A method and system for unmanned aerial vehicle relay location planning
Patent Information
- Application Number
- CN202611116320.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-27
- Publication Date
- 2026-11-03
Smart Images
Figure CN123035478A_ABST
Abstract
Description
Technical Field
[0001] This application generally relates to the field of unmanned aerial vehicle (UAV) communication and trajectory planning technology. More specifically, this application relates to a UAV relay location planning method and system. Background Technology
[0002] The application of UAV communication relay in complex terrains (such as mountains, hills, and urban canyons) is becoming increasingly widespread. The location of the relay node directly determines the quality and coverage of the communication link. Currently, the mainstream relay location planning methods mainly include the forward ray method, the grid search method, and the line-of-sight analysis method based on the digital elevation model (DEM).
[0003] However, the aforementioned existing technologies have the following significant drawbacks in practical applications: First, it suffers from high computational redundancy and poor adaptability to complex terrain: Traditional forward ray tracing emits a large number of rays outward from the observation source, which can easily fall into the invalid space outside the communication buffer, resulting in a serious waste of computing power. At the same time, when encountering multiple obstructions or complex uneven terrain, multiple iterations are required to determine the obstruction boundary, causing the algorithm complexity to increase exponentially.
[0004] Secondly, the storage overhead is high, which is not conducive to edge deployment: the grid search method relies heavily on absolute coordinate grids, which requires storing massive amounts of point coordinate data, resulting in extremely high memory consumption. This makes it difficult to deploy on resource-constrained UAV-borne embedded devices or edge computing nodes.
[0005] Third, the process is time-consuming and lacks dynamic response capability: existing methods often require point-by-point calculations or traversal of a large number of invalid areas, resulting in low planning efficiency. In scenarios such as UAV networking or emergency communication where target points and terrain change in real time, it is difficult to complete rapid replanning in a short period of time, failing to meet the real-time requirements of highly dynamic scenarios.
[0006] In view of this, there is an urgent need to provide a UAV relay location planning scheme to solve the above problems and achieve lightweight, high-efficiency dynamic replanning in complex terrain. Summary of the Invention
[0007] In order to at least address one or more of the technical problems mentioned above, this application proposes a UAV relay location planning scheme in several aspects.
[0008] In a first aspect, this application provides a method for planning the location of a UAV relay, comprising: acquiring a terrain obstacle model of a target area and a set of communication target points to be covered; constructing a closed buffer contour based on the set of communication target points; acquiring a set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour as the initial direction vector of a reverse ray; emitting a reverse ray into the interior of the closed buffer contour along the corresponding initial direction vector, starting from each sampling point in the sampling point set, and capturing topological relationship events occurring during the propagation of the reverse ray, wherein the topological relationship events include at least blocking events, opposite-side arrival events, and intersection events; performing vectorized reconstruction based on the topological relationship events to obtain an initial connected visual field, and removing heavily obstructed areas from the initial connected visual field to obtain an effective location area; and selecting target locations in the effective location area as UAV relay locations.
[0009] In some embodiments, during the process of constructing the closed buffer profile, the following steps are performed: extracting the coordinates of all communication target points in the communication target point set on the two-dimensional projection plane, and calculating the minimum bounding polygon of the communication target point set; taking each edge of the minimum bounding polygon as a reference, translating along the external normal direction by a preset extension distance, and performing arc transition or corner processing at the vertex to form a closed and continuous vector boundary line as the closed buffer profile.
[0010] In some embodiments, the following steps are performed during the process of obtaining the initial direction vector of the reverse ray: on the boundary of the closed buffer contour, multiple discrete points are selected to form the sampling point set by discretization with equal arc length or equal angle according to a preset step size; for each sampling point in the sampling point set, the inner normal vector of the boundary line segment of the closed buffer contour where it is located is calculated as the initial direction vector of the reverse ray emitted by the sampling point.
[0011] In some embodiments, during the process of capturing topological relationship events occurring during the propagation of a reverse ray, the following steps are performed: determining whether the reverse ray intersects the terrain obstacle model for the first time on the propagation path; in response to the reverse ray intersecting the terrain obstacle model for the first time on the propagation path, recording the determination result as a blocking event, storing the obstacle edge identifier at the intersection and the normalized position parameters of the intersection point along the obstacle edge, and terminating the propagation of the current reverse ray; in response to the reverse ray not intersecting the terrain obstacle model for the first time on the propagation path, determining whether the reverse ray reaches the opposite boundary of the closed buffer contour; in response to the reverse ray reaching the opposite boundary of the closed buffer contour, recording the determination result as an opposite arrival event, marking the corresponding reverse ray as a full-through ray, and terminating the propagation of the current reverse ray; in response to the reverse ray not reaching the opposite boundary of the closed buffer contour, controlling the reverse ray to continue propagating along its initial direction vector, and returning to the step of determining whether the reverse ray intersects the terrain obstacle model for the first time on the propagation path.
[0012] In some embodiments, during the process of capturing topological relationship events that occur during the propagation of reverse rays, the following steps are also performed: traversing and determining whether any two reverse rays from different sampling points meet inside the closed buffer contour; in response to the existence of two reverse rays from different sampling points meeting inside the closed buffer contour, recording the determination result as an intersection event, and storing the index information and topological intersection status of the intersecting reverse ray pair; in response to the existence of no two reverse rays from different sampling points meeting inside the closed buffer contour, confirming no intersection event, and ending the current topological relationship detection.
[0013] In some embodiments, during the process of obtaining the initial connected visible region by vectorization reconstruction based on the topological relationship events, the following steps are performed: extracting the obstacle edge identifiers and normalized position parameters recorded in the blocking events, and calculating the actual spatial coordinates of the intersection of the back ray and the terrain obstacle model; combining all the actual spatial coordinates into an ordered blocking point set according to the original arrangement order of the sampling points that generate the corresponding back rays on the boundary; connecting adjacent actual spatial coordinates within the ordered blocking point set according to the arrangement order of the ordered blocking point set to generate an occlusion front polygon for characterizing the spatial occlusion boundary; using the closed buffer contour as the minuend polygon and the occlusion front polygon as the subtrahend polygon to perform polygon difference operation; and determining the remaining vector region obtained after polygon difference operation as the initial connected visible region.
[0014] In some embodiments, during the process of removing heavily occluded areas from the initial connected visible field to obtain an effective addressing area, the following steps are performed: discretizing the two-dimensional space inside the closed buffer contour into a grid array; for each grid cell in the grid array, counting the total number of reverse rays passing through the grid cell and the number of reverse rays that have already experienced blocking events along the propagation path before reaching the grid cell; calculating the occlusion factor of the grid cell, wherein the expression for the occlusion factor of the grid cell is: R=M / N, where R is the occlusion factor of the grid cell, M is the number of reverse rays that have already experienced blocking events along the propagation path before reaching the grid cell, and N is the total number of reverse rays passing through the grid cell; Determine whether the occlusion factor of a grid cell is equal to a first occlusion threshold; if the occlusion factor of a grid cell is equal to the first occlusion threshold, classify the grid cell as a first-level visible area; if the occlusion factor of a grid cell is not equal to the first occlusion threshold, determine whether the occlusion factor of the grid cell is greater than the first occlusion threshold and less than a second occlusion threshold; if the occlusion factor of a grid cell is greater than the first occlusion threshold and less than the second occlusion threshold, classify the grid cell as a second-level visible area; if the occlusion factor of a grid cell is neither greater than the first occlusion threshold nor less than the second occlusion threshold, classify the grid cell as a severely occluded area; remove grid cells located within the initial connected visible area that belong to the severely occluded area, and output the effective addressing area.
[0015] In some embodiments, during the process of filtering target locations as UAV relay locations within the effective location area, target locations are filtered based on geometric features or topological features. Specifically, during the process of filtering target locations based on geometric features, the primary visible area and / or the secondary visible area are selected as candidate areas within the effective location area, and the center of the largest inscribed circle or the geometric centroid of the continuous area within the candidate area is calculated as the UAV relay location. During the process of filtering target locations based on topological features, recorded intersection events are retrieved, the number of communication target points corresponding to the reverse rays associated with each intersection point is counted, and the intersection point with the largest number of corresponding communication target points within the effective location area is selected as the UAV relay location.
[0016] In some embodiments, when the location of the communication target point moves, the following steps are performed: determining the contour of the local closed buffer zone affected by the movement of the communication target point, and obtaining the deformed region; extracting sampling points located within the deformed region, performing reverse ray propagation on the extracted sampling points, and updating the topology relationship events corresponding to the extracted sampling points; keeping the original topology relationship events of the region not affected by the movement of the target point unchanged, merging the updated topology relationship events with the original topology relationship events, and re-obtaining the initial connected visible area based on the merged topology relationship events.
[0017] In a second aspect, this application provides a UAV relay location planning system, which uses the UAV relay location planning method as described in any embodiment of the first aspect to plan UAV relay locations. The system includes: a data acquisition and construction module, used to acquire a terrain obstacle model of a target area and a set of communication target points to be covered, and to construct a closed buffer contour based on the set of communication target points; a sampling and initialization module, used to acquire a set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour, as the initial direction vector of the reverse ray; a ray propagation and event capture module, used to emit a reverse ray into the closed buffer contour along the corresponding initial direction vector, starting from each sampling point in the sampling point set, and capture topological relationship events occurring during the propagation of the reverse ray, wherein the topological relationship events include at least blocking events, opposite-side arrival events, and intersection events; a reconstruction and fine screening module, used to perform vectorized reconstruction based on the topological relationship events to obtain an initial connected visible field of view, and to remove heavily obstructed areas from the initial connected visible field of view to obtain an effective location selection area; and a location selection output module, used to filter target locations in the effective location selection area as UAV relay locations.
[0018] By introducing the UAV relay location planning scheme provided above, this embodiment transforms the relay site selection problem in complex environments into topological event detection based on boundary sampling points through the introduction of a reverse ray tracing mechanism. This significantly reduces spatial search redundancy and improves computational efficiency. Using key topological events such as blocking, arrival from the opposite side, and intersections for vectorized reconstruction of the visible field of view can accurately characterize the geometric features of complex obstacle boundaries, overcoming the accuracy limitations of traditional grid scanning methods. Simultaneously, by introducing an occlusion factor to eliminate heavily occluded areas, areas with extremely weak signals are avoided from the source, ensuring that the final selected effective site selection area not only possesses global connectivity but also robust communication links. This provides a reliable scientific basis for the rapid deployment of UAV relays in dynamic and constrained environments.
[0019] Furthermore, in some embodiments, a topological relationship event capture mechanism is established to achieve efficient mapping of complex environmental constraints. By recording blocking events and normalized position parameters, the occlusion boundaries of obstacles can be accurately located, providing a quantitative basis for visual field vector reconstruction. Arrival events from the opposite side ensure effective identification of the entire path, avoiding search blind spots. Intersection events establish spatial relationships between sampling points, laying the foundation for determining the optimal intersection area for multi-target coverage. This event-driven mechanism, through the timely termination of rays, effectively reduces invalid computational overhead, significantly improves the algorithm's response speed, and simultaneously ensures the rigor and accuracy of relay site selection under complex topologies.
[0020] Furthermore, in some embodiments, vectorized reconstruction is performed using polygon difference operations, effectively eliminating the jagged edge effect of traditional mesh methods and ensuring the geometric accuracy of the initial visible field. By introducing an occlusion factor to divide the visible area into multiple levels, severely occluded areas with weak signals are accurately eliminated, significantly improving the robustness of the communication link. In addition, a dual selection mechanism of geometry and topology is provided, which can meet safety clearance requirements through inscribed circles or centroids, and maximize coverage efficiency through intersection point statistics, ensuring that the site selection has optimal communication performance and deployment reliability in complex environments.
[0021] Furthermore, in some embodiments, an incremental update mechanism based on deformation regions is introduced, significantly improving the system's real-time response capability to dynamic targets. This mechanism accurately identifies local areas affected by target displacement, recalculating the inverse ray only for sampling points within the deformation region, avoiding repeated scanning of the entire field and greatly reducing computational overhead. Through a strategy of local refresh and global persistence, seamless merging of updated and historical data is achieved, ensuring that the location selection results can be iteratively corrected at high frequency as the target moves, providing robust communication assurance for UAVs in scenarios such as dynamic tracking and mobile collaboration. Attached Figure Description
[0022] The above and other objects, features, and advantages of exemplary embodiments of this application will become readily understood by reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of this application are illustrated by way of example and not limitation, and the same or corresponding reference numerals denote the same or corresponding parts, wherein: Figure 1 An exemplary flowchart of the UAV relay location planning method according to an embodiment of this application is shown; Figure 2 An exemplary flowchart of the environmental constraint quantification detection process according to an embodiment of this application is shown; Figure 3 An exemplary flowchart illustrating the process of obtaining the initial connected visible region according to an embodiment of this application is shown; Figure 4An exemplary flowchart illustrating the method for obtaining an effective location area according to an embodiment of this application is shown; Figure 5 An exemplary structural block diagram of an unmanned aerial vehicle (UAV) relay location planning system according to an embodiment of this application is shown. Detailed Implementation
[0023] 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, not all, of the embodiments of this application. 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.
[0024] It should be understood that the terms "comprising" and "including" used in the specification and claims of this application indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.
[0025] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application. As used in this specification and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this specification and claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes such combinations.
[0026] Figure 1 An exemplary flowchart of a drone relay location planning method 100 according to an embodiment of this application is shown.
[0027] like Figure 1 As shown, in step S110, the terrain obstacle model of the target area and the set of communication target points to be covered are obtained, and a closed buffer contour is constructed based on the set of communication target points.
[0028] In the embodiments of this application, during the initial data preparation phase, the system first inputs basic geographic information data of the target area, including terrain elevation data (e.g., DEM data) and the location contours of known obstacles (e.g., buildings, no-fly zones, etc.), which are then parsed to construct a terrain obstacle model for spatial occlusion calculation. Simultaneously, the system obtains the location information of all communication target points that need to be covered according to the task requirements, forming an initial set of communication target points.
[0029] In the embodiments of this application, during the process of constructing the closed buffer contour, firstly, the coordinates of all communication target points in the communication target point set on the two-dimensional projection plane are extracted, and the minimum bounding polygon of the communication target point set is calculated. Secondly, based on each edge of the minimum bounding polygon, a preset extension distance is translated along the external normal direction, and a circular arc transition or corner processing is performed at the vertex to form a closed and continuous vector boundary line, which serves as the closed buffer contour.
[0030] In the embodiments of this application, geometric calculation methods such as the convex hull algorithm are used to calculate and generate the minimum bounding polygon containing the set of communication target points.
[0031] In the embodiments of this application, the preset extension distance is set as the effective communication radius of the UAV.
[0032] In the embodiments of this application, the closed buffer contour effectively defines the global search space for subsequent location planning.
[0033] In one embodiment of this application, in a UAV relay location planning task in a mountainous scenario, during the data preparation phase, the system acquires the DEM data of the mountainous terrain to establish a three-dimensional terrain obstacle model. Assume the current task requires providing relay services to five communication target points on the ground in complex terrain, and the effective communication radius of the selected UAV is known to be 2km. Next, in the buffer zone construction phase, the system first extracts the two-dimensional coordinates of these five ground communication target points and calculates the minimum bounding polygon that exactly encloses these five points. Subsequently, based on the boundary of this minimum bounding polygon, it extends outwards by an equidistant distance of 2km (i.e., translation), and performs rounded corner processing at each sharp vertex. This generates a closed buffer zone contour that not only covers all communication target points but also extends outwards by a 2km communication redundancy range.
[0034] After completing step S110, in step S120, the set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour are obtained as the initial direction vector of the reverse ray.
[0035] In the embodiments of this application, in the process of obtaining the initial direction vector of the reverse ray, firstly, on the boundary of the closed buffer contour, multiple discrete points are selected to form a sampling point set by discretization with equal arc length or equal angle according to a preset step size. Secondly, for each sampling point in the sampling point set, the inner normal vector (i.e., the direction perpendicular to the boundary and strictly pointing into the buffer) of the boundary line segment (or tangent) of the closed buffer contour where it is located is calculated, and the inner normal vector is used as the initial direction vector for the corresponding sampling point to emit a reverse ray into the buffer.
[0036] In one embodiment of this application, on the defined closed buffer zone contour boundary, equidistant sampling (equi-arc length discretization) is performed according to a preset step size of a preset expansion distance (e.g., 10 meters), thereby obtaining a set of sampling points consisting of a series of uniformly distributed discrete points. Subsequently, the system traverses this set, and for each sampling point, calculates a unique corresponding inner normal vector based on the local boundary geometric features where it is located.
[0037] After completing step S120, in step S130, starting from each sampling point in the sampling point set, a reverse ray is emitted into the interior of the closed buffer contour along the corresponding initial direction vector, and the topological relationship events that occur during the propagation of the reverse ray are captured.
[0038] In embodiments of this application, topological relationship events include at least blocking events describing individual occlusion features, opposite-side arrival events describing path connectivity features, and intersection events describing multi-source spatial associations. This event-driven mechanism enables the transformation of continuous physical space exploration into discrete topological state records.
[0039] In the embodiments of this application, the process of capturing topological relationship events that occur during the propagation of reverse rays includes an environmental constraint quantification detection process and a multi-source topological association construction process.
[0040] The specific processes involved in the environmental constraint quantification detection process in the embodiments of this application can be found in [reference needed]. Figure 2 .
[0041] Figure 2 An exemplary flowchart of the environmental constraint quantification detection process according to an embodiment of this application is shown.
[0042] like Figure 2 As shown, in step S210, it is determined whether the reverse ray intersects the terrain obstacle model for the first time on its propagation path. In response to the reverse ray intersecting the terrain obstacle model for the first time on its propagation path, in step S220, the determination result is recorded as a blocking event, the obstacle edge identifier at the intersection and the normalized position parameters of the intersection point along the obstacle edge are stored, and the propagation of the current reverse ray is terminated. In response to the reverse ray not intersecting the terrain obstacle model for the first time on its propagation path, in step S230, it is determined whether the reverse ray reaches the opposite boundary of the closed buffer zone contour.
[0043] In response to the reverse ray reaching the opposite boundary of the closed buffer zone contour, in step S240, the determination result is recorded as an opposite arrival event, the corresponding reverse ray is marked as a full-pass ray, and the propagation of the current reverse ray is terminated. In response to the reverse ray not reaching the opposite boundary of the closed buffer zone contour, in step S250, the reverse ray is controlled to continue propagating along its initial direction vector, and the process returns to step S210, that is, returns to the step of determining whether the reverse ray intersects with the terrain obstacle model for the first time on the propagation path.
[0044] In the embodiments of this application, during step S210, complex absolute spatial coordinates are not recorded. Instead, the obstacle edge identifier (ID) at the intersection and the normalized position parameters of the intersection point along the obstacle edge (e.g., the proportional position of the intersection point on the obstacle line segment) are stored, and the propagation of the current reverse ray is then terminated. This recording method provides a lightweight and accurate quantitative basis for subsequent high-precision visual field vector reconstruction.
[0045] In the embodiments of this application, during the execution of step S240, the determination result is recorded as an arrival event on the opposite side, the corresponding reverse ray is marked as a full-pass ray, indicating that there is no terrain obstruction on the detection path, and the propagation of the current reverse ray is terminated.
[0046] In the embodiments of this application, during the construction of multi-source topology associations, iteratively determining whether any two reverse rays from different sampling points meet within the closed buffer contour. If two reverse rays from different sampling points meet within the closed buffer contour, the determination result is recorded as a crossover event, and the index information and topology crossover status of the intersecting reverse ray pair are stored. If no two reverse rays from different sampling points meet within the closed buffer contour, it is confirmed that there is no crossover event, and the current topology relationship detection ends.
[0047] In one embodiment of this application, during reverse ray detection in a mountainous environment, a ray list is first initialized. Each ray originates from a boundary sampling point, and its direction is set to the previously calculated inner normal. During propagation, the ray advances gradually in 1-meter increments. If a ray encounters a mountain (obstacle) model during its journey, the system immediately records the blocking event and marks the blocked mountain edge ID, using the normalized parameter t to lock the critical obstruction point. If the ray successfully passes through the entire buffer zone and reaches the opposite edge, it is confirmed as a fully accessible path. Simultaneously, the system continuously monitors the spatial relationships between different rays. Once two detection rays representing different coverage targets intersect, an intersection event recording is triggered.
[0048] After step S130 is completed, in step S140, vectorized reconstruction is performed based on topological relationship events to obtain the initial connected visible field, and heavily occluded areas are removed from the initial connected visible field to obtain the effective location area.
[0049] Based on step S140, the aforementioned discrete topological events are transformed into precise geometric entities, and the feasibility of the communication link is evaluated at multiple levels through probabilistic statistical indicators to ensure that the selected area has both geometric visibility and communication robustness.
[0050] In the embodiments of this application, the specific process involved in obtaining the initial connected visible region through vectorized reconstruction based on topological relationship events can be found in [reference needed]. Figure 3 .
[0051] Figure 3 An exemplary flowchart illustrating the process of obtaining an initial connected visible field according to an embodiment of this application is shown.
[0052] like Figure 3 As shown, in step S310, the obstacle edge identifiers and normalized position parameters recorded in the blocking event are extracted, and the actual spatial coordinates of the intersection between the reverse ray and the terrain obstacle model are calculated. In step S320, all actual spatial coordinates are combined into an ordered blocking point set according to the original arrangement order of the sampling points that generated the corresponding reverse ray on the boundary. In step S330, adjacent actual spatial coordinates within the ordered blocking point set are connected sequentially according to the arrangement order of the ordered blocking point set to generate an occlusion front polygon used to characterize the spatial occlusion boundary. In step S340, the closed buffer contour is used as the minuend polygon, and the occlusion front polygon is used as the subtrahend polygon to perform a polygon difference operation. In step S350, the remaining vector region obtained after the polygon difference operation is determined as the initial connected visible field.
[0053] In the embodiments of this application, during the execution of step S310, since the parameters recorded in the early stage are normalized parameters, the physical location where the occlusion occurs can be accurately restored by using a linear interpolation algorithm.
[0054] In the embodiments of this application, the zigzag effect of the traditional mesh method when dealing with complex boundaries is effectively eliminated by the vector polygon-based operation in step S350, ensuring the geometric accuracy of the visible area division.
[0055] In the embodiments of this application, the specific process involved in removing heavily occluded areas from the initial connected visible field to obtain an effective location area can be found in [reference needed]. Figure 4 .
[0056] Figure 4 An exemplary flowchart illustrating the process of obtaining an effective location area according to an embodiment of this application is shown.
[0057] as Figure 4 shown, in step S410, the two-dimensional space inside the closed buffer contour is discretized into a grid array. In step S420, for each grid cell in the grid array, the total number of reverse rays passing through the grid cell and the number of reverse rays that have undergone a blocking event along the propagation path before reaching the grid cell are counted respectively. In step S430, the occlusion factor of the grid cell is calculated. In step S440, it is determined whether the occlusion factor of the grid cell is equal to the first occlusion threshold.
[0058] In response to the fact that the occlusion factor of the grid cell is equal to the first occlusion threshold, in step S450, the grid cell is classified as a first-level visible area. In response to the fact that the occlusion factor of the grid cell is not equal to the first occlusion threshold, in step S460, it is determined whether the occlusion factor of the grid cell is greater than the first occlusion threshold and less than the second occlusion threshold. In response to the fact that the occlusion factor of the grid cell is greater than the first occlusion threshold and less than the second occlusion threshold, in step S470, the grid cell is classified as a second-level visible area. In response to the fact that the occlusion factor of the grid cell is not greater than the first occlusion threshold and less than the second occlusion threshold, in step S480, the grid cell is classified as a heavy occlusion area. Finally, in step S490, the grid cells located inside the initial connected visibility domain and belonging to the heavy occlusion area are eliminated, and an effective site selection area is output.
[0059] In the embodiment of the present application, the expression of the occlusion factor of the grid cell is: R=M / N, where R is the occlusion factor of the grid cell, M is the number of reverse rays that have undergone a blocking event along the propagation path before reaching the grid cell, and N is the total number of reverse rays passing through the grid cell.
[0060] In the embodiment of the present application, the first occlusion threshold is set to 0, and the second occlusion threshold is set to 0.5.
[0061] In the embodiment of the present application, when a grid cell is classified as a first-level visible area (occlusion-free area), it means that the location has excellent communication quality. When the grid cell is classified as a second-level visible area (slight occlusion area), it means that the communication link at the location may be affected by part of the terrain edge, but the influence is within an acceptable range.
[0062] In an embodiment of the present application, in a mountain environment, the total number N of rays passing through each grid and the number M of blocked rays are counted. If the occlusion rate R of a certain grid is 0, it is determined as an optimal area with excellent communication clearance. If 0<R<0.5, it is taken as an alternative sub-optimal area. If R≥0.5, that is, more than half of the potential communication paths are blocked by the mountain, it is determined as a heavy occlusion area and forcibly eliminated. This dual screening mechanism based on geometry and probability enables the finally output effective site selection area to provide the most robust signal coverage guarantee for the unmanned aerial vehicle.
[0063] After completing step S140, in step S150, target locations are selected in the effective location area to serve as drone relay locations.
[0064] In the embodiments of this application, during the process of selecting target locations as UAV relay locations in the effective site selection area, the target locations are selected based on geometric features or topological features.
[0065] In the embodiments of this application, during the process of selecting target locations based on geometric features as UAV relay locations, a primary visible area and / or a secondary visible area are selected as candidate areas in the effective location area, and the center of the largest inscribed circle or the geometric centroid of the continuous area within the candidate area is calculated as the UAV relay location.
[0066] In embodiments of this application, to ensure better spatial tolerance (i.e., away from obstacle edges) when the UAV performs relay tasks, the system calculates the center of the largest inscribed circle of a continuously connected region within the candidate area. This center is typically considered the safest center point because it is physically furthest from all surrounding obstacle boundaries. Alternatively, the geometric centroid of this connected region can be calculated based on actual needs and used as the UAV relay location to balance the overall spatial coverage ratio.
[0067] In the embodiments of this application, during the process of filtering target locations based on topological features as UAV relay locations, the recorded intersection events are retrieved, the number of communication target points corresponding to the reverse rays associated with each intersection point is counted, and within the effective location area, the intersection point with the largest number of corresponding communication target points is selected as the UAV relay location.
[0068] In the embodiments of this application, since each intersection is formed by the convergence of two or more reverse rays from different sampling points (representing different communication target directions), the system counts the number of communication target points corresponding to the reverse rays associated with each intersection. Subsequently, under the constraint of the effective location area, the intersection with the largest number of corresponding communication target points is selected as the UAV relay position. The advantage of this strategy is that it can find the point with the highest coverage efficiency for multiple targets with minimal computation, realizing the effective association of a single relay node with as many ground targets as possible.
[0069] In the embodiments of this application, after determining the optimal relay location, its latitude, longitude, and elevation coordinates are accurately output. Simultaneously, on the visual interactive interface, the system clearly marks the vectorized reconstructed connected view area, heavily occluded areas, and blocking front boundaries on the digital map, and marks the selected relay target point with a prominent icon.
[0070] In the embodiments of this application, if the position of the communication target point moves during the execution of steps S110-S150 (such as the dynamic displacement of ground targets during UAV swarm flight), the system will trigger an incremental update mechanism to achieve real-time replanning.
[0071] In the embodiments of this application, during the dynamic replanning process, firstly, the contour of the locally closed buffer zone affected by the movement of the communication target point is determined, and the deformed region is obtained. Secondly, sampling points located within the deformed region are extracted, reverse ray propagation is performed on the extracted sampling points, and the topology relationship events corresponding to the extracted sampling points are updated. Then, the original topology relationship events of the regions not affected by the movement of the target point remain unchanged, the updated topology relationship events are merged with the original topology relationship events, and the initial connected visible region is obtained again based on the merged topology relationship events.
[0072] In embodiments of this application, when the coordinates of some communication target points change, the system recalculates the minimum bounding polygon of all target points and generates an updated closed buffer profile. By performing geometric difference analysis on the buffer vector profiles before and after the update (e.g., using polygon XOR operation), local line segments whose boundaries have shifted, shrunk, or expanded are identified. These changed boundary segments and their neighborhoods are defined as deformation regions.
[0073] In the embodiments of this application, it is not necessary to resample the global boundary; instead, local updates are performed only on the boundary segments within the aforementioned deformation area, according to a preset step size (e.g., 10 meters). For these newly added or shifted sampling points, the reverse ray emission and detection process in steps S120 and S130 is re-executed. During propagation, the blocking events, opposite-side arrival events, or intersection events generated by these local rays are recaptured and recorded, generating a local incremental event list.
[0074] In the embodiments of this application, the principle of topological consistency is utilized to retain existing stable event data in unchanged regions while discarding outdated data. Subsequently, the locally updated incremental event list is seamlessly integrated with the retained stable event data to form a globally up-to-date topological event database. Based on this merged database, the system again executes the polygon difference operation and occlusion factor calculation in step S140, thereby quickly refreshing the initial connected visible field and effective addressable area.
[0075] Through dynamic replanning, it is possible to track changes in ground targets in real time and dynamically adjust the optimal hovering position of the UAV relay.
[0076] In summary, through the UAV relay location planning scheme provided above, this application's embodiment transforms the relay site selection problem in complex environments into topological event detection based on boundary sampling points by introducing a reverse ray tracing mechanism, significantly reducing spatial search redundancy and improving computational efficiency. Using key topological events such as blocking, arrival from the opposite side, and intersections for vectorized reconstruction of the visible field of view can accurately characterize the geometric features of complex obstacle boundaries, overcoming the accuracy limitations of traditional grid scanning methods. Simultaneously, by introducing an occlusion factor to eliminate heavily occluded areas, areas with extremely weak signals are avoided from the source, ensuring that the final selected effective site selection area not only possesses global connectivity but also robust communication links, providing a reliable scientific basis for the rapid deployment of UAV relays in dynamic and constrained environments.
[0077] Furthermore, in some embodiments, a topological relationship event capture mechanism is established to achieve efficient mapping of complex environmental constraints. By recording blocking events and normalized position parameters, the occlusion boundaries of obstacles can be accurately located, providing a quantitative basis for visual field vector reconstruction. Arrival events from the opposite side ensure effective identification of the entire path, avoiding search blind spots. Intersection events establish spatial relationships between sampling points, laying the foundation for determining the optimal intersection area for multi-target coverage. This event-driven mechanism, through the timely termination of rays, effectively reduces invalid computational overhead, significantly improves the algorithm's response speed, and simultaneously ensures the rigor and accuracy of relay site selection under complex topologies.
[0078] Furthermore, in some embodiments, vectorized reconstruction is performed using polygon difference operations, effectively eliminating the jagged edge effect of traditional mesh methods and ensuring the geometric accuracy of the initial visible field. By introducing an occlusion factor to divide the visible area into multiple levels, severely occluded areas with weak signals are accurately eliminated, significantly improving the robustness of the communication link. In addition, a dual selection mechanism of geometry and topology is provided, which can meet safety clearance requirements through inscribed circles or centroids, and maximize coverage efficiency through intersection point statistics, ensuring that the site selection has optimal communication performance and deployment reliability in complex environments.
[0079] Furthermore, in some embodiments, an incremental update mechanism based on deformation regions is introduced, significantly improving the system's real-time response capability to dynamic targets. This mechanism accurately identifies local areas affected by target displacement, recalculating the inverse ray only for sampling points within the deformation region, avoiding repeated scanning of the entire field and greatly reducing computational overhead. Through a strategy of local refresh and global persistence, seamless merging of updated and historical data is achieved, ensuring that the location selection results can be iteratively corrected at high frequency as the target moves, providing robust communication assurance for UAVs in scenarios such as dynamic tracking and mobile collaboration.
[0080] This application also provides a UAV relay location planning system, which can use the aforementioned UAV relay location planning method 100 to perform UAV relay location planning, or other methods can be used for UAV relay location planning, and this application does not impose any restrictions here.
[0081] Figure 5 An exemplary structural block diagram of an unmanned aerial vehicle (UAV) relay location planning system according to an embodiment of this application is shown.
[0082] like Figure 5 As shown, the system 500 includes a data acquisition and construction module 510, a sampling and initialization module 520, a ray propagation and event capture module 530, a reconstruction and fine screening module 540, and an addressing output module 550. In the embodiments of this application, the data acquisition and construction module 510, the sampling and initialization module 520, the ray propagation and event capture module 530, the reconstruction and fine screening module 540, and the addressing output module 550 may be separate units or may be located in the same controller; this application does not impose any restrictions here.
[0083] Specifically, the data acquisition and construction module 510 is used to acquire the terrain obstacle model of the target area and the set of communication target points to be covered, and to construct a closed buffer contour based on the set of communication target points.
[0084] Specifically, the sampling and initialization module 520 is used to obtain the set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour, so as to serve as the initial direction vector of the reverse ray.
[0085] Specifically, the ray propagation and event capture module 530 is used to emit reverse rays into the interior of the closed buffer contour along the corresponding initial direction vector, starting from each sampling point in the sampling point set, and capture the topological relationship events that occur during the propagation of the reverse rays. The topological relationship events include at least blocking events, opposite side arrival events, and intersection events.
[0086] Specifically, the reconstruction and fine screening module 540 is used to perform vectorized reconstruction based on topological relationship events to obtain an initial connected visible field, and to remove heavily occluded areas from the initial connected visible field to obtain an effective site selection area.
[0087] Specifically, the location output module 550 is used to filter target locations within the effective location area as relay locations for UAVs.
[0088] When system 500 performs UAV relay location planning using the aforementioned UAV relay location planning method 100, the data acquisition and construction module 510 executes the aforementioned step S110, the sampling and initialization module 520 executes the aforementioned step S120, the ray propagation and event capture module 530 executes the aforementioned step S130, the reconstruction and fine screening module 540 executes the aforementioned step S140, and the location selection output module 550 executes the aforementioned step S150. The specific execution process can be found above and will not be repeated here.
[0089] While numerous embodiments of this application have been shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will arise for those skilled in the art without departing from the spirit and intent of this application. It should be understood that various alternatives to the embodiments of this application described herein may be employed in the practice of this application. The appended claims are intended to define the scope of protection of this application and therefore cover equivalents or alternatives within the scope of these claims.
Claims
1. A method for unmanned aerial vehicle (UAV) relay location planning, characterized in that, include: Obtain the terrain obstacle model of the target area and the set of communication target points to be covered, and construct a closed buffer zone profile based on the set of communication target points; Obtain the set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour, as the initial direction vector of the reverse ray; Starting from each sampling point in the sampling point set, a reverse ray is emitted into the interior of the closed buffer contour along the corresponding initial direction vector, and topological relationship events occurring during the propagation of the reverse ray are captured. The topological relationship events include at least blocking events, opposite side arrival events, and intersection events. Based on the aforementioned topological relationship events, a vectorized reconstruction is performed to obtain an initial connected visible field of view, and heavily occluded areas are removed from the initial connected visible field of view to obtain an effective location area. Target locations are selected within the effective location area to serve as relay locations for unmanned aerial vehicles (UAVs).
2. The UAV relay location planning method according to claim 1, characterized in that, During the process of constructing the closed buffer outline, the following steps are performed: Extract the coordinates of all communication target points in the communication target point set onto the two-dimensional projection plane, and calculate the minimum bounding polygon of the communication target point set; Using each edge of the smallest circumscribed polygon as a reference, the polygon is translated along the direction of the outer normal by a preset extension distance, and a circular arc transition or corner treatment is performed at the vertex to form a closed and continuous vector boundary line, which serves as the closed buffer contour.
3. The UAV relay location planning method according to claim 1, characterized in that, In obtaining the initial direction vector of the reverse ray, the following steps are performed: On the boundary of the closed buffer zone contour, multiple discrete points are selected according to a preset step size and discretized by equal arc length or equal angle to form the sampling point set; For each sampling point in the sampling point set, calculate the inner normal vector of the boundary line segment of the closed buffer contour where it is located, and use it as the initial direction vector for the reverse ray emitted by the sampling point.
4. The UAV relay location planning method according to claim 1, characterized in that, In capturing topological events that occur during the backward propagation of rays, the following steps are performed: Determine whether the reverse ray intersects the terrain obstacle model for the first time on its propagation path; In response to the first intersection of the reverse ray with the terrain obstacle model on the propagation path, the determination result is recorded as a blocking event, the obstacle edge identifier at the intersection and the normalized position parameters of the intersection point along the obstacle edge are stored, and the propagation of the current reverse ray is terminated. In response to the fact that the reverse ray does not intersect the terrain obstacle model for the first time on the propagation path, it is determined whether the reverse ray reaches the opposite boundary of the closed buffer zone contour; In response to the reverse ray reaching the opposite boundary of the closed buffer contour, the determination result is recorded as an opposite arrival event, the corresponding reverse ray is marked as a full-pass ray, and the propagation of the current reverse ray is terminated. In response to the reverse ray failing to reach the opposite boundary of the closed buffer zone profile, the reverse ray is controlled to continue propagating along its initial direction vector, and the process returns to the step of determining whether the reverse ray intersects the terrain obstacle model for the first time on its propagation path.
5. The UAV relay location planning method according to claim 4, characterized in that, In capturing topological events that occur during the backward propagation of rays, the following steps are also performed: Iterate through the data to determine whether any two reverse rays from different sampling points meet inside the closed buffer contour; In response to the presence of two reverse rays from different sampling points meeting inside the closed buffer contour, the determination result is recorded as a crossover event, and the index information and topological crossover status of the intersecting reverse ray pair are stored. In response to the absence of two reverse rays from different sampling points meeting within the closed buffer contour, confirming no intersection event, the current topology detection ends.
6. The UAV relay location planning method according to claim 1, characterized in that, In the process of obtaining the initial connected visible region by vectorization reconstruction based on the aforementioned topological relationship events, the following steps are performed: Extract the obstacle edge identifiers and normalized position parameters recorded in the blocking event, and calculate the actual spatial coordinates of the intersection between the reverse ray and the terrain obstacle model; According to the original arrangement order of the sampling points that generate the corresponding reverse rays on the boundary, all actual spatial coordinate points are combined into an ordered set of blocking points; According to the arrangement order of the ordered blocking point set, connect adjacent actual spatial coordinate points within the ordered blocking point set in sequence to generate an occlusion front polygon used to characterize the spatial occlusion boundary; The closed buffer contour is used as the minuend polygon, and the occlusion front edge polygon is used as the subtrahend polygon to perform polygon difference operation. The remaining vector region obtained after polygon difference operation is determined as the initial connected visible region.
7. The UAV relay location planning method according to claim 1 or 6, characterized in that, In the process of removing heavily occluded areas from the initial connected visible field to obtain an effective location area, the following steps are performed: The two-dimensional space inside the closed buffer contour is discretized into a grid array; For each grid cell in the grid array, the total number of reverse rays passing through the grid cell and the number of reverse rays that have already experienced a blocking event before reaching the grid cell along the propagation path are counted. Calculate the occlusion factor of the grid cell, where the expression for the occlusion factor of the grid cell is: R=M / N, where R is the occlusion factor of the grid cell, M is the number of reverse rays that have already experienced blocking events before reaching the grid cell along the propagation path, and N is the total number of reverse rays passing through the grid cell. Determine whether the occlusion factor of the grid cell is equal to the first occlusion threshold; In response to the occlusion factor of a grid cell being equal to the first occlusion threshold, the grid cell is classified as a first-level visible area; In response to the occlusion factor of a grid cell not being equal to the first occlusion threshold, it is determined whether the occlusion factor of the grid cell is greater than the first occlusion threshold and less than the second occlusion threshold. In response to a grid cell having an occlusion factor greater than a first occlusion threshold and less than a second occlusion threshold, the grid cell is divided into a second-level visible area. In response to a grid cell's occlusion factor being neither greater than a first occlusion threshold nor less than a second occlusion threshold, the grid cell is classified as a heavily occluded area. Grid cells located within the initial connected visible area and belonging to the heavily occluded area are removed, and the effective addressing area is output.
8. The UAV relay location planning method according to claim 7, characterized in that, In the process of selecting target locations as UAV relay locations within the effective location area, target locations are selected based on geometric features or topological features. In the process of selecting target locations based on geometric features as UAV relay locations, the primary visible area and / or the secondary visible area are selected as candidate areas in the effective location area, and the center of the largest inscribed circle or the geometric centroid of the continuous area within the candidate area is calculated as the UAV relay location. In the process of selecting target locations based on topological features as UAV relay locations, the recorded intersection events are retrieved, the number of communication target points corresponding to the reverse rays associated with each intersection point is counted, and within the effective location area, the intersection point with the largest number of corresponding communication target points is selected as the UAV relay location.
9. The UAV relay location planning method according to claim 1, characterized in that, When the location of the communication target point moves, perform the following steps: Determine the contour of the local closed buffer zone affected by the movement of the communication target point, and obtain the deformation area; Extract sampling points located within the deformation region, perform reverse ray propagation on the extracted sampling points, and update the topological relationship events corresponding to the extracted sampling points; The original topological relationship events of the region unaffected by the target point's position movement are kept unchanged. The updated topological relationship events are merged with the original topological relationship events, and the initial connected visible region is obtained again based on the merged topological relationship events.
10. A UAV relay location planning system, characterized in that, The system employs the UAV relay location planning method as described in any one of claims 1-9 for UAV relay location planning, and comprises: The data acquisition and construction module is used to acquire the terrain obstacle model of the target area and the set of communication target points to be covered, and to construct a closed buffer contour based on the set of communication target points; The sampling and initialization module is used to obtain the set of sampling points on the boundary of the closed buffer contour and the inner normal vector inside the closed buffer contour, so as to serve as the initial direction vector of the reverse ray. The ray propagation and event capture module is used to emit reverse rays into the closed buffer contour along the corresponding initial direction vector, starting from each sampling point in the sampling point set, and capture the topological relationship events that occur during the propagation of the reverse rays. The topological relationship events include at least blocking events, opposite side arrival events, and intersection events. The reconstruction and fine screening module is used to perform vectorized reconstruction based on the topological relationship events to obtain an initial connected visible field, and to remove heavily occluded areas from the initial connected visible field to obtain an effective site selection area. The location selection output module is used to filter target locations within the effective location selection area as relay locations for UAVs.