Multi-unmanned aerial vehicle beam allocation and tracking method

Abstract

The invention relates to the technical field of unmanned aerial vehicles, aims to overcome the defects in the background technology, and provides a multi-unmanned aerial vehicle beam allocation and tracking method so as to realize continuous and efficient coverage of multi-target unmanned aerial vehicles under limited conditions. According to the technical scheme of the invention, the multi-unmanned aerial vehicle beam allocation and tracking method comprises the following steps: S1, input preparation; s2, predicting the flight path of the unmanned aerial vehicle; s3, vector conversion and candidate beam generation; S4, beam combination and pre-screening; s5, adjusting coincident beams in the combination; and S6, score calculation and beam allocation.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) technology, specifically to a method for beam allocation and tracking of multiple UAVs. Background Technology

[0002] With the widespread application of drones in logistics, inspection, emergency communication and other fields, the high-speed movement and collaborative operation of multiple drones have placed higher demands on the dynamic beam pointing of communication links.

[0003] Existing technologies mainly employ the following methods for beam adjustment:

[0004] 1. Chinese invention patent CN 120730318 A provides a beam adjustment method for multiple UAVs. The core of the method is to use a reinforcement learning framework to jointly optimize the beam-UAV association relationship. By acquiring the UAV position information, constructing a state vector, and using a policy network (such as a Transformer encoder) to output the association probability matrix and beam parameter vector, the pre-assigned beam scheme is dynamically adjusted.

[0005] The drawbacks of this method are: it relies on a pre-trained policy network, is sensitive to the initial state, and is difficult to achieve global optimal allocation under complex conditions such as the number of beams and the constraints of beam pointing changes; it does not consider the physical limitations of beam rotation and coverage continuity; it requires the use of reinforcement learning techniques, which makes model training difficult, computationally expensive, and output controllable, making it difficult to deploy in industrial applications with high reliability requirements.

[0006] 2. Chinese invention patent CN 120433810 A provides an adaptive beam tracking method for flight equipment based on multi-source information fusion. By fusing information from multiple sources such as GPS and attitude sensors, the relative position of the UAV and the antenna array is calculated, the position at the next moment is predicted based on the dynamic model, and the beam pointing angle is adjusted.

[0007] The advantage of this method is that it can achieve rapid beam alignment by using sensor data. However, it focuses on single-target tracking and does not involve the joint allocation of multiple beams and multiple targets. Therefore, it also lacks modeling of constraints such as beam overlap (transmit power competition).

[0008] It is evident that traditional beam adjustment methods are mostly based on static allocation or single target tracking, failing to fully utilize the flight path and trajectory information of UAVs. This makes it difficult to achieve continuous and efficient coverage of multiple UAVs under conditions of limited beam quantity and physical constraints on pointing changes. Summary of the Invention

[0009] The purpose of this invention is to overcome the shortcomings in the above-mentioned background technology and provide a method for beam allocation and tracking of multiple UAVs to achieve continuous and efficient coverage of multiple UAVs under limited conditions.

[0010] The technical solution of this invention is:

[0011] A method for beam allocation and tracking of multiple unmanned aerial vehicles (UAVs) includes the following steps:

[0012] S1: Input preparation;

[0013] S2: Drone trajectory prediction;

[0014] S3: Vector Transformation and Candidate Beam Generation

[0015] S4: Beamforming and pre-screening;

[0016] S5: Adjustment of overlapping beams within the combination;

[0017] S6: Fractional calculation and beam assignment.

[0018] S1 includes: acquiring antenna data, beam data, and UAV data; the antenna data includes antenna position, total number of beams, beam half-cone angle, beam azimuth angle limit, and beam elevation angle limit; the beam data includes the azimuth angle, elevation angle, and idle status at the current moment; the UAV data includes the total number of UAVs, UAV position, and UAV attitude at the current moment.

[0019] S2 includes: predicting the UAV position based on linear assumptions and planar approximation.

[0020] The vector transformation in S3 includes converting the azimuth and elevation angles of all beams at the current moment and the predicted position of the UAV into a three-dimensional unit vector in the antenna coordinate system; the candidate beam generation in S3 includes generating a UAV combination using a full combination method, taking the geometric center of the UAV combination as the candidate beam, removing duplicates from the candidate beams according to similarity, and recording the corresponding UAV list for the candidate beam.

[0021] S4 includes: combining candidate beams according to the total number of beams to obtain beam combinations; retaining beam combinations that can cover all UAVs, or retaining all beam combinations; grouping and filtering the beam combinations and calculating coverage scores, and retaining the beam combination with the highest score in each group.

[0022] The coverage score includes:

[0023] (1) Calculate the coverage score of a single drone.

[0024]

[0025] Where: subscore j Let θ be the coverage score of UAV j belonging to the k-th candidate beam; jLet α be the angle between the three-dimensional unit vector of UAV j and the geometric center vector of the k-th candidate beam; α is the half-cone angle of the beam.

[0026] (2) Calculate the coverage score of the candidate beams.

[0027]

[0028] Among them: Score k,cov The subscore represents the coverage score of candidate beam k. j Let $k$ be the coverage score of the $j$-th UAV in the list of UAVs corresponding to candidate beam $k$. n represents the number of drones corresponding to candidate beam k; UAV The total number of drones; w sum and w mul For weighting coefficients; ∈ base To ensure that the multiplier is not zero for small quantities.

[0029] (3) Summing up the coverage scores of all candidate beams within the beam combination yields the coverage score of the beam combination.

[0030] S5 includes: comparing candidate beams in the beam combination pairwise; when two candidate beams have overlapping angles, taking the geometric center of the two candidate beams as the new direction; if the beam centered on the new direction can cover all UAVs of the two candidate beams, then merging the candidate beams and taking the new direction as the new candidate beam.

[0031] S6 includes: for each beam combination, arranging all candidate beams to allocate them to the physical beams of the antenna. A score is calculated for each arrangement, and the arrangement with the highest total score is selected.

[0032] The calculation of the permutation allocation scores includes:

[0033] (1) Beam rotation cost:

[0034]

[0035] Δθ i The rotation angle is the current direction of the i-th beam. i The angle between the beam and the candidate beam arranged at the i-th position; σ i This represents the idle state of the i-th beam at the current moment, with 1 for on and 0 for off; σ′ i The state of the candidate beam assigned to the i-th position is 1 if it is on and 0 if it is off;

[0036] (2) Coverage score

[0037]

[0038] (3) Beam overlap penalty:

[0039]

[0040] Let n be the number of UAVs covered by a single candidate beam k within the beam combination. c,UAV The total number of drones covered by the beam combination;

[0041] (4) Beam switch state change penalty:

[0042]

[0043] (5) The weighted summation of beam rotation cost, coverage score, beam overlap penalty and beam switch state change penalty yields the total score for permutation allocation.

[0044] The beneficial effects of this invention are:

[0045] This invention enables automatic beam generation, allocation, and adjustment at a relatively low cost in complex scenarios involving multiple beams, multiple UAVs, and beam pointing variations subject to physical constraints, ensuring coverage continuity, beam resource utilization efficiency, and system stability.

[0046] This invention supports multi-target, multi-beam joint allocation, which avoids the limitation of some existing methods that are only applicable to single targets.

[0047] This invention employs global optimization: it achieves multi-objective optimization such as coverage continuity, rotation cost, and beam overlap penalty through a comprehensive scoring mechanism, and the allocation results have good smoothness in the time dimension;

[0048] This invention has low implementation cost: it can achieve near real-time calculations at the millisecond level without the need for pre-training a neural network model or a large amount of computing power. Attached Figure Description

[0049] The following describes some specific embodiments of the invention in a detailed manner by way of example and not limitation, with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar parts or components.

[0050] Figure 1 This is a flowchart of the present invention.

[0051] Figure 2a This is one of the schematic diagrams of Example 1.

[0052] Figure 2b This is the second schematic diagram of Example 1.

[0053] Figure 2c This is the third schematic diagram of Example 1.

[0054] Figure 2d This is the fourth schematic diagram of Example 1.

[0055] Figure 2e This is the fifth schematic diagram of Example 1.

[0056] Figure 3a This is one of the schematic diagrams of Example 2.

[0057] Figure 3b This is the second schematic diagram of Example 2.

[0058] Figure 3c This is the third schematic diagram of Example 2. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0060] A method for beam allocation and tracking of multiple unmanned aerial vehicles (UAVs) includes the following steps:

[0061] S1, Input Preparation

[0062] Acquire antenna data, beam data, and UAV data.

[0063] (1) Antenna data is antenna physical resource information, including: antenna location (longitude, latitude, altitude), total number of beams n beam , beam half-cone angle α (approximate), beam azimuth angle limit (maximum value, minimum value) and beam elevation angle limit (maximum value, minimum value).

[0064] (2) Beam data consists of all beam parameters (antenna coordinate system) at the current moment, including: azimuth angle az, elevation angle el, and idle state σ. i (1 for on, 0 for off). The antenna coordinate system is a Cartesian coordinate system (northeast-northeast coordinate system) established with the antenna as the origin.

[0065] (3) The UAV data includes the number of all online UAVs at the current moment, as well as their location and attitude information, specifically including: the total number of UAVs n UAV UAV location (longitude, latitude, altitude), UAV attitude (horizontal velocity v in the northeast coordinate system with the UAV as the origin). xy Vertical velocity v z , heading (due north)).

[0066] S2, UAV trajectory prediction

[0067] Based on the linear assumption and a planar approximation (northeast geodetic coordinate system), the position of each UAV at the prediction time is predicted using the following formula:

[0068] x′=x+v x Δt

[0069] y′=y+v y Δt

[0070] z′=z+v z Δt

[0071] Where: the current coordinates of the UAV in the northeast coordinate system are (x, y, z); the predicted coordinates of the UAV in the northeast coordinate system are (x′, y′, z′); v x v is the horizontal velocity component of the UAV in the direction of due north. y v is the horizontal velocity component of the UAV in the direction of due east. z Δt represents the vertical velocity of the UAV; Δt is the time difference between the predicted time and the current time, where Δt < 1s, with a preference for 10ms.

[0072] Convert the coordinates (x′, y′, z′) of each UAV at the predicted time into latitude and longitude.

[0073] S3, Vector Transformation and Candidate Beam Generation

[0074] S3-1, Vector Conversion

[0075] Convert the azimuth and elevation angles of all beams at the current moment into a three-dimensional unit vector Beam in the antenna coordinate system. i (i = 1, 2, ..., n) beam ).

[0076] Beam i =normalized(sinaz) i cosel i cosaz i ·cosel i sinel i )

[0077] Where: i is the beam number; normalized is the normalization operation.

[0078] Convert the predicted position of the UAV into a three-dimensional unit vector UAV in the antenna coordinate system. j (j = 1, 2, ..., n) UAV ).

[0079] UAV j =normalized(R·cosφ) ant ·(λ j -λant ),R·(φ j -φ ant ),h j -h ant )

[0080] Where: j is the UAV serial number; φ ant and φ j λ represents the longitude of the antenna and the UAV j, respectively; ant and λ j The latitudes of the antenna and the UAV j are respectively; h ant and h j Here, represents the height of the antenna and the drone j, respectively; R is the Earth's radius (approximately a constant, with a value of 6,371,000 meters).

[0081] S3-2, Candidate Beam Generation

[0082] (1) Generate drone assemblies using the full assembly method; drone assemblies can be one or more, and each drone assembly includes at least one drone;

[0083] (2) For each UAV combination, generate a candidate beam pointing towards its geometric center. k (k = 1, 2, ..., n) candidate_origin );

[0084] Where: k is the sequence number of the UAV combination and the candidate beam sequence number, n candidate_origin The number of candidate beams; the number of UAV combinations is the same as the number of candidate beams and corresponds one-to-one.

[0085] (3) Remove duplicates from candidate beams based on similarity; when the angle θ between two candidate beams candidate <∈ candidate Remove one of the candidate beams, ∈ candidate A predefined similarity threshold.

[0086] (4) Record the list of UAVs in the corresponding UAV combination for the candidate beam record, denoted as UAVList. k .

[0087] (5) If the number of candidate beams is less than the number of antenna beams (n) beam If the number of candidate beams is n, then the candidate beams are copied cyclically until the number of candidate beams equals n. beam This is to ensure consistency in subsequent calculations.

[0088] The full combinatorial approach includes the following steps:

[0089] (1) Enumerate all drone combinations;

[0090] (2) Calculate the geometric center vector center of all UAV combinations; center = norm(∑ j UAV j );

[0091] (3) Retain a combination of drones that can be completely covered by a single beam, with the following condition: all drones in the combination satisfy the condition. This indicates that the drone combination can be completely covered by a single beam; where: Let α be the angle between the three-dimensional unit vector of UAV j in the UAV combination and the geometric center vector of the combination, and let α be the half-cone angle of the beam.

[0092] After step S3, the number of candidate beams is uniformly represented as n. candidate .

[0093] S4, Beamforming and Pre-screening

[0094] S4-1, Based on the total number of beams n beam Candidate beam k By combining them, a beam combiner is obtained.

[0095] Each beam combination is represented as a Comb. c c is the beam combination number; the number of beam combinations is The number of candidate beams in each beam combination is n. beam .

[0096] S4-2. If a beam combination exists that can cover all drones, delete the beam combination that cannot cover all drones; otherwise, retain all beam combinations.

[0097] S4-3. Group and screen beam combinations, and exclude suboptimal combinations based on coverage scores.

[0098] The following is a detailed explanation of the suboptimal combination elimination method for S4-3.

[0099] (1) Beam combination grouping: For each beam combination, calculate its grouping key (GroupKey):

[0100] GroupKey=Hash(UnionAll(UAVList k k = 1, ..., n beam

[0101] Where: Hash is the hash operation, UnionAll is the union operation that retains duplicate elements, and UAVList kThis is the list of UAVs corresponding to the candidate beam k recorded in step S3-2.

[0102] Beam combinations with the same grouping key are grouped into the same group, denoted as CombGroup. g (g = 1, ..., n) group ).

[0103] Where n group This represents the number of groups after grouping based on the above rules.

[0104] (2) Elimination of suboptimal combinations within groups: Calculate the suboptimal combinations for each beam combination (Comb). c The coverage score is used to retain the beam combination with the highest coverage score within each group.

[0105] The coverage score is calculated as follows:

[0106] (1) Calculate the coverage score of a single drone.

[0107]

[0108] Where: subscore j Let θ be the coverage score of UAV j belonging to the k-th candidate beam; j Let θ be the angle between the three-dimensional unit vector of UAV j and the geometric center vector of the k-th candidate beam; j equal α is the half-cone angle of the beam;

[0109] (2) Calculate the coverage score of the candidate beams.

[0110]

[0111] Among them: Score k,cov The subscore represents the coverage score of candidate beam k. j Let $k$ be the coverage score of the $j$-th UAV in the list of UAVs corresponding to candidate beam $k$. n represents the number of drones corresponding to candidate beam k; UAV The total number of drones; w sum and w mul For weighting coefficients; ∈ base For small quantities where the multiplier is not zero, a value of 0.01 can be used.

[0112] (3) Calculate the coverage score of the beam combination.

[0113]

[0114] Among them: Score c,covis the total score of beam combination c, and is the sum of the coverage scores of all candidate beams in the beam combination;

[0115] After this step is completed, the remaining number of beam combinations is denoted as n. comb .

[0116] S5, Combined In-Beat Coincidence Adjustment

[0117] S5-1. Compare candidate beams within the beam combination pairwise to determine if there is any angular overlap; when θ is satisfied... combination <∈ combination When this occurs, it indicates that the two candidate beams have angular overlap, and the process proceeds to the next step of candidate beam merging.

[0118] Where: θ combination The angle between the two candidate beams, ∈ combination The threshold value is a predefined threshold.

[0119] S5-2. Take the geometric center of the two candidate beams as the new direction. If the beam centered on the new direction can cover all UAVs of the two candidate beams, then merge the candidate beams.

[0120] The merging process is as follows: the direction of one of the candidate beams is adjusted to the geometric center of the beam to be merged, the status of the other candidate beam is marked as "off", and the status of the remaining candidate beams is marked as "on" by default.

[0121] After this step is completed, the beam combination (Comb) c The number remains unchanged, still n. comb .

[0122] S6. Fractional Calculation and Beam Assignment

[0123] S6-1. Enumerate all beam-target assignment arrangements

[0124] For step S5, for each beam combination (Comb) c (c = 1, 2, ..., n) comb ), and select the candidate beams among them. k k = 1, ..., n beam The antenna is arranged in a permutation, and each permutation corresponds to a physical beam assigned to the antenna (beam number 1 to n). beam The method is as follows. Each beam combination can obtain... There are several ways to arrange and distribute the data, and the total number of arrangements is... Each permutation assignment is denoted as Perm. p .

[0125] S6-2. Calculate the score for each permutation / allocation method:

[0126] (1) Beam rotation cost:

[0127]

[0128] Where: Δθ i The rotation angle is the current direction of the i-th beam. i The angle between the beam and the candidate beam arranged at the i-th position; σ i This represents the idle state of the i-th beam at the current moment, with 1 for on and 0 for off; σ′ i The state of the candidate beam assigned to the i-th position is 1 if it is on and 0 if it is off;

[0129] (2) Coverage score (same as step S4):

[0130] Calculate the coverage score of drones

[0131]

[0132] Calculate the coverage score of candidate beams

[0133]

[0134] Calculate the coverage score of the beam combination

[0135]

[0136] (3) Beam overlap penalty:

[0137]

[0138] in: Let n be the number of UAVs covered by a single candidate beam k within beam combination c. c,UAV The total number of drones covered by beam combination c (excluding drones with duplicate coverage);

[0139] (4) Beam switch state change penalty:

[0140]

[0141] Where: σ i This represents the idle state of the i-th beam at the current moment, with 1 for on and 0 for off; σ′ i The state of the candidate beam assigned to the i-th position is 1 if it is on and 0 if it is off;

[0142] (5) Calculate the total score of the permutation and distribution.

[0143] Score p =w1Cost rot +w2Scorecov +w3Penalty overlap +w4Penalty switch

[0144] Among them: Score p The total score obtained by arranging and assigning p; w1, w2, w3, and w4 are weighting coefficients.

[0145] S6-3, Permutation and Allocation Selection

[0146] From the above arrangement and allocation methods, select the arrangement and allocation method with the highest total score.

[0147] S7, Pointing optimization adjustment

[0148] If a candidate beam exceeds the beam azimuth and elevation limits set in step S1, the azimuth and elevation of the beams exceeding the limits are corrected to ensure that the azimuth and elevation do not exceed the limits. Specifically, if the azimuth / elevation is less than the minimum value, it is adjusted to the minimum value; if it is greater than the maximum value, it is adjusted to the maximum value.

[0149] 8. Result Output

[0150] Output the final allocation and tracking information, including beam parameters, the set of covered drones, the total allocation score, and the set of uncovered drones. Adjust the beam according to the final allocation to track the drones.

[0151] Example 1

[0152] like Figures 2a to 2e As shown, there are 3 drones, a maximum of 2 beams, and the drones fly in two groups relative to each other, specifically including:

[0153] Time step 0: Three drones, two beams activated;

[0154] Time step 20: Three drones, activating one beam;

[0155] Time step 30: Three drones, two beams activated;

[0156] Time step 40: Three drones, two beams activated;

[0157] Time step 50: Three drones, two beams activated.

[0158] Example 2

[0159] like Figures 3a to 3c As shown, there are 6 drones, with a maximum of 2 beams, and the drones are divided into 3 groups flying in the same direction:

[0160] Time step 0: Four drones, activating one beam;

[0161] Time step 15: Six drones, two beams activated;

[0162] Time step 25: Six drones, two beams activated.

[0163] This invention achieves beam allocation and dynamic tracking in multi-beam and multi-UAV scenarios by automatically generating multi-target beam pointing candidates, integrating multi-dimensional scoring mechanisms and beam pointing adjustment under physical constraints. The beam allocation results have good smoothness in the time dimension.

[0164] The similarities between this invention and the prior art document CN 120730318 A are: both involve dynamic beam adjustment in multi-UAV scenarios, both use location information for beam parameter calculation and allocation decisions, and both consider the service relationship modeling between the beam and the UAV. The differences are: (1) Allocation mechanism: this invention automatically generates multi-target beam pointing candidates and performs combination screening, while the prior art document relies on a pre-trained policy network to output the correlation probability matrix; (2) Optimization dimensions: this invention comprehensively considers multi-dimensional scoring such as rotation cost, coverage continuity, beam overlap penalty, and switch state changes, while the prior art document mainly optimizes based on signal quality and coverage; (3) Computational cost and stability: the prior art document uses reinforcement learning methods for calculation, while this proposal uses spatial geometry and mathematical methods for calculation, which has significant advantages in terms of computation time, cost, and stability of computational results.

[0165] The similarity between this invention and prior art document CN 120433810 A is that both use position prediction to estimate the target position at the next moment. The difference is that this invention focuses on the allocation problem of multiple UAVs and multiple beams, while the prior art document mainly solves the beam tracking problem of a single flying device, which is quite different in technology.

[0166] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for beam allocation and tracking of multiple unmanned aerial vehicles (UAVs), comprising the following steps: S1: Input preparation; S2: Drone trajectory prediction; S3: Vector Transformation and Candidate Beam Generation S4: Beamforming and pre-screening; S5: Adjustment of overlapping beams within the combination; S6: Fractional calculation and beam assignment.

2. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 1, characterized in that: S1 includes: acquiring antenna data, beam data, and UAV data; the antenna data includes antenna position, total number of beams, beam half-cone angle, beam azimuth angle limit, and beam elevation angle limit; the beam data includes the azimuth angle, elevation angle, and idle status at the current moment; the UAV data includes the total number of UAVs, UAV position, and UAV attitude at the current moment.

3. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 2, characterized in that: S2 includes: predicting the UAV position based on linear assumptions and planar approximation.

4. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 3, characterized in that: The vector transformation in S3 includes converting the azimuth and elevation angles of all beams at the current moment and the predicted position of the UAV into a three-dimensional unit vector in the antenna coordinate system; the candidate beam generation in S3 includes generating a UAV combination using a full combination method, taking the geometric center of the UAV combination as the candidate beam, removing duplicates from the candidate beams according to similarity, and recording the corresponding UAV list for the candidate beam.

5. The multi-UAV beam allocation and tracking method according to claim 4, characterized in that: When the full combination method is used in S3, S4 includes: combining candidate beams according to the total number of beams to obtain beam combinations; retaining candidate beam combinations that can cover all UAVs, or retaining all beam combinations; grouping and filtering the beam combinations and calculating coverage scores, and retaining the beam combination with the highest score in each group.

6. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 5, characterized in that: The coverage score includes: (1) Calculate the coverage score of a single drone. subscore j Let θ be the coverage score of UAV j belonging to the k-th candidate beam; j Let θ be the angle between the three-dimensional unit vector of UAV j and the geometric center vector of the k-th candidate beam; j equal α is the half-cone angle of the beam; (2) Calculate the coverage score of the candidate beams score k,cov The subscore represents the coverage score of candidate beam k. j Let $k$ be the coverage score of the $j$-th UAV in the list of UAVs corresponding to candidate beam $k$. n represents the number of drones corresponding to candidate beam k; UAV The total number of drones; w sum and w mul For weighting coefficients; ∈ base To ensure that the multiplier is not zero, (3) sum the coverage scores of all candidate beams in the beam combination to obtain the coverage score of the beam combination.

7. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 6, characterized in that: S5 includes: comparing candidate beams in the beam combination pairwise; when two candidate beams have overlapping angles, the geometric center of the two candidate beams is taken as the new direction; if the beam centered on the adjusted direction can cover all UAVs of the two candidate directions, then the candidate beams are merged and the new direction is taken as the new beam candidate.

8. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 7, characterized in that: S6 includes: for each beam combination, arranging all candidate beams, calculating the score for each arrangement, and selecting the arrangement with the highest total score.

9. The method for beam allocation and tracking of multiple unmanned aerial vehicles according to claim 8, characterized in that: The calculation of the permutation allocation scores includes: (1) Beam rotation cost: Δθ i σ is the rotation angle; i This represents the idle state of the i-th beam at the current time; σ′ i The state of the candidate beam assigned to the i-th position; (2) Coverage score (3) Beam overlap penalty: n represents the number of UAVs covered by a single candidate beam within the beam combination. c,UAV The total number of drones covered by the beam combination; (4) Beam switch state change penalty: (5) The weighted summation of beam rotation cost, coverage score, beam overlap penalty and beam switch state change penalty yields the total score for permutation allocation.

Images