A method for identifying a swarm of bio-inspired drones

Abstract

The application discloses a kind of based on the identification method of biological simulation unmanned aerial vehicle cluster, method specifically includes: the signal processing of unmanned aerial vehicle cluster signal detected by radar obtains the spatial position of target unmanned aerial vehicle;By the spatial position obtains the measurement of target unmanned aerial vehicle;Based on the measurement, the cluster mode of unmanned aerial vehicle cluster is identified to obtain identification result;Wherein, the cluster mode includes flock mode, wolf pack mode and bee swarm mode.The present application solves the problem that the existing radar data processing technology does not have cluster mode perception and identification capability, by aiming at three common biological simulation unmanned aerial vehicle clusters, namely flock, wolf pack and bee swarm, three corresponding mode identification methods are proposed, which can well realize the identification of different cluster modes, and can be widely applied in the field of radar data processing technology.

Description

Technical Field

[0001] This invention relates to the field of radar data processing technology, and in particular to a method for identifying a swarm of unmanned aerial vehicles (UAVs) based on biomimetic techniques. Background Technology

[0002] A swarm target is defined as a set of multiple targets that maintain a relatively fixed spatial position over a sufficiently long period of time, satisfying a given target spacing criterion. Common swarm targets include: biological groups in nature, densely flying drone formations, mid-course ballistic missile swarms, multiple decoys releasing gold foil jamming, and satellite swarms. With the development of drone cooperative flight technology and the advancement of inexpensive 3D printing technology, large numbers of densely distributed swarm targets are frequently encountered in radar detection. Biologically inspired drone swarm targets are gradually becoming practical and have become a research hotspot in the field of radar target detection and tracking.

[0003] The primary task of traditional radar data processing is to detect and track swarm targets. For UAV swarms, determining the swarm pattern and sensing its changing patterns in real time is crucial. Based on the swarm pattern, the purpose of the UAV swarm can be preliminarily determined; further improvements in radar tracking accuracy can even be made based on the swarm pattern identification. Currently, existing radar data processing technologies lack the ability to perceive and identify swarm patterns, necessitating improvements to existing data processing algorithms and the addition of identification capabilities. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide an easy-to-operate and multi-identification method for identifying biomimetic drone swarms.

[0005] On one hand, embodiments of the present invention provide a method for identifying biomimetic drone swarms, including:

[0006] Signal processing is performed on the drone swarm signals detected by radar to obtain the spatial position of the target drone;

[0007] Measurements of the target UAV are obtained from the spatial location;

[0008] The clustering patterns of the drone swarm are identified based on the measurements, and the identification results are obtained; wherein, the clustering patterns include goose flocking, wolf flocking, and bee swarming.

[0009] Optionally, the identification result is obtained by identifying the swarm pattern of the UAV swarm based on the measurement; wherein, the swarm pattern includes goose flock pattern, wolf flock pattern, and bee swarm pattern, including:

[0010] Based on the measurements, the clustering pattern of the drone swarm is determined to be a flock of geese.

[0011] Based on the measurements, the cluster mode of the drone swarm was determined to be a wolf pack mode.

[0012] Based on the measurements, the cluster mode of the drone swarm is determined to be swarm mode.

[0013] Optionally, determining the swarm mode of the drone cluster as a goose flock based on the measurement includes:

[0014] The distance pair of the target UAV is obtained from the measurement;

[0015] Based on the distance pairs, valid distance pairs are filtered using the first, second, third, and fourth conditions;

[0016] The clustering mode of the drone swarm is determined to be the goose flock mode by the number of effective distance pairs and the preset expected distance pairs.

[0017] Optionally, the step of filtering valid distance pairs based on the distance pairs using a first condition, a second condition, a third condition, and a fourth condition includes:

[0018] Based on the distance pair, the target slope, target line segment, and target perpendicular foot are obtained;

[0019] Based on the distance filtering first condition of the distance pair, the distance pair is determined to be the nearby target of the measurement, and the first condition filtering result is obtained;

[0020] Based on the target slope, the preset slope threshold, the target perpendicular, the measurement, and the number of expected distance pairs, a second condition is selected to determine the symmetry between different distance pairs and the measurement, thus obtaining the second condition selection result.

[0021] Based on the target perpendicular and the third condition for measurement screening, it is determined that two points of the same distance pair are located on both sides of the measurement, and the third condition screening result is obtained.

[0022] Based on the target perpendicular and the fourth condition for measurement screening, it is determined that different distance pairs are located on the same side of the measurement, and the fourth condition screening result is obtained;

[0023] Valid distance pairs are obtained by filtering based on the first condition, the second condition, the third condition, and the fourth condition.

[0024] Optionally, determining the cluster mode of the drone swarm as a wolf pack mode based on the measurement includes:

[0025] Based on a preset first measurement radius and a first quantity threshold, the measurement is spatially clustered to obtain a first clustering result, which includes a first core point, a first boundary point, and a first noise point.

[0026] The clustering results were subjected to wolf pack-type filtering to obtain effective clusters;

[0027] Based on the effective clusters and the first quantity threshold, the cluster mode of the drone swarm is determined to be the wolf pack mode.

[0028] Optionally, the step of spatially clustering the measurements according to a preset first measurement radius and a first quantity threshold to obtain a first clustering result, wherein the first clustering result includes a first core point, a first boundary point, and a first noise point, including:

[0029] Obtain the number of midpoints in the first measurement radius;

[0030] When the number of points in the first measurement radius is greater than or equal to the first quantity threshold, the measurement is determined to be the first core point.

[0031] When the measurement is located within the first measurement radius of the first core point, and the number of points within the first measurement radius is less than the first quantity threshold, the measurement is determined to be the first boundary point.

[0032] When the measurement is neither the first core point nor the first boundary point, the measurement is determined to be the first noise point.

[0033] Optionally, determining the swarm mode of the drone cluster as a swarm mode based on the spatial location includes:

[0034] Based on the preset second measurement radius and second quantity threshold, the measurement is spatially clustered to obtain the second clustering result, which includes a second core point, a second boundary point, and a second noise point.

[0035] The second clustering results are subjected to bee colony filtering to determine the clustering mode of the drone cluster as the bee colony mode.

[0036] Optionally, the step of performing bee colony filtering on the second clustering results to determine the clustering mode of the drone swarm as the bee colony mode includes:

[0037] Based on the second clustering result, the nearest neighbor distance, second nearest neighbor distance, and farthest distance between the second core points are obtained;

[0038] Based on the nearest neighbor distance and the second nearest neighbor distance, it is determined that a certain distance is maintained between the second core points, and the fifth condition filtering result is obtained;

[0039] Based on the nearest neighbor distance and the farthest distance, the distribution of the second core points is determined to be a widely distributed set, thus obtaining the sixth condition filtering result;

[0040] Based on the results of the fifth and sixth conditional filtering, the cluster mode of the drone swarm is determined to be the identification result of the swarm mode.

[0041] On the other hand, embodiments of the present invention also provide an identification device based on a biomimetic drone swarm, comprising:

[0042] The first module is used to process the signals of the drone swarm detected by the radar to obtain the spatial position of the target drone;

[0043] The second module is used to obtain measurements of the target UAV from the spatial location;

[0044] The third module is used to identify the clustering pattern of the UAV swarm based on the measurements and obtain the identification result; wherein, the clustering pattern includes goose flocking pattern, wolf flocking pattern and bee swarming pattern.

[0045] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing a program that is executed by a processor to implement the aforementioned identification method for a biomimetic drone swarm.

[0046] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium and execute the computer instructions, causing the computer device to perform the aforementioned method.

[0047] The embodiments of the present invention include at least the following beneficial results: The present invention obtains identification results by identifying the clustering patterns of drone swarms based on measurements, wherein the clustering patterns include goose flocking, wolf flocking, and bee swarming. The present invention proposes three corresponding identification methods for the three clustering patterns of drone swarms, thereby completing the pattern identification of drone swarms and solving the problem that the prior art does not have the ability to perceive and identify the clustering patterns of drone swarms. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 This is a flowchart of a method for identifying a biomimetic drone swarm provided in an embodiment of the present invention;

[0050] Figure 2 This is a block diagram of a biomimetic drone swarm identification device provided in an embodiment of the present invention;

[0051] Figure 3 This is a simulation identification result diagram provided in the embodiment of the present invention;

[0052] Figure 4 This is a swarm pattern diagram of the UAV swarm mimicking a goose flock pattern provided in an embodiment of the present invention;

[0053] Figure 5 This is a cluster mode diagram of the drone swarm model mimicking a wolf pack, provided in an embodiment of the present invention.

[0054] Figure 6 This is a cluster mode diagram of the drone swarm simulation bee swarm mode provided in the embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0056] To address the problems existing in the prior art, embodiments of the present invention provide a method for identifying biomimetic drone swarms, such as... Figure 1 As shown, the method includes steps 100 to 300:

[0057] Step 100: Perform signal processing on the drone swarm signals detected by the radar to obtain the spatial position of the target drone.

[0058] Optionally, the spatial position of the target drone can be obtained by signal processing of the drone swarm signals detected by radar.

[0059] Step 200: Obtain the measurement of the target UAV from the spatial location.

[0060] Optionally, the measurement is the coordinates of the spatial position of a target UAV obtained by radar detection.

[0061] Step 300: Identify the clustering pattern of the UAV swarm based on the measurements to obtain the identification result; wherein, the clustering pattern includes goose flocking pattern, wolf flocking pattern and bee swarming pattern.

[0062] The above step 300 identifies the clustering pattern of the UAV swarm based on the measurement to obtain the identification result; wherein, the clustering pattern includes the goose flocking pattern, wolf flocking pattern and bee swarming pattern. This step may specifically include steps 310 to 330.

[0063] Step 310: Based on the measurement, determine that the clustering mode of the drone swarm is the goose flock mode.

[0064] Optionally, drone swarms in a goose-flock formation are primarily used for large-scale transportation. Geese typically migrate in a V-formation; drone swarms mimicking this migratory pattern can achieve longer-distance transportation. Figure 4 This is a diagram illustrating the swarm pattern of drones, mimicking the flocking of geese.

[0065] The step 310 above, which determines that the clustering mode of the UAV cluster is the goose flock mode based on the measurement, may specifically include steps 311 to 313.

[0066] Step 311: Obtain the distance pair of the target UAV from the measurement.

[0067] Optionally, the measurement is the coordinates of the spatial position of a target UAV obtained by radar detection; the distance between the target UAV and other UAVs can be obtained from the measurement. When the difference between the target UAV and the other two UAVs is not greater than a preset distance threshold, the other two UAVs form a distance pair, and all distance pairs of the target UAV form a distance pair set. For example, the measurement m i =(x i ,y i ), it is related to the measurement m j The distance between them is denoted as d. ij =||m i -m j ||2, Measurement m i The set of distances D between all other measurements i ={d i1 ,...,d i(i-1) ,d i(i+1) ,...,d in}, when D i Two distances in the middle satisfy: |d ij -d ik |<=ε d At that time, the m was weighed and measured. j and m k For measuring m i A distance pair, denoted as p l ={m j ,m k}, where ε d Given a preset distance pair threshold, the set of all distance pairs for measurement i is: P i For measuring m idistance pair set, where L is the number of distance pairs. The core of detecting whether a cluster mode is a goose flock lies in checking whether there are multiple distance pairs satisfying specific conditions in the measurements, so that a V-shaped formation similar to a "herringbone" appears in the measurements.

[0068] Step 312: based on said distance pairs, screen valid distance pairs through the first condition, the second condition, the third condition and the fourth condition.

[0069] Optionally, a target slope, a target line segment and a target foot of perpendicular are obtained based on said distance pairs first. A target line segment can be determined by two points of one distance pair, the slope of the target line segment is the target slope, and the foot of perpendicular from the measurement of the target UAV to the target line segment is the target foot of perpendicular, for example, measurement m i has one distance pair p l ={m j ,m k}, m j and m k form the target line segment l between each other, the target slope of the target line segment l is k l , and the foot of perpendicular from measurement i to line segment l is h l .

[0070] Optionally, the first condition is screened according to the distances of said distance pairs, so as to determine that said distance pairs are neighboring targets of said measurement, and obtain a first condition screening result. Sort the distances in the distance set D obtained in step 311 i in ascending order, and the elements in the distance pair are required to be smaller values in the distance set D i , so as to determine that the distance pair comes from neighboring targets of measurement m i .

[0071] Optionally, the second condition is screened according to said target slope, a preset slope threshold, said target foot of perpendicular, said measurement and the expected number of distance pairs, so as to determine that different distance pairs and said measurement form symmetry, and obtain a second condition screening result. Target slopes k formed by different distance pairs of one measurement are similar, and the distance from measurement m i to line segment l cannot be much larger than the length of line segment l, so as to ensure that the formation formed by different distance pairs and the measurement has symmetry. The screening discriminant of the second condition is:

[0072]

[0073] wherein, is the target slope determined by one distance pair, is the target slope determined by another distance pair, ε k is a preset slope threshold, m i is the measurement, h l is the target foot of perpendicular, ε p is the expected number of distance pairs, mj and m k For measuring m i Two points that are at a certain distance.

[0074] Optionally, based on the target perpendicular foot and the measurement screening third condition, two points of the same distance pair are determined to be located on opposite sides of the measurement, thus obtaining the third condition screening result. Perpendicular foot h k Within line segment l, to ensure that two points of the same distance pair are on opposite sides of the measurement point, the third condition's selection criterion is:

[0075] ||m j -h l ||2≤||m j -m k ||2

[0076] Where, ||m j -h l ||2 represents the distance measurement in meters. j to the target perpendicular h l The distance, ||m j -m k ||2 represents the distance between the two midpoints.

[0077] Optionally, based on the target perpendicular foot and the fourth measurement screening condition, it is determined that different distance pairs are located on the same side of the measurement, thus obtaining the fourth condition screening result. The slope direction of the perpendicular line from the measurement to the line segment l of different distance pairs is consistent to ensure that different distance pairs are on the same side of the measurement point. The fourth condition screening discriminant is:

[0078]

[0079] in, For measuring m i A distance to the foot of the perpendicular from the target forming the line segment. For measuring m i Another distance is the perpendicular to the target forming the line segment; "." indicates the vector dot product operation.

[0080] Optionally, valid distance pairs are obtained through the first condition filtering result, the second condition filtering result, the third condition filtering result, and the fourth condition filtering result. Distance pairs that simultaneously satisfy the filtering criteria of the first condition, the second condition, the third condition, and the fourth condition are considered valid distance pairs.

[0081] Step 313: Determine the clustering mode of the drone swarm as a goose flock mode by using the number of effective distance pairs and the preset expected distance pairs.

[0082] Optionally, when the number of effective distance pairs is greater than the preset expected number of distance pairs, the current drone swarm's swarm mode can be determined to be the goose flock mode.

[0083] Step 320: Based on the measurement, determine that the cluster mode of the drone swarm is wolf pack mode.

[0084] Optionally, wolves typically employ layered assistance and encirclement tactics during hunting. Drones simulating these wolf-like behaviors can improve the efficiency of attacks on enemy targets, making them suitable for close-range encirclement attacks using drone swarms, such as in anti-radiation attack modes. Figure 5 This is a diagram illustrating a swarm pattern for drone clusters that mimics a wolf pack.

[0085] The step 320 above, which determines that the cluster mode of the drone cluster is wolf pack mode based on the measurement, may specifically include steps 321 to 323.

[0086] Step 321: Based on the preset first measurement radius and first quantity threshold, perform spatial clustering on the measurement to obtain a first clustering result, the first clustering result including a first core point, a first boundary point and a first noise point.

[0087] Optionally, the spatial clustering method is DBSCAN clustering. Density-based noise spatial clustering (DBSCAN) is an unsupervised ML clustering algorithm. Unsupervised means that it does not use pre-labeled targets to cluster data points. Clustering refers to attempting to group similar data points into manually determined groups or clusters. The DBSCAN clustering method divides all points in a class into three types: core points, boundary points, and noise points.

[0088] Optionally, the number of midpoints of the first measurement radius can be obtained. The first measurement radius is set according to the actual situation, and the present invention does not impose any restrictions.

[0089] Optionally, when the number of points within the first measurement radius is greater than or equal to the first quantity threshold, the measurement is determined to be the first core point. The first quantity threshold is set according to the actual situation. When the number of points within the first measurement radius of the target measurement is greater than or equal to the first quantity threshold, the measurement is determined to be the first core point; wherein both the first measurement radius and the first quantity threshold are preset according to the actual situation.

[0090] Optionally, when the measurement is located within the first measurement radius of the first core point, and the number of points within the first measurement radius is less than the first quantity threshold, the measurement is determined to be the first boundary point. When a measurement point is located within the first measurement radius of the first core point, and the number of points within the first measurement radius of that measurement point is less than the first quantity threshold, the measurement is determined to be the first boundary point; when a measurement point is located within the first measurement radius of a point that has already been determined to be a first boundary point, and the number of points within the measurement radius of that point is less than the first quantity threshold, the measurement is also determined to be the first boundary point.

[0091] Optionally, when the measurement is neither the first core point nor the first boundary point, the measurement is determined to be the first noise point.

[0092] Step 322: Perform wolf pack class screening on the clustering results to obtain effective clusters.

[0093] Optionally, the core points and their adjacent boundary points in the clustering results obtained after completing step 321 are grouped into one cluster. The adjacent boundary points are boundary points located within the first measurement radius of the core point. When most of the points in a cluster fall within 1.5 times the first measurement radius, the cluster is considered a valid cluster. For example, the first core point p i Belongs to cluster C, Let be the set of points within the measurement radius of the first core point. The discriminant for an effective cluster is:

[0094] |{||p i -p j ||2≤1.5·ε d |p j ∈C}|≥80%·|C|

[0095] Where |·| represents the number of elements in the set.

[0096] Step 323: Based on the effective clusters and the first quantity threshold, determine that the cluster mode of the drone cluster is the wolf pack mode.

[0097] Optionally, when the number of valid clusters is greater than a first quantity threshold, the cluster mode of the drone cluster is determined to be the wolf pack mode.

[0098] Step 330: Based on the measurement, determine that the cluster mode of the drone swarm is swarm mode.

[0099] Optionally, bee colonies typically exhibit a concentrated, orderly, and forward-moving pattern during hive relocation. Drones can simulate this aggregation behavior, enabling transmission tasks even under communication constraints. (Refer to...) Figure 6This is a diagram illustrating a drone swarm model that mimics a bee swarm.

[0100] Optionally, the measurement is spatially clustered according to a preset second measurement radius and a second quantity threshold to obtain the second clustering result, which includes a second core point, a second boundary point, and a second noise point.

[0101] Optionally, the second clustering result is subjected to swarm class filtering to determine the cluster mode of the drone cluster as the swarm mode.

[0102] The step of performing swarm class filtering on the second clustering results to determine that the cluster mode of the drone cluster is the swarm mode can specifically include steps 331 to 334.

[0103] Step 331: Based on the second clustering result, obtain the nearest neighbor distance, second nearest neighbor distance and farthest distance between the second core points.

[0104] Optionally, the second core point and its adjacent boundary points are classified into a cluster. The nearest neighbor distance, second nearest neighbor distance and farthest distance between the core points adjacent to it in a cluster can be obtained from the measurement of the second core point.

[0105] Step 332: Based on the nearest neighbor distance and the second nearest neighbor distance, determine that the second core points maintain a certain distance to obtain the fifth condition filtering result.

[0106] Optionally, the nearest neighbor and second nearest neighbor distances of most of the second core points are similar to ensure that elements in the cluster maintain a certain spacing. The fifth conditional selection criterion is:

[0107]

[0108] Where C is the cluster after clustering, and the second core point subset in C is N = {p is a core point | p∈C}, and D is defined. N ={||p i -p j |||p i ∈N, p j ∈N} represents the distance between the second core points, and core point p i ∈C, sort(·) represents sorting the elements of the set in ascending order, d1 and sort(D) respectively i The 1st and εth p 1 element, i.e., d i | i＝1 and ε p This is the second quantity threshold.

[0109] Step 333: Based on the nearest neighbor distance and the farthest distance, determine that the set distribution of the second core points is a widely distributed set, and obtain the sixth condition screening result.

[0110] Optionally, the nearest distance between the second core points is much smaller than the farthest distance between the core points, to ensure that cluster C is a widely distributed set. The sixth conditional screening criterion is:

[0111] max(D N )≥1.5·ε p ·min(D N )

[0112] Among them, D N Let ε be the distance set between the core points. p The second threshold value is defined as max(·) and min(·), which represent the maximum and minimum values ​​of the elements in the set, respectively. N ) represents the nearest neighbor distance between the second core points, and max(·) represents the farthest distance between the second core points.

[0113] Step 334: Based on the results of the fifth and sixth condition screenings, determine that the cluster mode of the drone swarm is the identification result of the swarm mode.

[0114] Optionally, if a cluster satisfies both the fifth and sixth conditional screening criteria, then the cluster mode of the drone cluster is determined to be the swarm mode.

[0115] Optionally, considering the impact of noise and clutter on pattern recognition, for the identification of three swarming patterns—goose flock, wolf flock, and bee swarm—and considering the characteristic that radar target tracking requires the use of multiple frames of data, the following criteria can be used to improve the identification effect: if the measurement set is determined to be a certain swarming pattern for five consecutive time periods, then the target is considered to be flying in that swarming pattern; after a successful determination, the measurement set must not be determined to be a certain swarming pattern for five consecutive time periods before it is considered that the target is not flying in that swarming pattern.

[0116] The following example illustrates the application of a biomimetic drone swarm identification method provided by an embodiment of the present invention.

[0117] 1. First, the signal of the UAV swarm detected by the radar is processed to obtain the spatial position of the target UAV, and the measurement of the target UAV is obtained from the spatial position;

[0118] 2. Then, based on the measurements, the clustering patterns of the drone swarm are identified to obtain the identification results; wherein, the clustering patterns include goose flocking, wolf packing, and bee swarming.

[0119] 3. Based on the measurement, the clustering mode of the drone swarm is determined to be the goose flock mode; the distance pairs of the target drones are obtained from the measurement, and based on the distance pairs, effective distance pairs are filtered through the first condition, the second condition, the third condition, and the fourth condition. The clustering mode of the drone swarm is determined to be the goose flock mode by the number of effective distance pairs and the preset expected distance pairs.

[0120] 4. Based on the measurement, the cluster mode of the drone swarm is determined to be the wolf pack mode; based on the preset first measurement radius and first quantity threshold, the measurement is spatially clustered to obtain a first clustering result, the first clustering result including a first core point, a first boundary point and a first noise point; the clustering result is filtered by wolf pack to obtain effective clusters; based on the effective clusters and the first quantity threshold, the cluster mode of the drone swarm is determined to be the wolf pack mode;

[0121] 5. Based on the spatial location, determine that the cluster mode of the UAV cluster is a swarm mode; based on the preset second measurement radius and second quantity threshold, perform spatial clustering on the measurement to obtain the second clustering result, the second clustering result includes a second core point, a second boundary point and a second noise point, perform swarm class filtering on the second clustering result, and determine that the cluster mode of the UAV cluster is the swarm mode.

[0122] The following simulation experiment results verify the effectiveness of the identification method for a biomimetic drone swarm provided in this embodiment of the invention:

[0123] Simulations were used to verify the identification methods for three patterns: goose flock, wolf flock, and bee flock. The UAV was assumed to fly in multiple swarm patterns: mimicking a bee flock at t∈[0,200]s, a goose flock at t∈[200,400]s, and a wolf flock at t∈[400,500]s. The three pattern identification methods were then used for recognition. Table 1 shows the recognition accuracy of the three methods under different measurement error conditions. The recognition accuracy is the ratio of successful recognition time to flight time.

[0124] Table 1

[0125]

[0126] Where, σ m (m) represents different measurement errors.

[0127] Figure 3 The result image of pattern recognition is provided by Figure 3It can be seen that in the early stages of flight, the drone swarm cannot be successfully identified because it has not yet formed a formation. Once the formation stabilizes, the algorithm can effectively identify the swarm pattern of the drones. Comparing the results in the table, the accuracy when the measurement error is 0 is determined by the identification algorithm itself, and is also related to the formation time of the motion pattern. The larger the measurement error, the lower the pattern recognition accuracy, which is in line with expectations.

[0128] In summary, the identification method for biomimetic drone swarms according to embodiments of the present invention has the following advantages:

[0129] 1. This invention identifies the clustering patterns of UAV swarms based on measurements. The clustering patterns include goose flocking, wolf flocking, and bee swarming. This invention proposes three corresponding identification methods for the three types of UAV swarming patterns, thereby completing the pattern identification of UAV swarms. This solves the problem that the existing technology does not have the ability to perceive and identify the clustering patterns of UAV swarms, and can effectively identify different clustering patterns.

[0130] 2. Based on the identified cluster pattern, the purpose of the UAV cluster can be preliminarily determined, thereby further improving the accuracy of radar tracking.

[0131] Reference Figure 2 This invention also provides a biologically inspired drone swarm identification device, comprising:

[0132] The first module 201 is used to process the UAV swarm signals detected by the radar to obtain the spatial position of the target UAV;

[0133] The second module 202 is used to obtain measurements of the target UAV from the spatial location;

[0134] The third module 203 is used to identify the clustering pattern of the UAV swarm based on the measurement to obtain the identification result; wherein, the clustering pattern includes goose flocking pattern, wolf flocking pattern and bee swarming pattern.

[0135] This invention also provides a computer-readable storage medium storing a program that is executed by a processor to implement the aforementioned identification method for a biomimetic unmanned aerial vehicle swarm.

[0136] This invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, causing the computer device to perform... Figure 1 The method shown.

[0137] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is altered and sub-operations described as part of a larger operation are executed independently.

[0138] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the described functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0139] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, essentially, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0140] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0141] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0142] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0143] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0144] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0145] The above is a detailed description of the preferred embodiments of the present invention, but the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention, and these equivalent modifications or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A method for identifying biomimetic drone swarms, characterized in that, include: Signal processing is performed on the drone swarm signals detected by radar to obtain the spatial position of the target drone; Measurements of the target UAV are obtained from the spatial location; The distance pairs of the target UAV are obtained from the measurements; the target slope, target line segment, and target perpendicular foot are obtained based on the distance pairs; a first condition is selected based on the distance of the distance pairs, where the distance sets are sorted in ascending order to determine that the distance pairs are the neighboring targets of the measurements, resulting in a first condition selection result; a second condition is selected based on the target slope, a preset slope threshold, the target perpendicular foot, the number of measured and expected distance pairs, where the second condition uses a slope difference threshold and a ratio threshold between the measured distance to the line segment and the line segment length to determine symmetry, thus determining that different distance pairs form symmetry with the measurements, resulting in a second condition selection result; a third condition is selected based on the target perpendicular foot and the measurements, where the distance from one point in the distance pair to the perpendicular foot is less than or equal to the distance between the two points, determining that the two points of the same distance pair are located on opposite sides of the measurements, resulting in a third condition selection result; a fourth condition is selected based on the target perpendicular foot and the measurements, where the vector dot product is non-negative, determining that different distance pairs are located on the same side of the measurements, resulting in a fourth condition selection result; Valid distance pairs are obtained by filtering based on the first condition, the second condition, the third condition, and the fourth condition. The clustering mode of the drone swarm is determined to be a goose flock mode by the number of effective distance pairs and the preset expected distance pairs. Based on a preset first measurement radius and a first quantity threshold, the measurement is spatially clustered to obtain a first clustering result, which includes a first core point, a first boundary point, and a first noise point. The clustering results were subjected to wolf pack-type filtering to obtain effective clusters; Based on the effective clusters and the first quantity threshold, the cluster mode of the drone swarm is determined to be wolf pack mode; Based on the preset second measurement radius and second quantity threshold, the measurement is spatially clustered to obtain a second clustering result, which includes a second core point, a second boundary point, and a second noise point. The second clustering results are subjected to swarm class filtering to determine that the clustering mode of the drone cluster is swarm mode.

2. The identification method for a biomimetic unmanned aerial vehicle (UAV) swarm based on claim 1, characterized in that, The measurement is spatially clustered according to a preset first measurement radius and a first quantity threshold to obtain a first clustering result. The first clustering result includes a first core point, a first boundary point, and a first noise point, including: Obtain the number of midpoints in the first measurement radius; When the number of points in the first measurement radius is greater than or equal to the first quantity threshold, the measurement is determined to be the first core point. When the measurement is located within the first measurement radius of the first core point, and the number of points within the first measurement radius is less than the first quantity threshold, the measurement is determined to be the first boundary point. When the measurement is neither the first core point nor the first boundary point, the measurement is determined to be the first noise point.

3. The identification method for a biomimetic unmanned aerial vehicle (UAV) swarm based on claim 1, characterized in that, The step of performing bee colony filtering on the second clustering results to determine the cluster pattern of the drone swarm as the bee colony pattern includes: Based on the second clustering result, the nearest neighbor distance, second nearest neighbor distance, and farthest distance between the second core points are obtained; Based on the nearest neighbor distance and the second nearest neighbor distance, it is determined that a certain distance is maintained between the second core points, and the fifth condition filtering result is obtained; Based on the nearest neighbor distance and the farthest distance, the distribution of the second core points is determined to be a widely distributed set, thus obtaining the sixth condition filtering result; Based on the results of the fifth and sixth conditional filtering, the cluster mode of the drone swarm is determined to be the identification result of the swarm mode.

4. A biomimetic drone swarm identification device, characterized in that, include: The first module is used to process the signals of the drone swarm detected by the radar to obtain the spatial position of the target drone; The second module is used to obtain measurements of the target UAV from the spatial location; The third module is used to obtain distance pairs of the target UAV from the measurements; obtain the target slope, target line segment, and target perpendicular foot based on the distance pairs; filter a first condition based on the distance of the distance pairs, where the first condition determines that the distance pairs are the neighboring targets of the measurements by arranging the distance sets in ascending order, thus obtaining a first condition filtering result; filter a second condition based on the target slope, a preset slope threshold, the target perpendicular foot, the number of measured and expected distance pairs, where the second condition determines symmetry by using a slope difference threshold and a ratio threshold between the measured distance to the line segment and the line segment length, thus determining that different distance pairs form symmetry with the measurements, thus obtaining a second condition filtering result; filter a third condition based on the target perpendicular foot and the measurements, where the third condition determines that the two points of the same distance pair are located on opposite sides of the measurements by the distance from one point in the distance pair to the perpendicular foot being less than or equal to the distance between the two points, thus obtaining a third condition filtering result; filter a fourth condition based on the target perpendicular foot and the measurements, where the fourth condition determines that different distance pairs are located on the same side of the measurements by the non-negativity of the vector dot product, thus obtaining a fourth condition filtering result; Valid distance pairs are obtained by filtering based on the first condition, the second condition, the third condition, and the fourth condition. The clustering mode of the drone swarm is determined to be a goose flock mode by the number of effective distance pairs and the preset expected distance pairs. Based on a preset first measurement radius and a first quantity threshold, the measurement is spatially clustered to obtain a first clustering result, which includes a first core point, a first boundary point, and a first noise point. The clustering results were subjected to wolf pack-type filtering to obtain effective clusters; Based on the effective clusters and the first quantity threshold, the cluster mode of the drone swarm is determined to be a wolf pack mode; according to the preset second measurement radius and second quantity threshold, the measurement is spatially clustered to obtain a second clustering result, which includes a second core point, a second boundary point, and a second noise point; The second clustering results are subjected to swarm class filtering to determine that the clustering mode of the drone cluster is swarm mode.

5. A computer-readable storage medium, characterized in that, The storage medium stores a program that is executed by a processor to implement the method as described in any one of claims 1 to 3.

Images