Unmanned aerial vehicle cluster collaborative inspection route planning method and device

Abstract

The invention relates to a route planning method and device for unmanned aerial vehicle cluster collaborative inspection, and the method comprises the steps: obtaining the geographic information of a rail transit control and protection region, and constructing an environment model based on the geographic information; combining the environment model and the telemetry data of each unmanned aerial vehicle to construct an unmanned aerial vehicle individual twin with a motion state so as to obtain the pose of each unmanned aerial vehicle; constructing a corresponding cluster relation graph in combination with the pose and preset cooperation logic; and in combination with the environment model, the pose and the cluster relation graph, constructing a twin scene meeting a preset interaction condition, so as to output a route planning path of each unmanned aerial vehicle based on the twin scene. Therefore, the technical problems that in the related technology, the inspection coverage capacity of a single unmanned aerial vehicle is insufficient, multiple unmanned aerial vehicles lack effective cooperation, and balance between application requirements and economical efficiency is difficult to achieve are solved.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) technology, and in particular to a method and apparatus for route planning in collaborative inspection of UAV swarms. Background Technology

[0002] The rail transit control and protection zone plays a crucial role in ensuring the safe operation of rail transit, encompassing a complex array of equipment and facilities, including ground stations and ground lines, elevated stations, elevated line structures, underground stations, and tunnel structures.

[0003] Traditional inspections of controlled areas in rail transit mainly rely on manual inspections; however, this method suffers from low efficiency, limited coverage, and poor real-time performance. In contrast, drone monitoring offers advantages such as high efficiency, flexibility, and wide coverage.

[0004] However, while a single drone can compensate for the shortcomings of manual inspection to some extent, its endurance is limited. When dealing with complex routes, a single drone often cannot provide full coverage. Furthermore, when multiple drones operate simultaneously, effective collaboration is lacking, making it difficult to balance application needs with cost-effectiveness, which urgently requires improvement. Summary of the Invention

[0005] This application provides a method and apparatus for route planning in collaborative inspection of a swarm of unmanned aerial vehicles (UAVs), in order to solve the technical problems in related technologies, such as insufficient inspection coverage of a single UAV and lack of effective cooperation among multiple UAVs, making it difficult to achieve a balance between application requirements and economic efficiency.

[0006] The first aspect of this application provides a method for route planning in collaborative inspection of a swarm of unmanned aerial vehicles (UAVs), comprising the following steps: acquiring geographic information of a rail transit control zone and constructing an environment model based on the geographic information; combining the environment model and telemetry data of each UAV to construct a UAV twin with motion state to obtain the pose of each UAV; combining the pose and preset collaboration logic to construct a corresponding swarm relationship graph; combining the environment model, the pose, and the swarm relationship graph to construct a twin scene that satisfies preset interaction conditions, so as to output the route planning path of each UAV based on the twin scene.

[0007] Optionally, in one embodiment of this application, before constructing the corresponding cluster relationship graph by combining the pose and the preset cooperation logic, the method further includes: obtaining inspection task information of the UAV cluster; evaluating the status of each UAV using the telemetry data to obtain an evaluation result; and constructing the preset cooperation logic by combining the inspection task information and the evaluation result.

[0008] Optionally, in one embodiment of this application, after outputting the flight path planning path of each UAV based on the twin scenario, the method further includes: updating the telemetry data of each UAV to determine whether at least one UAV meets the preset obstacle conditions using the updated telemetry data; if at least one UAV meets the preset obstacle conditions, then using a preset fast exploration random tree-connection strategy to perform local path replanning to obtain a locally replanned path, so as to control the at least one UAV to inspect along the locally replanned path.

[0009] Optionally, in one embodiment of this application, after obtaining the local replanning path, the method further includes: sharing the local replanning path using a cluster communication network; marking corresponding obstacle information in the environment model based on the local replanning path; and adjusting the flight path planning paths of other UAVs besides the at least one UAV based on the local replanning path and the obstacle information.

[0010] Optionally, in one embodiment of this application, the method further includes: obtaining the communication link status of the cluster communication network; locating a faulty node based on the communication data of the cluster communication network when the communication link status does not meet preset communication conditions; and switching a communication protocol or communication path that meets preset backup conditions based on the faulty node, so that the link status meets the preset communication conditions.

[0011] Optionally, in one embodiment of this application, the method further includes: using the telemetry data and a pre-built energy consumption prediction model of the UAV swarm to predict the power consumption of each UAV; based on the power consumption, planning a return or charging path for any UAV that meets the preset charging conditions, and handing over the inspection task of any UAV based on the swarm relationship graph.

[0012] A second aspect of this application provides a route planning device for collaborative inspection of a swarm of unmanned aerial vehicles (UAVs), comprising: a first construction module for acquiring geographic information of a rail transit control zone and constructing an environment model based on the geographic information; a second construction module for combining the environment model and telemetry data of each UAV to construct a UAV twin with motion state to obtain the pose of each UAV; a third construction module for combining the pose and preset collaboration logic to construct a corresponding swarm relationship graph; and a planning module for combining the environment model, the pose, and the swarm relationship graph to construct a twin scene that satisfies preset interaction conditions, so as to output the route planning path of each UAV based on the twin scene.

[0013] Optionally, in one embodiment of this application, it further includes: a first acquisition module, used to acquire inspection task information of the UAV cluster; an evaluation module, used to evaluate the status of each UAV using the telemetry data and obtain an evaluation result; and a fourth construction module, used to combine the inspection task information and the evaluation result to construct the preset collaboration logic.

[0014] Optionally, in one embodiment of this application, it further includes: a judgment module, used to update the telemetry data of each UAV to determine whether at least one UAV meets the preset obstacle conditions using the updated telemetry data; and a replanning module, used to perform local path replanning using a preset fast exploration random tree-connection strategy when at least one UAV meets the preset obstacle conditions, to obtain a local replanned path, so as to control the at least one UAV to inspect along the local replanned path.

[0015] Optionally, in one embodiment of this application, the replanning module further includes: a sharing unit for sharing the local replanning path using a cluster communication network; a marking unit for marking corresponding obstacle information in the environment model based on the local replanning path; and an adjustment unit for adjusting the flight path planning paths of other UAVs besides the at least one UAV based on the local replanning path and the obstacle information.

[0016] Optionally, in one embodiment of this application, it further includes: a second acquisition module, used to acquire the communication link status of the cluster communication network; a positioning module, used to locate a faulty node based on the communication data of the cluster communication network when the communication link status does not meet the preset communication conditions; and a switching module, used to switch to a communication protocol or communication path that meets the preset backup conditions based on the faulty node, so that the link status meets the preset communication conditions.

[0017] Optionally, in one embodiment of this application, it further includes: a prediction module, used to predict the power consumption of each UAV using the telemetry data and a pre-built energy consumption prediction model of the UAV swarm; and a return-to-home planning module, used to plan a return-to-home or charging path for any UAV that meets preset charging conditions based on the power consumption, and to hand over the inspection task of any UAV based on the swarm relationship graph.

[0018] A third aspect of 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 route planning method for collaborative inspection of a drone swarm as described in the above embodiments.

[0019] A fourth aspect of this application provides a computer-readable storage medium storing computer instructions for causing the computer to execute the route planning method for collaborative inspection of a drone swarm as described in the above embodiments.

[0020] A fifth aspect of this application provides a computer program product, including a computer program that, when executed, is used to implement the above-mentioned method for flight path planning in collaborative inspection of unmanned aerial vehicle (UAV) swarms.

[0021] This application embodiment can construct an environmental model based on the geographic information of the rail transit control and protection area, and further combine it with the telemetry data of each UAV to construct individual UAV twins with motion states, obtaining the pose of each UAV. Then, according to the preset cooperation logic, a corresponding cluster relationship graph is constructed, thereby obtaining a twin scene that meets the preset interaction conditions. Based on the twin scene, the flight path planning path of each UAV is output. By constructing twin scenes based on the relationship between the environment, cluster pose, and cluster relationships, cooperation between multiple UAVs is realized, maintaining the order of the cluster while ensuring inspection coverage and economy. Thus, it solves the technical problem in related technologies that the inspection coverage of a single UAV is insufficient, while multiple UAVs lack effective cooperation, making it difficult to achieve a balance between application needs and economy.

[0022] Additional aspects and advantages of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of this application. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a flowchart of a route planning method for collaborative inspection of unmanned aerial vehicle (UAV) swarms according to an embodiment of this application; Figure 2 This is a schematic diagram illustrating the principle of a flight path planning method for collaborative inspection of unmanned aerial vehicle (UAV) swarms according to an embodiment of this application. Figure 3 This is a flowchart of a method for route planning in collaborative inspection of a drone swarm according to an embodiment of this application; Figure 4 Here is a flowchart of a cluster cooperative collision avoidance process according to an embodiment of this application; Figure 5 This is a flowchart of a dynamic task refactoring process according to an embodiment of this application; Figure 6 This is a schematic diagram of a route planning device for collaborative inspection of unmanned aerial vehicle (UAV) swarms according to an embodiment of this application. Figure 7 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application. Detailed Implementation

[0024] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0025] The following describes a method and apparatus for flight path planning in collaborative inspection of a drone swarm, based on embodiments of the present application, with reference to the accompanying drawings. Addressing the technical problems mentioned in the background art, such as insufficient inspection coverage of a single drone and a lack of effective collaboration among multiple drones, making it difficult to balance application requirements and economic efficiency, this application provides a flight path planning method for collaborative inspection of a drone swarm. In this method, an environmental model is constructed based on the geographical information of the rail transit control zone. Furthermore, by combining the telemetry data of each drone, a drone twin with motion states is constructed to obtain the pose of each drone. Then, according to preset collaboration logic, a corresponding swarm relationship graph is constructed, resulting in a twin scene that meets preset interaction conditions. The flight path planning path for each drone is output based on the twin scene. By constructing a twin scene based on the relationship between the environment, swarm pose, and swarm relationships, collaboration among multiple drones is achieved, maintaining swarm order while ensuring inspection coverage and economic efficiency. This solves the technical problem in the related art where the inspection coverage of a single drone is insufficient, and multiple drones lack effective collaboration, making it difficult to balance application requirements and economic efficiency.

[0026] Understandably, the controlled and protected areas of rail transit, as a crucial component of urban rail transit, are key to daily inspection work. With the continuous advancement of urban construction, external work activities that threaten the safety boundaries of rail transit are increasing. In this context, conducting daily inspections of the controlled and protected areas to prevent the subway from being affected by external construction becomes particularly important. Currently, the inspection work of the controlled and protected areas is gradually shifting from manual to unmanned intelligent inspection, utilizing equipment such as cameras, robots, and drones to achieve digital and intelligent inspection of the controlled and protected areas.

[0027] In the era of digitalization and intelligentization, digital twin technology, with its unique characteristics such as high simulation and real-time mapping, has rapidly become a powerful tool for various industries to achieve transformation and upgrading. In the field of inspection management, the application of digital twin technology has significantly improved inspection efficiency and effectively ensured the safe operation and maintenance of subway facilities. However, current inspection platforms based on digital twins still have significant shortcomings in setting inspection routes. Inspection routes set based on habits or experience are difficult to accurately match the actual conditions of the controlled and protected areas, greatly restricting the full release of system effectiveness.

[0028] Based on this, and considering the unique characteristics of urban rail transit lines, an intelligent inspection route planning system was designed and developed. This system can utilize big data technology to store and analyze massive amounts of data based on a 3D scene model of the controlled area along the rail line, providing richer and more accurate information support for setting inspection routes. Furthermore, it leverages artificial intelligence technology to autonomously optimize and dynamically adjust inspection routes, automatically generating the optimal inspection route, thereby improving the efficiency and quality of inspections.

[0029] However, in related technologies, there is a lack of effective dynamic collision avoidance mechanisms when conducting collaborative operations with drone swarms. How to achieve intelligent route planning within the controlled and protected areas of rail transit has become a pressing problem to be solved. In the complex environment of rail transit controlled and protected areas, there are various obstacles such as buildings and signal towers, and it is also necessary to avoid interfering with rail transit operations. Traditional route planning methods are difficult to meet the actual application requirements.

[0030] Furthermore, from a technical perspective, while digital twin technology enables real-time modeling and data collection from equipment, current systems are relatively weak in deep data analysis and intelligent decision-making. Digital twin modules face data compatibility issues when collecting this data, struggling to effectively integrate heterogeneous data from multiple sources. They cannot accurately process and clean operational route data, failing to provide reliable data support for automatic route planning, resulting in planned routes that do not meet actual inspection needs. Simultaneously, the lack of effective algorithms to intelligently correlate the actual condition of the track with the inspection route means that route optimization lacks a scientific basis. From a management perspective, enterprises' understanding and application of digital twin technology is not deep enough, still employing traditional inspection management thinking and failing to fully recognize the enormous potential of digital twin technology in dynamic inspection route planning. In addition, the imperfect data sharing and collaboration mechanisms between different departments also affect the scientific and rational nature of inspection route setting.

[0031] To address the aforementioned issues, the flight path planning method for collaborative inspection of drone swarms in this application can construct more accurate twin scenarios and fully leverage the advantages of drone swarm collaboration and digital twin technology to improve inspection efficiency and quality.

[0032] Specifically, Figure 1 This is a flowchart illustrating a method for route planning in collaborative inspection of a drone swarm provided in an embodiment of this application.

[0033] like Figure 1 As shown, the route planning method for collaborative inspection by this drone swarm includes the following steps: In step S101, the geographic information of the rail transit control and protection area is obtained, and an environmental model is constructed based on the geographic information.

[0034] To achieve digital and intelligent inspection, the first step in this application embodiment is to collect geographic information of the rail transit control and protection area that needs to be inspected in order to construct an environmental model.

[0035] For example, embodiments of this application can utilize technologies such as lidar, satellite imagery, or maps obtained from open-source platforms to acquire information on the terrain, buildings, and rail facilities of the protected area, perform data extraction, coordinate system transformation, and quality verification, in order to construct a three-dimensional spatial structure model of the rail transit protected area, i.e., an environmental model.

[0036] For example, when building a model, the detailed structural data of the building information model can be combined with the spatial analysis capabilities of the geographic information system to achieve integrated modeling of rail transit facilities (such as stations and tunnels) and their surrounding environment (such as pipelines and risk sources), resulting in a static environment model. This allows for three-dimensional visualization of buildings and structures along the subway line, supporting collision detection and construction simulation.

[0037] Based on this, the embodiments of this application can also embed IoT sensor interfaces in the environment model to access the real-time operating status (temperature, voltage, switch status) of track equipment (such as signal lights, power supply equipment) to output a dynamic environment twin (static structure + real-time equipment status).

[0038] An environment model is constructed using a static environment model and a dynamic environment twin.

[0039] In step S102, by combining the environmental model and the telemetry data of each UAV, an individual UAV twin with motion state is constructed to obtain the pose of each UAV.

[0040] Furthermore, embodiments of this application can create a dedicated digital twin for each drone, mapping its position, attitude, battery level, and other parameters in real time.

[0041] For example, in this application embodiment, a unique ID can be assigned to each drone, and pose, power, and speed data can be transmitted back in real time through onboard sensors (GPS / IMU / battery module) to drive the movement of the virtual twin.

[0042] This application embodiment can utilize a dynamic environment model (for collision detection benchmark) and real-time telemetry data from the UAV to obtain a UAV individual twin with motion state (position matrix, attitude angle, remaining battery power).

[0043] In step S103, the corresponding cluster relationship graph is constructed by combining the pose and the preset collaboration logic.

[0044] The embodiments of this application can construct a dynamic graph network, i.e. a cluster relationship graph, based on the real-time pose of the UAV and the preset cooperation logic, according to the distance and signal strength between the UAVs. This can include each UAV as a node, communication links as edges, and task status, etc.

[0045] The collaboration logic can store task allocation relationships (such as Leader-Follower), collision avoidance rules, and data relay paths in a graph database (such as Neo4j).

[0046] Optionally, in one embodiment of this application, before constructing the corresponding cluster relationship graph by combining the pose and preset cooperation logic, the method further includes: obtaining inspection task information of the UAV cluster; evaluating the status of each UAV using telemetry data to obtain evaluation results; and constructing preset cooperation logic by combining the inspection task information and evaluation results.

[0047] In some embodiments, efficient operation can be achieved by adjusting the task allocation mechanism. Based on reinforcement learning algorithms, combined with equipment status assessment and inspection targets, a preliminary task allocation strategy is generated, delineating the UAV inspection area and targets. An auction algorithm is introduced, allowing UAVs to "bid" for tasks based on their own endurance, payload, and other capabilities, achieving dynamic optimization allocation. In addition, priority coordination rules are established. When facing urgent tasks, resources and progress are quickly assessed through a digital twin model, and task arrangements are flexibly adjusted to ensure timely inspection.

[0048] By combining task allocation relationships, collision avoidance rules, and data relay paths, the embodiments of this application can construct corresponding collaborative logic so that the UAV swarm can adapt to various inspection conditions.

[0049] In step S104, a twin scene that meets preset interaction conditions is constructed by combining the environment model, pose and cluster relationship graph, so as to output the flight path planning path of each UAV based on the twin scene.

[0050] By combining the constructed environment model, UAV pose, and cluster relationship graph, the model accuracy is automatically switched according to the viewing distance, prioritizing the real-time update of the UAV cluster dynamic data. Under the premise of using block loading for the static environment, the embodiments of this application can construct a lightweight twin scene that can be interacted in real time.

[0051] Based on this, the embodiments of this application can adopt a hierarchical planning strategy to collaboratively plan routes at both the global and local levels. At the global level, an improved multi-agent path planning algorithm (such as MADDPG) is used, combined with environmental information from the digital twin model and task allocation results, to plan an initial global route that is conflict-free and meets the requirements.

[0052] Optionally, in one embodiment of this application, after outputting the flight path planning path for each UAV based on the twin scenario, the method further includes: updating the telemetry data of each UAV to determine whether at least one UAV meets the preset obstacle conditions using the updated telemetry data; if at least one UAV meets the preset obstacle conditions, then using a preset fast exploration random tree-connection strategy to perform local path replanning to obtain a locally replanned path, so as to control at least one UAV to inspect along the locally replanned path.

[0053] At the local level, when encountering dynamic obstacles, embodiments of this application can utilize the Rapid Exploration Random Tree-Connect (RRT-Connect) algorithm for local path replanning and share adjustment information through a cluster communication network to achieve collaborative obstacle avoidance among UAVs and maintain orderly cluster flight.

[0054] Optionally, in one embodiment of this application, after obtaining the local replanning path, the method further includes: sharing the local replanning path using a cluster communication network; marking the corresponding obstacle information in the environment model based on the local replanning path; and adjusting the flight path planning paths of other UAVs, excluding at least one UAV, based on the local replanning path and the obstacle information.

[0055] It is understandable that the data collected by the drone during the inspection process is constantly changing, and the environment of the rail transit control and protection area is also dynamically changing. Therefore, in this embodiment of the application, after at least one drone detects an obstacle, it can use the cluster communication network in the cluster relationship graph to share data in a timely manner and mark the obstacle in the environment model, so that other drones can adjust their flight path planning according to the marked obstacle to avoid the obstacle, or increase the inspection of the obstacle's blind spot, etc.

[0056] Optionally, in one embodiment of this application, the method further includes: obtaining the communication link status of the cluster communication network; locating the faulty node based on the communication data of the cluster communication network when the communication link status does not meet the preset communication conditions; and switching the communication protocol or communication path that meets the preset backup conditions based on the faulty node so that the link status meets the preset communication conditions.

[0057] In some embodiments, a digital twin communication network model can be constructed using a cluster relationship graph to monitor the status of communication links in real time, quickly locate nodes in case of failure, and automatically switch to backup protocols or paths to ensure uninterrupted communication. For UAV malfunctions, a diagnostic and emergency response mechanism is established to monitor operational status in real time, assess the impact range when a malfunction occurs, activate contingency plans, issue safe landing or return-to-base instructions, and reassign tasks to ensure that inspection missions are not significantly affected.

[0058] Optionally, in one embodiment of this application, the method further includes: predicting the power consumption of each drone using telemetry data and a pre-built energy consumption prediction model of the drone swarm; planning a return or charging path for any drone that meets the preset charging conditions based on the power consumption; and handing over the inspection task of any drone based on the swarm relationship graph.

[0059] In other embodiments, digital twin technology can be used to achieve dynamic monitoring and management of drone swarm resources and energy. An energy consumption prediction model is established, which predicts power consumption by considering factors such as flight speed, flight path, and payload. When power is insufficient, the optimal return or charging route is calculated, and task handover is coordinated. Simultaneously, payload resources are dynamically managed, rationally allocating camera shooting time and data storage capacity according to task requirements to improve resource utilization efficiency.

[0060] Combination Figures 2 to 5 As shown, the working principle of the unmanned aerial vehicle (UAV) swarm collaborative inspection route planning method of this application embodiment is explained in detail with an example.

[0061] like Figure 2 As shown in the embodiments of this application, data information of the track protection zone, i.e. the rail transit control zone, can be obtained, and a digital twin can be constructed by combining the corresponding environmental model and UAV data for route planning.

[0062] The physical world can be understood as the actual environment of the rail transit control and protection zone, encompassing tracks, bridges, tunnels, overhead contact lines, signaling equipment, and the swarms of drones flying within them. It serves as the source of data acquisition and the object of final command execution.

[0063] A digital twin is a high-fidelity, dynamic virtual mapping of the physical world, constructed using multi-source data such as integrated oblique photogrammetry (GIS), building information modeling (BIM), and laser point clouds. It can be continuously updated by receiving real-time data from the physical world (such as drone location, sensor readings, and meteorological information). It provides a highly realistic virtual sandbox for flight path planning and simulation testing.

[0064] The data from a digital twin can be used in intelligent algorithm models to complete cluster route planning, and can include three modules: The cluster task allocation module can be used to decompose the overall inspection task and allocate it to each drone in the cluster in an optimal way. For example, it can use an auction algorithm or a genetic algorithm to comprehensively consider the matching degree between the drone's location, battery power, payload capacity and sub-task area.

[0065] During the execution of a drone mission, the embodiments of this application can track its actual flight trajectory in real time and compare it with the planned route. Typically, Kalman filtering or model predictive control (MPC) algorithms are combined to achieve precise flight control and provide real-time data support for dynamic collision avoidance.

[0066] The intelligent algorithm model bases its decision-making on optimization objectives and logical rules. The cost function is a mathematical expression used to evaluate the merits of different flight routes, typically in the form of: Minimize(Total Flight Time + α * Total Energy Consumption + β * Safety Risk Value). The algorithm seeks the optimal solution by minimizing this function value. Reasoning is a logical judgment process based on preset rules (such as "within 3 meters of the overhead contact line is a no-fly zone") and real-time data.

[0067] This application embodiment can further divide the entire control and protection area into multiple logical sub-task areas based on the specific boundary range of the track control and protection zone and the actual spatial distribution of facilities along the line. This division method can effectively support multiple groups of UAVs to carry out inspection tasks in parallel, laying an important structural foundation for subsequent collaborative operations.

[0068] This application embodiment can comprehensively analyze each sub-region based on intelligent AI algorithms, and dynamically allocate the most suitable UAV execution units according to the region area, task complexity, and UAV performance parameters. This allocation strategy aims to achieve overall system load balancing, thereby improving task execution efficiency and reducing resource idle rates.

[0069] The dynamic collision avoidance planning module can be used to ensure the safety of drones with obstacles (such as overhead contact lines) and other drones. It generates a spatio-temporal corridor, allocating specific flight space and time windows for each drone to avoid conflicts at the source.

[0070] Furthermore, in addition to three-dimensional waypoints, time-dimensional constraints can be introduced to generate four-dimensional waypoints (i.e., "time control points") with strict temporal attributes. By constructing this flight path structure that integrates spatiotemporal information, key support is provided for achieving precise time synchronization and action coordination in large-scale UAV swarms.

[0071] Furthermore, when the UAV detects an unplanned near-field conflict risk during flight using its onboard sensors, this embodiment of the application can also utilize an inter-aircraft communication network to conduct real-time negotiation and autonomous decision-making based on a preset collision avoidance rule set (e.g., using the Boids group behavior model) to achieve dynamic obstacle avoidance and course adjustment.

[0072] The environmental adaptation decision module can be used to handle external dynamic changes, such as meteorological fusion adjustment (adjusting flight altitude and speed based on real-time wind speed and rainfall) and emergency situation statistics (such as temporary train additions or sudden obstacles), and make flight route adjustment decisions.

[0073] As the core output of the environmental adaptive decision-making model, the environmental adaptive decision-making module can integrate meteorological monitoring data (such as strong winds and rainfall) in real time and dynamically optimize flight routes. At the same time, it can activate statistical sensing and emergency response mechanisms to trigger customized inspection plans in response to sudden abnormal events (including foreign object intrusion and equipment malfunctions).

[0074] By integrating the outputs of all the above functional modules, this embodiment of the application can generate a system-level cluster flight path package. This data package contains the four-dimensional flight path information of all UAVs and is the final carrier form of the collaborative operation scheme.

[0075] Furthermore, the embodiments of this application can perform high-fidelity simulated flight tests on cluster flight routes, reproducing actual flight conditions in a virtual environment, which becomes the most critical safety and feasibility verification step before actual commands are issued.

[0076] For example, embodiments of this application can perform performance and collision detection evaluations. During the simulation process, comprehensive collision detection analysis (including distance monitoring between the UAV and static obstacles, and dynamic spacing control between UAVs) and comprehensive performance evaluation (covering indicators such as area coverage and endurance margin) are carried out. If the evaluation results do not meet the preset standards, they are fed back to the intelligent planning model for iterative optimization, forming a self-correcting closed loop.

[0077] The fully validated flight path command packets will be sent to the actual drone swarm via a highly reliable communication network (such as a 5G private network) to ensure the accurate transmission and synchronous execution of control commands.

[0078] During flight, the drone transmits real-time status data (including position, attitude, battery level, etc.) and collected inspection data (such as images, videos, and laser point clouds). This data is used to dynamically update the digital twin, maintaining consistency and real-time performance between the virtual model and the physical entity, while also providing support for subsequent intelligent data analysis and decision optimization.

[0079] When planning flight routes, this embodiment of the application needs to perform cluster task allocation, dynamic collision avoidance planning and environmental adaptation decision-making to output cluster flight route packages and perform twin simulation verification. After the verification is passed, the UAV is controlled to perform inspections along the flight route trajectory and the data is transmitted back during the inspection process to update the digital twin, thereby optimizing the flight route trajectory.

[0080] like Figure 3 As shown, embodiments of this application may include the following steps: Step S301: Construct a digital twin that integrates geographic information, facility models, and dynamic constraints.

[0081] This application's embodiments can construct a multi-dimensional digital twin model covering the environment, equipment, and drone swarms. On one hand, it accurately presents the three-dimensional spatial structure and equipment status of scenarios such as rail transit control zones; on the other hand, it establishes a dedicated digital twin for each drone, mapping its position, attitude, battery level, and other parameters in real time. Simultaneously, it constructs a swarm collaboration model, simulating communication links and collaboration modes between drones, and uses a graph database to store these relationships. Through Level of Detail (LOD) technology, the model is lightweighted, balancing data accuracy and computational resource consumption, ensuring that the model can reflect the real-world scenario in real time and provide a reliable basis for subsequent planning.

[0082] By constructing a multi-dimensional digital twin model, the system comprehensively and in real-time maps the environment, equipment, and drone status. Through graph databases and LOD technology, the flight path is dynamically adjusted, resulting in higher accuracy and adaptability to complex scenarios.

[0083] Step S302: Divide the inspection area according to the track topology and use an auction algorithm to allocate drones and tasks.

[0084] This application's embodiments can generate a preliminary task allocation strategy based on reinforcement learning algorithms, combined with equipment status assessment and inspection targets, to delineate UAV inspection areas and targets. An auction algorithm is introduced, allowing UAVs to "bid" for tasks based on their own endurance, payload, and other capabilities, achieving dynamic and optimized allocation. Furthermore, priority coordination rules are established; in the face of urgent tasks, a digital twin model is used to quickly assess resources and progress, flexibly adjusting task arrangements to ensure timely inspections.

[0085] An allocation mechanism combining reinforcement learning and auction algorithms is adopted to improve the rationality of task allocation and to quickly adjust resources in the face of urgent tasks.

[0086] Step S303: Generate a 4D flight path with timestamps and use a spatiotemporal corridor algorithm to avoid aircraft cluster conflicts. Step S304: Verify the safety of the flight path in a twin simulation environment.

[0087] Step S305: Update the twin based on the real-time feedback data and optimize subsequent tasks.

[0088] like Figure 4 As shown, the embodiments of this application can adopt a hierarchical planning strategy to collaboratively plan flight routes at both the global and local levels. At the global level, an improved multi-agent path planning algorithm (such as MADDPG) is used, combined with environmental information from the digital twin model and task allocation results, to plan an initial global flight route that is conflict-free and meets the requirements. At the local level, when encountering dynamic obstacles, the Rapid Exploration Random Tree-Connect (RRT-Connect) algorithm is used for local path replanning, and adjustment information is shared through the swarm communication network to achieve collaborative obstacle avoidance among UAVs and maintain orderly swarm flight.

[0089] By employing a hierarchical collaborative algorithm, conflict-free routes are pre-planned globally, and dynamic obstacles are addressed locally, enabling information sharing and collaborative adjustments.

[0090] For example, the cluster-cooperative collision avoidance process in this application embodiment may include the following steps: Step S1: Using high-precision sensors and real-time communication networks, continuously acquire key status information such as speed, position and heading of each UAV in the cluster, and conduct real-time conflict detection and prediction analysis of future spatiotemporal conflicts based on current motion trends and changes in the external environment.

[0091] Step S2: If a potential spatiotemporal conflict risk is predicted, a centralized resolution mechanism will be activated immediately. Through the embodiments of this application, unified calculations will be performed and coordination instructions will be issued, such as allocating altitude layers or adjusting speeds for specific UAVs, thereby eliminating the possibility of conflict at the global level. If there is no conflict at present, or if an unforeseen local conflict is encountered suddenly during flight (such as the sudden appearance of a dynamic obstacle), the distributed collaborative negotiation mechanism is activated. Each UAV autonomously broadcasts and receives its own intentions through the communication network, and executes the corresponding avoidance strategy according to the established collision avoidance criteria, and finally completes the fine adjustment of the local trajectory.

[0092] It is important to note that regardless of whether a centralized or distributed self-consistent execution path is adopted, it is ultimately necessary to reliably maintain a safe distance in order to ensure the overall operational safety and efficiency of UAV swarm flights in complex mission environments. In conflict prediction, this application embodiment can rely on the global dynamic information provided by the digital twin system to perform forward-looking calculations and multi-dimensional analysis of the flight trajectories of all UAVs within the cluster. This is primarily based on the current motion state of the UAVs (including real-time position, velocity vector, and heading angle) and pre-planned flight paths. High-precision 4D trajectory extrapolation (including three-dimensional spatial coordinates and time dimension) is performed in the digital twin to simulate the flight process over a future period. By calculating the relative positional relationship between any two UAVs at future moments, it is determined whether their preset safety envelopes will overlap or intersect at a certain point in time, thereby achieving accurate prediction of "spatiotemporal conflict."

[0093] The result of conflict prediction may include the existence of spatiotemporal conflict or no conflict. Based on the result, the embodiments of this application can trigger the corresponding conflict resolution mechanism and enter different decision paths.

[0094] Among them, the centralized resolution strategy can perform centralized optimization calculations and uniformly issue instructions when a systemic and foreseeable spatiotemporal conflict is predicted. It is suitable for the route collaborative planning stage before the mission begins, or for dynamic adjustments to routes on a global scale during mission execution due to unforeseen circumstances (such as temporary train additions or updates to airspace occupancy).

[0095] The specific implementation methods are as follows: Altitude layer allocation assigns different flight altitude layers to drones with potential conflicts, achieving physical isolation in the vertical direction. For example, in areas with intersecting orbits or overlapping airspace, it is clearly stipulated that drones performing different tasks should fly within designated altitude ranges to avoid mutual interference.

[0096] Speed ​​adjustment strategies, by slightly adjusting the flight speed of one or more drones, alter their arrival time at the conflict point, thereby achieving staggered passage in the time dimension. This method is particularly suitable for resolving cross-route conflicts, improving overall safety while minimizing changes to the flight path.

[0097] The distributed negotiation mechanism enables rapid conflict resolution during actual drone flight for unplanned, sudden close-range conflicts (such as trajectory deviations caused by wind disturbances or sensor errors). It employs an autonomous negotiation method based on inter-drone communication. This mechanism primarily addresses uncertainties in dynamic environments, effectively complementing centralized control and significantly enhancing the robustness and real-time response capabilities of the entire drone swarm system. Its specific process includes the following stages: 1. Intent broadcasting: UAVs use self-organizing communication links (such as ultra-wideband, wireless direct connection or other low-latency transmission protocols) to broadcast their real-time status information, including position, speed and short-term flight intent (such as "maintain current heading" "prepare to climb").

[0098] 2. Rule-based Collision Avoidance: Each drone has a built-in collision avoidance rule base that conforms to airspace operation regulations. Upon receiving intent information from a neighboring drone, the system quickly matches and makes decisions based on the rule base. Common rules are as follows: Priority rule: Drones with higher mission priority have priority passage.

[0099] Avoidance rules: Drawing on road traffic rules, it is stipulated that in a scenario of relative motion, one should veer to the right to avoid a collision.

[0100] Height difference rule: stipulates that height differences are created by actively climbing or descending in order to avoid conflict.

[0101] 3. Trajectory Fine-tuning: Based on rule-based decisions, the current flight trajectory is adjusted locally and slightly to achieve safe and efficient conflict resolution without affecting the overall mission as much as possible. The UAV flight control system will execute a local, slight trajectory correction command. This adjustment is precise and rapid to cope with sudden interference or dynamic obstacles that may occur during flight.

[0102] 4. Maintaining a safe distance: The ultimate goal of the above operations is to ensure that, throughout the entire drone swarm's coordinated flight, the actual physical distance between any two adjacent or intersecting drones is always strictly greater than the system's preset minimum safety threshold. For example, within the flight control protection zone, this threshold is typically set to no less than 5 meters to adequately address the collision risks that may arise from various uncertainties.

[0103] like Figure 5 As shown, this application embodiment can establish an energy consumption prediction model, comprehensively predicting power consumption based on factors such as flight speed, route, and payload. When power is insufficient, it calculates the optimal return or charging route and coordinates mission handover. Simultaneously, it dynamically manages payload resources, rationally allocating camera shooting time and data storage capacity according to mission requirements, thereby improving resource utilization efficiency. Through dynamic resource and energy management, it accurately predicts energy consumption, allocates payload resources, and automatically plans return or charging routes, reducing losses and costs.

[0104] For three typical types of emergencies, the system leverages digital twin technology to achieve rapid, safe, and autonomous reconfiguration of drone swarm missions, fully demonstrating the system's high robustness, adaptability, and intelligent decision-making capabilities.

[0105] For example, when dealing with unexpected tasks, embodiments of this application can perform cluster status monitoring. Through multi-source sensing and data fusion, the internal operating status of the cluster can be continuously and accurately perceived, and equipment anomalies and performance degradation can be identified in a timely manner, providing a reliable basis for intelligent decision-making. Unexpected tasks may include: drone malfunction and battery power falling below a certain threshold.

[0106] When a drone in the cluster exits a mission due to malfunction or insufficient power, this embodiment of the application can automatically reassign its unfinished inspection sub-tasks to other available drones, ensuring mission continuity. By utilizing intelligent algorithms for rapid simulation and calculation in a digital twin, and comprehensively considering constraints such as the location of remaining drones, battery capacity, and payload capacity, the optimal takeover strategy from a nearby drone is selected to achieve seamless mission transfer.

[0107] The system dynamically monitors battery voltage and remaining capacity, sets a low battery threshold (e.g., 30%) as an early warning condition, and initiates the energy scheduling process in advance. For drones with low battery, the system activates an autonomous energy scheduling and replenishment mechanism, plans a highest-priority return route, guides the drone to the nearest charging station quickly, and incorporates the original mission into the global mission migration process, which is then taken over by other units, balancing energy security and mission efficiency.

[0108] In response to sudden weather changes, flight paths can be collectively avoided. When facing widespread weather threats (such as strong winds or thunderstorms), unified decision-making can be implemented to generate coordinated avoidance routes, minimizing the impact of environmental risks on the overall mission. For example, in a digital twin, hazardous areas can be dynamically delineated based on real-time weather data, generating collective detour paths or new routes for all affected drones to descend to a safe altitude, ensuring the overall operational safety of the swarm.

[0109] In response to emergency scenarios requiring airspace clearance, such as fires or medical emergencies, the mandatory clearance mechanism is activated to ensure that critical airspace is free of drone activity during specific time periods, effectively avoiding collision risks, achieving rapid and orderly release of airspace, and comprehensively ensuring flight safety.

[0110] For sudden weather changes and unexpected scenarios, the embodiments of this application can perform global replanning, that is, after decision-making, global-level re-optimization is initiated to achieve overall coordination of resource, task allocation and path planning. The intelligent planning central system is triggered, and based on the latest task allocation strategy and newly added avoidance constraints, the four-dimensional flight paths (spatial-temporal corridors) of all UAVs are comprehensively recalculated. Through multi-objective optimization algorithms, it is ensured that the newly generated flight path schemes still maintain optimal performance at the global level and achieve completely conflict-free operation.

[0111] Before the reconstructed flight path plan is officially issued, this application embodiment can also perform rapid simulation verification in a digital twin to ensure the safe operation of the entire UAV swarm. The key to this is to perform high-speed real-time simulation and deduction of the new flight path data package generated after global replanning. The simulation process includes collision detection (covering collisions between UAVs and collisions between UAVs and obstacles) and performance evaluation (verifying whether the UAVs can complete the task efficiently and whether they violate preset constraints).

[0112] Furthermore, embodiments of this application can rapidly send new flight path instructions, which have undergone rigorous simulation verification, to every drone in the entire drone swarm with extremely high speed and accuracy. By fully utilizing the low latency and high reliability communication characteristics unique to 5G private networks, flight path instruction data packets, which have undergone efficient compression and encryption processing, are simultaneously sent to all relevant drones, thereby ensuring that the behavior of the entire drone swarm is highly synchronized and globally consistent after receiving the instructions.

[0113] In summary, the embodiments of this application can determine weather and scene changes based on the returned data, triggering situations such as collective flight path avoidance or emergency airspace clearing. In such cases, global planning can be re-performed.

[0114] By leveraging communication and emergency fault handling mechanisms, we ensure command transmission, quickly diagnose faults, and reassign tasks, ensuring that tasks can still be executed in case of emergencies.

[0115] In summary, the embodiments of this application can be applied to a rail transit control zone inspection project in a certain city. During the project implementation, the drone swarm can efficiently complete the inspection tasks of the control zone according to the intelligently planned flight path. Through real-time transmitted images and data, multiple potential hazards and illegal construction activities in rail facilities were promptly discovered, providing strong protection for the safe operation of rail transit. At the same time, the successful implementation of this project also proves the effectiveness and practicality of the intelligent flight path planning method for drone swarms in rail transit control zones based on digital twins.

[0116] The route planning method for collaborative inspection of UAV swarms proposed in this application can construct an environmental model based on the geographical information of the rail transit control area. Furthermore, by combining the telemetry data of each UAV, a UAV twin with motion status is constructed to obtain the pose of each UAV. Then, according to preset collaboration logic, a corresponding swarm relationship graph is constructed, resulting in a twin scene that meets preset interaction conditions. Based on this twin scene, the route planning path for each UAV is output. By constructing twin scenes based on the relationships between the environment, swarm pose, and swarm relationships, collaboration between multiple UAVs is achieved, maintaining swarm order while ensuring inspection coverage and economy. This solves the technical problem in related technologies where the inspection coverage of a single UAV is insufficient, while multiple UAVs lack effective collaboration, making it difficult to balance application needs and economy.

[0117] Next, referring to the accompanying drawings, a route planning device for collaborative inspection of unmanned aerial vehicle (UAV) swarms according to an embodiment of this application is described.

[0118] Figure 6 This is a block diagram of a route planning device for collaborative inspection of a drone swarm, according to an embodiment of this application.

[0119] like Figure 6 As shown, the route planning device 10 for collaborative inspection of the UAV swarm includes: a first construction module 100, a second construction module 200, a third construction module 300, and a planning module 400.

[0120] Specifically, the first construction module 100 is used to acquire geographic information of the rail transit control and protection area and construct an environmental model based on the geographic information.

[0121] The second construction module 200 is used to combine the environmental model and the telemetry data of each UAV to construct an individual UAV twin with motion state in order to obtain the pose of each UAV.

[0122] The third construction module 300 is used to combine pose and preset collaboration logic to construct the corresponding cluster relationship graph.

[0123] The planning module 400 is used to combine the environment model, pose and cluster relationship graph to construct a twin scene that meets the preset interaction conditions, so as to output the flight path planning path of each UAV based on the twin scene.

[0124] Optionally, in one embodiment of this application, the route planning device 10 for collaborative inspection of UAV swarms further includes: a first acquisition module, an evaluation module, and a fourth construction module.

[0125] The first acquisition module is used to acquire inspection task information of the drone cluster.

[0126] The evaluation module is used to assess the status of each drone using telemetry data and obtain evaluation results.

[0127] The fourth module is used to combine inspection task information and evaluation results to build a preset collaborative logic.

[0128] Optionally, in one embodiment of this application, the route planning device 10 for collaborative inspection of UAV swarms further includes: a judgment module and a replanning module.

[0129] The judgment module is used to update the telemetry data of each drone in order to determine whether at least one drone meets the preset obstacle conditions.

[0130] The replanning module is used to perform local path replanning using a preset fast exploration random tree-connection strategy when at least one drone meets the preset obstacle conditions, so as to obtain a locally replanned path and control at least one drone to inspect along the locally replanned path.

[0131] Optionally, in one embodiment of this application, the replanning module further includes: a sharing unit, a marking unit, and an adjustment unit.

[0132] The shared unit is used to share local replanning paths using the cluster communication network.

[0133] The labeling unit is used to label the corresponding obstacle information in the environment model based on the local replanning path.

[0134] The adjustment unit is used to adjust the flight path planning of other drones, excluding at least one drone, based on local replanning path and obstacle information.

[0135] Optionally, in one embodiment of this application, the route planning device 10 for collaborative inspection of UAV swarms further includes: a second acquisition module, a positioning module, and a switching module.

[0136] The second acquisition module is used to acquire the communication link status of the cluster communication network.

[0137] The location module is used to locate faulty nodes based on communication data from the cluster communication network when the communication link status does not meet the preset communication conditions.

[0138] The switching module is used to switch to a communication protocol or communication path that meets preset backup conditions based on the faulty node, so that the link status meets the preset communication conditions.

[0139] Optionally, in one embodiment of this application, the route planning device 10 for collaborative inspection of UAV swarms further includes: a prediction module and a return-to-home planning module.

[0140] The prediction module is used to predict the power consumption of each drone using telemetry data and a pre-built energy consumption prediction model for the drone swarm.

[0141] The return-to-home planning module is used to plan the return-to-home or charging path for any drone that meets the preset charging conditions based on power consumption, and to hand over the inspection tasks of any drone based on the cluster relationship graph.

[0142] It should be noted that the explanation of the above-mentioned embodiment of the flight path planning method for collaborative inspection of UAV swarms also applies to the flight path planning device for collaborative inspection of UAV swarms in this embodiment, and will not be repeated here.

[0143] The flight path planning device for collaborative inspection of UAV swarms proposed in this application can construct an environmental model based on the geographical information of the rail transit control zone, and further combine it with the telemetry data of each UAV to construct individual UAV twins with motion states, obtaining the pose of each UAV. Then, according to preset cooperation logic, a corresponding swarm relationship graph is constructed, thereby obtaining a twin scene that meets preset interaction conditions. Based on the twin scene, the flight path planning path of each UAV is output. By constructing twin scenes based on the relationship between the environment, swarm pose, and swarm relationships, cooperation between multiple UAVs is realized, maintaining swarm order while ensuring inspection coverage and economy. Thus, it solves the technical problem in related technologies where the inspection coverage of a single UAV is insufficient, and multiple UAVs lack effective cooperation, making it difficult to achieve a balance between application needs and economy.

[0144] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: The memory 701, the processor 702, and the computer program stored on the memory 701 and executable on the processor 702.

[0145] When the processor 702 executes the program, it implements the route planning method for collaborative inspection of UAV clusters provided in the above embodiments.

[0146] Furthermore, electronic devices also include: Communication interface 703 is used for communication between memory 701 and processor 702.

[0147] The memory 701 is used to store computer programs that can run on the processor 702.

[0148] The memory 701 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0149] If the memory 701, processor 702, and communication interface 703 are implemented independently, then the communication interface 703, memory 701, and processor 702 can be interconnected via a bus to complete communication between them. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of representation, Figure 7 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0150] Optionally, in a specific implementation, if the memory 701, processor 702, and communication interface 703 are integrated on a single chip, then the memory 701, processor 702, and communication interface 703 can communicate with each other through an internal interface.

[0151] The processor 702 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of this application.

[0152] This embodiment also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described method for route planning in collaborative inspection of a drone swarm.

[0153] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the route planning method for collaborative inspection of unmanned aerial vehicle (UAV) swarms provided in this embodiment of the invention.

[0154] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0155] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "N" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0156] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or N executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as should be understood by those skilled in the art to which embodiments of this application pertain.

[0157] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (electronic device), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Alternatively, the computer-readable medium may be paper or other suitable media on which the program can be printed, since the program can be obtained electronically by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in a computer memory.

[0158] It should be understood that the various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0159] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0160] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0161] The storage medium mentioned above can be a read-only memory, a disk, or an optical disk, etc. Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of this application.

Claims

1. A method for route planning in collaborative inspection of unmanned aerial vehicle (UAV) swarms, characterized in that, Includes the following steps: Obtain geographic information of the rail transit control and protection zone, and construct an environmental model based on the geographic information; By combining the environmental model and the telemetry data of each UAV, an individual UAV twin with motion state is constructed to obtain the pose of each UAV. By combining the pose and the preset collaboration logic, a corresponding cluster relationship graph is constructed; By combining the environment model, the pose, and the cluster relationship graph, a twin scene that meets preset interaction conditions is constructed, so as to output the flight path planning path of each UAV based on the twin scene.

2. The method according to claim 1, characterized in that, Before constructing the corresponding cluster relationship graph by combining the pose and preset collaboration logic, the following steps are also included: Obtain inspection task information from the drone swarm; The status of each UAV is assessed using the telemetry data to obtain assessment results; The preset collaboration logic is constructed by combining the inspection task information and the evaluation results.

3. The method according to claim 1, characterized in that, After outputting the flight path planning route for each UAV based on the twin scenario, the process also includes: Update the telemetry data of each UAV to determine whether at least one UAV meets the preset obstacle conditions. If at least one drone meets the preset obstacle conditions, a preset fast exploration random tree-connection strategy is used to perform local path replanning to obtain a locally replanned path, so as to control the at least one drone to inspect along the locally replanned path.

4. The method according to claim 3, characterized in that, After obtaining the local replanning path, the following is also included: The local replanning path is shared using a cluster communication network; Based on the locally replanned path, the corresponding obstacle information is marked in the environmental model; Based on the local replanning path and the obstacle information, the flight path planning paths of other UAVs besides the at least one UAV are adjusted.

5. The method according to claim 4, characterized in that, Also includes: Obtain the communication link status of the cluster communication network; If the communication link status does not meet the preset communication conditions, the faulty node is located based on the communication data of the cluster communication network; Based on the faulty node, switch to a communication protocol or communication path that meets the preset backup conditions so that the link status meets the preset communication conditions.

6. The method according to claim 1, characterized in that, Also includes: The power consumption of each drone is predicted using the telemetry data and a pre-built energy consumption prediction model for the drone swarm. Based on the power consumption, a return or charging path is planned for any drone that meets the preset charging conditions, and the inspection task handover of any drone is carried out based on the cluster relationship graph.

7. A route planning device for collaborative inspection of unmanned aerial vehicle (UAV) swarms, characterized in that, include: The first construction module is used to acquire the geographic information of the rail transit control and protection area, and to construct an environmental model based on the geographic information; The second construction module is used to combine the environmental model and the telemetry data of each UAV to construct a UAV individual twin with motion state in order to obtain the pose of each UAV. The third construction module is used to combine the pose and preset collaboration logic to construct the corresponding cluster relationship graph; The planning module is used to combine the environment model, the pose, and the cluster relationship graph to construct a twin scene that meets preset interaction conditions, so as to output the flight path planning path of each UAV based on the twin scene.

8. An electronic device, characterized in that, include: The system includes 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 route planning method for collaborative inspection of a drone swarm as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to implement the route planning method for collaborative inspection of UAV swarms as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed, it is used to implement the route planning method for collaborative inspection of unmanned aerial vehicle swarms as described in any one of claims 1-6.

Images