DE102018108204B4 - Intervention system for countering a swarm of unmanned aerial vehicles and method for countering a swarm of unmanned aerial vehicles using such an intervention system - Google Patents

Abstract

Intervention system (1) with a sensor unit (2), with a computing unit (3) and with an effector unit (4), wherein by means of the sensor unit (2) a swarm (5) of several unmanned missiles (6) with at least one guidance drone (7 ) can be detected, with individual unmanned missiles (6) being able to be tracked by means of the sensor unit (2), with the information recorded by the sensor unit (2) being able to be forwarded to the processing unit (3), with the processing unit (3) having a computer with a processor , having a memory and software, wherein the computing unit (3) can be used to analyze the information supplied by the sensor unit (2), characterized in that the at least one guide drone (7) can be identified by the computing unit, the at least a guide drone (7) coordinates and/or controls the swarm (5), the movement behavior and/or the composition and/or the arrangement of the missiles (6) being able to be analyzed using the computing unit (3), the computing unit (3) a combat pattern for combating the guide drone (7) can be generated and this can be forwarded to the effector unit (4), with a movement pattern of drones (6) in the swarm (5) being able to be analyzed by means of the computing unit (3), with it being analyzed which drone (6) is the first to initiate a change of direction, with this drone (6) being identified as the guide drone (7) and/or that the center of gravity of the swarm (5) can be determined by means of the computing unit (3) in order to guide the guide drone (7) in the to identify the virtual center of gravity.

Description

The invention relates to an intervention system for defending against a swarm of unmanned missiles with the features of the preamble of patent claim 1.

Furthermore, the invention relates to a method for defending against a swarm of unmanned missiles by means of such an intervention system with the features of patent claim 7.

The intervention system has a sensor unit, a computing unit and an effector unit. The swarm and/or the individual missiles of the swarm can be detected and/or tracked by means of the sensor unit. The sensor unit can be formed by a multi-sensor platform. In the following, the terms drone and unmanned aerial vehicle are used synonymously.

From the US 2017/0059333 A1 a system for defense against a swarm of smaller ships is known, the course setting and the choice of weapons being optimized by the appropriate control system. A graphical user interface is provided to navigate a ship around multiple hazards. The ship can be attacked by a swarm. In a first step, the ship's radar and an electro-optical sensor system are used to detect the position of the enemy attackers. In a second step, the data relating to the ship's weapons system is compiled. This information may include the maximum and minimum ranges of the weapons and the accuracy and maximum and minimum azimuth angles of the weapons. Based on this information, routes are now calculated using a computing unit. Based on the determined, several possible paths, the best path is determined using a Monte Carlo simulation.

Here, the number of destroyed threats and the use of the weapon system are simulated as parameters.

From the U.S. 2015/0377596 A1 a method and apparatus for optimizing a resource allocation plan is known. A sensor unit, for example in the form of a radar, is used in this case. Particle swarm optimization is used for optimization.

From the U.S. 2015/0301529 A1 a method for controlling and coordinating a swarm of drones is known. The drones form around a central point in which they maintain specified distances from each other. The drone swarm can consist of several sub-swarms.

From the US 2012/0000349 A1 discloses a system and method for associating a plurality of weapons with at least one of a plurality of threats. The algorithms used use neural networks.

From the DE 10 2015 011 579 A1 a defense system for defending against foreign drones is known, the defense system having a defense drone with a capture device. Two defense systems are also shown. The second defense system should have a laser weapon system to destroy the foreign drone. At least one defense unit and/or at least one sensor unit is selected by means of a control unit. The defense unit is selected based on the number and/or size of the foreign drones. A different defense system is deployed when a swarm of foreign drones is detected in the surveillance room instead of just a single foreign drone. The control unit is set up to use data from the sensor unit to classify detected third-party drones. In particular, a degree of criticality in the combating of the individual third-party drones can be determined on the basis of this classification. A size, a speed, a flight altitude and/or a current course of the third-party drone are advantageously included in this classification. The identification of a lead drone is not disclosed. Only characteristic sizes of the individual third-party drones are recognized, but the swarm is not analyzed in its entirety.

From the U.S. 2014/0 251 123 A1 an air burst weapon system for repelling UAV's is known. The weapon system is linked to a number of electro-optical sensors via a computer system. If a swarm is detected by the sensors, individual drones are prioritized. A combat pattern is thus created. Characteristics of the drones, such as their size and shape, can be analyzed. This reference does not show the identification of a lead drone. Furthermore, it is not shown that a change of direction initiated first by one of the drones and/or a center of gravity of the swarm is determined.

From the DE 10 2009 032 293 A1 a jet turret device, in particular a water turret device, is known. This should serve in particular to defend merchant ships or passenger ships against pirate boats. Now, if a ship is attacked by several pirate boats, the jet launcher device should carry out an autonomous multiple attack can. The pirate boats or the objects are prioritized according to a threat risk.

From the U.S. 2010/0177936 A1 a video surveillance system for monitoring, for example, public places, train stations, streets, buildings, etc. is known. Using the video surveillance system, movement patterns of monitored persons can be identified.

The prior art has the disadvantage that no specific method for effectively combating a swarm of drones is known.

The object of the invention is therefore to improve the combating of a swarm of drones.

This object on which the invention is based is now achieved by an intervention system with the features of patent claim 1 and by a corresponding method with the features of patent claim 6 .

A guide drone is identified by the processing unit, with the guide drone coordinating and/or controlling the swarm, with the processing unit being able to analyze the behavior and/or the composition and/or the extent of the swarm, with the processing unit generating a combat pattern and this being forwarded to the effector unit. The computing unit can be used to assign probabilities to the drones that the respective drone is the lead drone.

A movement pattern of drones in the swarm can be analyzed by means of the computing unit, with it being analyzed which drone initiates a change of direction first, with this drone being identified as the lead drone and/or the center of gravity of the swarm can be determined by means of the computing unit in order to identify the lead drone in the virtual center of gravity.

The information available about the swarm of drones is analyzed by the computing unit, with at least one guide drone being identified which coordinates and/or controls the swarm of drones. The computing unit identifies the drones that are essential for the coordination and control of the swarm. These drones can be referred to as guide drones. Based on this, a combat pattern is generated by the computing unit and this is forwarded to the effector unit. The combat pattern serves to primarily combat the lead drones. The effector unit then uses the specified combat pattern to specifically attack the lead drone in order to disrupt and prevent the coordination of the drone swarm. If the coordination of the swarm is disturbed, it can no longer carry out its task. Optionally, the other drones can also be fought afterwards.

The movement behavior and/or the composition and/or the extent of the swarm and/or its effect on the object to be protected is analyzed and classified by means of the computing unit. This is made possible by pattern recognition.

In a preferred embodiment, the computing unit uses algorithms from the field of artificial intelligence for pattern recognition, in particular for identifying, classifying and/or analyzing the swarm of drones. In a particularly preferred embodiment, pattern recognition and preferably movement behavior analysis can be carried out to identify and classify the drones.

The computing unit is advantageously connected to a drone database or has a drone database in which properties or patterns of known drones and/or swarms of drones are stored. In particular, designs of drones or typical optical or radar-typical signatures of known drone types are stored in the database.

The sensor unit preferably has a radar, a LIDAR and/or an electro-optical sensor and an infrared sensor. It is conceivable that the sensor unit has a combination of these sensors.

In a particularly preferred embodiment, the computing unit is pre-trained using the machine learning method. Artificial neural networks and in particular the method of deep learning can be used for this. Deep learning refers to a class of optimization methods for artificial neural networks that have numerous intermediate layers between an input layer and an output layer and thus have an extensive inner structure. This improves learning success. The learning is carried out before the computing unit is used and not during use. The combat patterns remain clear and predeterminable after the learning process depending on the inputs. The combat patterns do not develop dynamically. It is a deterministic system.

The drone swarm can consist of at least one lead drone and several subordinate drones or subswarms with several lead drones and subordinate drones. The guide drones can be recognized, for example, by their movement behavior within the swarm. In particular, the lead drones can be the first to change direction, with the rest of the swarm following with a minimal time delay. The computing unit preferably examines the movement pattern of the drones in the drone swarm. In particular, the computing unit can be used to identify which drone is the first to initiate a change of direction. This drone is identified as a lead drone.

The lead drones can also be found in a virtual focal point within the swarm or at a location with a particularly high drone density within the swarm. The center of gravity of the swarm is determined by means of the computing unit. In a preferred embodiment, a drone density is determined by means of the computing unit.

It is conceivable that the lead drones will fly ahead and then be found at a tip of the swarm. A peak of the swarm is therefore preferably identified with the arithmetic unit.

Furthermore, it is conceivable that the guide drones differ visually from other drones, e.g. by a different size or design. In a preferred embodiment, the size of the drones is analyzed with the computing unit and the size of the drones is compared with the computing unit. The designs of the drones are preferably analyzed by means of the computing unit.

In the case of swarms of drones that reorganize themselves independently if a lead drone fails, for example, the system can analyze the reorganization and specifically combat the new lead drones. Preferably, after a guide drone has been recognized by the computing unit, the other drones are assigned different probabilities that, after the guide drone has been switched off, they will take over the function of a new guide drone. These drones can be detected with increased priority after eliminating the old lead drone.

The effector unit is formed by a weapon system which can preferably attack a target precisely and with pinpoint accuracy. The weapon system can have automatic cannons and/or laser weapons, for example.

As an alternative or in addition to recognizing the guide drone, the intervention system can use the computing unit to analyze the movement behavior and/or the composition and/or the arrangement of the missiles, with a combat pattern being able to be generated using the computing unit and this being forwarded to the effector unit, the combat pattern comprising a subset of the unmanned missiles which are necessary for the coordination and/or control of the swarm due to the number and/or arrangement of the missiles in the swarm. Even if no lead drone is detected, the coordination and/or control of the swarm can be disrupted. The analysis of the processing unit is pre-trained as explained above.

The combat pattern preferably comprises a subset of less than or equal to 50%, in particular 10-35% of the drones, which are spatially distributed as evenly as possible within the swarm. For example, every second, third, fourth or fifth drone can be selected in the combat pattern.

In a further embodiment, the combat pattern comprises a subset of less than or equal to 50%, in particular 10-35% of the drones, with these being arranged in areas of the highest swarm density.

• 1 ein Interventionssystem zur Abwehr eines Schwarms von unbemannten Flugkörpern.

There are now a large number of possibilities for designing and developing the method according to the invention and the intervention system in more detail. For this purpose, reference may first be made to the patent claims subordinate to patent claims 1 and 6 . A preferred embodiment of the invention is explained in more detail below with reference to the drawing and the associated description. In the drawing shows:

• 1 an intervention system to counter a swarm of unmanned missiles.

In 1 an intervention system 1 is clearly visible. The intervention system 1 has a sensor unit 2 , a computing unit 3 and an effector unit 4 . The sensor unit 2 preferably has a radar. The sensor unit 2 also preferably has LIDAR and/or an electro-optical sensor and/or an infrared sensor. The sensor unit 2 can be formed by a multiplatform sensor unit. The intervention system is designed in particular as an air defense system.

The processing unit 3 has a computer with a memory and a processor. Software is stored in the memory and is executed by the processor. The computing unit 3 receives signals or data derived from the signals of the sensor unit 2 .

In particular, a swarm 5 with a plurality of unmanned missiles 6 can be detected by means of the sensor unit 2 . These unmanned missiles 6 can also be referred to as 6 drones. The intervention system 1 is now used to defend against the swarm 5. The information recorded about the swarm 5, i.e. the sensor data and/or the information derived therefrom, is analyzed by means of the computing unit 3.

In particular, at least one guide drone 7 is identified by means of the computing unit 3 . A pattern analysis is carried out by means of the arithmetic unit 3 . The processing unit 3 analyzes and classifies the behavior and/or the composition and/or the extent of the swarm 5 and/or assesses its effect on an object to be protected that is not shown in detail. The guidance drone 7 is essential for the coordination and control of the swarm 5 .

In particular, a movement pattern analysis is carried out by means of the computing unit 3 . The distances between the drones are monitored.

Furthermore, a drone type analysis is carried out with the computing unit 3 . A behavioral analysis is also carried out.

A control pattern is generated based on the analysis results. The combat pattern generated in this way is forwarded to the effector unit 4 . The effector unit 4 specifically combats the lead drone 7 by means of the predetermined combat pattern in order to disrupt and prevent the coordination of the swarm 5 . If the coordination of the swarm 5 is disturbed, it can no longer carry out its task. In a further subsequent step, the other drones 6 can be combated.

The software running on the processing unit 3 uses in particular algorithms from the field of artificial intelligence for pattern recognition and thus for identifying the guide drone 7. The processing unit 3 uses algorithms from the field of artificial intelligence to carry out the corresponding analysis and thus to recognize the guide drone 7. The arrangement of the drones 6 in particular is analyzed in the pattern recognition. The distances between the drones 6 are recorded. The change in the arrangement over time is analyzed. The guidance drone 7 can be recognized, for example, by the movement behavior over time within the swarm 5 . The lead drone 7 can be the first to change direction, with the rest of the swarm 5 following this change of direction with a time delay. The processing unit 3 thus examines the movement pattern of the drones 6 in the swarm 5. The movement pattern recognition makes it possible to identify which drone 6, namely the guide drone 7, initiates a change of direction first. This drone 6 is identified as a lead drone 7 .

Further patterns for recognizing the guide drone 7 can be that the guide drone 7 is located at a location within the swarm 5 with a particularly high drone density. The drone density within the swarm 5 is thus preferably determined by means of the computing unit 3 . It is also conceivable that the guide drone 7 flies ahead and can therefore be found at a tip of the swarm 5 . In this case, the tip of the swarm is preferably identified with the arithmetic unit 3 as a pattern.

Another possible recognition pattern is that the design of the drones 6 is analyzed by the computing unit 3 . For this purpose, the computing unit 3 can be connected to a database 8, with the database 8 storing designs of drones 6 or typical optical or radar-typical signatures of known drones 6.

In an alternative embodiment, the computing unit 3 can include the database 8 . In a particularly preferred embodiment, the computing unit 3 is pre-trained using the machine learning method. Artificial neural networks can be used for this. In particular, the computing unit 3 has been pre-trained using the deep learning method. The learning is carried out before the computing unit 3 is used and in particular not during use. This has the advantage that it ensures that the system is deterministic, i.e. that with the same inputs, i.e. with the same signals from the sensor unit 2, the same control patterns are also passed on to the effector unit 4.

It is possible for a swarm 5 to consist of several sub-swarms, each of the sub-swarms having a guide drone 7 . During the analysis, the computing unit 3 will also recognize as a pattern whether it is a swarm with a guide drone 7 or several sub-swarms with several guide drones 7 .

Reference List

1

intervention system

2

sensor unit

3

unit of account

4

effector unit

5

swarm

6

unmanned aerial vehicle / drone

7

guide drone

8th

Database

Claims (8)

Intervention system (1) with a sensor unit (2), with a computing unit (3) and with an effector unit (4), with a swarm (5) of several unmanned missiles (6) with at least one guidance drone (7) being detectable by means of the sensor unit (2), with individual unmanned missiles (6) being able to be tracked by means of the sensor unit (2), with the information recorded by the sensor unit (2) being able to be forwarded to the computing unit (3), with the computing unit (3) having a Computer with a processor, with a memory and software, whereby the information supplied by the sensor unit (2) can be analyzed by means of the computing unit (3),characterizedthat the at least one guide drone (7) can be identified by the computing unit, with the at least one guiding drone (7) coordinating and/or controlling the swarm (5), with the computing unit (3) being able to analyze the movement behavior and/or the composition and/or the arrangement of the missiles (6), with the computing unit (3) being able to generate a combat pattern for combating the guiding drone (7) and this being forwardable to the effector unit (4), with the computing unit (3) a movement pattern of drones (6) in the swarm (5) can be analyzed, with it being analyzed which drone (6) initiates a change of direction first, with this drone (6) being identified as the guide drone (7) and/or that the center of gravity of the swarm (5) can be determined by means of the computing unit (3) in order to identify the guide drone (7) in the virtual center of gravity. intervention system claim 1 , characterized in that the size and/or the design of the drones (6) can be analyzed by means of the computing unit (3). Intervention system according to one of the preceding claims, characterized in that after a guide drone (7) has been identified, the computing unit (3) can allocate different probabilities to the other drones (6) that these take over the function of a new guide drone (7) after the guide drone (7) has been switched off. Intervention system according to one of the preceding claims, characterized in that the effector unit (4) has at least one automatic cannon and/or at least one laser weapon. Intervention system (1) according to one of the preceding claims, characterized in that the movement behavior and/or the composition and/or the arrangement of the missiles (6) can be analyzed by means of the computing unit (3), wherein a combat pattern can be generated by means of the computing unit (3) and this can be forwarded to the effector unit (4), the combat pattern comprising a subset of the unmanned missiles (6) which is determined by the number and/or arrangement of the missiles (6) in the swarm for coordination and/or control of the swarm are necessary. Method for defending against a swarm of unmanned missiles (6) by means of an intervention system (1) according to one of the preceding claims, wherein the intervention system (1) has a sensor unit (2), a computing unit (3) and an effector unit (4), with the sensor unit (2) being used to detect a swarm (5) of a plurality of unmanned missiles (6) with at least one guide drone (7), with individual unmanned missiles (6) being tracked by the sensor unit (2), with the sensors being tracked by the sensor unit (2) the information recorded is forwarded to a computing unit (3), the computing unit (3) having a computer with a processor, a memory and software, with the computing unit (3) carrying out an analysis of the information supplied by the sensor unit (2),characterizedthat the at least one guide drone (7) is identified by the computing unit, the at least one guiding drone (7) coordinating and/or controlling the swarm (5), the movement behavior and/or the composition and/or the arrangement of the missiles (6) being analyzed using the computing unit (3), a combat pattern for combating the at least one guiding drone (7) being generated using the computing unit (3) and this being forwarded to the effector unit (4), the computing unit (3) generating a movement pattern of the drone n (6) is analyzed in the swarm (5), whereby it is analyzed which drone (6) initiates a change of direction first, with this drone (6) being identified as the guide drone (7) and/or the center of gravity of the swarm (5) being determined by means of the computing unit (3) in order to identify the guide drone (7) in the virtual center of gravity. procedure after claim 6 , characterized in that the size and/or the design of the drones (6) are analyzed by the computing unit (3). procedure after claim 6 or 7 , characterized in that by means of the computing unit (3) after recognizing a guide drone (7), the other drones (6) are assigned different probabilities that these other drones (6) take over the function of a new guide drone (7) after the guide drone (7) has been switched off.

Images