Unmanned aerial vehicle cluster envelope extraction and number estimation method based on multi-view fusion

Abstract

The application provides a kind of unmanned aerial vehicle cluster envelope extraction and quantity estimation method based on multi-view fusion, and relates to low-altitude detection technical field.The application arranges three radars in azimuth direction at equal intervals, 360° covers the monitored area, then identifies targets in cluster according to target echo signals of the three radars and calculates target distance, speed, angle and other information.Accordingly, the relative coordinates of all detected targets are calculated and converted into absolute coordinates, obtaining the target point set of the cluster targets detected by each radar under a certain view angle.The target point sets measured by the three radars are fused to obtain the complete point set of the cluster, and finally the envelope extraction and quantity estimation of the complete point set of the cluster are carried out.The application can effectively perform envelope extraction and accurate quantity estimation of cluster targets through multi-view data fusion and super-resolution angle measurement algorithm, even in the case of high cluster target density and serious internal occlusion.

Description

Technical Field

[0001] This invention relates to the field of low-altitude detection technology, specifically to a method for extracting the envelope and estimating the number of unmanned aerial vehicle (UAV) swarms based on multi-view fusion. Background Technology

[0002] Improving radar's ability to detect multiple targets in dense formations is a key challenge in the field of low-altitude detection technology. Conventional radars primarily distinguish multiple targets based on characteristics such as range, angle, and Doppler shift. By increasing the radar's operating bandwidth, range resolution can be improved, allowing targets that are close in distance to be differentiated and detected.

[0003] Currently, super-resolution algorithms such as MUSIC and ESPRIT are generally used to improve the azimuth resolution, and by accumulating more pulses, the velocity resolution is improved, thus distinguishing targets from the velocity dimension.

[0004] However, due to the small scattering cross-section of individual UAV swarm targets and significant low-altitude clutter interference, the signal-to-noise ratio (SNR) of the target echo signals is low. Currently used super-resolution algorithms such as MUSIC and ESPIRT require high SNR and a sufficient number of input pulses, resulting in poor performance in UAV swarm target detection and inability to accurately extract the envelope and estimate the number of swarm targets. Researching new angle measurement algorithms to improve angular resolution can improve the detection accuracy of dense individual targets to some extent, but it cannot solve the problem of decreased detection accuracy caused by occlusion within the swarm. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, which solves the technical problem of low accuracy in the application of existing methods for UAV swarm target detection.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] In a first aspect, the present invention provides a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, comprising:

[0010] S1. Obtain the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° azimuth direction;

[0011] S2. Acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, identify targets within the cluster based on the target echo signal, and calculate the target distance, speed, and angle.

[0012] S3. Calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars to obtain the point set of cluster targets detected by each radar under a certain view.

[0013] S4. Calculate the distance between any two points in each point set to obtain the distance set. Cluster the distance sets to obtain the minimum distance in each of the three point sets. Average the three minimum distances to obtain the minimum distance between the targets in the cluster.

[0014] S5. Based on the minimum distance between targets within the cluster, the point sets are fused to obtain the complete point set of the cluster;

[0015] S6. Extract the envelope and estimate the number of points in the complete point set of the cluster.

[0016] Preferably, S2 includes:

[0017] S201. Obtain the echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars. Store the echo signal collected by each array unit of the three radars in rows to obtain a data cube composed of three dimensions: range, velocity, and azimuth.

[0018] S202. Based on the data cube, perform azimuth dimension averaging to obtain two-dimensional data. After performing 2D-FFT processing on the two-dimensional data, perform CFAR calculation and peak aggregation to obtain the target's velocity and target distance.

[0019] S203. Based on the data cube, calculate the angle of the target using an angle super-resolution algorithm.

[0020] Preferably, S203 includes:

[0021] The virtual array radar signal model is established based on the data cube as follows:

[0022]

[0023] In the formula, L represents the number of signal sources, n represents the system noise, and s represents the spatial steering vector. d m θ represents the distance of the m-th array element relative to the reference point; θ is the range of the search angle. l Let λ be the angle value of the signal source l, and λ be the wavelength of the transmitted signal;

[0024] Optimize the radar signal model:

[0025]

[0026] The dual vector c further expresses equation (2) as:

[0027]

[0028] Let b(θ) = s(θ) H c, and expressed using Fourier series:

[0029]

[0030] α m [k] can be derived from a m (θ) is obtained through DFT calculation:

[0031]

[0032] Where P = 2N + 1, equation (3) is further expressed using a finite-dimensional matrix:

[0033]

[0034] in, i and j are indices for traversing the elements in the finite-dimensional matrix H, m refers to the m-th element of the antenna array, M is the total number of elements, N is the number of DFT transformation sampling points, δ is the convergence criterion, Equation (6) is a semidefinite programming problem, and its optimization result is b(θ), from which the angle of the target is calculated.

[0035] Preferably, S3 includes:

[0036]

[0037]

[0038]

[0039] Among them, (X) i ,Y i Z i Let R be the relative coordinates of the i-th target. i Let be the relative distance between the i-th target and the radar. Let be the elevation angle value of the i-th target relative to the radar. Let be the azimuth angle value of the i-th target relative to the radar. The target's relative coordinates plus the corresponding radar absolute coordinates are the target's absolute coordinates.

[0040] The set of all target points detected by a single radar is the set of targets that the radar can detect from a certain viewpoint when measuring the cluster of targets.

[0041] Preferably, S4 includes:

[0042] Calculate the Euclidean distance between any two points in the point sets Set1, Set2, and Set3 of the three radars respectively to obtain the spacing sets D1, D2, and D3;

[0043] Perform K-MEANS clustering on sets D1, D2, and D3 respectively, and sort the mean of each cluster by size to obtain the minimum intervals M1, M2, and M3;

[0044] The average of M1, M2, and M3 is used as the minimum distance between targets within the cluster.

[0045] Preferably, S5 includes:

[0046] First, merge the point sets Set1 and Set2. Iterate through the points in Set1 and calculate the Euclidean distance d between each point and all points in Set2. t ;

[0047] If d t <w*d mean Calculate the mean of the two points and add them to a new set Set. new middle;

[0048] If d t w*d mean Then add both points to the Set. new middle;

[0049] After merging Set1 and Set2, Set new The Set3 is further merged using the same method as Set1 and Set2 to obtain the complete point set of the cluster after multi-view fusion.

[0050] Where Set1, Set2, and Set3 are the point sets of the three radars; w is an empirical value, 0 <w<1;d mean This represents the minimum distance between targets within the cluster.

[0051] Preferably, S6 includes:

[0052] Using the alphaShape function in MATLAB to plot the complete point set of a cluster l Edge extraction is performed and a corresponding envelope is established. The extracted envelope is matched with the formation structure commonly used in UAV swarms. The points in the complete point set are refitted using the matched formation structure and the minimum distance between targets within the swarm. The number of UAVs in the swarm is estimated based on the fitted new formation.

[0053] Secondly, the present invention provides a system for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, characterized in that it includes:

[0054] The data acquisition module is used to acquire the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° in the azimuth direction;

[0055] The first calculation module is used to acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, and to identify targets within the cluster based on the target echo signal and to calculate the target distance, speed and angle.

[0056] The second calculation module is used to calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars, so as to obtain the point set of cluster targets detected by each radar under a certain view.

[0057] The third calculation module is used to calculate the distance between any two points in each point set, obtain the distance set, cluster the distance set to obtain the minimum distance in each of the three point sets, and average the three minimum distances to obtain the minimum distance of the target in the cluster.

[0058] The fusion module is used to fuse point sets based on the minimum distance between targets within the cluster to obtain the complete point set of the cluster;

[0059] The extraction and estimation module is used to extract the envelope and estimate the number of points in the complete point set of the cluster.

[0060] Thirdly, the present invention provides a computer-readable storage medium storing a computer program for a method for extracting and estimating the envelope of a UAV swarm based on multi-view fusion, wherein the computer program causes a computer to execute the method for extracting and estimating the envelope of a UAV swarm based on multi-view fusion as described above.

[0061] Fourthly, the present invention provides an electronic device, comprising:

[0062] One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including methods for performing UAV swarm envelope extraction and number estimation based on multi-view fusion as described above.

[0063] (III) Beneficial Effects

[0064] This invention provides a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion. Compared with existing technologies, it has the following advantages:

[0065] This invention first acquires the system parameters and absolute coordinates of three radars evenly spaced in the azimuth direction, covering the monitored area in a 360° azimuth direction. Then, it acquires the target echo signals obtained by the three radars simultaneously transmitting linear frequency modulated continuous wave signals. Based on the target echo signals, it identifies targets within the cluster and calculates their distance, velocity, and angle. Based on the distance and angle, it calculates the relative coordinates of all detected targets and converts them to absolute coordinates according to the corresponding radars, obtaining a set of cluster targets detected by each radar from a certain perspective. It calculates the distance between any two points within each set, obtaining a set of distances. The distance sets are then clustered to obtain the minimum distances in each of the three set sets. The average of these three minimum distances yields the minimum distance between targets within the cluster. Based on the minimum distance between targets within the cluster, the set of points is fused to obtain the complete set of points for the cluster. Finally, it performs envelope extraction and quantity estimation on the complete set of points for the cluster. This invention, through the fusion of multi-view data, can effectively extract the envelope of cluster targets and accurately estimate their quantity even when the cluster target density is high and internal occlusion is severe. Attached Figure Description

[0066] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0067] Figure 1 This is a block diagram of a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, according to an embodiment of the present invention.

[0068] Figure 2 This is an overall flowchart of a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, according to an embodiment of the present invention. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are described clearly and completely. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0070] This application provides a method for extracting the envelope and estimating the number of UAV swarm targets based on multi-view fusion, which solves the technical problem of low accuracy of existing methods in UAV swarm target detection and achieves accurate extraction of the envelope and estimation of the number of UAV swarm targets.

[0071] The technical solution in this application is to solve the above-mentioned technical problems, and the general idea is as follows:

[0072] Currently used super-resolution algorithms such as MUSIC and ESPIRT require high signal-to-noise ratios and a sufficient number of input pulses, resulting in poor performance in UAV swarm target detection. Therefore, this invention employs a meshless beamforming super-resolution angle estimation method, which can achieve accurate target angle estimation under low signal-to-noise ratios and has lower requirements for input data. Furthermore, considering the large number and dense distribution of UAV swarm targets, and the problem of mutual occlusion between targets, it is very difficult for a single radar to identify their formation and estimate their numbers. This invention uses three radars to detect UAV swarm targets from different perspectives and fuses the results from the three radars to achieve envelope extraction and accurate estimation of the number of UAV swarm targets.

[0073] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0074] This invention provides a method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, including:

[0075] S1. Obtain the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° azimuth direction;

[0076] S2. Acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, identify targets within the cluster based on the target echo signal, and calculate the target distance, speed, and angle.

[0077] S3. Calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars to obtain the point set of cluster targets detected by each radar under a certain view.

[0078] S4. Calculate the distance between any two points in each point set to obtain the distance set. Cluster the distance sets to obtain the minimum distance in each of the three point sets. Average the three minimum distances to obtain the minimum distance between the targets in the cluster.

[0079] S5. Based on the minimum distance between targets within the cluster, the point sets are fused to obtain the complete point set of the cluster;

[0080] S6. Extract the envelope and estimate the number of points in the complete point set of the cluster.

[0081] This invention, through the fusion of multi-view data and angle super-resolution algorithms, can still effectively extract the envelope of cluster targets and accurately estimate their number even when the cluster target density is high and internal occlusion is severe.

[0082] The following is combined with, for example Figure 2 The overall flowchart shown provides a detailed explanation of each step:

[0083] In step S1, the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area 360° in the azimuth direction are obtained. The specific implementation process is as follows:

[0084] The three radar systems have the same parameter settings. They are all virtual array radars with the same array length. The pulse signal is a linear frequency modulated continuous wave. The absolute coordinates of the three radars are Ra1, Ra2, and Ra3, respectively.

[0085] In step S2, the target echo signal is acquired by synchronously transmitting linear frequency modulated continuous wave signals from three radars. Based on the target echo signal, targets within the cluster are identified, and their distance, velocity, and angle are calculated. The specific implementation process is as follows:

[0086] S201. Obtain the echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from the three radars. Store the echo signal collected by each array unit of the three radars row by row to obtain a data cube composed of three dimensions: range, velocity, and azimuth.

[0087] S202. Based on the data cube, the azimuth dimension is first averaged, and the resulting two-dimensional data is then subjected to 2D-FFT calculation. CFAR calculation and peak aggregation are then performed to obtain the target's velocity and distance.

[0088] S203. Based on the data cube, the angle of the target is calculated using an angle super-resolution algorithm, specifically including:

[0089] First, a virtual array radar signal model is established as follows:

[0090]

[0091] in, L represents the number of signal sources, n represents the system noise, and s represents the spatial steering vector. d m θ represents the distance of the m-th array element relative to the reference point (generally, the outermost array element is chosen as the reference point); θ is the range of the search angle. l Let λ be the angle value of the signal source l, and λ be the wavelength of the transmitted signal;

[0092] Considering the sparsity of the signal source, x(θ) can be reconstructed from y, which can be specifically expressed as the following optimization problem:

[0093]

[0094] For ease of solving, the above equation can be further expressed using the dual vector c as follows:

[0095]

[0096] Let b(θ) = s(θ) H c, and expressed using Fourier series:

[0097]

[0098] α m [k] can be derived from a m (θ) is obtained through DFT calculation:

[0099]

[0100] Where P = 2N + 1, equation (3) is further expressed using a finite-dimensional matrix:

[0101]

[0102] in, i and j are the indices for traversing the elements in the finite-dimensional matrix H, m refers to the m-th element of the antenna array, M is the total number of elements, N is the number of DFT transformation sampling points, δ is the convergence criterion, Equation (6) is a semidefinite programming problem, and its optimization result is b(θ), from which the angle of the target can be calculated.

[0103] In step S3, the relative coordinates of all detected targets are calculated based on distance and angle, and the detected targets are converted into absolute coordinates according to the absolute coordinates of the corresponding radars, thus obtaining the point set of cluster targets detected by each radar from a certain viewpoint. The specific implementation process is as follows:

[0104] Calculate the target's three-dimensional coordinates relative to the radar based on the target's distance and angle. Assume the relative coordinates of the i-th target are (X... i ,Y i Z i The calculation formula is as follows:

[0105]

[0106]

[0107]

[0108] Among them, Ri is the relative distance of the i-th target from the radar, is the pitch angle value of the i-th target relative to the radar, is the azimuth angle value of the i-th target relative to the radar. The absolute coordinates of a target are obtained by adding the relative coordinates of the target to the corresponding absolute coordinates of the radar. The set of all target points detected by a single radar is the targets that can be detected by the radar from a certain perspective for the cluster targets. The target point sets measured by the three radars are denoted as Set1, Set2, and Set3.

[0109] In step S4, the distance between any two points in each point set is calculated to obtain a distance set, the distance set is clustered to obtain the respective minimum distances in the three point sets, and the three minimum distances are averaged to obtain the minimum target distance within the cluster. The specific implementation process is as follows:

[0110] The Euclidean distance between any two points in Set1, Set2, and Set3 is calculated respectively to obtain new sets D1, D2, and D3. K-MEANS clustering is performed on sets D1, D2, and D3 respectively, and the mean value of each category is sorted by size to obtain the minimum distances M1, M2, and M3. M1, M2, and M3 are averaged, and the result is used as the minimum target distance within the cluster, denoted as d mean .

[0111] In step S5, the point sets are fused according to the minimum target distance within the cluster to obtain a complete point set of the cluster. The specific implementation process is as follows:

[0112] First, point set fusion is performed on Set1 and Set2. Traverse the points in Set1, and calculate the Euclidean distance d between the current point and all points in Set2 t , if d t <w*d mean (0<w<1 is an empirical value obtained by comprehensively considering factors such as system distance, angle resolution, and noise level), the mean value of the two points is calculated and added to the new set Set new , if d t >w*d mean then both points are added to Set new . After the merging of Set1 and Set2 is completed, Set new and Set3 are further merged according to the above method, and finally, a complete point set of the cluster after multi-perspective fusion is obtained, denoted as Set total .

[0113] In step S6, envelope extraction and quantity estimation are performed on the complete point set of the cluster. The specific implementation process is as follows:

[0114] The alphaShape function in MATLAB is used for the complete point set Set of the clustertotal Edge extraction is performed and corresponding envelopes are established. The extracted envelopes are then matched with commonly used formations in UAV swarms, and the matched formations and d are used as the basis for the analysis. mean For Set total The points in the data are refitted, and the number of drones in the swarm is estimated based on the new fitted formation.

[0115] This invention also provides a system for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, comprising:

[0116] The data acquisition module is used to acquire the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° in the azimuth direction;

[0117] The first calculation module is used to acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, and to identify targets within the cluster based on the target echo signal and to calculate the target distance, speed and angle.

[0118] The second calculation module is used to calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars, so as to obtain the point set of cluster targets detected by each radar under a certain view.

[0119] The third calculation module is used to calculate the distance between any two points in each point set, obtain the distance set, cluster the distance set to obtain the minimum distance in each of the three point sets, and average the three minimum distances to obtain the minimum distance of the target in the cluster.

[0120] The fusion module is used to fuse point sets based on the minimum distance between targets within the cluster to obtain the complete point set of the cluster;

[0121] The extraction and estimation module is used to extract the envelope and estimate the number of points in the complete point set of the cluster.

[0122] It is understood that the UAV swarm envelope extraction and quantity estimation system based on multi-view fusion provided in this embodiment of the invention corresponds to the above-mentioned UAV swarm envelope extraction and quantity estimation method based on multi-view fusion. The explanation, examples, and beneficial effects of the relevant content can be referred to the corresponding content in the UAV swarm envelope extraction and quantity estimation method based on multi-view fusion, and will not be repeated here.

[0123] This invention also provides a computer-readable storage medium storing a computer program for extracting and estimating the number of UAV swarms based on multi-view fusion, wherein the computer program causes a computer to execute the UAV swarm envelope extraction and number estimation method based on multi-view fusion as described above.

[0124] This invention also provides an electronic device, comprising:

[0125] One or more processors;

[0126] Memory; and

[0127] One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including methods for performing UAV swarm envelope extraction and number estimation based on multi-view fusion as described above.

[0128] In summary, compared with existing technologies, it has the following beneficial effects:

[0129] 1. The embodiments of the present invention, through the fusion of multi-view data and angle super-resolution algorithm, can still effectively extract the envelope of cluster targets and accurately estimate their number even when the cluster target density is high and the internal occlusion is severe.

[0130] 2. The embodiments of the present invention employ a meshless beamforming super-resolution angle estimation algorithm, which has a low dependence on the number of pulses when calculating the angle. Even when the input data is correlated and noisy, the angle resolution remains high, greatly improving the accuracy of individual target detection.

[0131] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0132] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, characterized in that, include: S1. Obtain the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° azimuth direction; S2. Acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, identify targets within the cluster based on the target echo signal, and calculate the target distance, speed, and angle. S3. Calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars to obtain the point set of cluster targets detected by each radar under a certain view. S4. Calculate the distance between any two points in each point set to obtain the distance set. Cluster the distance sets to obtain the minimum distance in each of the three point sets. Average the three minimum distances to obtain the minimum distance between the targets in the cluster. S5. Based on the minimum distance between targets within the cluster, the target point set is fused to obtain the complete point set of the cluster; S6. Perform envelope extraction and count estimation on the complete point set of the cluster, including: In MATLAB, the alphaShape function is used to extract edges from the complete point set of the swarm and establish the corresponding envelope. The extracted envelope is then matched with the formation structure commonly used by UAV swarms. The points in the complete point set are then refitted using the matched formation structure and the minimum distance between targets within the swarm. The number of UAVs in the swarm is estimated based on the new fitted formation.

2. The method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion as described in claim 1, characterized in that, S2 includes: S201. Obtain the echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars. Store the echo signal collected by each array unit of the three radars in rows to obtain a data cube composed of three dimensions: range, velocity, and azimuth. S202. Based on the data cube, perform azimuth dimension averaging to obtain two-dimensional data. After performing 2D-FFT processing on the two-dimensional data, perform CFAR calculation and peak aggregation to obtain the target's velocity and target distance. S203. Based on the data cube, calculate the angle of the target using an angle super-resolution algorithm.

3. The method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion as described in claim 2, characterized in that, S203 includes: The virtual array radar signal model is established based on the data cube as follows: （1） In the formula, , L The number of signal sources, n For system noise, s For spatial guidance vectors, , d m For the first m The distance between each array element and the reference point; For the range of search angles, Let l be the angle value of the signal source. The wavelength of the transmitted signal; Optimize the radar signal model: （2） The dual vector c further expresses equation (2) as: （3） make And expressed using Fourier series: （4） can be The results were obtained through DFT calculation: （5） Where P = 2N + 1, equation (3) is further expressed using a finite-dimensional matrix: （6） in, , , i , j For finite-dimensional matrices H The index used to iterate over the inner elements. m This refers to the first antenna array m M is the total number of array elements. N It is the number of sampling points for the DFT transform. δ The convergence criterion is given by equation (6), which is a semidefinite programming problem. The optimization result is given by equation (6). b(θ) ,Depend on b(θ) can Calculate the angle of the target.

4. The method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion as described in any one of claims 1 to 3, characterized in that, S3 includes: in, For the first i The relative coordinates of each target For the first i The relative distance of each target from the radar. For the first i The elevation angle of a target relative to the radar. For the first i The target's azimuth angle relative to the radar, plus the corresponding radar absolute coordinates, gives the target's absolute coordinates. The set of all target points detected by a single radar is the set of targets that the radar can detect from a certain viewpoint when measuring the cluster of targets.

5. The method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion as described in any one of claims 1 to 3, characterized in that, S4 includes: Calculate the Euclidean distance between any two points in the point sets Set1, Set2, and Set3 of the three radars respectively to obtain the spacing sets D1, D2, and D3; For sets D1 and D respectively 2、 D3 performs K-MEANS clustering and sorts the mean of each category by size to obtain the minimum intervals M1, M2, and M3; The average of M1, M2, and M3 is used as the minimum distance between targets within the cluster.

6. The method for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion as described in any one of claims 1 to 3, characterized in that, S5 includes: First, merge the point sets Set1 and Set2. Iterate through the points in Set1 and calculate the Euclidean distance d between each point and all points in Set2. t ; If d t <w*d mean Calculate the mean of the two points and add them to a new set Set. new middle; If d t w*d mean Then add both points to the Set. new middle; After merging Set1 and Set2, Set new The Set3 is further merged using the same method as Set1 and Set2 to obtain the complete point set of the cluster after multi-view fusion. Where Set1, Set2, and Set3 are the point sets of the three radars; w is an empirical value, 0 <w<1; d mean This represents the minimum distance between targets within the cluster.

7. A system for extracting the envelope and estimating the number of UAV swarms based on multi-view fusion, characterized in that, include: The data acquisition module is used to acquire the system parameters and absolute coordinates of three radars that are equally spaced in the azimuth direction and cover the monitored area in 360° in the azimuth direction; The first calculation module is used to acquire the target echo signal obtained by the synchronous transmission of linear frequency modulated continuous wave signals from three radars, and to identify targets within the cluster based on the target echo signal and to calculate the target distance, speed and angle. The second calculation module is used to calculate the relative coordinates of all detected targets based on distance and angle, and convert the detected targets into absolute coordinates based on the absolute coordinates of the corresponding radars, so as to obtain the point set of cluster targets detected by each radar under a certain view. The third calculation module is used to calculate the distance between any two points in each point set, obtain the distance set, cluster the distance set to obtain the minimum distance in each of the three point sets, and average the three minimum distances to obtain the minimum distance of the target in the cluster. The fusion module is used to fuse point sets based on the minimum distance between targets within the cluster to obtain the complete point set of the cluster; The extraction and estimation module is used to perform envelope extraction and count estimation on the complete point set of the cluster, including: In MATLAB, the alphaShape function is used to extract edges from the complete point set of the swarm and establish the corresponding envelope. The extracted envelope is then matched with the formation structure commonly used by UAV swarms. The points in the complete point set are then refitted using the matched formation structure and the minimum distance between targets within the swarm. The number of UAVs in the swarm is estimated based on the new fitted formation.

8. A computer-readable storage medium, characterized in that, It stores a computer program for a method of extracting and estimating the envelope of a drone swarm based on multi-view fusion, wherein the computer program causes a computer to execute the method of extracting and estimating the envelope of a drone swarm based on multi-view fusion as described in any one of claims 1 to 6.

9. An electronic device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including methods for performing multi-view fusion-based UAV swarm envelope extraction and number estimation as described in any one of claims 1 to 6.

Images