Aircraft collaborative coverage configuration optimization method based on probability Voronoi segmentation method

Abstract

The invention discloses an aircraft collaborative coverage configuration optimization method based on a probability Voronoi segmentation method. The method comprises the following steps: 1) calculating a potential range of a target group according to fuzzy target pointing information; 2) calculating discretized potential probability value distribution based on the target potential area; 3) according to the current position of the aircraft, distributing a to-be-covered area of the aircraft by using a Voronoi partitioning algorithm; and 4) according to the potential probability value distribution condition in the distribution area, giving the optimal expected position of each aircraft corresponding to the minimum overall coverage cost, thereby obtaining the optimal expected configuration of the aircraft cluster. According to the method, the problem of uncertainty of a low, small and slow target radar signal weak target potential area is solved, and the cooperative coverage detection capability of multiple aircrafts is guaranteed.

Description

Technical Field

[0001] This invention belongs to the field of UAV cooperative control technology, specifically relating to a method for optimizing aircraft cooperative coverage configuration based on probabilistic Vino segmentation. Background Technology

[0002] With the increasing demand for collaborative detection and search by multiple UAVs, ground-based radar often exhibits significant ambiguity in the initial localization of low-altitude, small, and slow-moving target groups due to limited resolution, detection blind spots, and multipath interference, failing to provide accurate spatial distribution information. To compensate for this deficiency, it is necessary to utilize the collaborative stitching of individual UAVs' fields of view as the formation approaches the target area to achieve reconnaissance coverage of the potential target region.

[0003] Current methods mostly rely on geometric distances or fixed formation templates, making it difficult to simultaneously consider target position uncertainties and multi-aircraft field-of-view fusion performance. In actual flight, sensor errors, flight attitude deviations, or communication delays can all lead to insufficient field-of-view overlap or gaps, lacking an adaptive balance mechanism between coverage accuracy and resource utilization. Therefore, there is an urgent need for a novel formation configuration design method that takes into account target probability distribution, aircraft dynamic maneuvering characteristics, and field-of-view stitching effectiveness to achieve high-precision coverage and collaborative detection of low-altitude, small, and slow-moving target groups. Summary of the Invention

[0004] The purpose of this invention is to provide a method for optimizing the cooperative coverage configuration of aircraft based on the probabilistic Vino segmentation method, which solves the problem of uncertainty in the potential area of ​​small, slow targets with weak radar signals, and ensures the cooperative coverage detection capability of multiple aircraft.

[0005] To achieve the above objectives, this invention provides a method for optimizing the cooperative coverage configuration of aircraft based on the probabilistic Vino partitioning method, comprising: 1) calculating the potential range of the target group based on fuzzy target pointing information; 2) calculating the discretized potential probability value distribution based on the potential target region; 3) allocating the area to be covered by the aircraft to the current position of the aircraft using the Vino partitioning algorithm; 4) based on the potential probability value distribution within the allocated area, giving the optimal expected position of each aircraft corresponding to the minimum overall coverage cost, thereby obtaining the optimal expected configuration of the aircraft cluster.

[0006] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation method, wherein step 1) includes: based on the azimuth angle θ of target i in polar coordinates obtained by ground radar measurement. ti Elevation angle φ ti and relative position d ti By transforming coordinates, the horizontal and vertical measurement positions p of target i in the Cartesian coordinate system are calculated. ti =(y ti , zti ),

[0007]

[0008] In the formula, y ti For the position of target i in Cartesian coordinate system, measure the Y-axis component of the position, z ti The Z-axis direction component of the measured position of target i in the Cartesian coordinate system;

[0009] Considering the radar azimuth measurement error Δ θi And pitch angle measurement error Δ φi The corresponding projection errors on the YOZ plane are as follows:

[0010]

[0011] In the formula, Δr yi The position measurement error in the Y-axis direction is Δr. zi This refers to the position measurement error along the Z-axis.

[0012] The radius of uncertainty r of the possible location of target i ti for:

[0013]

[0014] Therefore, the potential region of target i is modeled as a circular region, and the center point of this circular region is (y ti ,z ti ), with a radius of r ti The potential range of the target group; Expressed as:

[0015]

[0016] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation uses multiple circular regions on the YOZ plane to represent the potential range of the target group.

[0017] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation, wherein step 2) includes: establishing a normalized detection plane Q:

[0018]

[0019] In the formula, y min y max Let z be the minimum and maximum boundary coordinates along the Y-axis, respectively. min z max These are the minimum and maximum boundary coordinates along the Z-axis, representing the horizontal and vertical detection range of the radar, respectively.

[0020] The normalized probe plane Q is divided into N. y ×N z A uniform grid, the coordinate set of grid points The definition is as follows:

[0021]

[0022] In the formula, Indicates rounding down;

[0023] Calculate grid point p g =(y g , z g The potential probability value v for target i gi ,

[0024]

[0025] The total potential probability value v of the grid point g The potential probability value of each target is superimposed from the grid points.

[0026]

[0027] For each grid point, the potential probability value of the grid point for target i and the total potential probability value of the grid point can be calculated using formulas (7) and (8), thus obtaining the discretized potential probability value distribution.

[0028] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation, wherein step 3) includes: Let N m A swarm system of aircraft, given the current position p of aircraft j. j =(y j ,z j ),have:

[0029]

[0030] Where, p k Let k represent the current position of aircraft k, and k ≠ j; Let q represent the area assigned to aircraft j in the normalized detection plane Q, i.e., the Vino region of aircraft j. In this region, the distance from all points q to aircraft j is less than or equal to the distance from point q to any other aircraft. In this way, each aircraft is only responsible for its nearest region, ensuring that the regions responsible for each aircraft do not overlap, and the normalized detection plane Q is completely divided; q represents the points in the region responsible for aircraft j.

[0031] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation, wherein step 4) includes: designing an objective function based on weighted squared distance. Place aircraft j in a location Above, making that position To the Vino region The weighted squared distance between all points within the range is minimized;

[0032]

[0033] In formula (10), the coverage cost The current position p of the spacecraft was quantified. j The degree of deviation from the high-probability value region; by minimizing Guide the aircraft towards a high-probability value v g The grid point q is close;

[0034] To calculate the optimal desired position for each aircraft, Regarding the position of the aircraft Taking the partial derivative, the optimal expected position of spacecraft j at the current moment is:

[0035]

[0036] In formula (11), φ(q) represents the probability value of the continuous domain, corresponding to the probability value v of the discretized grid points. g Calculate the desired position of the current aircraft j, i.e., the corresponding Vino region. probability-weighted centroid By replacing the continuous domain integration with grid discretization and summation, we obtain the following equation:

[0037]

[0038] In formula (12), the aircraft j is in the Vino region. The position of any grid point (y g ,z g The total potential probability value v corresponding to this grid point. g ; For the Vino region of the aircraft Summing all grid points in the middle;

[0039] The optimal expected position is calculated for each aircraft, which is the most reasonable deployment position under the current potential probability value distribution, thereby obtaining the optimal expected configuration of the aircraft cluster.

[0040] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation further includes: 5) calculating the comprehensive coverage rate based on the rectangular field-of-view projection splicing range of each aircraft in the current optimal expected configuration of the aircraft cluster, and verifying the effectiveness of the configuration design.

[0041] The above-mentioned aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation, wherein, in step 5), the azimuth angle θ of the seeker's field of view of aircraft j is used. mj Pitch angle φ mj and projection distance d mj The field of view of aircraft j can be simplified to a rectangular region, d yj and d zj Let the horizontal width and vertical length of the simplified rectangular field of view be respectively, then the area formed by stitching together the simplified rectangular field of view of each aircraft... Represented as:

[0042]

[0043] stitching multiple fields of view Potential range of the target group The intersection is defined as the effective coverage area.

[0044]

[0045] On the discretized grid, area coverage and value coverage metrics are defined. Area coverage, denoted as η, represents how many grid points of the target potential area are covered by the stitched region. a ,

[0046]

[0047] Value Coverage η v After considering the weighted importance of the potential probability value distribution to different regions, the extent to which the stitched region covers the high-value target area is measured.

[0048]

[0049] By using the above two types of indicators, the coverage performance of the current configuration in the target potential area under the actual sensor range can be quantitatively evaluated, thereby verifying the actual effectiveness of the designed configuration.

[0050] Compared with the prior art, the beneficial technical effects of the present invention are:

[0051] (1) This invention characterizes the detection error of ground radar detection as the potential area of ​​the target, transforms the detection coverage problem in an uncertain environment into an analytical configuration optimization problem, and reduces the problem dimension and solution complexity;

[0052] (2) The present invention introduces the Vino segmentation and weighted probability centroid optimization mechanism in the design of collaborative detection array positions. Compared with the traditional design method based on a specific order arrangement, this method makes full use of the limited detection field of view, prioritizes the coverage of high-value areas, and improves the utilization rate of detection resources. Attached Figure Description

[0053] The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation of the present invention is given by the following embodiments and figures.

[0054] Figure 1 This is a flowchart of the aircraft cooperative coverage configuration optimization method based on the probabilistic Vino segmentation method according to an embodiment of the present invention.

[0055] Figure 2 This is a schematic diagram of the collaborative detection scenario in an embodiment of the present invention.

[0056] Figure 3 This is a schematic diagram illustrating the potential range modeling of the target group in an embodiment of the present invention.

[0057] Figure 4 This is a discretized grid and potential probability value distribution map in an embodiment of the present invention.

[0058] Figure 5 This is a schematic diagram illustrating the division of the aircraft's area of ​​responsibility in an embodiment of the present invention.

[0059] Figure 6 This is a schematic diagram of the optimal desired configuration of the aircraft group in an embodiment of the present invention.

[0060] Figure 7 This is a simplified matrix field-of-view diagram of the aircraft seeker head in an embodiment of the present invention.

[0061] Figure 8 This is a schematic diagram of the field of view stitching and coverage of the aircraft group formation in an embodiment of the present invention.

[0062] Figure 9 This is a schematic diagram of the convergence process curve of the overall coverage in an embodiment of the present invention. Detailed Implementation

[0063] The following will combine Figures 1-9 The method for optimizing aircraft cooperative coverage configuration based on probabilistic Vino segmentation according to the present invention will be described in further detail.

[0064] The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation of the present invention includes:

[0065] 1) Calculate the potential range of the target group based on fuzzy target orientation information;

[0066] Based on ground radar detection information (including angle and range information), the center and radius of each target potential area are defined, and the potential range of the target group is represented by multiple circular areas on a two-dimensional working plane;

[0067] In a scenario involving target detection and coordinated aircraft detection, assuming the ground radar detects N... tThe location information of each target includes azimuth, elevation, and relative distance (distance between the target and the radar). Due to the inherent error in radar angle measurement accuracy, the center and radius of the potential area for each target are defined based on the location information detected by the radar, as follows:

[0068] The azimuth angle θ of target i in polar coordinates obtained from ground radar measurements. ti Elevation angle φ ti and relative position d ti Through coordinate transformation, the horizontal and vertical measurement positions p of target i in the Cartesian coordinate system (with the origin at the radar detection center) are calculated. ti =(y ti , z ti ),

[0069]

[0070] In the formula, y ti For the position measurement of target i in Cartesian coordinate system, the Y-axis direction component (i.e., the horizontal component) is z. ti The Z-axis direction component (i.e., vertical component) of the target i is measured in the Cartesian coordinate system.

[0071] Considering the radar azimuth measurement error Δ θi And pitch angle measurement error Δ φi The corresponding projection errors on the two-dimensional plane (YOZ plane, where O is the origin of the Cartesian coordinate system) are as follows:

[0072]

[0073] In the formula, Δr yi The position measurement error in the Y-axis direction is Δr. zi This refers to the position measurement error along the Z-axis.

[0074] The radius of uncertainty r of the possible location of target i ti for:

[0075]

[0076] Therefore, the coverage area of ​​the potential location of target i (i.e., the potential region of target i) can be modeled as a circular region, with the center point of this circular region being (y... ti ,z ti ), with a radius of r ti Then the potential range of all targets (target group) Expressed as:

[0077]

[0078] 2) Calculate the discretized potential probability value distribution based on the target potential region;

[0079] Potential range of target group This constitutes the set of potential target areas that the spacecraft needs to cover for its mission. Furthermore, to establish a unified reference system for subsequent grid division, a normalized detection plane Q needs to be established.

[0080]

[0081] In the formula, y min y max Let z be the minimum and maximum boundary coordinates along the Y-axis, respectively. min z max These are the minimum and maximum boundary coordinates along the Z-axis, with the origin of the coordinate system being the radar detection center, representing the radar's horizontal and vertical detection ranges, respectively.

[0082] After determining the detection area, the normalized detection plane Q is divided into N. y ×N z A uniform grid, the coordinate set of grid points The definition is as follows:

[0083]

[0084] In the formula, Indicates rounding down;

[0085] Calculate grid point p g =(y g , z g The potential probability value v for target i gi ,

[0086]

[0087]

[0088] The total potential probability value v of the grid point g The potential probability value of each target is superimposed from the grid points.

[0089]

[0090] For each grid point, the potential probability value of that grid point for target i and the total potential probability value of that grid point can be calculated using formulas (7) and (8), thus obtaining the discretized potential probability value distribution; v g This provides a crucial basis for subsequent region segmentation, centroid solving, and configuration control.

[0091] 3) Based on the current location of the aircraft, the Vino partitioning algorithm is used to allocate the area to be covered by the aircraft (the area of ​​responsibility of the aircraft);

[0092] After obtaining the discretized potential probability value distribution, this invention is inspired by the Vino partitioning algorithm and automatically partitions the normalized detection plane Q according to the current position of the aircraft (i.e., divides the normalized detection plane Q), with each aircraft responsible for its nearest region.

[0093] Let N m A swarm system of aircraft, given the current position p of aircraft j. j =(y j ,z j ), its partition

[0094]

[0095] Where, p k Let k represent the current position of aircraft k, and k ≠ j; Let q represent the region assigned to aircraft j within the normalized detection plane Q, i.e., the Vino region of aircraft j. Within this region, the distance from all points q to aircraft j is less than or equal to the distance from point q to any other aircraft. In this way, each aircraft is only responsible for its nearest region, ensuring that the regions responsible for each aircraft do not overlap, and the normalized detection plane Q is completely divided. q represents the points within the region responsible for aircraft j. This allocation strategy provides the basis for subsequent target value weighting and position updates.

[0096] 4) Based on the distribution of potential probability values ​​within the allocated area, the optimal expected positions of each aircraft corresponding to the minimum overall coverage cost are given, thereby obtaining the optimal expected configuration of the aircraft cluster;

[0097] Assume there are N m A spacecraft, spacecraft J, is responsible for the area (Vino area). Grid points Value weight v g By designing an objective function based on weighted squared distance Place aircraft j in a location Above, making that position The weighted squared distance to all points within the Vino region is minimized;

[0098]

[0099] In formula (10), the coverage cost The current position p of the spacecraft was quantified. j The degree of deviation from the high-probability value region is minimized. Guide the aircraft towards a high-probability value v gThe grid point q is close;

[0100] To calculate the optimal desired position for each aircraft, Regarding the position of the aircraft Taking the partial derivative, the optimal expected position of spacecraft j at the current moment is:

[0101]

[0102] In formula (11), φ(q) represents the probability value of the continuous domain, corresponding to the probability value v of the discretized grid points. g Calculate the desired position of the current aircraft j, i.e., the corresponding Vino region. probability-weighted centroid By replacing the continuous domain integration with grid discretization and summation, we obtain the following equation:

[0103]

[0104] In formula (12), the aircraft j is in the Vino region. The position of any grid point (y g ,z g The total potential probability value v corresponding to this grid point. g ; For the Vino region of the aircraft Summing all grid points in the middle;

[0105] Finally, the optimal expected position for each aircraft is calculated, which is the most reasonable deployment position under the current potential probability value distribution. This calculation is the key to the entire cooperative configuration design and provides a directional basis for subsequent position evolution and convergence.

[0106] 5) Calculate the overall coverage based on the rectangular field-of-view projection splicing range of each aircraft in the current optimal expected configuration of the aircraft cluster, and verify the effectiveness of the configuration design;

[0107] Based on the azimuth angle θ of the aircraft's seeker's field of view mj Pitch angle φ mj and projection distance d mj The field of view of aircraft j can be simplified to a rectangular region, d yj and d zj Let the horizontal width and vertical length of the simplified rectangular field of view be respectively, then the area formed by stitching together the simplified rectangular field of view of each aircraft... Represented as:

[0108]

[0109] stitching multiple fields of view Potential range of the target group The intersection is defined as the effective coverage area.

[0110]

[0111] On the discretized grid, area coverage and value coverage metrics are defined. Area coverage, denoted as η, represents how many grid points of the target potential area are covered by the stitched region. a ,

[0112]

[0113] Value Coverage η v After considering the weighted importance of the potential probability value distribution to different regions, the extent to which the stitched region covers the high-value target area is measured.

[0114]

[0115] By using the above two types of indicators, the coverage performance of the current configuration in the target potential area within the actual sensor reach range can be quantitatively evaluated, thereby verifying the actual effectiveness of the designed configuration.

[0116] Example:

[0117] like Figure 1 The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation in this embodiment includes:

[0118] S1: According to Figure 2 The scenario shown is a collaborative detection scenario, which includes the possible distribution area of ​​the target group, the deployment location and flight path of the aircraft, and the field of view coverage area.

[0119] Target potential region modeling: Based on the target azimuth, elevation, and relative distance information measured by radar, and combined with the angle measurement accuracy error range, the two-dimensional boundary of the target's possible location is calculated. Each potential target region is modeled as a circular region, with its center and radius determined by the measured values ​​and the upper limit of the error. In this embodiment, as shown... Figure 3 As shown, if the measured azimuth and elevation angles of multiple targets are [0.0°, 0.3°, 0.9°, -2.8°] and [-1.4°, -0.53°, 0.53°, -1.1°] respectively, and the relative distance is 2500 meters, multiple circular boundaries with a radius of 77 meters can be obtained in the combat airspace, representing the uncertain areas of each target's location. The marker points (i.e., the centers of each circular boundary) are the target locations detected by radar, namely (0, -45), (32, -18), (94, 18), and (-104, -36). By performing spatial abstraction and mathematical modeling of the combat airspace, a two-dimensional spatial model representing the possible locations of targets is generated based on sensor information, providing a foundation for subsequent probability distribution calculations and area division.

[0120] S2: Discretized Potential Probability Value Distribution Calculation: Project the possible distribution region of the target group (i.e., the potential range of the target group) onto a two-dimensional working plane (Cartesian coordinate system YOZ plane) to obtain the normalized detection plane Q. Divide the normalized detection plane Q into N... y ×N z A uniform grid is used, and each grid point is assigned a potential probability value, thus achieving a quantitative representation of the target distribution. Each grid point is assigned a potential probability value (potential probability weight value) v based on its geometric relationship with the potential target region. g This indicates the probability that a target will appear at that location. For example... Figure 4 As shown, let y min =-285, y max =285, z min =-245,z max =245, taking the union of the uncertain regions of all targets, i.e., the maximum boundary covering all targets. Set the grid resolution parameter N. y =300 and N z =300, calculate the planar position (y) of the grid points. g ,z g Then, for each grid point, the distance to each target is calculated, and the potential probability value of each target at that grid point is calculated according to formula (7), and finally superimposed. This potential probability distribution model successfully transforms the ambiguous radar measurements into a quantized spatial distribution map. In the heat map, the bright areas correspond to high-probability target areas, providing accurate numerical basis for the subsequent field of view optimization of aircraft cluster formation.

[0121] S3: Aircraft Responsibility Area Delineation Based on Vino Partitioning Algorithm: Based on aircraft distribution, the operational airspace is divided into multiple responsibility areas, and each aircraft is assigned to a corresponding responsibility area to reduce detection blind spots and improve coverage efficiency. Based on the current positions of each aircraft, the normalized detection plane Q is partitioned using Vino partitioning to obtain the responsibility area for each aircraft.

[0122] S4: Optimal Desired Position Calculation for Each Aircraft: The centroid position of each responsibility area is calculated based on weighted probabilities, with the optimization objective being to minimize the weighted squared distance from the target point to the aircraft within that area. For example... Figure 5 As shown, taking the initial positions of the three aircraft as p1 = (150, 70), p2 = (-90, 100), and p3 = (-47, -150) as an example, white, black, and gray represent the responsibility area division results of different aircraft, and the gray dot is the current position of the aircraft. The probability centroid position is calculated according to formulas (11) and (12). These are the optimal expected positions of the three aircraft, such as... Figure 6As shown, black dots represent the optimal expected positions of each aircraft, thus obtaining the optimal expected configuration of the aircraft cluster, forming a cooperative detection array layout with high coverage and reasonable distribution.

[0123] S5: Field of View Stitching Coverage Verification and Performance Evaluation: Based on the sensor parameters of each aircraft, a rectangular field of view model is established and projected and stitched in space to form the overall perception area. For example... Figure 7 As shown, assuming the aircraft's seeker has a field of view azimuth of 11 degrees and a pitch angle of 8 degrees, an effective projection plane distance of 1 kilometer, and a detection area (sensor field of view) of length d... zj 190 meters, width d yj The target is a rectangle with a radius of 140 meters. Within a range of 1 kilometer to 500 meters, the seeker's azimuth angle gradually increases from 11 degrees to 22 degrees, and its elevation angle gradually increases from 8 degrees to 16 degrees. When the relative distance to the target aircraft exceeds 1 kilometer, the detection area is reduced, or the azimuth and elevation angles are decreased. The specific azimuth and elevation angles can be obtained by inversely solving the above formula, thus keeping the rectangular detection area unchanged. Figure 8 As shown, the overall perception coverage shape and distribution formed by multiple aircraft stitching together are intuitively displayed. By sharing and stitching the field of view information among multiple aircraft, continuous coverage of the target area is achieved, eliminating blank areas caused by insufficient detection range of a single aircraft, and improving the overall detection integrity and accuracy. The area coverage rate and value coverage rate are calculated by formulas (18) and (19) respectively. Figure 9 The dynamic changes in area coverage and value coverage as iterations progress are presented. In the initial stage, the two indicators are approximately 40% and 50%, respectively. With continuous optimization of the configuration and adjustment of the aircraft position, both indicators gradually converge to over 90%. After 15 iterations, the final value coverage reaches 96%, verifying the effectiveness and convergence of the proposed configuration optimization strategy.

Claims

1. A method for optimizing the cooperative coverage configuration of aircraft based on probabilistic Vino segmentation, characterized in that, include: 1) Calculate the potential range of the target group based on fuzzy target orientation information; 2) Calculate the discretized potential probability value distribution based on the target potential region; 3) Based on the current location of the aircraft, the Vino partitioning algorithm is used to allocate the area to be covered by the aircraft; 4) Based on the distribution of potential probability values ​​within the allocated area, the optimal expected position of each aircraft corresponding to the minimum overall coverage cost is given, thereby obtaining the optimal expected configuration of the aircraft cluster.

2. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 1, characterized in that, Step 1) includes: The azimuth angle θ of target i in polar coordinates obtained from ground radar measurements. ti Elevation angle φ ti and relative position d ti By transforming coordinates, the horizontal and vertical measurement positions p of target i in the Cartesian coordinate system are calculated. ti =(y ti ,z ti ), In the formula, y ti For the position of target i in Cartesian coordinate system, measure the Y-axis component of the position, z ti The Z-axis direction component of the measured position of target i in the Cartesian coordinate system; Considering the radar azimuth measurement error Δ θi And pitch angle measurement error Δ φi The corresponding projection errors on the YOZ plane are as follows: In the formula, Δr yi The position measurement error in the Y-axis direction is Δr. zi This refers to the position measurement error along the Z-axis. The radius of uncertainty of the possible location of target i ti for: Therefore, the potential region of target i is modeled as a circular region, and the center point of this circular region is (y ti ,z ti ), with a radius of r ti The potential range of the target group; Expressed as:

3. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 2, characterized in that, The potential range of the target group is represented by multiple circular regions on the YOZ plane.

4. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 2, characterized in that, Step 2) includes: Establish a standardized detection plane Q: In the formula, y min y max Let z be the minimum and maximum boundary coordinates along the Y-axis, respectively. min z max These are the minimum and maximum boundary coordinates along the Z-axis, representing the horizontal and vertical detection range of the radar, respectively. The normalized probe plane Q is divided into N. y ×N z A uniform grid, the coordinate set of grid points The definition is as follows: In the formula, Indicates rounding down; Calculate grid point p g =(y g , z g The potential probabilistic value v for target i gi , The total potential probability value v of the grid point g The potential probability value of each target is superimposed from the grid points. For each grid point, the potential probability value of the grid point for target i and the total potential probability value of the grid point can be calculated using formulas (7) and (8), thus obtaining the discretized potential probability value distribution.

5. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 4, characterized in that, Step 3) includes: Let N m A swarm system of aircraft, given the current position p of aircraft j. j =(y j ,z j ),have: Where, p k Let k represent the current position of aircraft k, and k ≠ j; Let q represent the area assigned to aircraft j in the normalized detection plane Q, i.e., the Vino region of aircraft j. In this region, the distance from all points q to aircraft j is less than or equal to the distance from point q to any other aircraft. In this way, each aircraft is only responsible for its nearest region, ensuring that the regions responsible for each aircraft do not overlap, and the normalized detection plane Q is completely divided; q represents the points in the region responsible for aircraft j.

6. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 5, characterized in that, Step 4) includes: Design an objective function based on weighted squared distance. Place aircraft j in a location Above, making that position To the Vino region The weighted squared distance between all points within the range is minimized; In formula (10), the coverage cost The deviation of the current aircraft position pj from the high-probability value region was quantified; by minimizing Guide the aircraft towards a high-probability value v g The grid point q is close; To calculate the optimal desired position for each aircraft, Regarding the position of the aircraft Taking the partial derivative, the optimal expected position of spacecraft j at the current moment is: In formula (11), φ(q) represents the probability value of the continuous domain, corresponding to the probability value v of the discretized grid points. g Calculate the desired position of the current aircraft j, i.e., the corresponding Vino region. probability-weighted centroid By replacing the continuous domain integration with grid discretization and summation, we obtain the following equation: In formula (12), the aircraft j is in the Vino region. The position of any grid point (y g ,z g The total potential probability value v corresponding to this grid point. g ; For the Vino region of the aircraft Summing all grid points in the middle; The optimal expected position is calculated for each aircraft, which is the most reasonable deployment position under the current potential probability value distribution, thereby obtaining the optimal expected configuration of the aircraft cluster.

7. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 6, characterized in that, The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation also includes: 5) calculating the comprehensive coverage rate based on the rectangular field-of-view projection splicing range of each aircraft in the current optimal expected configuration of the aircraft cluster, and verifying the effectiveness of the configuration design.

8. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 7, characterized in that, In step 5), the azimuth angle θ of the seeker's field of view is based on the aircraft j. mj Pitch angle φ mj and projection distance d mj The field of view of aircraft j can be simplified to a rectangular region, d yj and d zj Let the horizontal width and vertical length of the simplified rectangular field of view be respectively, then the area formed by stitching together the simplified rectangular field of view of each aircraft... Represented as:

9. The aircraft cooperative coverage configuration optimization method based on probabilistic Vino segmentation as described in claim 8, characterized in that, In step 5), stitching multiple fields of view Potential range of the target group The intersection is defined as the effective coverage area. On the discretized grid, area coverage and value coverage metrics are defined. Area coverage, denoted as η, represents how many grid points of the target potential area are covered by the stitched region. a , Value Coverage η v After considering the weighted importance of the potential probability value distribution to different regions, the extent to which the stitched region covers the high-value target area is measured. By using the above two types of indicators, the coverage performance of the current configuration in the target potential area under the actual sensor range can be quantitatively evaluated, thereby verifying the actual effectiveness of the designed configuration.

Images