Method for identifying boundary nodes of two-dimensional unmanned aerial vehicle cluster based on extreme value envelope iteration

Abstract

The application discloses a two-dimensional unmanned aerial vehicle cluster boundary node identification method based on an extreme value envelope iteration, belongs to the field of information technology calculation, and comprises the following steps: step 1, projecting an aerial unmanned aerial vehicle cluster to a specific two-dimensional plane, regarding each unmanned aerial vehicle as a particle, and obtaining a two-dimensional plane with unmanned aerial vehicle points distributed with the number of clusters; step 2, according to the position information of the unmanned aerial vehicle points in the two-dimensional plane, iteratively finding out extreme value unmanned aerial vehicles, and obtaining initial boundary nodes; step 3, based on the position information of the unmanned aerial vehicle points in the two-dimensional plane and the initial boundary nodes, judging the states of all the unmanned aerial vehicle points relative to the initial boundary; and step 4, according to the states of the unmanned aerial vehicle points relative to the initial boundary, finding new boundary nodes in the unmanned aerial vehicle points not enveloped by the initial boundary, updating the boundary, and continuing until all the unmanned aerial vehicle points are inside the boundary. The application adopts simple and understandable mathematical principles to identify boundary points, and has a wide application prospect.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of information technology computing, and particularly relates to a two-dimensional unmanned aerial vehicle cluster boundary node identification method based on extreme value envelope iteration. BACKGROUND

[0002] Unmanned aerial vehicle cluster combat has become a new mode of operation on the modern battlefield, and developing unmanned aerial vehicle cluster technology and anti-unmanned aerial vehicle cluster technology is a new requirement of the times. Accurate identification of the configuration of the unmanned cluster is the premise of the anti-unmanned cluster. By accurately identifying the configuration of the enemy unmanned aerial vehicle cluster, important information about its technical specifications, combat capabilities and intentions can be obtained, which helps to develop more accurate coping strategies and combat plans, and provides an information basis and reaction time for anti-unmanned cluster. Unmanned aerial vehicle cluster boundary node identification is a key step in identifying the configuration of the unmanned aerial vehicle cluster.

[0003] At present, there is a blank in the boundary node identification technology for unmanned aerial vehicle clusters, and most of them use general boundary node identification algorithms to identify on simplified unmanned aerial vehicle cluster models. In the document (Han Fenglin, Li Weijian, Su Bin, etc. Pose recognition of aviation blades based on 3D point cloud boundary point features [J / OL]. Computer integrated manufacturing systems: 1-24 [2024-01-11]. http: / / kns.cnki.net / kcms / detail / 11.5946.TP.20231020.1756.016.html.), for point cloud data, a tangent plane is made through the points in the point cloud, and all points in its neighborhood are projected onto the tangent plane. By analyzing the vector angle between the point and its projection point, it is determined whether the point is a boundary point. If the point is a boundary point, there must be a large angle between the two projection vectors that is much larger than the angle between other vectors. This method requires the designer to set a threshold for the angle. When the maximum angle in the neighborhood of the point to be judged is greater than the threshold, the point to be judged is considered a boundary point, so the effectiveness of the method is sensitive to the angle threshold. In the document (Li Guangxing. A boundary point recognition algorithm for grid k-neighbor set [J]. Shipboard electronic engineering, 2015, 35(07): 132-135+164.), according to the feature that there is a density difference around the boundary, a boundary recognition algorithm (BGN) for grid k-neighbor set is proposed. This technology determines the boundary degree by calculating the quality of the k-neighbor set of the grid cell and its nearest neighbor cell and the center distance between the cells in the grid space. Compare the boundary degree with the boundary threshold to determine whether each grid cell is a boundary cell or a noise cell. Then extract the data closer to the edge from the boundary cells as the final boundary points. This method is suitable for dense data point distribution, but for sparse data, it cannot guarantee the accuracy of boundary node identification.

[0004] When a ground base station observes a cluster of unmanned aerial vehicles (UAVs) at a sufficient distance, two-dimensional information of the cluster of UAVs projected on a sensor observation plane is generally obtained. Therefore, a two-dimensional UAV cluster boundary node identification method based on extreme value envelope iteration is needed. SUMMARY

[0005] The purpose of the present application is to provide a two-dimensional UAV cluster boundary node identification method based on extreme value envelope iteration, which solves the problem of low accuracy and slow speed of boundary node identification in the prior art.

[0006] To achieve the above purpose, the present application provides a two-dimensional UAV cluster boundary node identification method based on extreme value envelope iteration, comprising the following steps:

[0007] Step 1: Project the cluster of UAVs in the air onto a specific two-dimensional plane, and regard each UAV as a particle to obtain a two-dimensional plane with a number of UAV points distributed;

[0008] Step 2: According to the position information of the UAV points in the two-dimensional plane, find the extreme value UAV by traversal to obtain the initial boundary node;

[0009] Step 3: Based on the position information of the UAV points in the two-dimensional plane and the initial boundary node, determine the state of all UAV points relative to the initial boundary. For all UAV points, determine the relative position relationship of each UAV point and the boundary formed by the initial boundary node to obtain a cluster UAV state set I0={IN1,IN2,···,IN N}, wherein INi is the relative position relationship of the i-th UAV point and the boundary formed by the initial boundary node, IN1=1 indicates that the i-th UAV point is inside the boundary formed by the initial boundary node, IN1=0 indicates that the i-th UAV point is outside the boundary formed by the initial boundary node, the initial boundary does not envelope the i-th UAV point, i=1,2,···,N, and N is the total number of cluster UAVs;

[0010] Step 4: According to the state of the UAV points relative to the initial boundary, find new boundary nodes in the UAV points not enveloped by the initial boundary, update the boundary, and continue until all UAV points are inside the boundary.

[0011] Preferably, in step 1, when observing the aerial drone swarm using sensors on a ground base station, the three-dimensional moving drone swarm is considered as a series of points moving in the sensor's observation plane. A point on the observation plane corresponds to a drone in the swarm. Using positioning technology, the position information of the drone swarm is obtained, and the coordinate information of the points on the observation plane is acquired. The i-th drone point represents the point of the i-th drone in the swarm on the observation plane, where i = 1, 2, ..., N, and N is the total number of drones in the swarm. The coordinate information of the drone point corresponding to the drone swarm is {X1, X2, ..., X...}. N}, where X i The coordinate information of the i-th drone point is a two-dimensional data, i = 1, 2, ..., N, where N is the total number of drones in the cluster.

[0012] Preferably, in step 2, the specific process of traversing and finding the extreme value UAV based on the position information of the UAV points in the two-dimensional plane to obtain the initial boundary nodes is as follows: In {X1,X2,...,X... N The algorithm iterates through the data to find the drone points corresponding to the minimum, maximum, minimum, and maximum values ​​of the x-coordinate, and the y-coordinate, respectively. These points are then used as the extreme value drones, forming the extreme value drone set. in The coordinates of the j-th extreme UAV point, j = 1, 2, ..., n, 2 ≤ n ≤ 4, are the initial set of boundary nodes. in Let k be the coordinate information of the k-th boundary node, k = 1, 2, ..., p, 2 ≤ p ≤ N.

[0013] Preferably, the specific process of step 3 is as follows:

[0014] Step 301: Determine whether the point to be judged is a boundary node. If the point to be judged is a boundary node, then the point to be judged is an internal point; if the point to be judged is not a boundary node, proceed to step 302.

[0015] Step 302: Determine whether the point to be determined is on the horizontal edge of the boundary polygon. If the point to be determined is on the horizontal edge of the boundary polygon, then the point to be determined is an internal point; if the point to be determined is not on the horizontal edge of the boundary polygon, proceed to step 303.

[0016] Step 303: Determine the number of intersections between the horizontal ray emanating from the point to be judged and the polygon's edge. If the horizontal ray intersects the polygon's edge twice, then the point to be judged is an external point. If the horizontal ray intersects the polygon once and the point to be judged has the same ordinate as a vertex of the edge that generates the intersection, then move the point to be judged by 0.0001 units and continue to determine the relative position of the moved point to the polygon according to steps 301-303.

[0017] Preferably, the specific process of step 4 is as follows:

[0018] Step 401, according to the cluster unmanned aerial vehicle state set I0, the unmanned aerial vehicle point set O0 outside the boundary B0 is obtained = {OU1, OU2,..., OU q}, wherein OU s is the s unmanned aerial vehicle point outside the boundary B0, s = 1, 2,..., q, 0 ≤ q ≤ N-2, N is the total number of cluster unmanned aerial vehicle frames, and the extreme unmanned aerial vehicle point is found again in O0, which is added to the boundary node set B0, the concave-convexity of the kth boundary node in the boundary node set B0 is judged, and the concave boundary point is deleted, until the boundary polygon formed by the boundary node set is a convex polygon, at this time the boundary node set is updated as B1, wherein k = 1, 2,..., p, 2 ≤ p ≤ N;

[0019] Step 402, for the boundary node set, the boundary polygon sketching algorithm is adopted, and the boundary nodes are sequentially connected to form a convex polygon;

[0020] Step 403, according to the method of step 401, the boundary node set B is updated constantly until all unmanned aerial vehicle points are inside the boundary polygon, IN i = 1, i = 1, 2,..., N.

[0021] Preferably, the concave-convexity of the polygon vertex is judged by the cross product of the adjacent edge vectors of the polygon, according to the mathematical basis knowledge, all interior angles of the convex polygon are less than 180°, and the cross products of all adjacent edge vectors are of the same sign, since the extreme unmanned aerial vehicle point selected at the beginning forms a convex polygon, at this time the signs of the adjacent cross products correspond to the convex boundary points, and the boundary nodes with different signs of the adjacent cross products are recorded as the concave boundary points.

[0022] Preferably, the specific process of the boundary polygon sketching algorithm is as follows: first, the two unmanned aerial vehicle points with the maximum horizontal coordinate value and the minimum horizontal coordinate value in the boundary node set B are found, the straight line formed by the two points divides the vertices of the boundary polygon into an upper left part B l and a lower right part B r , the boundary nodes in the upper left part are arranged in ascending order of horizontal coordinate value, the boundary nodes in the lower right part B r are arranged in descending order of horizontal coordinate value, and the boundary nodes in the upper left part and the lower right part after sorting are combined to form a sorted boundary node set, and the nodes in the boundary node set are sequentially connected to form a polygon.

[0023] Therefore, the two-dimensional unmanned aerial vehicle cluster boundary node recognition method based on the extreme envelope iteration has the following beneficial effects:

[0024] (1) Project the UAV cluster onto the observation plane of the ground base station, simplify the UAV cluster boundary node identification problem to a series of point boundary identification problem in two-dimensional plane;

[0025] (2) The boundary point identification is carried out by using simple and understandable mathematical principles, the algorithm is simple and efficient; (3) The boundary polygon outlining algorithm is designed to quickly connect the boundary nodes to form the cluster boundary.

[0026] The technical solutions of the present application will be further described in detail below by means of the accompanying drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 is the overall flowchart of the two-dimensional UAV cluster boundary node identification method of the present application based on extreme value envelope iteration;

[0028] Figure 2 is the flowchart of the present application using the ray method to determine the relative position relationship between the point and the polygon;

[0029] Figure 3 is the simulated UAV cluster diagram on the two-dimensional plane of the present application;

[0030] Figure 4 is the simulated initial boundary diagram of the UAV cluster of the present application;

[0031] Figure 5 is the simulated final UAV cluster boundary diagram of the present application. DETAILED DESCRIPTION

[0032] The following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative labor are within the scope of protection of the present application.

[0033] Please refer to Figures 1-5 , the two-dimensional UAV cluster boundary node identification method based on extreme value envelope iteration includes the following steps:

[0034] Step 1: Project the aerial drone swarm onto a specific two-dimensional plane, treating each drone as a point mass, resulting in a two-dimensional plane containing a number of drone points distributed throughout the swarm. When a ground base station observes an aerial drone swarm at a sufficiently long distance, the drones' external features are not obvious and can be approximated as point masses. When using sensors on a ground base station to observe an aerial drone swarm at a sufficiently long distance, the three-dimensional moving drone swarm can be considered as a series of points moving within the sensor's observation plane. A point on the observation plane corresponds to a drone in the swarm. Using positioning technology, the position information of the drone swarm can be obtained, further yielding the coordinate information of points on the observation plane. The i-th drone point represents the i-th drone in the swarm on the observation plane, where i = 1, 2, ..., N, and N is the total number of drones in the swarm. The coordinate information of the drone point corresponding to the swarm is {X1, X2, ..., X...}. N}, where Xi is the coordinate information of the i-th drone point, which is a two-dimensional data, i = 1, 2, ..., N, and N is the total number of drones in the cluster;

[0035] Step 2: Based on the position information of the UAV points in the two-dimensional plane, traverse and find the extreme value UAVs to obtain the initial boundary nodes; the specific process is as follows: in {X1,X2,...,X... N The algorithm iterates through the data to find the drone points corresponding to the minimum, maximum, minimum, and maximum values ​​of the x-coordinate, and the y-coordinate, respectively. These points are then used as the extreme value drones, forming the extreme value drone set. in The coordinates of the j-th extreme UAV point, j = 1, 2, ..., n, 2 ≤ n ≤ 4, are the initial set of boundary nodes. in The coordinate information of the k-th boundary node, k = 1, 2, ..., p, 2 ≤ p ≤ N;

[0036] Step 3: Based on the position information of the UAV points in the two-dimensional plane and the initial boundary nodes, determine the state of all UAV points relative to the initial boundary; for all UAV points, determine the relative positional relationship between each UAV point and the boundary formed by the initial boundary nodes, and obtain the cluster UAV state set I0 = {IN1, IN2, ..., IN...} N}, where IN i IN1 represents the relative positional relationship between the i-th drone point and the boundary formed by the initial boundary node. IN1 = 1 indicates that the i-th drone point is inside the boundary formed by the initial boundary node, and IN1 = 0 indicates that the i-th drone point is outside the boundary formed by the initial boundary node. The initial boundary does not enclose the i-th drone point. i = 1, 2, ..., N, where N is the total number of drones in the cluster.

[0037] In the algorithm application scene, the boundary formed by the boundary nodes is a polygon, which is called boundary polygon. The boundary polygon is generally a convex polygon. The relative position of the UAV point to the boundary polygon is determined by using the ray method. The basic idea is that a ray is drawn from the point to be determined in any direction, and then the number of intersection points of the ray and the polygon is calculated. According to the parity of the number of intersection points, the relative position of the point to the polygon can be determined. If a point is inside the polygon, the ray drawn at any angle will definitely have one intersection point with the polygon or overlap with the boundary line of the polygon. If a point is outside the polygon, the ray drawn at any angle will have one intersection point with the polygon, or two intersection points, or no intersection point, or overlap with the boundary line of the polygon.

[0038] In order to distinguish the case that the ray has only one intersection point with the polygon, when the ray of the external point has one intersection point with the polygon, the intersection point must be a vertex. If the ray is moved up or down by 0.0001 units, it will either have two intersection points or no intersection point. The number of intersection points of the ray and the polygon will not change when the internal point is moved. In order to distinguish the case that the ray overlaps with the boundary line of the polygon, when the ray of the internal point overlaps with the boundary line of the polygon, the internal point must be on the edge of the polygon.

[0039] The specific process of determining the state of all UAV points relative to the initial boundary is as follows:

[0040] Step 301, determine whether the point to be determined is a boundary node. If the point to be determined is a boundary node, the point to be determined is an internal point. If the point to be determined is not a boundary node, execute step 302;

[0041] Step 302, determine whether the point to be determined is on the horizontal edge of the boundary polygon. If the point to be determined is on the horizontal edge of the boundary polygon, the point to be determined is an internal point. If the point to be determined is not on the horizontal edge of the boundary polygon, execute step 303;

[0042] Step 303, determine the number of intersection points of the horizontal ray drawn by the point to be determined and the polygon edge. If the horizontal ray has two intersection points with the polygon edge, the point to be determined is an external point. If the horizontal ray has one intersection point with the polygon and the vertical coordinate of the point to be determined is the same as that of one vertex of the edge generating the intersection point, move the point to be determined by 0.0001 units, and continue to determine the relative position of the moved point to the polygon according to steps 301-303.

[0043] Step 4, according to the state of the UAV point relative to the initial boundary, find new boundary nodes in the UAV points not enveloped by the initial boundary, update the boundary, and until all UAV points are inside the boundary; the specific process is as follows:

[0044] Step 401: Based on the state set I0 of the clustered drones, obtain the set of drone points O0 outside the boundary B0 = {OU1, OU2, ..., OU...} q}, where OU s Let s be the s-th drone point outside the boundary B0, where s = 1, 2, ..., q, 0 ≤ q ≤ N-2, and N is the total number of drones in the cluster. Then, find the extreme value drone point in O0 and add it to the boundary node set B0. In the boundary node set B0, for the k-th boundary node... The concavity / convexity of the polygon is determined by deleting concave boundary points until the boundary polygon formed by the set of boundary nodes is a convex polygon. At this point, the set of boundary nodes is updated to B1, where k = 1, 2, ..., p, 2 ≤ p ≤ N. The concavity / convexity of the polygon vertices is determined by the cross product of the adjacent edge vectors of the polygon. According to basic mathematical knowledge, all interior angles of a convex polygon are less than 180°, and the cross product of all adjacent edge vectors has the same sign. Since the polygon formed by the initially selected extreme point of the drone must be a convex polygon, the sign of the cross product of the adjacent edge vectors at this point corresponds to the convex boundary point. The sign of the cross product of the adjacent edge vectors at this point is recorded. Boundary nodes with different signs from the cross product of the adjacent edge vectors at this point are concave boundary points.

[0045] Step 402: For the boundary node set, use the boundary polygon delineation algorithm to connect the boundary nodes in sequence to form a convex polygon. The specific process of the boundary polygon delineation algorithm is as follows: First, find the two drone points with the largest and smallest x-coordinate values ​​in the boundary node set B. The straight line formed by these two points divides the vertices of the boundary polygon into the upper left part B. l B in the lower right corner r For the boundary nodes in the upper left part, sort them in ascending order of their x-coordinate values; for the lower right part B... r The boundary nodes are arranged in descending order of their horizontal coordinates; the boundary nodes in the upper left and lower right parts of the sorted boundary nodes are combined into one to form a sorted set of boundary nodes, and the nodes in the set of boundary nodes are connected in sequence to form a polygon.

[0046] Step 403: Following the method in step 401, continuously update the boundary node set B until all UAV points are inside the boundary polygon. i =1, i=1,2,···,N.

[0047] Example

[0048] The effectiveness of the two-dimensional planar UAV swarm boundary node identification method designed in this invention was verified through simulation.

[0049] The cluster is composed of 30 unmanned aerial vehicles, adjacent unmanned aerial vehicle individuals are spaced about 0.3 km, the unmanned aerial vehicle cluster is distributed less than 7 km in the lateral direction and less than 3 km in the longitudinal direction. The unmanned aerial vehicle cluster is projected onto a two-dimensional observation plane, and the coordinate information of the 30 unmanned aerial vehicle points is as shown in Table 1:

[0050] Table 1: Unmanned aerial vehicle point coordinates

[0051] Number Abscissa Ordinate Number Abscissa Ordinate 1 3.579 2.923 16 4.879 1.428 2 6.151 0.430 17 1.466 1.467 3 4.162 1.011 18 5.387 1.374 4 6.341 1.434 19 1.579 0.258 5 2.800 1.025 20 0.101 2.624 6 1.429 2.263 21 2.491 0.732 7 1.959 1.346 22 1.202 0.505 8 4.654 0.869 23 0.008 2.174 9 6.266 2.967 24 5.510 2.182 10 0.115 0.373 25 6.356 1.990 11 0.437 2.746 26 0.893 1.848 12 6.527 2.598 27 5.945 0.965 13 3.233 1.977 28 1.927 0.217 14 3.757 2.322 29 1.815 0.646 15 1.675 1.771 30 5.156 1.849

[0052] The 30 unmanned aerial vehicle points are traversed to find the unmanned aerial vehicle points corresponding to the minimum horizontal coordinate, the maximum horizontal coordinate, the minimum vertical coordinate and the maximum vertical coordinate, which are extreme unmanned aerial vehicle points. The extreme unmanned aerial vehicle points are initial boundary nodes. In this example, the initial boundary nodes are shown in Table 2:

[0053] Table 2: Initial boundary node coordinates

[0054] Number Abscissa Ordinate 9 6.266 2.967 12 6.527 2.598 23 0.008 2.174 28 1.927 0.217

[0055] According to the initial boundary, the relative positions of all unmanned aerial vehicle points to the initial boundary are determined to obtain a cluster unmanned aerial vehicle state set, as shown in Table 3 (where 1 represents that the unmanned aerial vehicle point is within the initial boundary, and 0 represents that the unmanned aerial vehicle point is outside the initial boundary):

[0056] Table 3: Relative position relationship of cluster unmanned aerial vehicle and initial boundary

[0057]

[0058]

[0059] At this time, the initial boundary of the unmanned aerial vehicle cluster found by the algorithm is shown in Figure 4 . The extreme unmanned aerial vehicles in the set of unmanned aerial vehicle points not enveloped by the initial boundary are continuously found and added to the boundary node set. In the boundary node set, the concave-convex nature of each boundary node is judged, and the concave boundary points are deleted until the boundary polygon formed by the boundary node set is a convex polygon, at which time the boundary node set is updated. The boundary node set is continuously updated in this way until all unmanned aerial vehicle points are inside the boundary polygon. The finally identified unmanned aerial vehicle cluster boundary nodes are shown in Table 4:

[0060] Table 4: Initial boundary node coordinates

[0061] Number Abscissa Ordinate 1 3.579 2.923 2 6.151 0.430 4 6.341 1.434 9 6.266 2.967 10 0.115 0.373 11 0.437 2.746 12 6.527 2.598 20 0.101 2.624 23 0.008 2.174 28 1.927 0.217

[0062] The unmanned aerial vehicle cluster boundary identified by the final algorithm is shown in Figure 5 . It can be found through observation and calculation that the boundary polygon at this time is a convex polygon, and all unmanned aerial vehicle points are inside the boundary. This result verifies the effectiveness of the two-dimensional plane unmanned aerial vehicle cluster boundary node identification method designed by the present application.

[0063] Therefore, the application adopts the above-mentioned two-dimensional UAV cluster boundary node identification method based on the extreme value envelope iteration, simplifies the UAV cluster boundary node identification problem into a boundary identification problem of a series of points in a two-dimensional plane. A simple and understandable mathematical principle is used for boundary point identification, and a boundary polygon outlining algorithm is designed to quickly connect the boundary nodes to form the cluster boundary, which has a broad application prospect.

[0064] Finally, it should be noted that: the above examples are only used to illustrate the technical solutions of the present application, but not to limit them. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can still be modified or replaced by equivalents, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present application.

Claims

1. A method for identifying boundary nodes of a two-dimensional UAV swarm based on extreme value envelope iteration, characterized in that, Includes the following steps: Step 1: Project the aerial drone swarm onto a specific two-dimensional plane, treating each drone as a point mass, resulting in a two-dimensional plane containing a number of drone points distributed throughout the swarm. When observing the aerial drone swarm using sensors on a ground base station, the three-dimensional moving drone swarm is considered as a series of points moving within the sensor's observation plane. A point on the observation plane corresponds to a drone in the swarm. Using positioning technology, the position information of the drone swarm is obtained, and the coordinate information of the points on the observation plane is acquired. The i-th drone point represents the i-th drone in the swarm on the observation plane, where i = 1, 2, ..., N, and N is the total number of drones in the swarm. The coordinate information of the drone points corresponding to the drone swarm is {X1, X2, ..., X...}. N }, where X i The coordinate information of the i-th drone point is a two-dimensional data, i = 1, 2, ..., N, where N is the total number of drones in the cluster. Step 2: Based on the position information of the UAV points in the two-dimensional plane, traverse and find the extreme UAVs to obtain the initial boundary nodes; Step 3: Based on the position information of the UAV points in the two-dimensional plane and the initial boundary nodes, determine the state of all UAV points relative to the initial boundary; For all drone points, determine the relative positional relationship between each drone point and the boundary formed by the initial boundary node, and obtain the cluster drone state set I0 = {IN1, IN2, ..., IN...} N }, where IN i This represents the relative positional relationship between the i-th drone point and the boundary formed by the initial boundary node. IN1 = 1 indicates that the i-th drone point is inside the boundary formed by the initial boundary node, and IN1 = 0 indicates that the i-th drone point is outside the boundary formed by the initial boundary node. The initial boundary does not enclose the i-th drone point. i = 1, 2, ..., N, where N is the total number of drones in the swarm. The specific process is as follows: Step 301: Determine whether the point to be judged is a boundary node. If the point to be judged is a boundary node, then the point to be judged is an internal point; if the point to be judged is not a boundary node, proceed to step 302. Step 302: Determine whether the point to be determined is on the horizontal edge of the boundary polygon. If the point to be determined is on the horizontal edge of the boundary polygon, then the point to be determined is an internal point; if the point to be determined is not on the horizontal edge of the boundary polygon, proceed to step 303. Step 303: Determine the number of intersections between the horizontal ray emanating from the point to be judged and the edge of the polygon. If the horizontal ray intersects the edge of the polygon twice, then the point to be judged is an external point. If the horizontal ray intersects the polygon once and the ordinate of the point to be judged is the same as that of a vertex of the edge that produces the intersection, then move the point to be judged by 0.0001 units and continue to judge the relative position of the moved point to the polygon according to steps 301-303. Step 4: Based on the state of the drone points relative to the initial boundary, find new boundary nodes among the drone points not enclosed by the initial boundary, update the boundary, until all drone points are inside the boundary. The specific process is as follows: Step 401: Based on the state set I0 of the clustered drones, obtain the set of drone points O0 outside the boundary B0 = {OU1, OU2, ..., OU...} q }, where OU s Let s be the s-th drone point outside the boundary B0, where s = 1, 2, ..., q, 0 ≤ q ≤ N-2, and N is the total number of drones in the cluster. Then, find the extreme value drone point in O0 and add it to the boundary node set B0. In the boundary node set B0, for the k-th boundary node... The concavity and convexity of the boundary nodes are judged, and concave boundary points are deleted until the boundary polygon formed by the boundary node set is a convex polygon. At this time, the boundary node set is updated to B1, where k = 1, 2, ..., p, 2 ≤ p ≤ N. Step 402: For the set of boundary nodes, use the boundary polygon delineation algorithm to connect the boundary nodes in sequence to form a convex polygon; Step 403: Following the method in step 401, continuously update the boundary node set B until all UAV points are inside the boundary polygon. i =1, i=1,2,···,N.

2. The two-dimensional UAV swarm boundary node identification method based on extreme value envelope iteration according to claim 1, characterized in that: In step 2, based on the position information of the UAV points in the two-dimensional plane, the extreme value UAV is found through traversal, and the specific process of obtaining the initial boundary nodes is as follows: In {X1,X2,...,X... N The algorithm iterates through the data to find the drone points corresponding to the minimum, maximum, minimum, and maximum values ​​of the x-coordinate, and the y-coordinate, respectively. These points are then used as the extreme value drones, forming the extreme value drone set. in The coordinates of the j-th extreme UAV point, j = 1, 2, ..., n, 2 ≤ n ≤ 4, are the initial set of boundary nodes. in Let k be the coordinate information of the k-th boundary node, k = 1, 2, ..., p, 2 ≤ p ≤ N.

3. The two-dimensional UAV swarm boundary node identification method based on extreme value envelope iteration according to claim 1, characterized in that: The concavity or convexity of a polygon vertex is determined by the cross product of adjacent edge vectors. According to basic mathematical knowledge, all interior angles of a convex polygon are less than 180°, and the cross product of all adjacent edge vectors has the same sign. Since the polygon formed by the initially selected extreme value drone points must be a convex polygon, the sign of the cross product of adjacent edges corresponds to the convex boundary point. The sign of the cross product of adjacent edges is recorded at this time. Boundary nodes with different signs from the cross product of adjacent edges at this time are concave boundary points.

4. The two-dimensional UAV swarm boundary node identification method based on extreme value envelope iteration according to claim 3, characterized in that, The specific process of the boundary polygon delineation algorithm is as follows: First, find the two drone points with the largest and smallest x-coordinate values ​​in the boundary node set B. The straight line formed by these two points divides the vertices of the boundary polygon into the upper left part B. l B in the lower right corner r For the boundary nodes in the upper left part, sort them in ascending order of their x-coordinate values; for the lower right part B... r The boundary nodes are arranged in descending order of their horizontal coordinates; the boundary nodes in the upper left and lower right parts of the sorted boundary nodes are combined into one to form a sorted set of boundary nodes, and the nodes in the set of boundary nodes are connected in sequence to form a polygon.

Images