental perception foundation for UAV patrols.

[0062] This embodiment's innovative design not only solves the data processing and modeling accuracy problems of traditional methods but also establishes a continuously optimizing environmental modeling framework. Through improvements in data processing strategies and optimization of modeling methods, the system can continuously enhance its ability to represent complex road network environments, providing strong support for intelligent road network monitoring. This intelligent modeling mechanism ensures that the system maintains high processing efficiency and reliable modeling results when facing complex and ever-changing urban road networks. In particular, the fusion analysis of multi-source data significantly improves the accuracy and completeness of environmental perception.

[0063] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S301: Obtain the center coordinates of the event location, construct a circular search space area with the center coordinates as the center, divide the circular area into grid cells, calculate the center point coordinates of each grid cell, determine whether the center point of each grid cell overlaps with the boundary point of the obstacle based on the obstacle distribution map, filter the center points of the grid cells that do not overlap with the obstacle as a candidate return point set, calculate the distance and height difference from each return point in the candidate return point set to the event location, sort the candidate return point set according to the distance from near to far, and select the return point with the highest sorting as the emergency return point; Step S302: Construct a three-dimensional coordinate system for the road network topology map, project the coordinates of the emergency return point and the event location onto the three-dimensional coordinate system, calculate the initial flight path from the emergency return point to the event location based on the A-star search algorithm, read the boundary point data of the obstacle distribution map, calculate the collision detection distance between the initial flight path and the obstacle, adjust the flight path nodes to generate obstacle avoidance segments when a collision risk is detected, combine the obstacle avoidance segments with the flight path segments that have not collided to generate a patrol flight path, and smooth the patrol flight path according to the turning radius and climb angle constraints of the UAV.

[0064] Optionally, this embodiment addresses the problems of unreasonable return point selection, unsafe route planning, and insufficient obstacle avoidance capabilities in traditional road network inspections by innovatively designing a route planning scheme based on spatial analysis. This embodiment first constructs a search space with the event location as the core, and uses a polar coordinate system to calculate the search range: Search_Radius = f(Event_Type, Area_Size), where Event_Type is the event type and Area_Size is the area of ​​influence, ensuring that the search range covers all possible return points.

[0065] This embodiment deeply optimizes the grid partitioning mechanism. The circular search area is discretized using an adaptive grid partitioning algorithm, and the grid size is dynamically adjusted according to the complexity of the area. The system uses a quadtree structure to manage grid cells and improves query efficiency through spatial indexing technology. For each grid cell, the center point coordinates are determined through centroid calculation, and a grid coding system is established to facilitate spatial retrieval. Especially in densely built-up areas, fine-grained partitioning improves the accuracy of return point selection.

[0066] This embodiment innovatively implements an obstacle detection strategy. Based on a pre-constructed obstacle distribution map, the spatial relationship between the grid center point and obstacles is detected using ray projection. The system ensures the safety of selected points through multi-level collision detection, taking into account the influence of aerial obstacles such as building height and power lines. For example, in areas with high-rise buildings, the system calculates the shadow range of the buildings to avoid selecting occluded locations.

[0067] This embodiment optimizes the return-to-home point evaluation mechanism. The selected candidate return-to-home points are evaluated from multiple dimensions, calculating the Euclidean distance and elevation difference to the target location. The system considers terrain undulations and building distribution, determining the priority of return-to-home points through weighted scoring. For patrol tasks on critical road sections, factors such as visibility and communication signal strength are specifically considered.

[0068] This embodiment innovatively designs a route planning scheme. A path search space is constructed in a three-dimensional coordinate system, and an improved A* algorithm is used for initial route planning: Path_Cost = g(n) + h(n), where g(n) is the actual cost from the starting point to the current node, and h(n) is the estimated cost from the current node to the target point. The system optimizes search efficiency through a heuristic function, taking into account altitude changes and turning costs.

[0069] This embodiment features a deeply optimized collision detection mechanism. Dynamic collision detection is performed on the planned flight path, quickly identifying potential collision risks using bounding box technology. The system employs a hierarchical detection strategy, first conducting a coarse detection, then performing a refined detection on potentially risky flight segments. Particularly for complex grade-separated areas, three-dimensional spatial analysis ensures sufficient flight safety margins.

[0070] This embodiment innovatively implements an obstacle avoidance path generation strategy. When a collision risk is detected, the system adjusts the route nodes using an artificial potential field method to generate a safe obstacle avoidance path. It simulates the repulsive effect of obstacles using a virtual force field while simultaneously considering the attractive effect of the target point, achieving dynamic route planning. For multi-obstacle scenarios, a path optimization algorithm is used to find the optimal obstacle avoidance solution.

[0071] This embodiment achieves safe design of UAV flight paths through intelligent planning technology. It demonstrates strong planning capabilities and obstacle avoidance, particularly when dealing with complex urban environments. Through scientific return-to-home point selection and precise flight path planning, the system ensures the safe execution of UAV patrol missions, providing reliable support for road network monitoring.

[0072] This embodiment's innovative design not only solves the safety and reliability issues of traditional methods but also establishes a continuously optimized flight path planning framework. Through continuous improvement of planning strategies and optimization of obstacle avoidance methods, the system can continuously enhance its adaptability to complex scenarios, providing strong support for intelligent road network patrol. This intelligent planning mechanism ensures that the system maintains efficient planning capabilities and reliable obstacle avoidance effects when facing complex and ever-changing urban environments. In particular, through multi-level safety assessments, the reliability and efficiency of UAV patrols are significantly improved.

[0073] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S401: Receive telemetry data packets from each UAV from the UAV management platform, parse the telemetry data packets to obtain the UAV identification code, read the corresponding UAV's position coordinates, remaining battery power, flight speed, and mission payload type based on the identification code, calculate the remaining flight range based on the remaining battery power, determine the camera specifications and image transmission equipment performance of the UAV based on the mission payload type, and write the position coordinates, remaining flight range, flight speed, camera specifications, and image transmission equipment performance into the UAV status information table; Step S402: Establish a task fitness score calculation model, using the straight-line distance between the UAV's position coordinates and the location of the event, the ratio of the UAV's remaining range to the length of the patrol route, the matching degree between the UAV's flight speed and the expected patrol duration, and the adaptability of the camera specifications to the patrol task requirements as scoring indicators. Determine the weight coefficients of each scoring indicator based on the analytic hierarchy process, and calculate the task fitness score by weighting the scoring indicators and the weight coefficients. Select the UAV with the highest task fitness score as the target UAV.

[0074] Optionally, this embodiment addresses the problems of inaccurate state perception, unreasonable task matching, and low decision-making efficiency in traditional UAV task scheduling by innovatively designing an intelligent scheduling scheme based on multi-dimensional evaluation. This embodiment first establishes a real-time communication link with the UAV management platform, employing an efficient telemetry data transmission protocol to ensure the real-time nature and reliability of data acquisition. The telemetry data packets are processed through a data parsing engine: Telemetry = Parse(ID, Position, Battery, Speed, Payload), where each parameter represents the UAV identifier, location information, battery status, flight parameters, and payload information, respectively.

[0075] This embodiment deeply optimizes the status information processing mechanism. Based on real-time data from the battery management system, a mapping model between battery charge and flight range is established: Range = f(Battery, Weight, Wind), where Battery is the current battery percentage, Weight is the takeoff weight, and Wind is the wind speed factor. The system establishes an energy consumption prediction model through historical data analysis, considering the impact of weather conditions, flight attitude, and other factors on flight range. Different energy consumption parameters are used for different aircraft models to ensure the accuracy of flight range prediction.

[0076] This embodiment innovatively implements a payload evaluation strategy. A device performance evaluation system is constructed for different types of task payloads. The system quantitatively evaluates parameters such as camera resolution, focal length, and aperture, while also considering performance indicators of the image transmission equipment, such as bandwidth, latency, and anti-interference capability. For example, for traffic accident investigation, the focus is on evaluating zoom capability and image stability; for disaster monitoring, greater emphasis is placed on thermal imaging and wide-angle coverage capabilities.

[0077] This embodiment optimizes the status information management mechanism. A distributed database is used to store UAV status information, and a real-time update mechanism ensures data timeliness. The system establishes a complete status tracking chain, recording the UAV's historical performance and maintenance records, providing a reliable basis for task allocation decisions. In particular, for critical status parameters, data verification ensures the accuracy of the information.

[0078] This embodiment innovatively designs a fitness scoring model. The scoring system is constructed based on the analytic hierarchy process (AHP): Score = w1 D+w2 R + w3 V+w4 C represents the distance score, R the range score, V the speed score, C the payload score, and w1 to w4 the corresponding weight coefficients. The system determines the initial weights through expert experience and historical data analysis, and dynamically optimizes the weight configuration using machine learning methods.

[0079] This embodiment deeply optimizes the performance evaluation mechanism. For distance scoring, the actual flight distance is calculated considering road topology; for endurance scoring, the remaining battery capacity is assessed in conjunction with mission duration; for speed scoring, the optimal flight speed is determined by analyzing mission urgency; and for payload scoring, the equipment's suitability is evaluated based on mission requirements. Through this multi-dimensional comprehensive evaluation, the most suitable execution unit is selected.

[0080] This embodiment innovatively implements a decision optimization strategy. It employs a dynamic programming algorithm to optimize task allocation, considering the synergistic effects of multiple drones. The system assesses task execution risk through state prediction, maximizing resource utilization efficiency while ensuring task completion quality. For urgent tasks, a rapid decision-making mechanism shortens response time.

[0081] This embodiment achieves intelligent scheduling of UAV missions through multi-dimensional evaluation technology. It demonstrates strong decision-making capabilities and matching efficiency, particularly when handling complex mission scenarios. Through precise state perception and a scientific evaluation mechanism, the system can quickly select the most suitable UAV to perform patrol tasks, providing reliable technical support for road network monitoring.

[0082] This embodiment's innovative design not only solves the state assessment and task matching problems in traditional methods but also establishes a continuously optimizing scheduling framework. Through continuous optimization of the evaluation model and improvement of decision-making strategies, the system can continuously enhance its adaptability to various task scenarios, providing strong support for intelligent road network inspection. This intelligent scheduling mechanism ensures that the system maintains efficient decision-making capabilities and reliable execution results when facing complex and ever-changing road conditions. In particular, through multi-dimensional comprehensive evaluation, the accuracy of task allocation and execution efficiency are significantly improved.

[0083] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S501: Read the preset mapping relationship between event types and patrol rules in the task template library, match the corresponding patrol rule identifier based on the event type, extract the patrol start point coordinates, patrol end point coordinates, flight path sampling interval, flight altitude parameters, flight speed parameters, camera pitch angle parameters, and image acquisition frequency parameters according to the patrol rule identifier, and combine the patrol rule parameters with the event occurrence location information and the target UAV identification code to generate a task configuration file; Step S502: Construct a UAV mission control instruction set, encode the parameters in the mission configuration file according to the communication protocol specification, generate mission initialization instructions, flight path setting instructions, obstacle avoidance path instructions, and image acquisition instructions, establish a data communication link with the target UAV, send the control instruction set to the target UAV through the data communication link, receive the instruction verification code returned by the target UAV, and verify the reliability of the mission instructions by verifying the verification code.

[0084] Optionally, this embodiment addresses the problems of incomplete patrol rules, unreliable command issuance, and inaccurate task configuration in traditional UAV mission execution by innovatively designing a template-based task configuration scheme. This embodiment first constructs a structured task template library, employing a layered design to ensure rule integrity. Rule mapping is achieved through a template parsing engine: Rule = Map(Event_Type, Rule_ID), where Event_Type is the event type code and Rule_ID is the corresponding rule identifier, enabling precise matching between event types and patrol rules.

[0085] This embodiment deeply optimizes the rule parameter extraction mechanism. Based on rule identifiers, the system uses a parameterized configuration method to extract the key parameters required for task execution. Different parameter combinations are set for different types of patrol tasks. For example, traffic accident patrol requires a lower flight altitude and a higher image acquisition frequency to obtain clear details of the accident scene; while natural disaster monitoring requires a larger flight path sampling interval and a wider coverage area to comprehensively assess the scope of disaster impact.

[0086] This embodiment innovatively implements a task configuration generation strategy. It intelligently fuses extracted rule parameters with real-time event information to construct a standardized configuration file: Config = Combine(Rules, Event, UAV_ID), where Rules is the set of rule parameters, Event is the event information, and UAV_ID is the UAV identifier. The system ensures the completeness and rationality of the configuration through parameter validation, paying particular attention to the logical correlation between parameters, such as the matching relationship between flight altitude and camera pitch angle.

[0087] This embodiment optimizes the instruction set construction mechanism. Based on standard communication protocols, task configurations are converted into executable control instructions. The system adopts a hierarchical instruction structure to ensure the orderly execution of tasks. For initialization instructions, basic task parameters and execution conditions are included; for flight path setting instructions, flight paths and control points are defined in detail; for obstacle avoidance path instructions, safe flight paths and avoidance strategies are included; and for image acquisition instructions, the acquisition timing and image parameters are specified.

[0088] This embodiment innovatively designs a communication link management scheme. An encrypted communication protocol is used to ensure the security of command transmission: Command = Encode(Config, Protocol), where Config is the configuration information and Protocol is the communication protocol specification. The system ensures communication reliability through multi-channel backup and automatically switches communication channels when signal interference occurs. Especially for critical commands, an acknowledgment mechanism is used to ensure reliable delivery of commands.

[0089] This embodiment deeply optimizes the instruction verification mechanism. A complete verification process is constructed, ensuring instruction integrity through cyclic redundancy verification. The system is designed with a multi-level verification strategy, including instruction format verification, parameter range verification, and execution condition verification. For complex task instructions, segmented verification improves verification efficiency and ensures the accuracy of each execution stage.

[0090] This embodiment innovatively implements a task confirmation strategy. A two-way communication mechanism is established to monitor the instruction execution status in real time. The system assesses task execution risks through status feedback and adjusts the instruction sequence promptly when anomalies are detected. For multi-stage tasks, state machine management ensures task continuity and integrity.

[0091] This embodiment achieves precise execution of UAV missions through template-based configuration technology. It demonstrates strong configuration capabilities and execution efficiency, particularly when handling complex patrol scenarios. Through standardized command management and reliable communication mechanisms, the system ensures accurate issuance and reliable execution of mission commands, providing strong support for road network patrols.

[0092] This embodiment's innovative design not only solves the configuration accuracy and command reliability issues of traditional methods but also establishes a continuously optimizeable task execution framework. Through continuous improvement of rule templates and optimization of the command system, the system can continuously enhance its adaptability to various patrol scenarios, providing reliable assurance for intelligent road network monitoring. This intelligent configuration mechanism ensures that the system maintains high efficiency and reliable task results when facing complex and ever-changing patrol tasks. In particular, multi-level command verification significantly improves the accuracy and reliability of task execution.

[0093] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S601: Receive the image data stream transmitted back by the target UAV, parse the image frames and shooting timestamps from the image data stream, store the image frames and shooting timestamps into the image cache queue, read the target UAV position data to record the shooting position coordinates of the image frames, perform geometric correction and image enhancement processing on the image frames, use the target detection model to identify the vehicle distribution density, road traffic status, and traffic control facilities in the image, extract the accident scene range, vehicle backlog length, and road blockage location based on the scene recognition model, and generate a target detection result table and a scene recognition result table; Step S602: Fuse the spatial location information of the target detection result table and the scene recognition result table to construct the boundary contour of the event's influence range, calculate the geometric center and coverage radius of the boundary contour, establish a local coordinate system with the geometric center as the origin, mark the locations of information boards, display screens, and broadcasting equipment along the road in the local coordinate system, calculate the distance from each device location to the geometric center, and filter devices with a distance smaller than the coverage radius to generate the roadside device distribution area.

[0094] Optionally, this embodiment addresses the problems of untimely image processing, inaccurate scene understanding, and imprecise impact range assessment in traditional road network patrols by innovatively designing an intelligent analysis scheme based on deep learning. This embodiment first establishes a high-speed image transmission channel and uses real-time stream processing technology to receive image data. Key information is extracted through a data stream parsing engine: Frame = Parse(Image, Timestamp, Position), where Image is the image data, Timestamp is the shooting time, and Position is the location information.

[0095] This embodiment deeply optimizes the image preprocessing mechanism. Geometric correction is performed based on UAV attitude information to eliminate image distortion and deformation. The system employs an adaptive enhancement algorithm to improve image quality: Enhancement = f(Brightness, Contrast, Noise), where each parameter represents brightness, contrast, and noise level, respectively. Particularly for images under different weather conditions, scene-adaptive methods are used to optimize image features, ensuring the accuracy of subsequent analysis.

[0096] This embodiment innovatively implements an object detection strategy. An improved YOLOv5 model is used for multi-object detection, and model performance is optimized through transfer learning. The system focuses on key targets such as vehicles, pedestrians, and facilities, tailored to the characteristics of traffic scenes. The model input is a preprocessed image, and the output includes information such as object category, location, and confidence level. An attention mechanism is used to improve detection accuracy, especially for small object detection in dense scenes.

[0097] This embodiment optimizes the scene recognition mechanism. A scene understanding model is built based on a deep learning network, and recognition accuracy is improved through multi-scale feature fusion. The system performs semantic segmentation on key scenes such as accident scenes, vehicle queues, and road blockages, and assesses the degree of impact by combining spatial location information. For example, traffic congestion status is analyzed through vehicle density distribution, and traffic capacity is assessed through road occupancy.

[0098] This embodiment innovatively designs a data fusion scheme. It spatially registers the results of target detection and scene recognition to construct a unified analysis framework. The system integrates target location, scene features, and spatial relationships through multi-source information fusion technology to generate a complete event description. Particularly for complex scenes, it assesses the event's development trend through time-series analysis.

[0099] This embodiment deeply optimizes the influence range analysis mechanism. An event influence model is constructed based on the fused data, and the influence range is determined through a boundary extraction algorithm. The system uses a convex hull algorithm to generate the contour of the influence region and calculates its geometric center and coverage radius. For irregularly shaped influence regions, polygon approximation is used to improve the accuracy of representation.

[0100] This embodiment innovatively implements a device distribution analysis strategy. A device location index is established in a local coordinate system, and affected devices are identified through spatial queries. The system considers road topology and evaluates device coverage and information transmission efficiency. For critical road sections, multi-level filtering ensures the effective dissemination of early warning information.

[0101] This embodiment achieves accurate assessment of road network events through intelligent analysis technology. It demonstrates particularly strong perception capabilities and analytical efficiency when handling complex traffic scenarios. Through the application of deep learning models, the system can quickly understand the scene's state, accurately assess the impact of events, and provide a reliable basis for issuing early warnings.

[0102] This embodiment's innovative design not only solves the problems of scene understanding and range assessment in traditional methods, but also establishes a continuously optimizing analytical framework. Through continuous optimization of the analytical model and improvement of processing strategies, the system can continuously enhance its adaptability to various scenarios, providing strong support for intelligent road network monitoring. This intelligent analytical mechanism ensures that the system maintains efficient perception capabilities and reliable assessment results when facing complex and ever-changing road conditions. In particular, through multi-dimensional data fusion, the accuracy of event impact assessment and the timeliness of early warning issuance are significantly improved.

[0103] In one embodiment of the UAV mission scheduling method for intelligent road network inspection of this application, it may further include the following: Step S701: Construct a warning information classification template, divide the warning level threshold based on the straight-line distance between the vehicle and the location of the event, read the communication address and device type identifier of each device in the distribution area of ​​roadside devices, match the corresponding information release format according to the device type identifier, generate device control instructions according to the information release format, establish communication connection with each device, send the device control instructions to the corresponding electronic information board and LED display screen, and receive the instruction response confirmation information returned by the device; Step S702: Obtain real-time location data of driving vehicles from the vehicle network platform, calculate the straight-line distance between each vehicle and the location where the event occurred, compare the straight-line distance with the warning level threshold to determine the warning information level, select the corresponding information push template according to the warning information level, write the event handling status information into the information push template to generate vehicle terminal prompt information, push the prompt information to the vehicle terminal of the target vehicle through the vehicle network communication gateway, monitor the event handling progress status, and regenerate the warning information content when a progress update is received.

[0104] Optionally, this embodiment addresses the problems of untimely information distribution, inaccurate warning levels, and uncoordinated equipment linkage in traditional road network early warning systems by innovatively designing an intelligent early warning release scheme based on multi-level warnings. This embodiment first constructs a hierarchical early warning template library and uses adaptive threshold technology to classify warning levels: Level = f(Distance, Event_Type), where Distance is the distance from the vehicle to the event location, and Event_Type is the event type, ensuring the targeted nature of the warning information.

[0105] This embodiment deeply optimizes the device control mechanism. A unified device management framework is established based on device type identifiers. The system designs a standardized information publishing format for different types of display devices: Format = Template(Device_Type, Content_Type), where Device_Type is the device type and Content_Type is the content type. For example, for electronic information boards, a partitioned display strategy is adopted to rationally arrange the display positions of text and symbols; for LED displays, information scrolling and brightness adjustment are considered.

[0106] This embodiment innovatively implements a communication management strategy. It employs multi-protocol adapters to connect various roadside devices, and uses a unified communication interface to issue commands. The system establishes a reliable communication link and monitors device status through a heartbeat detection mechanism. For critical node devices, dual-channel backup ensures communication reliability. Especially under severe weather conditions, adaptive power control improves communication quality.

[0107] This embodiment optimizes the device response mechanism. A complete command confirmation process is constructed, and the command execution status is verified through a response mechanism. The system is designed with an exception handling strategy, automatically switching to a backup device when a device failure is detected. For the information publishing process, status feedback ensures the accuracy and completeness of the displayed content.

[0108] This embodiment innovatively designs a vehicle-to-everything (V2X) interaction scheme. It acquires vehicle location information in real time from the V2X platform and evaluates the relative relationship between vehicles and events using a spatial computing engine. The system employs efficient spatial indexing technology to quickly filter target vehicles requiring warnings. For vehicles traveling at high speeds, it sends warning information in advance through predictive analysis.

[0109] This embodiment deeply optimizes the information push mechanism. The warning level is dynamically adjusted based on vehicle distance, employing a differentiated push strategy: Warning = Select(Level, Template), where Level is the warning level and Template is the corresponding push template. The system determines the urgency of the information through scenario analysis, using a strong prompt method for urgent warnings and a regular push method for general warnings.

[0110] This embodiment innovatively implements a status update strategy. An event handling status monitoring mechanism is established to track the progress of handling in real time. The system assesses changes in the situation through a status analysis engine and updates the warning content promptly when significant progress is detected. For ongoing events, timeliness of information is ensured through periodic updates.

[0111] This embodiment achieves accurate dissemination of road network events through multi-level early warning technology. It demonstrates particularly strong information distribution capabilities and efficient coordination, especially in handling emergencies. Through scientific early warning classification and reliable equipment control, the system can quickly respond to various emergencies, providing strong support for road safety management.

[0112] This embodiment's innovative design not only solves the information dissemination and device linkage problems of traditional methods but also establishes a continuously optimizable early warning framework. Through continuous improvement of early warning strategies and optimization of the dissemination mechanism, the system can continuously enhance its response capabilities to various events, providing reliable assurance for intelligent road network management. This intelligent early warning mechanism ensures that the system maintains efficient dissemination capabilities and reliable linkage effects when facing complex and ever-changing road conditions. In particular, the multi-channel information push significantly improves the coverage and timeliness of early warning information.

[0113] To effectively address the shortcomings of traditional technologies in task planning, drone scheduling, and early warning linkage, and to significantly improve the intelligence level and emergency response effectiveness of road network patrol, this application provides an embodiment of an intelligent road network patrol drone task scheduling device for implementing all or part of the aforementioned intelligent road network patrol drone task scheduling method. See [link to embodiment]. Figure 2 The intelligent road network patrol drone task scheduling device specifically includes the following components: The path planning module 10 is used to receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system, extract the event type, location, and event level attributes from the event information, construct a three-dimensional road network topology model based on the road network electronic map, mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map, calculate the available emergency return points for UAVs around the event location, and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. The task inspection module 20 is used to obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location through the inspection route, select the UAV with the highest task fitness score as the target UAV, extract inspection rules from the task template library based on the event type, and send the inspection rules and the obstacle avoidance path to the target UAV to execute the inspection task. The abnormal alarm module 30 is used to collect image data obtained by the target drone during the patrol, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded early warning information to electronic information boards and LED displays in the distribution area, push different levels of early warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the early warning information in real time according to the progress of event handling.

[0114] As described above, the intelligent road network patrol drone task scheduling device provided in this application embodiment can accurately plan drone patrol routes by constructing a three-dimensional road network topology model and obstacle distribution map. A task-adaptability-based drone scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively solves the shortcomings of traditional technologies in task planning, drone scheduling, and early warning linkage, significantly improving the intelligence level of road network patrol and emergency response effectiveness.

[0115] From a hardware perspective, in order to effectively address the shortcomings of traditional technologies in task planning, drone scheduling, and early warning linkage, and significantly improve the intelligence level and emergency response effectiveness of road network patrol, this application provides an embodiment of an electronic device for implementing all or part of the drone task scheduling method for intelligent road network patrol. The electronic device specifically includes the following components: The system comprises a processor, memory, a communications interface, and a bus; wherein the processor, memory, and communications interface communicate with each other via the bus; the communications interface is used to realize information transmission between the intelligent road network patrol drone task scheduling device and core business systems, user terminals, and related databases and other related devices; the logic controller can be a desktop computer, tablet computer, or mobile terminal, etc., and this embodiment is not limited to these. In this embodiment, the logic controller can be implemented with reference to the embodiments of the intelligent road network patrol drone task scheduling method and the intelligent road network patrol drone task scheduling device, the contents of which are incorporated herein by reference, and repeated details will not be described again.

[0116] It is understood that the user terminal may include smartphones, tablet computers, network set-top boxes, portable computers, desktop computers, personal digital assistants (PDAs), in-vehicle devices, smart wearable devices, etc. Among these, the smart wearable devices may include smart glasses, smartwatches, smart bracelets, etc.

[0117] In practical applications, the drone task scheduling method for intelligent road network patrol can be partially executed on the electronic device side as described above, or all operations can be completed in the client device. The choice can be made based on the processing power of the client device and the limitations of the user's usage scenario. This application does not impose any limitations on this. If all operations are completed in the client device, the client device may further include a processor.

[0118] The aforementioned client device may have a communication module (i.e., a communication unit) that can communicate with a remote server to achieve data transmission with the server. The server may include a server on the task scheduling center side; in other implementation scenarios, it may also include a server on an intermediate platform, such as a server on a third-party server platform that has a communication link with the task scheduling center server. The server may include a single computer device, a server cluster consisting of multiple servers, or a distributed server structure.

[0119] Figure 3 This is a schematic block diagram illustrating the system configuration of the electronic device 9600 according to an embodiment of this application. Figure 3 As shown, the electronic device 9600 may include a central processing unit 9100 and a memory 9140; the memory 9140 is coupled to the central processing unit 9100. It is worth noting that... Figure 3 This is an example; other types of structures can also be used to supplement or replace this structure to achieve telecommunications functions or other functions.

[0120] In one embodiment, the intelligent road network patrol drone task scheduling method can be integrated into the central processing unit 9100. The central processing unit 9100 can be configured to perform the following control: Step S101: Receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system; extract the event type, location, and event level attributes from the event information; construct a three-dimensional road network topology model based on the road network electronic map; mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map; calculate the available emergency return points for UAVs around the event location; and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. Step S102: Obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location via the patrol route, select the UAV with the highest task fitness score as the target UAV, extract patrol rules from the task template library based on the event type, and send the patrol rules and the obstacle avoidance path to the target UAV to execute the patrol task. Step S103: Collect image data acquired by the target drone during the patrol process, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded early warning information to electronic information boards and LED displays in the distribution area, push different levels of early warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the early warning information in real time according to the progress of event handling.

[0121] As described above, the electronic device provided in this application embodiment achieves precise planning of UAV patrol routes by constructing a three-dimensional road network topology model and obstacle distribution map. A task-adaptability-based UAV scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively solves the shortcomings of traditional technologies in task planning, UAV scheduling, and early warning linkage, significantly improving the intelligence level of road network patrol and emergency response effectiveness.

[0122] In another embodiment, the drone task scheduling device for intelligent road network patrol can be configured separately from the central processing unit 9100. For example, the drone task scheduling device for intelligent road network patrol can be configured as a chip connected to the central processing unit 9100, and the drone task scheduling method function for intelligent road network patrol can be realized through the control of the central processing unit.

[0123] like Figure 3 As shown, the electronic device 9600 may further include: a communication module 9110, an input unit 9120, an audio processor 9130, a display 9160, and a power supply 9170. It is worth noting that the electronic device 9600 does not necessarily need to include these components. Figure 3 All components shown; in addition, the electronic device 9600 may also include Figure 3 For components not shown, please refer to existing technologies.

[0124] like Figure 3 As shown, the central processing unit 9100, sometimes also referred to as a controller or operating control, may include a microprocessor or other processor device and / or logic device, which receives inputs and controls the operation of various components of the electronic device 9600.

[0125] The memory 9140 may be, for example, one or more of a cache, flash memory, hard drive, removable media, volatile memory, non-volatile memory, or other suitable devices. It may store the aforementioned failure-related information, and also store a program for executing that information. The central processing unit 9100 may execute the program stored in the memory 9140 to perform information storage or processing, etc.

[0126] Input unit 9120 provides input to central processing unit 9100. Input unit 9120 may be, for example, a keypad or touch input device. Power supply 9170 provides power to electronic device 9600. Display 9160 displays images and text. Display may be, for example, an LCD display, but is not limited thereto.

[0127] The memory 9140 can be a solid-state memory, such as a read-only memory (ROM), random access memory (RAM), a SIM card, etc. It can also be a memory that retains information even when power is off, can be selectively erased, and contains more data; examples of this type of memory are sometimes referred to as EPROMs. The memory 9140 can also be some other type of device. The memory 9140 includes a buffer memory 9141 (sometimes referred to as a buffer). The memory 9140 may include an application / function storage unit 9142 for storing application programs and function programs or processes for executing the operation of the electronic device 9600 via the central processing unit 9100.

[0128] The memory 9140 may also include a data storage unit 9143 for storing data, such as contacts, digital data, pictures, sounds, and / or any other data used by the electronic device. The driver storage unit 9144 of the memory 9140 may include various drivers for the electronic device for communication functions and / or for performing other functions of the electronic device (such as messaging applications, address book applications, etc.).

[0129] The communication module 9110 is a transmitter / receiver that sends and receives signals via the antenna 9111. The communication module 9110 (transmitter / receiver) is coupled to the central processing unit 9100 to provide input signals and receive output signals, which is the same as in a conventional mobile communication terminal.

[0130] Based on different communication technologies, multiple communication modules 9110 can be configured in the same electronic device, such as cellular network modules, Bluetooth modules, and / or wireless LAN modules. The communication module 9110 (transmitter / receiver) is also coupled to a speaker 9131 and a microphone 9132 via an audio processor 9130 to provide audio output via the speaker 9131 and receive audio input from the microphone 9132, thereby realizing typical telecommunications functions. The audio processor 9130 may include any suitable buffer, decoder, amplifier, etc. Additionally, the audio processor 9130 is coupled to a central processing unit 9100, enabling on-device recording via the microphone 9132 and on-device playback of stored audio via the speaker 9131.

[0131] Embodiments of this application also provide a computer-readable storage medium capable of implementing all steps of the intelligent road network patrol drone task scheduling method with a server or client as the execution subject in the above embodiments. The computer-readable storage medium stores a computer program that, when executed by a processor, implements all steps of the intelligent road network patrol drone task scheduling method with a server or client as the execution subject in the above embodiments. For example, when the processor executes the computer program, it implements the following steps: Step S101: Receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system; extract the event type, location, and event level attributes from the event information; construct a three-dimensional road network topology model based on the road network electronic map; mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map; calculate the available emergency return points for UAVs around the event location; and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. Step S102: Obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location via the patrol route, select the UAV with the highest task fitness score as the target UAV, extract patrol rules from the task template library based on the event type, and send the patrol rules and the obstacle avoidance path to the target UAV to execute the patrol task. Step S103: Collect image data acquired by the target drone during the patrol process, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded early warning information to electronic information boards and LED displays in the distribution area, push different levels of early warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the early warning information in real time according to the progress of event handling.

[0132] As described above, the computer-readable storage medium provided in this application enables precise planning of UAV patrol routes by constructing a three-dimensional road network topology model and obstacle distribution map. A task-adaptability-based UAV scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively addresses the shortcomings of traditional technologies in task planning, UAV scheduling, and early warning linkage, significantly improving the intelligence level of road network patrols and the effectiveness of emergency response.

[0133] Embodiments of this application also provide a computer program product capable of implementing all steps in the intelligent road network patrol drone task scheduling method described above, where the execution subject is a server or client. When executed by a processor, this computer program / instruction implements the steps of the intelligent road network patrol drone task scheduling method. For example, the computer program / instruction implements the following steps: Step S101: Receive traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system; extract the event type, location, and event level attributes from the event information; construct a three-dimensional road network topology model based on the road network electronic map; mark the locations of buildings, overpasses, and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map; calculate the available emergency return points for UAVs around the event location; and plan the UAV patrol route and obstacle avoidance path based on the obstacle distribution map. Step S102: Obtain the real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, calculate the task fitness score of each UAV reaching the event location via the patrol route, select the UAV with the highest task fitness score as the target UAV, extract patrol rules from the task template library based on the event type, and send the patrol rules and the obstacle avoidance path to the target UAV to execute the patrol task. Step S103: Collect image data acquired by the target drone during the patrol process, perform target detection and scene recognition on the image data, extract event impact range information, determine the distribution area of ​​roadside equipment that needs to be linked based on the event impact range information, send graded early warning information to electronic information boards and LED displays in the distribution area, push different levels of early warning information to the vehicle terminal based on the distance of the vehicle from the event location, and update the content of the early warning information in real time according to the progress of event handling.

[0134] As described above, the computer program product provided in this application enables precise planning of UAV patrol routes by constructing a three-dimensional road network topology model and obstacle distribution map. A task-adaptability-based UAV scheduling strategy is designed, combining real-time location, remaining range, and payload status to extract patrol rules from a task template library for intelligent scheduling. A tiered early warning mechanism is introduced, using target detection and scene recognition to analyze the impact range of events and achieve differentiated early warnings for roadside equipment and vehicle-mounted terminals. This method effectively addresses the shortcomings of traditional technologies in task planning, UAV scheduling, and early warning linkage, significantly improving the intelligence level of road network patrols and the effectiveness of emergency response.

[0135] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, apparatus, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (devices), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0137] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0139] Specific embodiments have been used to illustrate the principles and implementation methods of this invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of this invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this invention. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A method for scheduling unmanned aerial vehicle (UAV) tasks for intelligent road network inspection, characterized in that, The method includes: The system receives traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system. It extracts the event type, location, and level attributes from the event information. Based on the road network electronic map, it constructs a three-dimensional road network topology model. In the three-dimensional road network topology model, it marks the locations of buildings, overpasses, and tunnels to construct an obstacle distribution map. It obtains the center coordinates of the event location and constructs a circular search space with the center coordinates as the center. The circular area is divided into grid cells, and the center point coordinates of each grid cell are calculated. Based on the obstacle distribution map, it determines whether the center point of each grid cell overlaps with the obstacle boundary point. The center points of grid cells that do not overlap with obstacles are selected as a set of candidate return points. The distance and height difference from each return point in the candidate return point set to the event location are calculated. The candidate return point set is sorted from closest to furthest in distance, and the return point with the highest ranking is selected as the emergency return point. A three-dimensional coordinate system for constructing a road network topology map is established. The coordinates of the emergency return point and the location of the incident are projected onto the three-dimensional coordinate system. The initial flight path from the emergency return point to the location of the incident is calculated based on the A-star search algorithm. The boundary point data of the obstacle distribution map is read, and the collision detection distance between the initial flight path and the obstacle is calculated. When a collision risk is detected, the flight path nodes are adjusted to generate an obstacle avoidance segment. The obstacle avoidance segment is combined with the flight path segment that has not collided to generate a patrol route. In the constructed virtual flight environment, the safety of the patrol route is evaluated using the Monte Carlo method. The patrol route is smoothed according to the turning radius and climb angle constraints of the UAV. The real-time location information, remaining range, and payload status information of each drone are obtained from the drone management platform. A task fitness scoring calculation model is established, using the straight-line distance between the drone's location coordinates and the location of the event, the ratio of the drone's remaining range to the length of the patrol route, the matching degree between the drone's flight speed and the expected patrol duration, and the adaptability of the camera specifications to the patrol task requirements as scoring indicators. The weight coefficients of each scoring indicator are determined based on the analytic hierarchy process (AHP). The task fitness score of each drone is obtained by weighting the scoring indicators and the weight coefficients. The drone with the highest task fitness score is selected as the target drone. Patrol rules are extracted from the task template library based on the event type. The patrol rules and obstacle avoidance paths are sent to the target drone to execute the patrol task. The payload status information includes at least the camera's resolution, focal length, and aperture, and at least the bandwidth, latency, and anti-interference capability of the image transmission device. Collect image data acquired by the target unmanned aerial vehicle during the inspection process, perform target detection and scene recognition on the image data, extract event impact range information, and determine the distribution area of road-side devices that need to be linked based on the event impact range information, which specifically includes: receiving the image data stream returned by the target unmanned aerial vehicle, parsing image frames and shooting timestamps from the image data stream, storing the image frames and the shooting timestamps into an image buffer queue, reading the position data of the target unmanned aerial vehicle and recording the shooting position coordinates of the image frames, performing geometric correction and image enhancement processing on the image frames, identifying vehicle distribution density, road traffic status and traffic control facilities in the images by using a target detection model, extracting accident site range, vehicle queue length and road blocking position based on a scene recognition model, and generating a target detection result table and a scene recognition result table; fusing the spatial position information in the target detection result table and the scene recognition result table, constructing a boundary contour of the event impact range, calculating the geometric center and coverage radius of the boundary contour, establishing a local coordinate system with the geometric center as the origin, marking the positions of variable message signs, display screens and broadcasting equipment along the road in the local coordinate system, calculating the distance from each device position to the geometric center, and screening devices with a distance smaller than the coverage radius to generate a road-side device distribution area; Sending hierarchical early warning information to electronic variable message signs and LED display screens in the distribution area, pushing early warning information of different levels to on-board terminals based on the distance between vehicles and the event occurrence position, and updating the content of the early warning information in real time according to the progress of event disposal.

2. The UAV mission scheduling method for intelligent road network inspection according to claim 1, characterized in that, Receiving traffic accident, natural disaster and road abnormal event information collected by a road network monitoring system, extracting event type, occurrence position and event level attributes from the event information, constructing a three-dimensional road network topology model based on an electronic road network map, and marking positions of buildings, overpasses and tunnels in the three-dimensional road network topology model to construct an obstacle distribution map, which includes: Reading an event information data stream from the road network monitoring system, establishing an event information buffer queue to store the data stream, parsing a data packet header of the data stream to extract an event code and a timestamp, matching a preset event type mapping table according to the event code to obtain an event type identifier, extracting the longitude and latitude coordinates of the event occurrence position and an event level value from a message body of the data stream, and writing the event type identifier, the longitude and latitude coordinates, and the event level value into an event attribute table; Reading vector data of the electronic road network map, constructing a road network skeleton line matrix based on the vector data, calculating three-dimensional spatial coordinates of road network nodes to generate a road network node set, constructing the three-dimensional road network topology model by using the road network skeleton line matrix and the road network node set, retrieving building contour data, overpass elevation data and tunnel boundary data from a geographic information database, converting the contour data, the elevation data and the boundary data into an obstacle boundary point set, and marking the obstacle boundary point set in the three-dimensional road network topology model to generate an obstacle distribution map.

3. The UAV mission scheduling method for intelligent road network inspection according to claim 1, characterized in that, The process of obtaining real-time location information, remaining range, and payload status information of each drone from the drone management platform, calculating the task fitness score of each drone reaching the event location via the patrol route, and selecting the drone with the highest task fitness score as the target drone includes: The system receives telemetry data packets from each drone from the drone management platform, parses the telemetry data packets to obtain the drone identification code, reads the corresponding drone's position coordinates, remaining battery power, flight speed, and mission payload type based on the identification code, calculates the remaining flight range based on the remaining battery power, determines the camera specifications and image transmission equipment performance of the drone based on the mission payload type, and writes the position coordinates, remaining flight range, flight speed, camera specifications, and image transmission equipment performance into the drone status information table. A task fitness score calculation model is established, using the straight-line distance between the UAV's location coordinates and the location of the event, the ratio of the UAV's remaining flight range to the length of the patrol route, the matching degree between the UAV's flight speed and the expected patrol duration, and the adaptability of the camera specifications to the patrol task requirements as scoring indicators. The weight coefficients of each scoring indicator are determined based on the analytic hierarchy process (AHP). The task fitness score is obtained by weighting the scoring indicators with the weight coefficients, and the UAV with the highest task fitness score is selected as the target UAV.

4. The UAV mission scheduling method for intelligent road network inspection according to claim 1, characterized in that, The step of extracting patrol rules from the task template library based on the event type, and issuing the patrol rules and the obstacle avoidance path to the target drone to execute the patrol task includes: The system reads the preset mapping relationship between event types and patrol rules in the task template library, matches the corresponding patrol rule identifier based on the event type, and extracts the patrol start point coordinates, patrol end point coordinates, flight route sampling interval, flight altitude parameters, flight speed parameters, camera pitch angle parameters, and image acquisition frequency parameters according to the patrol rule identifier. The system then combines the patrol rule parameters with the event occurrence location information and the target UAV identification code to generate a task configuration file. A drone mission control instruction set is constructed. The parameters in the mission configuration file are encoded according to the communication protocol specification to generate mission initialization instructions, flight path setting instructions, obstacle avoidance path instructions, and image acquisition instructions. A data communication link with the target drone is established. The control instruction set is sent to the target drone through the data communication link. The instruction verification code returned by the target drone is received, and the verification code is verified to confirm the reliability of the mission instructions.

5. The UAV mission scheduling method for intelligent road network inspection according to claim 1, characterized in that, The system sends tiered early warning information to electronic information boards and LED displays within the distribution area, pushes different levels of early warning information to vehicle terminals based on the distance of the vehicle from the location of the incident, and updates the content of the early warning information in real time according to the progress of the incident handling, including: A graded template for early warning information is constructed. The threshold for early warning level is divided based on the straight-line distance between the vehicle and the location of the incident. The communication address and device type identifier of each device in the distribution area of ​​roadside equipment are read. The corresponding information release format is matched according to the device type identifier. The early warning information content is generated into device control instructions according to the information release format. Communication connection with each device is established. The device control instructions are sent to the corresponding electronic information board and LED display screen. The instruction response confirmation information returned by the device is received. The system obtains real-time location data of vehicles from the vehicle-to-everything (V2X) platform, calculates the straight-line distance between each vehicle and the location of the event, compares the straight-line distance with a warning level threshold to determine the warning information level, selects the corresponding information push template based on the warning information level, writes the event handling status information into the information push template to generate a vehicle terminal prompt message, pushes the prompt message to the vehicle terminal of the target vehicle through the V2X communication gateway, monitors the progress of event handling, and regenerates the warning information content when a progress update is received.

6. A drone task scheduling device for intelligent road network patrol, characterized in that, The device includes: The path planning module receives traffic accident, natural disaster, and road surface anomaly information collected by the road network monitoring system. It extracts event type, location, and event level attributes from the event information, constructs a 3D road network topology model based on the road network electronic map, marks the locations of buildings, overpasses, and tunnels in the 3D road network topology model to create an obstacle distribution map, obtains the center coordinates of the event location, constructs a circular search space centered on these center coordinates, divides the circular area into grid cells, calculates the center point coordinates of each grid cell, determines whether the center points of each grid cell overlap with obstacle boundary points based on the obstacle distribution map, and selects the center points of grid cells that do not overlap with obstacles as a candidate return point set. Finally, it calculates the distance from each return point in the candidate return point set to the event location. The distance and altitude difference are used to sort the candidate return-to-home points from nearest to farthest, and the return-to-home points with the highest ranking are selected as emergency return-to-home points. A three-dimensional coordinate system of the road network topology map is constructed, and the coordinates of the emergency return-to-home points and the location of the event are projected onto the three-dimensional coordinate system. The initial flight path from the emergency return-to-home point to the location of the event is calculated based on the A-star search algorithm. The boundary point data of the obstacle distribution map is read, and the collision detection distance between the initial flight path and the obstacle is calculated. When a collision risk is detected, the flight path nodes are adjusted to generate an obstacle avoidance segment. The obstacle avoidance segment is combined with the flight path segment that has not collided to generate a patrol route. In the constructed virtual flight environment, the safety of the patrol route is evaluated using the Monte Carlo method, and the patrol route is smoothed according to the turning radius and climb angle constraints of the UAV. The task inspection module is used to obtain real-time location information, remaining range, and payload status information of each UAV from the UAV management platform, establish a task fitness scoring calculation model, and use the straight-line distance between the UAV's position coordinates and the event location, the ratio of the UAV's remaining range to the inspection route length, the matching degree between the UAV's flight speed and the expected inspection duration, and the adaptability of the camera specifications to the inspection task requirements as scoring indicators. The weight coefficients of each scoring indicator are determined based on the analytic hierarchy process (AHP), and the task fitness score of each UAV is obtained by weighting the scoring indicators and the weight coefficients. The UAV with the highest task fitness score is selected as the target UAV. Inspection rules are extracted from the task template library based on the event type, and the inspection rules and obstacle avoidance paths are sent to the target UAV to execute the inspection task. The payload status information includes at least the camera's resolution, focal length, and aperture, and also at least the bandwidth, latency, and anti-interference capability of the image transmission device. The anomaly alarm module is used to collect image data acquired by the target drone during its patrol, perform target detection and scene recognition on the image data, extract event impact range information, and determine the distribution area of ​​roadside equipment requiring linkage based on the event impact range information. Specifically, it includes: receiving image data streams transmitted from the target drone; parsing image frames and shooting timestamps from the image data stream; storing the image frames and shooting timestamps in an image cache queue; reading the target drone's position data to record the shooting position coordinates of the image frames; performing geometric correction and image enhancement processing on the image frames; using a target detection model to identify vehicle distribution density, road traffic status, and traffic control facilities in the image; and extracting the accident scene range, vehicle backlog length, and road blockage position based on a scene recognition model. The system generates a target detection result table and a scene recognition result table. It then fuses the spatial location information from these tables to construct the boundary contour of the event's impact range. The geometric center and coverage radius of this boundary contour are calculated. A local coordinate system is established with the geometric center as the origin. The locations of roadside information boards, displays, and broadcasting equipment are marked in this local coordinate system. The distance from each device to the geometric center is calculated. Devices with a distance less than the coverage radius are selected to generate a roadside equipment distribution area. Tiered warning information is sent to the electronic information boards and LED displays within this distribution area. Different levels of warning information are pushed to vehicle terminals based on the distance between the vehicle and the event location. The content of the warning information is updated in real time according to the progress of the event handling.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the UAV task scheduling method for intelligent road network patrol as described in any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the UAV mission scheduling method for intelligent road network patrol as described in any one of claims 1 to 5.

Images