Carrier delivery point and target grouping collaborative decision-making method for task planning in carrier delivery mode

Abstract

The invention discloses a carrier delivery point and target grouping collaborative decision-making method for task planning in a carrier delivery mode, and belongs to the technical field of unmanned equipment guarantee. The method comprises the following steps: 1, sorting all targets; step 2, selecting a feasible delivery point set, judging whether all target points can be covered or not, if so, executing step 5, otherwise, executing step 3; step 3, recording uncovered target points, and selecting delivery points from all recorded target points; 4, fixing the selected delivery points, removing the recorded target points from the total target point set, updating the target point set, sorting, and returning to the step 2; 5, inputting the selected delivery point to obtain a target point grouping set with the delivery point as the center; and 6, fusing the isolated target points into other groups to obtain target point classifications and corresponding delivery points. The task execution efficiency and safety can be improved, the risk of being found and intercepted by enemies is reduced, and the vehicle navigation distance and execution time are shortened.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned equipment support, and relates to a collaborative decision-making method for vehicle delivery points and target grouping for mission planning in the vehicle delivery mode. Background Art

[0002] Generally speaking, the target points in the battlefield are widely distributed, and different targets are in different environments. Considering the limited flight range of unmanned aerial vehicles (UAVs), how to deploy UAVs safely and quickly into the battlefield has become a widely concerned issue in modern warfare. A feasible idea is to use high-flight-range and high-capacity vehicles to deliver UAVs. The core of this problem lies in how to select feasible delivery points on the vast battlefield to implement the UAV delivery mission. In order to quickly plan the delivery path of the vehicle and the UAV mission execution plan, a feasible idea is to use clustering algorithms to divide the target points and select the delivery points. Currently, the mainstream clustering algorithms can be divided into three categories: partitioning methods represented by K-means and KNN; hierarchical methods represented by AGNES and DIANA; density-based methods represented by DBSCAN and DPC. Existing improvements span all the main aspects of algorithm design, including algorithm input, process, output, and concept modification. However, most improvements focus more on the information of the data itself and the solution speed of the algorithm, and rarely pay attention to the constraints in the clustering process. For example, a large-scale UAV mission planning method and system (CN119045505A) in a Chinese invention patent uses a spectral clustering algorithm to divide multiple initial mission sequences into multiple groups when facing large-scale UAV missions, but it does not consider the actual constraints that will occur during the mission execution and is not applicable to the actual combat process. Another example is the UAV and manned aircraft joint search and rescue path planning method under airspace stratification (CN119759042A) in a Chinese invention patent, which proposes a rule-based airspace stratification model to provide an initial range for the joint path planning of manned aircraft and UAVs. This invention considers the environmental division in the rescue scenario and ensures the rescue efficiency under the condition of meeting the constraints of heterogeneous aircraft types. However, due to the heterogeneous constraints in different scenarios, it cannot be migrated to the battlefield for use. For example, a millimeter-wave anti-blocking multi-UAV deployment method based on building geometric analysis (CN114039652A) in a Chinese invention patent proposes a BT-K-means multi-UAV clustering algorithm, which performs multiple iterations according to the minimum path loss principle to obtain the clustering results of users and the initial positions of UAVs. Similarly, although this invention fully considers the actual constraints between cities, it is still not applicable to the battlefield environment. In order to respond to the actual combat requirements, when selecting the delivery point, it is necessary to consider the actual constraints such as the flight range of UAVs, no-fly zones, and radars to ensure the vehicle can execute the delivery combat mission safely and efficiently. Summary of the Invention

[0003] To solve the above-mentioned technical problems, the present invention proposes a collaborative decision-making method for vehicle delivery points and target grouping in the task planning under the vehicle delivery mode. It is mainly applied to the task planning and delivery point selection of unmanned aerial vehicle (UAV) swarms in the battlefield environment. This method comprehensively considers multiple factors such as enemy radar threats, no-fly zone restrictions, and UAV flight ranges. Through inputting the position information of target points and calculating the local density and relative distance for preliminary screening, and then proposes a multi-constraint fusion strategy to select delivery points, ensuring that the delivery points are outside the radar coverage range and avoiding no-fly zones. At the same time, the problem of uneven target point density is solved by the principle of regional equal division. For target points within the radar coverage range and no-fly zones, the map grid is traversed, and the grid with the highest value is selected as the new delivery point. By fusing isolated target points into other groups, the number of delivery points is reduced, and the vehicle navigation distance and potential risks are decreased. This method designs a logic-based clustering method, effectively solving the limitations of traditional clustering algorithms in practical applications and improving the task execution efficiency and safety of UAV swarms in complex battlefield environments.

[0004] To achieve the above object, the specific technical solution adopted by the present invention is as follows:

[0005] A collaborative decision-making method for vehicle delivery points and target grouping in the task planning under the vehicle delivery mode, the collaborative decision-making method for vehicle delivery points and target grouping includes the following steps:

[0006] Step 1, input the position information of target points, use statistical formulas to solve and obtain the local density and relative distance of each point, and sort all targets in descending order according to the product of the local density and the relative distance;

[0007] Step 2, based on the sorted target points, propose a constraint fusion strategy considering multiple constraints to select a set of feasible delivery points. According to the selected delivery points, judge whether the delivery points can cover all target points. If yes, execute Step 5; otherwise, execute Step 3;

[0008] Step 3, record the target points not covered in Step 2, use the grid map, traverse the map grid, and select delivery points for all recorded target points according to the value of the grid;

[0009] Step 4, fix the delivery points selected in Step 3, remove the target points recorded in Step 3 from the total target point set, and update the target point set. Sort the updated target point set in descending order according to the product of the local density and the relative distance, and return to Step 2;

[0010] Step 5, input the selected delivery points, adopt the minimum distance principle to conduct target grouping, and obtain a set of target point groups centered on the delivery points;

[0011] Step 6: Input the set of grouped target points, fuse the groups that contain only one target point, and merge the isolated target points into other groups. Finally, obtain the target point classification scheme and the corresponding delivery points.

[0012] Further, the specific steps of Step 1 are as follows:

[0013] Input the position information of the target points. First, calculate the local density of each point. The local density calculation uses the Gaussian kernel formula:

[0014] (1)

[0015] In formula (1), and represent two different target points, represents the local density of target point , represents target point and is the Euclidean distance between is the preset neighborhood truncation distance. To fit the flight range limit of the drone, in the present invention, the truncation distance is set to , where is a parameter for adjusting the flight range, and its value is less than , represents the flight range of the drone.

[0016] After calculating the local density of all target points, calculate the relative distance. The relative distance refers to the distance between target point and the nearest target point with a higher local density than . If the local density of target point is the highest among all target points, let be the local density of target point , then the expression formula for its relative distance is:

[0017] (2)

[0018] Otherwise, the relative distance of target point is expressed as:

[0019] (3)

[0020] After obtaining the local density and relative distance of each target point, calculate the product of the local density and relative distance of each target point, and use this as an index to sort the target points from largest to smallest.

[0021] Further, the specific steps of Step 2 are as follows:

[0022] Step 2.1: Select delivery points using the constraint fusion strategy;

[0023] After sorting the target points, the process of selecting delivery points is carried out. Based on the sorted target points, the present invention proposes a constraint fusion strategy considering multiple constraints to select a set of feasible delivery points. According to the sorted order, the target points are input one by one for retrieval. If the current target point meets the constraint fusion strategy, it is selected as a delivery point and stored in the delivery point set. . After retrieving one target point, the retrieval of the next target point is carried out. When the selected delivery points can cover all target points or all target points have been retrieved once, the retrieval process stops. In the constraint fusion strategy, the present invention considers radar constraints, no-fly zone constraints, screening constraints, and UAV flight range constraints, and corresponding solutions and delivery point screening measures are given around the above three constraints. First, the screening conditions for the three constraints are introduced. When a target point meets the three screening conditions, it will be selected as a delivery point. Specifically as follows:

[0024] 1) Construct screening conditions around radar constraints;

[0025] In this strategy, only when the target point is outside the radar coverage range will it be selected as a delivery point. For the convenience of description, in the constraint fusion strategy, the target point is represented as , and its position is represented as , where represents the abscissa of point , and represents the ordinate of point . The radar is represented as , then its position is represented as , respectively represent the abscissa and ordinate of the radar. Then, if the target point can be selected as a delivery point, it needs to meet the condition:

[0026] (4)

[0027] where is the detection coverage range of radar . Formula (4) means that the distance between the target point and any radar needs to be greater than the detection coverage range of the radar, otherwise it cannot be selected as a delivery point.

[0028] 2) Construct screening conditions around no-fly zone constraints;

[0029] In addition, the selection of delivery points needs to consider the no-fly zone in the battlefield. The delivery points cannot be located within the no-fly zone range. The constraint is expressed as follows:

[0030] (5)

[0031] Among them, represents the position of the target point ; represents the coverage range of the th no-fly zone; represents the number of no-fly zones; is used to describe the relationship of not belonging in mathematics.

[0032] 3) Construct screening conditions around the screening constraints and the UAV flight range constraints;

[0033] This setting is used to solve the limitations of the algorithm under the condition of uneven target point density and the UAV flight range constraints. The specific form is as follows:

[0034] For the convenience of description, assume that several delivery points have been selected currently. Then, let the set of current delivery points be . Among them, represents the first selected delivery point in this set, and let be the set of all target points within the coverage of a circle with the delivery point as the center and as the radius. In the set , let the th delivery point be , and let be the set of target points covered by a circle with the center and the radius ; Let the newly added clustering center in be , and the currently retrieved point be , then there is:

[0035] (6)

[0036] Formula (6) means that the currently retrieved target point cannot be located within the coverage of the existing delivery points, otherwise it cannot be selected as a new delivery point. Among them, represents the number of delivery points in the set . When the retrieved target point satisfies the three constraints of Formula (4) to Formula (6), it is added to the delivery point set .

[0037] Step 2.2: Determine whether there are remaining target points that cannot be covered by the current delivery points;

[0038] After Step 2.1 is completed, the delivery point set is obtained. Then, calculate the number of target points covered by the delivery points and make a judgment. Let the target point set be , if all the elements in the set are target points, then we have:

[0039] (7)

[0040] Among them, represents the set of target points not covered by the current delivery point. For the convenience of description, it is named the set of unassigned points, and the target points not covered are unassigned points. If the set has 0 elements, then execute step 5. If the number of its elements is not 0, then execute step 3.

[0041] Further, the specific steps of step 3 are as follows:

[0042] In this step, the corresponding delivery point selection will be carried out for the unassigned points obtained in step 2. From step 2, the set of unassigned points is obtained. Since this algorithm will perform multiple rounds of iteration, when executing step 3 each time, is different. In order to record the unassigned points that appear during the entire algorithm iteration process, another global set of unassigned points is set as , which is initially an empty set. is the coverage radius control parameter for the delivery point selection of unassigned points, and is used to control the coverage range size of the selected delivery point in this step later. Its size limit is , and the execution steps are as follows:

[0043] Step 3.1, generate a 2D battlefield environment floor plan and grid the battlefield. Extract all unassigned points and record the grids that intersect with the circular domain with a radius of where each unassigned point is located, and record the unassigned points of this round: ;

[0044] Step 3.2, calculate the value of the recorded grids. The value function is: , among which, represents the number of unassigned points covered within the circular domain with this grid as the center and as the radius; represents the radar threat weight; represents the radar threat received by this grid, and the specific form is:

[0045] (8)

[0046] Among them, represents the parameter of the acceptable degree of the distance between the delivery point and the radar. When the grid is within an unacceptable range, its radar threat is infinite. represents the radar and the current grid the distance between; represents an infinite number; represents the radar set.

[0047] Step 3.3, after calculating the values of all grids, select the grid with the highest value as the new delivery point, store this grid in the set , and update the set of points without attribution :

[0048] (9)

[0049] wherein, represents the points without attribution within the circular domain where the grid with the maximum value is located; represents the grid with the maximum value; represents the maximum flight range of the UAV; represents the coverage radius control parameter for selecting the delivery point for the points without attribution;

[0050] Finally, repeat Step 3-2 to Step 3-3 until there are no points without attribution in .

[0051] The specific steps of Step 4 are as follows:

[0052] Store the delivery points selected in Step 3 , since the points without attribution cannot find delivery points through the constraint fusion strategy, and the corresponding delivery points are found through Step 3. Also, because the target points will affect each other's local density and relative distance, it is necessary to discard the points without attribution, update the local density and relative distance of the remaining target points, and reorder them. First, remove the points without attribution recorded in Step 3 from the total target point set , and update the target point set. Then, according to the formula in Step 1, recalculate the local density and relative distance of each target point in the set , and sort the target points in descending order according to the product of the local density and relative distance. Finally, let , and return to Step 2.

[0053] The specific steps of Step 5 are as follows:

[0054] This step is established after all target points have their corresponding delivery points. At this time, there is a final delivery point set , and each element in the set represents a delivery point, Indicates the number of delivery points. Then, a target grouping process is carried out, that is, which drone dispatched by the delivery point will execute the task of the target point. This step uses the principle of minimum distance for grouping, that is, the target point is assigned to the group of the delivery point closest to it, and finally a grouping set is obtained. , where each element in the set represents a grouping. Taking as an example, is the target point grouping with as the delivery point.

[0055] The specific content of step 6 is as follows:

[0056] After step 5 completes the target point grouping, there may be some isolated points (a grouping contains only one target point). This step is to enable such target points to be integrated into other groupings by changing the delivery points of other groupings, thereby reducing the navigation distance of the vehicle by reducing the number of delivery points and reducing the potential risks that the vehicle may suffer. The specific process is as follows:

[0057] First, retrieve the set , and add the elements with a modulus length of 1 in to the set . For convenience of description, it is named the isolated point set. After the screening is completed, the elements in are selected in turn for judgment. Let the currently selected point be , which is the th element in the isolated point set . Perform the following operations:

[0058] Step 5.1, select the isolated point , and retrieve the circular domain centered on the isolated point with a radius of . Classify all the target points within the circular domain except according to their belonging groups, which is expressed as:

[0059]

[0060] Among them, all the points covered within the circular domain belong to different clusters, represents the set of points within the circular domain and belonging to the th group.

[0061] Step 5.2, select the target points included in each element in in turn. If the following two conditions are met: (1) If this point is used as the new delivery point of the group it belongs to, it can make the isolated point Fusing into this group (2) does not violate the radar and no-fly zone constraints in the constraint fusion strategy. Then, the target points that meet the two conditions are recorded into a new set .

[0062] Step 5.3, when after all elements are retrieved, if the number of elements in is greater than one, then select the target point closest to the outlier in, and fuse it into the group to which the target point belongs, and update the delivery point and group information of this group.

[0063] Execute the above process until all points in the outlier set are retrieved, and finally obtain a new set of delivery points and the corresponding group set .

[0064] The beneficial effects of the present invention are as follows:

[0065] (1) The present invention comprehensively considers actual combat constraints such as enemy radar threats, no-fly zone restrictions, and UAV flight ranges, and proposes a constraint fusion strategy to ensure that the selected delivery points avoid high-risk areas and meet the UAV mission execution conditions, significantly improving the safety of vehicle delivery and the feasibility of mission planning.

[0066] (2) By dynamically calculating the local density and relative distance of target points and combining a multi-round screening mechanism, the present invention can efficiently cover all target points in the battlefield, avoid coverage blind spots caused by traditional clustering algorithms ignoring actual constraints, and ensure that the mission is executed without omission.

[0067] (3) By introducing a strategy for fusing outliers, the present invention effectively reduces the number of redundant delivery points, thereby reducing the vehicle's navigation distance and mission execution cost, while improving the utilization rate of UAVs and achieving efficient resource allocation.

[0068] (4) The method proposed by the present invention is not only applicable to the mission planning of UAV swarms in complex battlefield environments, but can also be extended to other fields that require multi-constraint collaborative decision-making, such as logistics distribution, disaster relief, etc., and has strong versatility and scalability.

[0069] In summary, the present invention can improve the mission execution efficiency and safety of UAV swarms in complex battlefield environments, reduce the risk of UAV swarms being detected and intercepted by the enemy, reduce the vehicle's navigation distance and mission execution time, and provide an efficient, safe and reliable solution for the mission planning of UAV swarms in battlefield environments. Description of the Drawings

[0070] Figure 1 It is a flowchart of the invention.

[0071] Figure 2 It is a schematic diagram of the battlefield.

[0072] Figure 3 It is a schematic diagram for simplifying the algorithm.

[0073] Figure 4 It is a schematic diagram of points without attribution after the execution of the first-round constraint fusion strategy.

[0074] Figure 5 It is a schematic diagram for selecting the delivery points to which the points without attribution in the first round belong.

[0075] Figure 6 It is a schematic diagram of grouping each target point after the execution of the second-round constraint fusion strategy.

[0076] Figure 7 It is a schematic diagram of points without attribution after the execution of the second-round constraint fusion strategy.

[0077] Figure 8 It is a schematic diagram for selecting the delivery points to which the points without attribution in the second round belong.

[0078] Figure 9 It is a schematic diagram of grouping each target point and isolated points after the execution of the third-round constraint fusion strategy.

[0079] Figure 10 It is a schematic diagram of grouping after the fusion of isolated points.

[0080] Figure 11 It is the final effect diagram of the algorithm. Specific implementation manners

[0081] The present invention will be further described below in conjunction with specific embodiments.

[0082] Figure 2 It represents the initial schematic diagram of the battlefield, where the battlefield area is 435 km × 435 km, there are 200 homogeneous targets, 14 obstacles of various sizes and shapes, 10 radars, showing the distribution of targets, obstacles, radars and the starting point of the vehicle. The obstacles are depicted in gray in the figure, the target points are marked as dots on the figure, and the starting point of the vehicle is a diamond point. The radar is a red triangle, and the area within the solid circle represents the absolute infeasible region of the vehicle, and the area within the red dashed line represents the coverage radius of the radar. A method for collaborative decision-making of vehicle delivery points and target grouping in vehicle delivery mode task planning provided in this embodiment includes the following steps:

[0083] The specific content of step 1 is as follows:

[0084] Input the position information of the target points. First, calculate the local density of each point, and the Gaussian kernel formula is used for the local density calculation:

[0085] (1)

[0086] In formula (1), and represent two different target points, represents the local density of the target point of represents the target point and is the Euclidean distance between is a preset neighborhood truncation distance. To fit the flight range limit of the drone, the present invention sets the truncation distance to take a value of , where is a parameter for adjusting the flight range, and its value is less than , represents the flight range of the drone.

[0087] After calculating the local density of all target points, the relative distance calculation is carried out. The relative distance refers to the distance between the target point and the nearest target point with a higher local density than . If the local density of the target point is the highest among all target points, let be the local density of the target point , then the expression formula for its relative distance is:

[0088] (2)

[0089] Otherwise, the relative distance of the target point is expressed as:

[0090] (3)

[0091] After obtaining the local density and relative distance of each target point, calculate the product of the local density and relative distance of each target point, and use this as an index to sort the target points from largest to smallest.

[0092] The specific steps of step 2 are as follows:

[0093] Step 2.1: Use the constraint fusion strategy to screen the delivery points;

[0094] After sorting the target points, the selection process of the delivery points is carried out. Based on the sorted target points, the present invention proposes a constraint fusion strategy considering multiple constraints to select a set of feasible delivery points. According to the sorted order, the target points are input one by one for retrieval in sequence. If the current target point meets the constraint fusion strategy, it is selected as the delivery point and stored in the delivery point set After retrieving one target point, the retrieval of the next target point is carried out. When the selected delivery point can cover all target points or all target points have been retrieved once, the retrieval process stops. In the constraint fusion strategy, the present invention takes into account radar constraints, no-fly zone constraints, screening constraints, and UAV flight range constraints, and corresponding solutions and delivery point screening measures are given around the above three constraints. First, the screening conditions for the three constraints are introduced. When a target point meets the three screening conditions, it will be selected as a delivery point.

[0095] (1) Construct screening conditions around radar constraints;

[0096] In this strategy, a target point is only selected as a delivery point when it is outside the radar coverage. For the convenience of description, in the constraint fusion strategy, the target point is represented as and its position is represented as where represents the abscissa of point and represents the ordinate of point . The radar is represented as then its position is represented as , represent the abscissa and ordinate of the radar respectively. Then if the target point can be selected as a delivery point, it needs to meet the condition:

[0097] (4)

[0098] where is the detection coverage range of radar . Formula (4) means that the distance between the target point and any radar needs to be greater than the detection coverage range of the radar, otherwise it cannot be selected as a delivery point.

[0099] (2) Construct screening conditions around no-fly zone constraints;

[0100] In addition, the selection of the delivery point needs to consider the no-fly zone in the battlefield. The delivery point cannot be within the no-fly zone range. The constraint is expressed as follows:

[0101] (5)

[0102] where represents the position of the target point ; represents the coverage range of the th no-fly zone; represents the number of no-fly zones; is used to describe the relationship of not belonging in mathematics.

[0103] (3)Construct screening conditions around the screening constraints and the UAV flight range constraints;

[0104] This setting is used to solve the limitations of the algorithm under the condition of uneven target point density and the UAV flight range constraints. The specific form is as follows:

[0105] For the convenience of description, assume that several delivery points have been selected currently. Let the set of current delivery points be . Among them represents the first selected delivery point in this set, and let be the set of all target points within the range covered by the circle with the delivery point as the center and as the radius. In the set , let the th delivery point be , and let be the set of target points covered by the circle with the center and the radius ; Let be the latest added clustering center in , and the currently retrieved point be , then there is:

[0106] (6)

[0107] Formula (6) indicates that the currently retrieved target point cannot be located within the coverage range of the existing delivery points, otherwise it cannot be selected as a new delivery point. Among them represents the number of delivery points in the set . When the retrieved target point satisfies the three constraints of formula (4) to formula (6), it is added to the delivery point set .

[0108] Complete the initial screening;

[0109] Step 2.2: Judge whether there are remaining target points that cannot be covered by the current delivery points;

[0110] After step 2-1 is completed, the delivery point set is obtained. Then calculate the number of target points covered by the delivery points and make a judgment. Let the target point set be , and the elements in the set are all target points. Then there is:

[0111] (7)

[0112] Among them, Denote the set of target points not covered by the current delivery point as the set of points without attribution for convenience of description. The target points not covered are points without attribution. If the number of elements in the set is 0, step 5 is executed; if the number of its elements is not 0, step 3 is executed.

[0113] The specific steps of step 3 are as follows:

[0114] In this step, delivery points will be selected for the points without attribution obtained in step 2. From step 2, the set of points without attribution is obtained. Since this algorithm will perform multiple rounds of iteration, when step 3 is executed each time, is different. In order to record the points without attribution that appear during the entire algorithm iteration process, another global set of points without attribution is set as , which is initially an empty set. is the coverage radius control parameter for selecting delivery points for points without attribution, and is used to control the coverage range size of the selected delivery points in this step later. Its size limit is , and the steps are as follows:

[0115] Step 3.1: Generate a 2D battlefield environment floor plan and grid the battlefield. Extract all points without attribution and record the grids that intersect with the circular domain with a radius of where each point without attribution is located, and record the points without attribution for this round: ;

[0116] Step 3.2: Calculate the value of the recorded grids. The value function is: , where represents the number of points without attribution covered within the circular domain centered on this grid with a radius of ; represents the radar threat weight; represents the radar threat received by this grid, and the specific form is:

[0117] (8)

[0118] where is a parameter representing the acceptable degree of the distance between the delivery point and the radar. When the grid is within an unacceptable range, its radar threat is infinite. represents the radar and the current grid distance between them; represents an infinite number; represents the set of radars.

[0119] After calculating the values of all grids, select the grid with the highest value As a new delivery point, store this grid in the set , and update the set of points without attribution :

[0120] (9)

[0121] Among them, represents the points without attribution within the circular area where the grid with the maximum value is located ; represents the grid with the maximum value represents the maximum flight range of the UAV represents the coverage radius control parameter for selecting delivery points for points without attribution

[0122] Finally, repeat steps 3-2 to 3-3 until there are no points without attribution in .

[0123] The specific steps of step 4 are as follows:

[0124] Store the delivery points selected in step 3 . Since the points without attribution cannot find delivery points through the constraint fusion strategy, and the corresponding delivery points are found through step 3, and because the target points will affect each other's local density and relative distance, it is necessary to discard the points without attribution, update the local density and relative distance of the remaining target points, and reorder them. First, remove the points without attribution recorded in step 3 from the total target point set , and update the target point set. Then, according to the formula in step 1, recalculate the local density and relative distance of each target point in the set , and sort the target points in descending order according to the product of the local density and relative distance. Finally, let , and return to step 2

[0125] The specific steps of step 5 are as follows:

[0126] This step is established after all target points have corresponding delivery points. At this time, there is a final delivery point set . Each element in the set represents a delivery point represents the number of delivery points. Then, the target grouping process is carried out, that is, which UAV delivering the task of the target point. This step uses the minimum distance principle for grouping, that is, dividing the target points into the group of the nearest delivery point, and finally obtaining the grouping set , where each element in the set represents a group. Taking as an example to illustrate is Grouping of target points as delivery points.

[0127] The specific steps of Step 6 are as follows:

[0128] After the target points are grouped in Step 5, there may be some isolated points (only one target point in the group). This step is to enable such target points to be incorporated into other groups by changing the delivery points of other groups, thereby reducing the sailing distance of the vehicle by reducing the number of delivery points and reducing the potential risks that the vehicle may encounter. The specific process is as follows:

[0129] First, retrieve the set Retrieve, and add the elements with a modulus length of 1 in to the set . For the convenience of narration, name it the isolated point set. After the screening is completed, sequentially select the elements in for judgment. Let the currently selected point be , which is the th element in the isolated point set . Perform the following operations:

[0130] Step 5.1, select the isolated point . Retrieve the circular domain centered on the isolated point with a radius of , and classify all the target points within the circular domain except according to their respective groups, expressed as:

[0131]

[0132] where all the points covered within the circular domain belong to different clusters, represents the set of points within the circular domain and belonging to the th group.

[0133] Step 5.2, sequentially select the target points included in each element in . If the following two conditions are met: (1) If this point is used as the new delivery point of its group, it can incorporate the isolated point into this group; (2) It does not violate the radar and no-fly zone constraints in the constraint fusion strategy. Then record the target points that meet the two conditions into a new set .

[0134] Step 5.3, when all elements have been retrieved, if the number of elements in is greater than one, select the target point in that is closest to the isolated point , and Integrate it into the group to which the target point belongs, and update the delivery points and group information of this group.

[0135] Execute the above process until the retrieval of each point in the set of isolated points is completed, and finally a new set of delivery points and the corresponding set of groups are obtained.

[0136] To more intuitively see the implementation process of the algorithm proposed by the present invention, an example is shown as follows: The simplified schematic of the battlefield environment is as Figure 3 shown. First, use the constraint fusion strategy to select preliminary delivery points for the target points, and retrieve whether there are unassigned points. In Figure 4 , we mark the unassigned points with light gray; then perform grid sampling on the points that cannot be delivered above, and comprehensively consider the radar threat and the number of target points covered to find delivery points, that is, Figure 5 the dots connected to each gray point shown in; Based on the delivery points selected in the previous step, execute the constraint fusion strategy to select strong constraint delivery points outside the radar coverage range, and allocate the target points according to the allocation rules. Figure 6 In, the black crosses represent the selected delivery points, and at the same time, each target point is connected to its corresponding delivery point by a dotted line. On the other hand, we observe that the generation of the delivery points obtained by grid sampling causes the decision value of the target points to change. So that the delivery points selected by the constraint fusion strategy in this round are different from the preliminary delivery points in 4. This in turn causes some assigned points within the radar coverage range in the initial screening to become non-cluster points.

[0137] To make it more obvious for everyone to see, we use boxes to locate the points that have changed in Figure 6 and 7 ; when such points appear, perform grid sampling again to find the corresponding delivery points, and the solution result is as Figure 8 ; Similarly, based on all the selected delivery points, execute the constraint fusion strategy and allocate the target points according to the allocation rules. At this time, all target points have corresponding delivery points that can be responsible for them, and the result is as Figure 9 shown. Considering that there are target points in a relatively isolated position, forming a group that only contains one target, in order to reduce the task execution cost, fuse such points, and the comparison results are circled with boxes of different colors in Figure 9 and 10 respectively; the final clustering allocation result is given by Figure 11 .

[0138] The above-described embodiments merely represent the implementation modes of the present invention, but should not be construed as limiting the scope of the present invention. It should be noted that for those skilled in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention.

Claims

1. A collaborative decision-making method for vehicle delivery points and target grouping in task planning under the vehicle delivery mode, characterized in that The vehicle delivery point and target grouping collaborative decision-making method includes the following steps: Step 1: Input the target point location information, use statistical formulas to solve for the local density and relative distance of each point, and sort all the targets in descending order according to the product of the local density and relative distance. Step 2: Based on the sorted target points, a constraint fusion strategy considering multiple constraints is proposed to select a set of feasible delivery points; according to the selected delivery points, judge whether the delivery points can cover all the target points; if so, execute Step 5, otherwise execute Step 3. Step 3: Record the target points not covered in Step 2, use a grid map to traverse the map grid, and select delivery points for all the recorded target points according to the value of the grid. Step 4: Fix the delivery points selected in Step 3, remove the target points recorded in Step 3 from the total target point set, and update the target point set; sort the updated target point set in descending order according to the product of the local density and relative distance, and return to Step 2. Step 5: Input the selected delivery points, and use the minimum distance principle to perform target grouping to obtain a set of target point groupings centered on the delivery points. Step 6: Input the set of target point groupings, fuse the groups that only include one target point, and fuse the isolated target points into other groups to finally obtain a target point classification scheme and the corresponding delivery points.

2. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 1, wherein The specific content of Step 1 is as follows: Input the target point location information. First, calculate the local density of each point. The local density calculation uses the Gaussian kernel formula: （1） In formula (1), and represent two different target points, represents the local density of the target point ; represents the Euclidean distance between the target points and ; is a preset neighborhood truncation distance; to fit the flight range limit of the UAV, let the truncation distance take the value of , where is a parameter for adjusting the flight range, and its value is less than , represents the flight range of the UAV; After calculating the local density of all target points, the relative distance calculation is carried out; the relative distance refers to the distance between the target point and the nearest target point with a higher local density than ; if the local density of the target point is the highest among all target points, let be the local density of the target point , then its relative distance expression formula is: （2） Otherwise, the target point is represented by the relative distance as follows: （3） After obtaining the local density and relative distance of each target point, calculate the product of the two for each target point, and use this as an index to sort the target points from large to small.

3. The collaborative decision-making method for vehicle delivery points and target groups in mission planning under the vehicle delivery mode according to claim 2, wherein, The specific content of Step 2 is as follows: Step 2.1: Use the constraint fusion strategy to screen the delivery points. After sorting the target points, the selection process of the delivery points is carried out; based on the sorted target points, a constraint fusion strategy considering multiple constraints is proposed to select a set of feasible delivery points; according to the sorted order, the target points are input one by one for retrieval in sequence. If the current target point meets the constraint fusion strategy, it is selected as a delivery point and stored in the delivery point set ; after retrieving one target point, the retrieval of the next target point is carried out; when the selected delivery points can cover all target points or all target points have been retrieved once, the retrieval process stops; in the constraint fusion strategy, radar constraints, no-fly zone constraints, screening constraints and UAV flight range constraints are considered, and corresponding solutions and delivery point screening measures are given around the above three constraints; when the target point meets the three screening conditions, it will be selected as a delivery point Step 2.2: Judge whether there are remaining target points that cannot be covered by the current delivery points. After step 2.1 is completed, a set of delivery points is obtained , and then the number of target points covered by the delivery points is calculated and judged. Let the set of target points be , and the elements in the set are all target points. Then there is （7） Among them, represents the set of target points not covered by the current delivery point. For the convenience of description, it is named the set of points without attribution, and the target points not covered are points without attribution. If the set has 0 elements, then step 5 is executed. If the number of its elements is not 0, then step 3 is executed.

4. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 3, wherein, In Step 2.1, the screening conditions for the three constraints are as follows: 1) Construct screening conditions around the radar constraint. The target point will only be selected as the delivery point when it is outside the radar coverage area; in the constraint fusion strategy, the target point is represented as , and its position is represented as , where represents the abscissa of point , represents the ordinate of point ; the radar is represented as , then its position is represented as , respectively represent the abscissa and ordinate of the radar; then if the target point can be selected as the delivery point, the condition to be satisfied is: （4） Among them, is the detection coverage range of the radar The formula (4) indicates that the distance between the target point and any radar needs to be greater than the detection coverage range of the radar, otherwise it cannot be selected as the delivery point; 2) Construct screening conditions around the no-fly zone constraint. The selection of delivery points needs to consider the no-fly zone in the battlefield. The delivery points cannot be located within the no-fly zone range. The constraint is expressed as follows: （5） Among them, represents the position of the target point ; represents the coverage range of the th no-fly zone; represents the number of no-fly zones; is used to describe the relationship of not belonging to in mathematics; 3) Construct screening conditions around the screening constraint and the UAV flight range constraint. Assume that several delivery points have been selected currently, and let the set of current delivery points be ; where represents the first selected delivery point in this set, and let be the set of all target points within the circle with the delivery point as the center and as the radius; in the set , let the th delivery point be , and let be the set of target points covered by the circle with the center and the radius ; additionally, let the newly added clustering center in be , and the currently retrieved point be , then there is: （6） Formula (6) represents the target point currently retrieved cannot be located within the coverage range of the existing delivery points; otherwise, it cannot be selected as a new delivery point, where represents the set the number of delivery points in; When the target point to be retrieved meets the three constraints of Formula (4) to Formula (6), it is added to the delivery point set .

5. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 3, wherein, The specific content of Step 3 is as follows: The delivery point selection will be carried out for the unassigned points obtained in step 2; from step 2, the set of unassigned points is obtained ; Since this algorithm will perform multiple rounds of iteration, when step 3 is executed each time is different. In order to record the unassigned points that appear during the entire algorithm iteration process, another global set of unassigned points is set as , initially an empty set; is the coverage radius control parameter for the delivery point selection of unassigned points, and will be used later to control the coverage range of the delivery points selected in this step. Its size limit is .

6. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 5, characterized in that, The execution steps of Step 3 are as follows: Step 3.1, generate a 2D battlefield environmental plan view and grid the battlefield; extract all unassigned points and record the grids that intersect with the circular domains with a radius of where each unassigned point is located, and record the unassigned points in this round: ; Step 3.2, calculate the value of the recorded grid, and the value function is: , where represents the number of unassigned points covered within the circular domain centered on this grid with as the radius; represents the radar threat weight; represents the radar threat received by this grid, and the specific form is: （8） Among them, A parameter indicating the acceptable degree of the distance between the delivery point and the radar. When the grid is within an unacceptable range, its radar threat is infinite; Indicates the radar And the current grid The distance between; Represents an infinite number; Indicates the radar set; Step 3.3, after calculating the values of all grids, select the grid with the highest value as the new delivery point, and store this grid in the set , and update the set of points without attribution : （9） Among them, represents the unassigned points within the circular area where the maximum value grid is located; represents the maximum value grid; represents the maximum flight range of the drone; represents the coverage radius control parameter for selecting delivery points for unassigned points; Finally, repeat Step 3-2 to Step 3-3 until there are no unassigned points in , and finally obtain the set of unassigned point delivery points .

7. A collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 5, characterized in that The specific content of Step 4 is as follows: Delivery point selected in step 3 , since the point without an attribution point cannot find a delivery point through the constraint fusion strategy, and the corresponding delivery point is found through step 3. Also, because the target points will affect each other's local density and relative distance, it is necessary to discard the point without an attribution point, update the local density and relative distance of the remaining target points, and reorder them. First, the point without an attribution point recorded in step 3 is removed from the total target point set to update the target point set. Then, according to the formula in step 1, recalculate the local density and relative distance of each target point in the set , and sort the target points in descending order according to the product of the local density and the relative distance. Finally, let , and return to step 2.

8. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 7, characterized in that, The specific content of Step 5 is as follows: Step 5 is established after all target points have their corresponding delivery points, and at this time, there is a final delivery point set , and each element in the set represents a delivery point, represents the number of delivery points; After that, the target grouping process is carried out, that is, which drone delivered from the delivery point will execute the task of the target point; the minimum distance principle is used for grouping, that is, the target point is assigned to the group of the delivery point closest to it, and finally the grouping set is obtained , where each element in the set represents a grouping is the target point grouping with as the delivery point 9. The collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 8, characterized in that, The specific content of Step 6 is as follows: First, retrieve the set , and add the elements with a modulus length of 1 in to the set . For convenience of description, name it the isolated point set. After the screening is completed, select the elements in in sequence for judgment. Let the currently selected point be , and be the th element in the isolated point set . Perform the following operations: Step 5.1, select the isolated points , for the isolated points as the center, and the radius is to retrieve the circular domain, and classify all the target points within the circular domain except according to their respective groups, expressed as: , where all the points covered within the circular domain respectively belong to different clusters, represents the set of points that are within the circular domain and belong to the th group; Step 5.2, sequentially select the target points included in each element in, if two conditions are satisfied, record the target points that satisfy the two conditions into a new set ; Step 5.3, when after all elements are retrieved, if the number of elements in is greater than one, select the target point closest to the outlier and merge it into the group to which the target point belongs, and update the delivery point and group information of the group; Execute the above process until the retrieval of each point in the set of isolated points is completed, and a new set of delivery points and the corresponding grouping set are finally obtained.

10. A collaborative decision-making method for vehicle delivery points and target groups in task planning under the vehicle delivery mode according to claim 9, characterized in that The two conditions in step 5.2 are as follows: (1) If this point serves as a new delivery point for the affiliated group, it can integrate the isolated point into this group; (2) It does not violate the radar and no-fly zone constraints in the constraint integration strategy.

Images