A method for countering a drone "swarm"

Abstract

The application discloses a method for countering a UAV "swarm", which comprises the following steps: S1, acquiring three-dimensional spatial position distribution information of an enemy UAV swarm by a radar, and obtaining an augmented cluster order of the enemy UAV swarm based on a clustering algorithm; S2, obtaining a clustering result of the enemy UAV swarm by an OPTICS clustering algorithm according to the acquired information and the augmented cluster order parameter; S3, obtaining an optimized clustering result according to relationship data between the enemy UAV swarm and a target to be protected; and S4, optimizing and distributing the countermeasure according to the optimized clustering result and a priority level. The method for countering the UAV "swarm" can realize the acquisition of a target clustering result, the matching of a corresponding countermeasure priority, and the reasonable task distribution of a countermeasure UAV that can be dispatched according to the number and characteristics of each cluster, for the UAV swarm three-dimensional spatial density clustering.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of human-computer countermeasures, in particular to a method for countering a UAV "swarm". BACKGROUND

[0002] There have been many UAV safety incidents around the world. The public has become increasingly aware of the potential harm caused by UAVs. Currently, the main countermeasures against UAVs include fire interception, helicopter aerial interception and forced landing, net capture, electronic interference, microwave damage, and high-power laser damage. Among these countermeasures, electronic interference has a long range, a large angle, and strong ability to deal with multiple targets, and has become a development hotspot.

[0003] A UAV "swarm" system composed of multiple UAVs is a large-scale combat platform carrying different payloads. It relies on distributed intelligent group technology to complete complex tasks such as warning, search, target positioning, direction guidance, and attack in various environments through the perception interaction of individual UAVs and the information exchange between them, achieving low cost, low consumption, and high efficiency. The UAV "swarm" system has the characteristics of large scale, small size, low price, decentralization, and autonomous decision-making, and has new features that are much better than traditional air threats, which can put great pressure on any existing air defense system.

[0004] The countermeasures against the UAV "swarm" mainly include soft destruction and hard killing based on detection and perception. The former mainly includes interference suppression and induction of enemy UAV "swarm", and the latter mainly includes attack by artillery and laser directional energy weapons, swarm counterattack, and physical capture. In view of the "saturation" attack of enemy UAV "swarm", the use of modern weapons for attack has a low cost-effectiveness ratio, and may also lead to the exposure, depletion, and even paralysis of the combat resources of the air defense system.

[0005] At this time, using a "swarm" to counter enemy "swarm" is an "equivalent" and "efficient" countermeasure. It can deploy electronic countermeasures UAVs with large-area soft-killing capability to interfere with and suppress enemy UAVs, or deploy reconnaissance and attack integrated UAVs for counterattack, or set up obstacles to block the flight of enemy UAVs, or even deploy manned / unmanned aircrafts for close-range precision strikes. SUMMARY

[0006] The present application overcomes the shortcomings of the prior art and provides a method for countering a UAV "swarm". It can obtain target clustering results, match corresponding countermeasures priority, and reasonably allocate tasks to countermeasures UAVs that can be deployed according to the number and characteristics of each cluster.

[0007] To achieve the above object, the technical scheme adopted by the present application is as follows: a method for opposing a UAV "swarm" of enemy UAVs, comprising the following steps:

[0008] In step S1, three-dimensional spatial position distribution information of the enemy UAV swarm is collected by a radar, and information of the enemy UAV swarm is acquired based on a clustering algorithm to obtain an augmented cluster order of the enemy UAV swarm.

[0009] In step S2, the clustering result of the enemy UAV swarm is obtained by an OPTICS clustering algorithm based on the acquired information and the augmented cluster order parameter.

[0010] In step S3, an optimized clustering result is obtained based on the relationship data between the enemy UAV swarm and the target to be protected.

[0011] In step S4, the optimized clustering result and the priority level are used to optimize the opposition allocation.

[0012] In a preferred embodiment of the present application, in step S2, the eps value is calculated according to the size and distribution of the enemy UAV swarm to be opposed, and the eps value is a neighborhood radius value; the clustering result of the UAV swarm is obtained based on the reachability distance ordering graph and the core distance matrix obtained by the clustering algorithm; or / and, in step S3, the shortest distance index between the UAVs in the enemy UAV swarm and the target to be protected is calculated, the clustering result obtained based on the clustering algorithm is optimized, the number of clusters is greater than or equal to minPts, minPts is the minimum number of points in the Eps neighborhood, and the UAVs that are close to the target to be protected and have a close opposition distance can be added to the cluster as an expansion node; the cluster parameters related to the expansion node are calculated; or / and, in step S4, the priority of each cluster is set according to the shortest distance between the cluster center of each cluster and the target to be protected; the opposition task allocation is performed according to the principle of maximum combat effectiveness and the principle of high-value target defense of the target to be protected.

[0013] In a preferred embodiment of the present application, the three-dimensional spatial position distribution information includes longitude, latitude and altitude; wherein the longitude and latitude data are converted into radians, and the altitude is expressed in meters; or / and, the cluster parameters include one or more of the cluster center, the number of members, the average core density and the neighborhood radius of each cluster.

[0014] In a preferred embodiment of the present application, in step S1, minPts is set as the minimum number of neighbors within a domain radius ε, the target position to be detected is three-dimensional data, and minPts is greater than or equal to 4; the enemy UAV swarm is subjected to density clustering based on the density clustering algorithm OPTICS, wherein the distance between the enemy UAVs in the swarm is calculated as follows:

[0015] The spatial positions of the node one and the node two are respectively (lon_1, lat_1, alt_1) and (lon_12, lat_2, alt_2), wherein lon represents longitude, lat represents latitude, and alt represents altitude; the Haversine distance of the longitude and latitude of the node one and the node two is formula haversine ((lon_1, lat_1), ((lon_12, lat_2)).m, the unit of measurement is meter; the distance calculation method of the altitude of the node one and the node two is:

[0016] dist = sqrt (haversine ((lon_1, lat_1), ((lon_12, lat_2)).m**2, (alt_1-alt_2)**2), wherein '.m' represents that the distance is measured by meters; after the enemy UAV swarm is subjected to density clustering, the clustering structure of the enemy UAV swarm density, that is, the augmented cluster order, can be obtained.

[0017] In a preferred embodiment of the application, in step S2, the acquisition of the OPTICS clustering result is based on the clustering cluster formed by the valley part of the eps value in the reachability distance graph, the eps value is calculated according to the scale of the enemy UAV swarm to be implemented against; according to the calculated eps value, in the reachability distance graph obtained by the OPTICS clustering, the clustering result of the UAV swarm is obtained according to the reachability distance ordering graph and the core distance matrix obtained by the OPTICS clustering algorithm. Specifically, for the detected enemy UAV swarm with a scale of n, n is the number of enemy swarm UAVs; x UAVs are expected to be attacked (n>x), then the reachability distance matrix calculated by the OPTICS clustering algorithm is sorted in ascending order, and the xth reachability distance value is taken as its eps value, at this time the number of clustered UAVs will be ≤x; wherein part of the nodes are noise points, the value can be adjusted according to the number of noise points and the number of clusters, the value of x / n is between 0.4 and 0.7, that is, the proportion of multiple UAVs expected to be attacked is between 40% and 70% of the enemy swarm. Specifically, if the value of x is too small, since the UAV swarm is widely distributed, a too small eps value will result in that many UAVs are noise points and cannot be clustered, on the contrary, if the value of x is too large, most of the UAVs will be placed in the cluster, which will result in that the adjacent distance of the UAVs in the cluster is too large and loses the practical significance of the confrontation.

[0018] In a preferred embodiment of the present application, in step S3, the shortest distance between each UAV in the enemy swarm and the target to be protected is calculated, and the distance value is quantized to the interval [0.9, 1.1], representing the distance index of the target to be protected, if the distance index value < 1, it represents that the distance to the important target is relatively close, otherwise it is relatively far; when the number of nodes in the cluster is less than minPts, the nodes in the cluster need to be processed.

[0019] In a preferred embodiment of the present application, the number of enemy swarm UAVs is n, which is denoted as enemy swarm UAV set U = {u1, u2, u n}, and the number of targets to be protected is q, which is denoted as target set O = {o1, o2, o n}; then the distance between any UAV u i and any target q j in the target set Q is d ij , and the distance matrix D n×q can be obtained, where 1≤i≤n, 1≤j≤q. Let d min = min(D) and d max = max(D) represent the nearest distance and the farthest distance between each UAV in the enemy swarm and the target to be protected. For any UAV u i , define the distance min(d ij ) between u i and the nearest target as d i , and the quantized value l i of the distance between u i and the target to be protected is

[0020] In a preferred embodiment of the present application, in step S3, the node processing includes: first, obtaining the distance index of each node in the cluster, which is greater than 1, indicating that each UAV in the cluster is far away from the important target, and then all the nodes in the cluster are treated as noise points.

[0021] Otherwise, take the nearest minPts unclustered adjacent nodes of each node to join the list of nodes to be processed, and for each node in the list, if the weighted core distance of the adjacent node is less than eps, that is, the core distance multiplied by the distance index is less than eps, then the adjacent node is added to the cluster as an extended node, otherwise the node is still treated as a noise node.

[0022] After all the nodes to be processed are processed in turn, if the total number of nodes in the expanded cluster is greater than or equal to minPts, the cluster is retained, otherwise all the nodes in the cluster are treated as noise points.

[0023] In a preferred embodiment of the present application, in step S3, the clustering results of the enemy UAV swarm are further analyzed after optimization, the cluster center, the number of members, the shortest distance from the clustering center to the target to be protected, the average core density of the cluster, and the maximum radius of the cluster are calculated, and the priority of each cluster is set according to the shortest distance from the clustering center to the target to be protected, that is, the node closest to the target to be protected has the highest priority, and the priority is lower as the distance increases. Define the number of cluster classes as p, then the priority can be allocated from 0 to p-1, and the priority weight is represented by 1, (p-1) / p, 1 / p respectively.

[0024] In a preferred embodiment of the present application, in step S4, the counter task list of the UAV swarm that can be dispatched to counter the enemy UAV swarm can be obtained according to the clustering results of the enemy UAV swarm, and the UAV swarm that can be dispatched to counter is assigned tasks according to the principle of maximizing combat effectiveness and the principle of best defense of important high-value targets.

[0025] In a preferred embodiment of the present application, in step S4, the method for distributing the combat effectiveness value of different types of UAVs in the UAV swarm dispatched to counter the enemy UAV swarm to each UAV in the enemy UAV swarm includes: if the distance between the UAV and the cluster center exceeds 1 / 2 of the maximum combat distance of the UAV, or the height of the cluster center exceeds the maximum flight height of the UAV, the combat effectiveness index value e is 0, that is, the UAV cannot perform the task at all; otherwise, the task matching degree is calculated, and the task matching degree between the UAV and the task is k, and the combat effectiveness index value is calculated as 100-(4-k)*25; the combat effectiveness index value is divided by grade, and when the combat effectiveness index of different types of UAVs to the same task is the same, the priority order of different types of UAVs is defined as electronic warfare UAV>attack UAV>capture UAV, that is, the electronic countermeasure UAV is preferred under the same condition;

[0026] Or / and, in step S4, the combat task allocation method includes: setting the number of clusters of the enemy UAV swarm clustering as t, and the number of UAVs that can be dispatched as u, u is a non-zero natural number, then there are two cases t<u or t≥u; the combat task allocation is to arrange all the UAVs that can be dispatched to perform combat tasks, and each UAV can only perform one task; first, the defense principle of important targets is met, and the cluster of the enemy UAV swarm with the highest priority and the largest threat distance is assigned tasks, and then the unassigned tasks are sequentially assigned;

[0027] When t >= u, all UAV task assignments are completed, and the process ends, and the farther and lower level long-distance cluster can be processed as the next combat target; or, when t < u, after all tasks are assigned, the u-t remaining UAVs are calculated for combat effectiveness values with each task, and all UAVs are assigned tasks according to the principle of the maximum combat effectiveness value.

[0028] The present application solves the defects in the background art, and has the following beneficial effects:

[0029] A UAV "swarm" confrontation method, which can obtain target clustering results, match corresponding confrontation priorities, and reasonably assign tasks to the countermeasure UAVs that can be dispatched according to the number and characteristics of each cluster, so as to weaken the combat capability of enemy UAV swarms, reduce the difficulty of low-altitude defense, and effectively counterattack the saturation attack of enemy UAV swarms. BRIEF DESCRIPTION OF DRAWINGS

[0030] The present application will be further described below in conjunction with the drawings and examples.

[0031] Figure 1 is a schematic diagram of the confrontation system structure of the preferred embodiment of the present application;

[0032] Figure 2 is a schematic diagram of the system structure of the combat effectiveness index calculation in the preferred embodiment of the present application;

[0033] Figure 3 is a schematic diagram of the system structure of the UAV confrontation task assignment in the preferred embodiment of the present application;

[0034] Figure 4 is a schematic diagram of the reachability distance map in the preferred embodiment of the present application (divided according to the threshold line (eps), there are three valleys, which means that there are three clusters, and the coordinate values of the reachable distance are only a segmented diagram, and the specific parameters of the actual reachable distance are defined according to the actual operation process). DETAILED DESCRIPTION

[0035] The present application will be further described below in conjunction with the drawings and examples, which are all simplified schematic diagrams, and only illustrate the basic structure of the present application in a schematic manner, and therefore only show the configurations related to the present application.

[0036] Example 1

[0037] As shown in Figures 1-4 , a UAV "swarm" confrontation method includes the following steps:

[0038] Step S1, according to the radar, the three-dimensional spatial position distribution information of the enemy UAV swarm is collected, the information of the enemy UAV swarm is acquired based on the clustering algorithm, and the augmented cluster order of the enemy UAV swarm is obtained. The three-dimensional spatial position distribution information includes longitude, latitude and altitude; wherein, the longitude and latitude data are converted into radian, and the altitude is expressed in meters; the clustering algorithm adopts the OPTICS clustering algorithm, and the cluster clustering parameters of the OPTICS clustering algorithm include the cluster center, the member quantity, the average core density and the maximum radius of each cluster.

[0039] Specifically, set minPts as the minimum number of neighbors within the domain radius ε, the target position to be detected is three-dimensional data, and minPts takes a value ≥4; in this embodiment, minPts takes a value of 4. The enemy UAV swarm is subjected to density clustering based on the density clustering algorithm OPTICS, wherein the calculation method of the distance between the enemy UAV swarm is as follows: set the spatial positions of node one and node two as (lon_1, lat_1, alt_1) and (lon_12, lat_2, alt_2), wherein lon represents longitude, lat represents latitude, and alt represents altitude; set the Haversine distance of the longitude and latitude of node one and node two as formula haversine ((lon_1, lat_1), ((lon_12, lat_2)).m, the unit of measurement is meter; the distance calculation method of the altitude of node one and node two is as follows: dist=sqrt(haversine((lon_1,lat_1),((lon_12,lat_2)).m**2,(alt_1-lat_2)**2); the function is the calling method of the Haversine distance formula, and “.m” represents that the distance is measured by meters; after the enemy UAV swarm is subjected to density clustering, the clustering structure of the enemy UAV swarm density, i.e. the augmented cluster order, can be obtained.

[0040] Step S2, according to the acquired information and augmented cluster order parameters, the clustering result of the enemy UAV swarm is acquired through the clustering algorithm. According to the size and distribution of the enemy UAV swarm to be counteracted, the value of eps is calculated, and the value of eps is the value of the neighborhood radius; according to the reachability distance ordering diagram and the core distance matrix acquired by the clustering algorithm, the clustering result of the UAV swarm is acquired.

[0041] Specifically, the clustering algorithm adopts the OPTICS clustering algorithm; and the OPTICS clustering result is obtained according to the clustering cluster formed by the valley part of the eps value in the reachability distance graph; the eps value is calculated according to the scale of the enemy unmanned aerial vehicle swarm to be implemented against; for the enemy unmanned aerial vehicle swarm of the scale n detected, it is expected to implement attack against x unmanned aerial vehicles (n > x), first, the reachability distance matrix calculated by the OPTICS clustering algorithm is sorted in ascending order, and then the mth reachability distance value is taken as the eps value, at this time, the number of clustered unmanned aerial vehicles obtained is ≤ x; a small number of nodes are noise points, and the value of x is moderately increased, preferably, the value of x / n is 0.4-0.7. Specifically, when the value of x is too small, because the unmanned aerial vehicle swarm is widely distributed, too small eps value will result in that many unmanned aerial vehicles are noise points and cannot be clustered, on the contrary, when the value of m is too large, most of the unmanned aerial vehicles will be placed in the cluster, which will result in that the adjacent distance of the unmanned aerial vehicles in the cluster is too large and loses the practical significance of the confrontation.

[0042] According to the calculated eps value, in the reachability distance graph obtained by the OPTICS clustering, the clustering result of the unmanned aerial vehicle swarm is obtained according to the reachability distance sorting graph and the core distance matrix obtained by the OPTICS clustering algorithm.

[0043] Step S3, obtaining the optimized clustering result according to the relationship data between the enemy unmanned aerial vehicle swarm and the target to be protected.

[0044] Specifically, the shortest distance index of the unmanned aerial vehicles in the enemy unmanned aerial vehicle swarm and the target to be protected is calculated, the clustering result obtained based on the clustering algorithm is optimized, the number of clusters is greater than or equal to minPts, minPts is the minimum number of points in the Eps neighborhood, and the unmanned aerial vehicles which are close to the target to be protected and have a close confrontation distance can be added to the cluster as expansion nodes; the cluster parameters related to the expansion nodes are calculated.

[0045] Further, the shortest distance between each unmanned aerial vehicle in the enemy swarm and the target to be protected is calculated, and the distance value is quantized to the interval [0.9, 1.1], representing the distance index of the target to be protected, if the distance index value <1, it represents that the distance to the important target is relatively close, otherwise it is relatively far; when the number of nodes in the clustering cluster obtained according to the calculated eps value is less than minPts, the nodes in the clustering cluster need to be processed. Specifically, the number of enemy swarm unmanned aerial vehicles is n, which is counted as the enemy swarm unmanned aerial vehicle set U = {u1, u2, u n}, the number of targets to be protected by our side is q, which is counted as the target to be protected by our side set O = {o1, o2, o n}; then for any unmanned aerial vehicle u i , and any target qj The distance between the ith node and the jth node is d ij The distance matrix D is obtained as follows n×q where 1≤i≤n, 1≤j≤q. Let d min =min(D) and d max =max(D) represent the nearest distance and the farthest distance between each UAV in the enemy swarm and the target to be protected, respectively. For any UAV u i , define the distance between u i and the nearest target to be protected as d ij =min(d i ). Then, the distance quantization value l i of u i to the target to be protected is

[0046] The node processing includes: first, obtaining the distance index of each node in the cluster, all of which are greater than 1, indicating that each UAV in the cluster is far away from the important target, and then all the nodes in the cluster are treated as noise points; otherwise, the nearest minPts unclustered neighboring nodes of each node are added to the list of nodes to be processed, and for each node in the list, if the weighted core distance of its neighboring node is less than eps, i.e., the core distance multiplied by the distance index is less than eps, then the neighboring node is added to the cluster as an expansion node of the cluster, otherwise the node is still treated as a noise node; after all the nodes to be processed are processed in turn, if the total number of nodes in the expanded cluster is greater than or equal to minPts, the cluster is retained, otherwise all the nodes in the cluster are set as noise points. After the cluster result is optimized, the cluster result of the enemy UAV swarm is further analyzed, and the cluster center, the number of members, the shortest distance from the cluster center to the target to be protected, the average core density of the cluster, and the maximum radius of the cluster are calculated. Then, the priority of each cluster is set according to the shortest distance from the cluster center to the target to be protected, i.e., the node closest to the target to be protected has the highest priority, and the priority decreases with the increase of the distance. If the number of clusters is p, the priority can be allocated from 0 to p-1, and the priority weight is represented by 1, (p-1) / p, and 1 / p, respectively.

[0047] In step S4, the optimized cluster result and the priority level are used to optimize the confrontation allocation. The priority of each cluster is set according to the shortest distance from the cluster center to the target to be protected, and the task allocation is performed according to the principle of maximizing combat effectiveness and the principle of best defense of important high-value targets.

[0048] Specifically, the cluster result of the enemy UAV swarm can be used to obtain the confrontation task list of the UAV swarm that can be dispatched for confrontation. According to the type and number of UAVs in the UAV swarm that can be dispatched for confrontation, the task allocation is performed on the UAV swarm that can be dispatched for confrontation according to the principle of maximizing combat effectiveness and the principle of best defense of important high-value targets.

[0049] Further, the method for distributing the combat effectiveness value of different types of unmanned aerial vehicles in the dispatched opposing unmanned aerial vehicle swarm against each unmanned aerial vehicle in the enemy unmanned aerial vehicle swarm comprises: Figure 2 As shown in the formula, if the distance between the unmanned aerial vehicle and the cluster center exceeds 1 / 2 of the maximum combat distance of the unmanned aerial vehicle, or the height of the cluster center exceeds the maximum flight height of the unmanned aerial vehicle, the combat effectiveness index value e of the unmanned aerial vehicle is 0, that is, the unmanned aerial vehicle cannot perform the task at all; otherwise, the task matching degree between the unmanned aerial vehicle and the task in the task matching degree calculation is k, and the matching values from high to low are 4, 3, 2, and 1; the task matching degree in the embodiment is k. Then, the combat effectiveness index value is calculated according to 100-(4-k)*25; the combat effectiveness index value is divided according to the grade, and in the case that the combat effectiveness index of different types of unmanned aerial vehicles against the same task is the same, the priority order of different types of unmanned aerial vehicles is defined as electronic warfare unmanned aerial vehicle > attack unmanned aerial vehicle > capture unmanned aerial vehicle, that is, the electronic countermeasure unmanned aerial vehicle is preferentially used in the same case.

[0050] In step S4, as shown in the formula, Figure 3 The combat task distribution method comprises: setting the number of clusters of the enemy unmanned aerial vehicle swarm cluster as t, and the number of the unmanned aerial vehicles that can be dispatched as u, u is a non-zero natural number, then there are two cases t < u or t ≥ u; the combat task distribution is to arrange all the unmanned aerial vehicles that can be dispatched with combat tasks, and each unmanned aerial vehicle can only perform one task; first, the defense principle of important targets is met, the cluster of the enemy unmanned aerial vehicle swarm with the highest priority and the largest threat distance is distributed with tasks, and then the tasks are distributed in sequence; when t ≥ u, all the unmanned aerial vehicles are distributed with tasks, and the process ends; the low-level long-distance cluster far away can be processed as the next combat target; or, when t < u, after all the tasks are distributed, the u-t unmanned aerial vehicles that are not distributed with tasks are calculated for the combat effectiveness value with each task, and then the unmanned aerial vehicles are distributed with tasks according to the principle of the largest combat effectiveness value.

[0051] Further, the small unmanned aerial vehicle "swarm" is usually released in the low and medium altitude near the target area cluster, and the countermeasure method of the "swarm" against the "swarm" proposed in the present application is mainly aimed at the unmanned aerial vehicle cluster in the low and medium altitude and close to the important target area of our side, and the low-altitude early warning value can be selected as: the flight height is between 1000 meters and 5000 meters, and the distance from the high-value important target of our side is less than 200 km. But it is not limited to this, in other embodiments, the specific parameters can be adjusted according to the actual use demand.

[0052] Embodiment two

[0053] As shown in the formula, Figures 1-3 A method for unmanned aerial vehicle "swarm" confrontation, comprising the following steps:

[0054] Step S1, according to the radar, the three-dimensional spatial position distribution information of the enemy UAV swarm is collected, the information of the enemy UAV swarm is acquired based on the clustering algorithm, and the augmented cluster order of the enemy UAV swarm is obtained. The three-dimensional spatial position distribution information includes longitude, latitude and altitude; wherein the longitude and latitude data are converted into radians, and the altitude is expressed in meters; the clustering algorithm adopts the OPTICS clustering algorithm or the DBCSAN clustering algorithm; in this embodiment, the OPTICS clustering algorithm is preferred. The cluster clustering parameters include the cluster center, the member quantity, the average core density and the maximum radius of each cluster.

[0055] Specifically, set minPts as the minimum number of neighbors within the domain radius epsilon, the target position to be detected is three-dimensional data, and minPts takes a value greater than or equal to 4; in this embodiment, minPts takes a value of 4. The enemy UAV swarm is subjected to density clustering based on the density clustering algorithm OPTICS, wherein the calculation method of the distance between the enemy UAV swarm is as follows: set the spatial positions of node one and node two as (lon_1, lat_1, alt_1) and (lon_12, lat_2, alt_2), wherein lon represents longitude, lat represents latitude, and alt represents altitude; set the Haversine distance of the longitude and latitude of node one and node two as formula haversine((lon_1, lat_1), ((lon_12, lat_2)).m, the unit of measurement is meter; the distance calculation method of the altitude of node one and node two is as follows: dist=sqrt(haversine((lon_1, lat_1), ((lon_12, lat_2)).m**2, (alt_1-lat_2)**2); wherein “.m” represents the distance measured in meters; after the enemy UAV swarm is subjected to density clustering, the clustering structure of the enemy UAV swarm density, i.e. the augmented cluster order, can be obtained.

[0056] Step S2, according to the acquired information and augmented cluster order parameters, the clustering result of the enemy UAV swarm is acquired through the clustering algorithm. According to the scale and distribution of the enemy UAV swarm to be counteracted, the value of eps is calculated, and the value of eps is the value of the neighborhood radius; according to the reachability distance ordering diagram and the core distance matrix acquired by the clustering algorithm, the clustering result of the UAV swarm is acquired.

[0057] Specifically, the clustering algorithm adopts the OPTICS clustering algorithm; and the OPTICS clustering result is obtained according to the clustering cluster formed by the valley part of the eps value in the reachability distance graph; the eps value is calculated according to the scale of the enemy unmanned aerial vehicle swarm to be implemented against; for the enemy unmanned aerial vehicle swarm of the scale n detected, it is expected to implement attack against x unmanned aerial vehicles (n>x), then the reachability distance matrix calculated by the OPTICS clustering algorithm is sorted in ascending order, and the xth reachability distance value is taken as the eps value, at this time the number of clustered unmanned aerial vehicles obtained is ≤x; a small number of nodes are noise points, and the value of x is moderately increased, preferably, the value of x / n is 0.4-0.7. Specifically, when the value of x is too small, because the unmanned aerial vehicle swarm is widely distributed, too small eps value will result in that many unmanned aerial vehicles are noise points and cannot be clustered, on the contrary, when the value of m is too large, most unmanned aerial vehicles will be placed in the cluster, which will result in that the adjacent distance of the unmanned aerial vehicles in the cluster is too large and loses the practical significance of the confrontation.

[0058] According to the calculated eps value, in the reachability distance graph obtained by the OPTICS clustering, the clustering result of the unmanned aerial vehicle swarm is obtained according to the reachability distance sorting graph and the core distance matrix obtained by the OPTICS clustering algorithm.

[0059] Step S3, obtaining the optimized clustering result according to the relationship data between the enemy unmanned aerial vehicle swarm and the target to be protected.

[0060] Specifically, the shortest distance index of the unmanned aerial vehicles in the enemy unmanned aerial vehicle swarm and the target to be protected is calculated, the clustering result obtained based on the clustering algorithm is optimized, the number of clusters is greater than or equal to minPts, minPts is the minimum number of points in the Eps neighborhood, and the unmanned aerial vehicles which are close to the target to be protected and have a close confrontation distance can be added to the cluster as expansion nodes; the cluster parameters related to the expansion nodes are calculated.

[0061] Further, the shortest distance between each unmanned aerial vehicle in the enemy swarm and the target to be protected is calculated, and the distance value is quantized to the interval [0.9, 1.1], representing the distance index of the target to be protected, if the distance index value <1, it represents that the distance between the unmanned aerial vehicle and the important target is relatively close, otherwise it is relatively far; when the number of nodes in the clustering cluster is less than minPts, the nodes in the clustering cluster need to be processed according to the clustering result of the enemy unmanned aerial vehicle swarm obtained by the calculated eps value. Let the number of enemy swarm unmanned aerial vehicles be n, and the enemy swarm unmanned aerial vehicle set U={u1, u2, u n}, the number of targets to be protected by our side is q, and the target to be protected set O={o1, o2, o n}; then for any unmanned aerial vehicle u i in the set, and any target q jThe distance between the ith node and the jth node is d ij The distance matrix D is obtained n×q Where 1≤i≤n, 1≤j≤q. Let d min = min(D) and d max =max(D) represent the nearest distance and the farthest distance between each UAV in the enemy swarm and the target to be protected. For any UAV u i , define u i 's nearest target distance min(d ij ) as d i , then u i 's distance quantization value l i to the target to be protected is

[0062] The node processing includes: first, obtaining the distance index of each node in the cluster, all greater than 1, which means that each UAV in the cluster is far away from the important target, then all the nodes in the cluster are treated as noise points; otherwise, take the nearest minPts unclustered adjacent nodes of each node to join the to-be-processed node list, for each node in the list, if the weighted core distance of its adjacent node is less than eps, that is, the core distance multiplied by the distance index is less than eps, then its adjacent node is added to the cluster as an expansion node of the cluster, otherwise the node is still treated as a noise node; after all the to-be-processed nodes are processed in turn, if the total number of nodes in the expanded cluster is greater than or equal to minPts, the cluster is retained, otherwise all the nodes in the cluster are set as noise points. After the cluster result is optimized, the clustering result of the enemy UAV swarm is further analyzed, the cluster center, the number of members, the shortest distance from the cluster center to the target to be protected, the average core density of the cluster, and the maximum radius of the cluster are calculated, and then the priority of each cluster is set according to the shortest distance from the cluster center to the target to be protected, that is, the node closest to the target to be protected has the highest priority, and the priority decreases with the increase of the distance, and the number of cluster classes is defined as p, then the priority can be allocated from 0 to p-1, and the priority weight is represented as 1, (p-1) / p, 1 / p respectively.

[0063] Table 1 shows an example of analysis data of each cluster

[0064]

[0065]

[0066] Step S4, according to the optimized clustering result and the priority level, the countermeasure is optimized. According to the shortest distance from the cluster center to the target to be protected, the priority of each cluster is set; according to the principle of maximum combat effectiveness and the principle of high-value target defense of the target to be protected, the countermeasure task is allocated.

[0067] Specifically, the counter task list of the UAV swarm capable of being dispatched for confrontation can be obtained according to the clustering result of the enemy UAV swarm, and the UAV swarm capable of being dispatched for confrontation is allocated tasks according to the principle of maximum combat effectiveness and the principle of best defense of important high-value targets, in view of the type and number of UAVs capable of being dispatched for confrontation.

[0068] The UAV capable of being dispatched for confrontation generally includes electronic countermeasure UAV, reconnaissance and attack integrated UAV, UAV carrying combat attack weapons, UAV with the ability to capture and block the flight of enemy UAV, etc., but is not limited thereto, and in its embodiments, can also be manned aircraft. The combat capability of the dispatched UAV is mainly represented by the combat radius, attack and confrontation capability, etc., and is limited by the flight distance, flight height, etc. For the electronic countermeasure UAV, including radar countermeasure and communication countermeasure UAV, the combat radius thereof is the maximum range of the electronic interference thereof, and the corresponding attack and confrontation capability can be set to a maximum value, indicating that the UAVs within the combat radius will be attacked without quantity limitation. For the UAV for attack and confrontation, the maximum combat radius thereof is generally limited by the weapon attack range thereof, and due to the mobility thereof, the appropriate combat radius thereof can be set to 3-10 times of the weapon attack range thereof. For the UAV with the ability to capture and block the flight of enemy UAV by launching capture nets, etc., the combat radius thereof depends on the actual area of the capture net and the flight speed, etc. The area of the capture net is generally several square meters, the flight speed thereof is generally greater than 50 km / h, the combat radius thereof is about 1000 meters, and the attack and confrontation capability thereof is related to the number of UAVs distributed within the range thereof, and the value thereof can be set to a maximum value, which is generally suitable for confrontation with low, slow and small UAV clusters.

[0069] The main technical indicators of typical UAVs for confrontation are shown in the following table:

[0070]

[0071] The counter task list of the UAV swarm for implementing the counter task can be obtained according to the clustering result of the enemy UAV swarm, and the counter UAV swarm for implementing the counter task can be allocated tasks according to the principle of maximum combat effectiveness and the principle of best defense of important high-value targets, in view of the type and number of UAVs capable of being dispatched for confrontation.

[0072] The task allocation method is as follows:

[0073] Let the number of enemy UAV swarm clustering be a, and the counter tasks thereof can be represented as T1-Ta.

[0074] The representation method for each task Ti is as follows:

[0075] Ti={priority, center, member, short_dist, mean_density, max_radius}, respectively represent its priority, cluster center position (longitude, latitude, altitude), the number of members, the shortest distance (m) from the cluster center to the important target, the average core density (m) of the cluster, and the maximum radius (m) of the cluster.

[0076] The b unmanned aerial vehicles that can be dispatched to implement the countermeasure task can be respectively represented as U1~Ub. For each unmanned aerial vehicle Uj, the representation method is:

[0077] Uj={type, max_radius, max_capacity, max_dist, max_height}, respectively represent the type, maximum combat radius, attack countermeasure ability, maximum flight distance (m) and maximum flight height of each unmanned aerial vehicle.

[0078] The combat effectiveness parameter effectiveness of the task and the unmanned aerial vehicle matching is difficult to calculate uniformly by an absolute formula. Therefore, first, the information of each cluster analysis result of each cluster is converted:

[0079] The number of members in the cluster can be divided into three levels, represented by 0-2 respectively, representing few, general, and many members; the average core density of each cluster can be divided into three levels, represented by 0-2 respectively, representing small, general, and large cluster density; the maximum radius of each cluster can be divided into three levels, represented by 0-2 respectively, representing small, general, and large cluster density. Secondly, for each type of task unmanned aerial vehicle, the adaptability of the unmanned aerial vehicle group style can be represented by grading according to its combat ability, which can be divided into five levels, represented by 0-4 respectively, representing complete mismatch, mismatch, general, very match, and complete match.

[0080] The task matching degree table of different types of unmanned aerial vehicles for different countermeasure unmanned aerial vehicle swarms is shown in the following table:

[0081]

[0082] The distribution method of the combat effectiveness value of different types of unmanned aerial vehicles in the dispatched countermeasure unmanned aerial vehicle swarm to each unmanned aerial vehicle in the enemy unmanned aerial vehicle swarm includes: Figure 2As shown, if the distance between the UAV and the cluster center exceeds 1 / 2 of the maximum combat distance of the UAV, or the height of the cluster center exceeds the maximum flight height of the UAV, the combat effectiveness index value e of the UAV is 0, that is, the UAV cannot perform the task at all; otherwise, the task matching degree is calculated between the UAV and the task, and k is the task matching degree between the UAV and the task; the matching values from high to low are 4, 3, 2, and 1 respectively. Then, the combat effectiveness index value is calculated according to 100-(4-k)*25; the combat effectiveness index value is divided according to the grade, and the priority order of different types of UAVs is defined in the case that the combat effectiveness index of the same task of different types of UAVs is the same, that is, the electronic warfare UAV is preferred to the attack UAV, and the capture UAV, that is, the electronic countermeasure UAV is preferred under the same condition.

[0083] In step S4, as shown in Figure 3 The combat task allocation method includes: setting the number of clusters of enemy UAV swarms as t, the number of deployable UAVs as u, u is a non-zero natural number, then including two cases t

[0084] Further, the small UAV "swarm" is usually released in the low altitude near the target area, and the countermeasure method of the "swarm" against the "swarm" proposed in the application is mainly aimed at the UAV swarm in the low altitude and close to the important target area of the home side, and the early warning value of the low altitude early warning can be selected as: the flight height is between 1000m and 5000m, and the distance from the high value important target of the home side is less than 200km. But it is not limited to this, in other embodiments, the specific parameters can be adjusted according to the actual use demand.

[0085] Working principle:

[0086] The anti-countermeasure method obtains the augmented cluster order of the enemy UAV swarm based on OPTICS by density clustering of the enemy UAV swarm, and can obtain the clustering result of the UAV swarm according to the enemy swarm target. The neighborhood radius eps is calculated according to the enemy swarm target, the clustering result of the UAV swarm to be counteracted is obtained, and the clustering result is adjusted and optimized according to the distance of the UAV swarm to be counteracted from the important target. According to the clustering result, the center node position, node quantity, average core distance of clustering, maximum radius of clustering, etc. of each cluster of the UAV swarm to be counteracted can be obtained, and the combat priority of each cluster is set according to the shortest distance between the center of each cluster of the UAV swarm to be counteracted and the important target to be protected. The UAVs available for counteracting are reasonably assigned tasks according to the number and characteristics of each cluster of the UAV swarm to be counteracted, so as to maximize the combat capability of the enemy UAV swarm, reduce the difficulty of low-altitude defense, and effectively counteract the saturation attack of the enemy UAV swarm. The unknown source UAVs in the military management area or other institutions or areas involving security can be safely and efficiently interfered and counteracted.

[0087] The above is based on the ideal embodiment of the application, and through the above description, relevant personnel can make various changes and modifications without deviating from the technical idea of the application. The technical scope of the application is not limited to the content of the specification, and must be determined according to the scope of the claims.

Claims

1. A method of counteracting a swarm of drones, characterized in that, Comprise the following steps: Step S1, according to the radar acquisition enemy unmanned aerial vehicle swarm three-dimensional space position distribution information, based on clustering algorithm to enemy unmanned aerial vehicle swarm information acquisition, obtain the augmented cluster ordering of enemy unmanned aerial vehicle swarm; Step S2, according to the information obtained and augmented cluster ordering parameter, through OPTICS clustering algorithm to obtain the clustering result of enemy unmanned aerial vehicle swarm; Step S3, according to the relationship data between enemy unmanned aerial vehicle swarm and the target to be protected to obtain the optimized clustering result; Step S4, according to the optimized clustering result and priority level optimization countermeasure allocation; In step S2, the acquisition of OPTICS clustering result is according to the clustering cluster formed in the valley part of the reachability distance graph according to the eps value, the scale and distribution of the enemy unmanned aerial vehicle swarm to be implemented countermeasure are calculated, and the eps value is the neighborhood radius value; For the detected enemy unmanned aerial vehicle swarm with a scale of n, x unmanned aerial vehicles are expected to be attacked, n>x, then the reachability distance matrix calculated by the OPTICS clustering algorithm is sorted in ascending order, and the xth reachability distance value is taken as its eps value, at this time the number of clustered unmanned aerial vehicles obtained will be less than or equal to x; Wherein part of the nodes are noise points, which can be adjusted according to the number of noise points and the number of clusters, the value of x / n is between 0.4 and 0.7, that is, the proportion of multiple unmanned aerial vehicles expected to be attacked is between 40% and 70% of the enemy swarm; According to the calculated eps value, in the reachability distance graph obtained by OPTICS clustering, the clustering result of the unmanned aerial vehicle swarm is obtained according to the reachability distance ordering graph and the core distance matrix obtained by the OPTICS clustering algorithm; In step S3, the shortest distance between each unmanned aerial vehicle in the enemy swarm and the target to be protected is calculated, and the distance value is quantized to the interval [0.9, 1.1], representing the distance index of the target to be protected, if the distance index value <1, it represents that the distance between the important target and our important target is relatively close, otherwise it is relatively far; According to the clustering result of the enemy unmanned aerial vehicle swarm obtained by the calculated eps value, when the number of nodes in the clustering cluster is less than minPts, the nodes in these clustering clusters need to be processed, minPts is the minimum number of Eps neighborhood points; In step S3, after the clustering result is optimized, the clustering result of the enemy unmanned aerial vehicle swarm is further analyzed, the cluster center, member number, shortest distance from the clustering center to the target to be protected, clustering average core density and clustering maximum radius of each cluster are calculated, and then the priority of each cluster is set according to the shortest distance from the clustering center to the target to be protected, that is, the node closest to the target to be protected has the highest priority, and the priority is lower with the increase of distance, define the cluster number as p, then the priority can be allocated according to 0~p-1, and the priority weight is represented as 1, (p-1) / p, 1 / p respectively; In step S4, according to the clustering result of the enemy unmanned aerial vehicle swarm, the countermeasure task list of the unmanned aerial vehicle swarm that can be dispatched for countermeasure can be obtained, according to the combat effectiveness maximization principle and the best defense principle of important high value target, the unmanned aerial vehicle swarm that can be dispatched for countermeasure is allocated according to the type and number of unmanned aerial vehicles.

2. The method of claim 1, wherein: The three-dimensional spatial position distribution information includes longitude, latitude and altitude; wherein, the longitude and latitude data are converted into radian, and the altitude is expressed in meters; Or / and, the cluster clustering parameters include one or more of the cluster center, the number of members, the average core density and the neighborhood radius of each cluster.

3. The method of claim 2, wherein: In step S1, minPts is set as the minimum number of neighbors within the domain radius ε, the target position to be detected is three-dimensional data, and minPts takes a value greater than or equal to 4; The enemy UAV swarm is subjected to density clustering based on the density clustering algorithm OPTICS, wherein the calculation method of the distance between enemy UAVs in the swarm is as follows: The spatial positions of node one and node two are set as (lon_1, lat_1, alt_1) and (lon_12, lat_2, alt_2), wherein lon represents longitude, lat represents latitude, and alt represents altitude; The Haversine distance of the longitude and latitude of node one and node two is set as formula haversine((lon_1, lat_1), ((lon_12, lat_2)).m, the unit of measurement is meters; The distance calculation method of the altitude of node one and node two is as follows: dist = sqrt(haversine((lon_1, lat_1), (lon_2, lat_2)).m 2 + (alt_1 - alt_2) 2 ), where ".m" means the distance is measured in meters; After the enemy UAV swarm is subjected to density clustering, the clustering structure of the enemy UAV swarm density, i.e., the augmented cluster ordering, can be obtained.

4. The method of claim 3, wherein: Let the number of enemy drone swarms be n, denoted as enemy drone swarm set , and the number of our protected targets be q, denoted as our protected target number set ; then the distance between any one of the enemy drone swarms and any target in our key target set Q is , and the distance matrix is obtained as , wherein ; Let and denote the nearest distance and the farthest distance between each UAV in the enemy swarm and the target to be protected by us respectively; for any UAV , define the distance between the nearest key target to us as , then the distance between the target to be protected by us as .

5. The method of claim 4, wherein: In step S3, the node processing includes: first, obtaining the distance index of each node in the clustering cluster, all greater than 1, indicating that each UAV in the cluster is far away from the important target of our side, and then all the nodes in the clustering cluster are taken as noise points; Otherwise, the nearest minPts unclustered adjacent nodes of each node in the list are taken to join the list of nodes to be processed, and for each node in the list, if the weighted core distance of the adjacent node of the node is less than eps, i.e., the core distance multiplied by the distance index is less than eps, the adjacent node of the node is taken as an extended node of the cluster and is added to the cluster, otherwise the node is still taken as a noise node; After all the nodes to be processed are processed in turn, if the total number of nodes of the extended cluster is greater than or equal to minPts, the clustering cluster is retained, otherwise all the nodes in the clustering cluster are taken as noise points.

6. The method of claim 2, wherein: In step S4, The distribution method of the combat effectiveness value of each UAV in the enemy UAV swarm by the different types of UAVs in the dispatched counter UAV swarm includes: If the distance between the UAV and the cluster center exceeds 1 / 2 of the maximum combat distance of the UAV, or the height of the cluster center exceeds the maximum flight height of the UAV, the combat effectiveness index value e of the UAV is 0, i.e., the UAV cannot perform the task at all; otherwise, the task matching degree is calculated, the task matching degree between the UAV and the task is k, and then the combat effectiveness index value is calculated as 100-(4-k)*25; The combat effectiveness index value is divided according to the level, and when the combat effectiveness index of the same type of UAV for the same task is the same, the priority order of different types of UAVs is defined, i.e., electronic warfare UAV > attack UAV > capture UAV, i.e., the electronic countermeasure UAV is preferred under the same condition; Or / and, In step S4, the combat task allocation method includes: setting the number of clusters of enemy UAV swarms as t, the number of available UAVs as u, u being a non-zero natural number, then including two cases t < u or t ≥ u; the combat task allocation is to arrange all available UAVs with combat tasks, and each UAV can only execute one task; first, the defense principle of important targets must be met, and the cluster of the enemy UAV swarm with the highest priority and the largest threat distance is allocated with tasks, and then the unallocated tasks are allocated in turn; When t ≥ u, the process ends after all UAVs are allocated with tasks, and the far and low-level long-distance clusters are processed as the next combat target; Or, when t < u, after all tasks are allocated, the remaining u-t UAVs need to calculate their combat effectiveness values with each task, and then allocate tasks to all UAVs according to the principle of the largest combat effectiveness value.

Images