Wolf-pack hunting method for cooperative control of UAV swarm

Abstract

The application discloses a kind of unmanned aerial vehicle cluster cooperative control methods of imitating wolf pack encirclement and capture, which comprises the following steps: each unmanned aerial vehicle in cluster independently obtains the motion state information of target;Unmanned aerial vehicle in cluster obtains unified desired attack time;Each unmanned aerial vehicle in cluster determines whether the desired attack time obtained is reasonable, if the desired attack time of any unmanned aerial vehicle is unreasonable, then new desired attack time is reacquired, until all unmanned aerial vehicles determine that the desired attack time obtained is reasonable;Based on desired attack time, each unmanned aerial vehicle in cluster obtains respective control instruction, controls the flight state of itself, and realizes hitting target at desired time.The unmanned aerial vehicle cluster cooperative control method of imitating wolf pack encirclement and capture disclosed by the application can enable unmanned aerial vehicles in cluster to track and attack target quickly and stably under the premise of meeting actual flight speed limit of itself.

Description

Technical Field

[0001] This invention relates to a method for collaborative control of a drone swarm mimicking wolf pack encirclement, belonging to the field of guidance technology. Background Technology

[0002] Drones use numerical superiority, or saturation attacks, to encircle and capture targets. Because this method borrows from the hunting behavior of wolf packs, it is also known as wolf pack-like drone swarm collaborative control.

[0003] The key to saturation attacks is constraining the attack time, but traditional guidance laws can only satisfy the positional constraint of hitting the target, not the time constraint. Currently, there are two control methods for constraining time: closed-loop time-coordinated control and open-loop time-coordinated control.

[0004] Closed-loop time-coordinated guidance and control is a method based on the communication network between UAV swarms. Each UAV transmits information such as its relative distance to the target and its own speed to its neighbors, allowing them to simultaneously strike targets by adjusting their speed or direction. However, this method relies entirely on the self-organizing network of inter-UAV communication. In certain environments, if the communication network is interfered with or a link is broken, the entire UAV swarm will collapse. Therefore, this method has low fault tolerance, strong dependence on communication, and is difficult to implement in engineering.

[0005] Open-loop time-coordinated control is a mission-oriented approach that requires each UAV to independently complete its task without relying on inter-aircraft communication, while still ensuring simultaneous target engagement. This method requires pre-setting the same expected attack time for each UAV, or different expected attack times to allow for staggered strikes in different batches. Traditional open-loop time-coordinated control was developed for missiles with fixed speeds. However, with the advent of variable-speed aircraft such as UAVs and loitering munitions, traditional open-loop time-coordinated control fails to leverage the variable-speed characteristics and results in energy waste.

[0006] Therefore, it is necessary to conduct in-depth research on the collaborative control method of drone swarms that mimics wolf pack encirclement in order to solve the above problems. Summary of the Invention

[0007] To overcome the above problems, the inventors conducted in-depth research and provided a method for collaborative control of a drone swarm mimicking wolf pack encirclement, which includes the following steps:

[0008] S1. Each drone in the cluster independently acquires the target's motion status information;

[0009] S2, the drones in the cluster obtain a unified expected attack time;

[0010] S3. Each drone in the cluster determines whether the expected attack time it has acquired is reasonable. If the expected attack time of any drone is unreasonable, a new expected attack time is acquired again until all drones determine that the expected attack time they have acquired is reasonable.

[0011] S4. Based on the expected attack time, each drone in the cluster receives its own control commands to control its own flight status and achieve the goal of hitting the target at the expected time.

[0012] In a preferred embodiment, in S1, the motion status information of the target is obtained in real time through an optoelectronic pod carried on the UAV.

[0013] In a preferred embodiment, in step S1, the motion state information of the target includes:

[0014] The line-of-sight angle of the target relative to the pitch direction of the mission UAV, the line-of-sight angle of the target relative to the yaw direction of the mission UAV, and the distance of the target relative to the mission UAV.

[0015] In a preferred embodiment, it is determined whether the selection of the desired attack time based on target motion information, the upper speed limit of the UAV itself, and the relative distance between the target and the UAV is reasonable.

[0016] In a preferred embodiment, in S3, "reasonable" means that the expected attack time obtained satisfies the following conditions:

[0017]

[0018] Among them, V c,0 t represents the target's initial velocity relative to the drone. s Indicates the tangential control command switching time, t d To obtain the expected attack time, V c,max This represents the maximum value of the drone's own speed constraint, i.e., the speed upper limit, where k is the control convergence coefficient.

[0019] The tangential control command switching time t s Obtained through the following formula:

[0020]

[0021]

[0022] Where x(0) represents the initial state of the UAV, and r0 is the relative distance between the UAV and the target at the initial moment.

[0023] In a preferred embodiment, in S4, the control command includes a normal acceleration command and a radial acceleration command, the radial acceleration command being set based on the desired attack time.

[0024] In a preferred embodiment, an angle constraint is set in the normal acceleration command.

[0025] In a preferred embodiment, the normal acceleration command includes a Y-axis acceleration command in the line-of-sight frame. Acceleration command along the Z-axis in the line-of-sight frame Represented as:

[0026]

[0027] Among them, V c This represents the projection of the target's velocity relative to the drone along the line of sight; Ω Los q represents the target's angular rotation rate relative to the UAV's line of sight; y q is the angle of view pitch. d For the desired line-of-sight pitch angle, a tn This represents the normal acceleration of the target in the line-of-sight frame.

[0028] In a preferred embodiment, the radial acceleration command a cx Represented as:

[0029]

[0030] Among them, a tx The acceleration component of the target along the line of sight in the line-of-sight frame is represented by t; k represents the convergence control coefficient. d The expected attack time is represented by t, and the flight time of the drone is represented by t. s This indicates the switching time of the tangential control command.

[0031] In a preferred embodiment, the control command a is represented as:

[0032]

[0033] in, The coordinate rotation matrix from the line-of-sight frame to the inertial frame is represented as:

[0034]

[0035] The beneficial effects of this invention include:

[0036] (1) It can estimate the target's motion and navigate autonomously accordingly, enabling rapid and stable tracking and attack of the target in the shortest possible time;

[0037] (2) Based on the speed variability of UAVs, an attack time constraint control method considering its own speed constraint was designed in its tangential channel;

[0038] (3) Achieve fully autonomous operation and, under the premise of meeting its own actual flight speed limit, achieve all-round and multi-angle encirclement and attack on any moving target. Attached Figure Description

[0039] Figure 1 A schematic flowchart of a drone swarm cooperative control method for simulated wolf pack hunting according to a preferred embodiment of the present invention is shown.

[0040] Figure 2 The motion trajectories of different drones and targets in Example 1 are shown;

[0041] Figure 3 The relative distance between different UAVs and targets in Example 1 is shown as a function of time.

[0042] Figure 4 The speed variation curves of different UAVs in Example 1 are shown;

[0043] Figure 5 The curves showing the line-of-sight angle variation and the terminal line-of-sight angle values ​​for different UAVs in Example 1 are shown. Detailed Implementation

[0044] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Through these descriptions, the features and advantages of the present invention will become clearer and more apparent.

[0045] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0046] According to the present invention, a method for collaborative control of a drone swarm mimicking wolf pack encirclement is provided, such as... Figure 1 As shown, the method includes the following steps:

[0047] S1. Each drone in the cluster independently acquires the target's motion status information;

[0048] S2, the drones in the cluster obtain a unified expected attack time;

[0049] S3. Each drone in the cluster determines whether the expected attack time it has acquired is reasonable. If the expected attack time of any drone is unreasonable, a new expected attack time is acquired again until all drones determine that the expected attack time they have acquired is reasonable.

[0050] S4. Based on the expected attack time, each drone in the cluster receives its own control commands to control its own flight status and achieve the goal of hitting the target at the expected time.

[0051] According to the present invention, the drones in the drone swarm have different initial positions.

[0052] Preferably, the drones in the drone swarm have different maximum speed constraints, that is, the drones in the drone swarm are of different models, in order to achieve better coordination.

[0053] Preferably, the UAV is a rotary-wing UAV, specifically a UAV with four or more rotors. The real-time acceleration command is input to the UAV's rotor control system. The rotor control system can adjust the rotor speed according to the input acceleration command. The change in speed causes a change in rotor lift, thereby controlling the attitude and position of the mission UAV.

[0054] It should be noted that the velocity, position, and acceleration information of the target in this invention refers to velocity and position information in an inertial frame. This inertial frame is defined with the mission UAV's center of mass as the origin, geographical east as the x-axis, geographical north as the y-axis, and the z-axis perpendicular to the x and y axes, with upward being positive, conforming to the right-hand rule. The line-of-sight frame is defined with the mission UAV's center of mass as the origin, and the direction of the line connecting the mission UAV and the target as the x-axis. L Axis, Z L The axis is perpendicular to X in the vertical plane. L The axis, pointing upwards is positive, Y L The axis is determined by the right-hand rule.

[0055] In a preferred embodiment, in S1, the motion status information of the target is obtained in real time through an optoelectronic pod carried on the UAV.

[0056] The optoelectronic pod can be any existing optoelectronic pod in the field, capable of performing target tracking and outputting the line-of-sight angle and line-of-sight angular rate of the target relative to the multirotor aircraft.

[0057] In this invention, the specific method for obtaining the target motion state information is not limited. Those skilled in the art can choose any known method based on experience, such as obtaining it through the existing Kalman filter algorithm.

[0058] Machine vision technology using airborne optoelectronic pods to detect and track space targets offers advantages over radar and vehicle-mounted optoelectronic detection equipment, including higher accuracy, greater intuitiveness, lower cost, and immunity to clutter interference.

[0059] In a preferred embodiment, in step S1, the motion state information of the target includes:

[0060] The line-of-sight angle q of the target relative to the pitch direction of the mission UAV y The line-of-sight angle q of the target relative to the yaw direction of the mission UAV z And the distance R between the target and the mission drone.

[0061] In a preferred embodiment, prior to S1, the target is photographed by an electro-optical pod carried on the mission UAV to obtain continuous images containing the target; by estimating the state of the target in the continuous images, the angular rate control command for the rotation of the electro-optical pod is calculated, thereby controlling the electro-optical pod to continuously observe the target.

[0062] The specific calculation method is not limited in this invention. Any airborne machine vision algorithm can be used to process the continuous images to give the state estimate of the target, which can then serve as the control information source for the electro-optical pod on the multirotor aircraft.

[0063] In S2, the expected attack time can be preset and stored in each drone. For example, multiple expected attack times can be stored in the drone's onboard computer in the form of a sequence. Alternatively, it can be set temporarily and then transmitted to each drone via wireless communication.

[0064] According to a preferred embodiment of the present invention, it is reasonable to determine the selection of the desired attack time based on target motion information, the upper limit of the UAV's own speed, and the relative distance between the target and the UAV.

[0065] In a preferred embodiment, in S3, "reasonable" means that the expected attack time obtained satisfies the following conditions:

[0066]

[0067] Among them, V c,0 t represents the target's initial velocity relative to the drone. s Indicates the tangential control command switching time, t d To obtain the expected attack time, V c,max This represents the maximum speed constraint of the UAV, i.e., the speed limit. k is the control convergence coefficient, which is a constant and preferably set to 1.

[0068] In this invention, by introducing a tangential control command switching time, the guidance process of each UAV takes into account its own speed constraints and attack time constraints, thus better leveraging the speed variability of the UAV.

[0069] In a preferred embodiment, the tangential control command switching time t s Obtained through the following formula:

[0070]

[0071]

[0072] Where x(0) represents the initial state of the UAV, and r0 is the relative distance between the UAV and the target at the initial moment.

[0073] In a preferred embodiment, in S4, the control command includes a normal acceleration command and a radial acceleration command, the radial acceleration command being set based on the desired attack time.

[0074] In a preferred embodiment, an angle constraint is set in the normal acceleration command to achieve the lowest possible energy consumption.

[0075] In a preferred embodiment, the normal acceleration command includes a Y-axis acceleration command in the line-of-sight frame. Acceleration command along the Z-axis in the line-of-sight frame Represented as:

[0076]

[0077] Among them, V c This represents the projection of the target's velocity relative to the drone along the line of sight; Ω Los q represents the target's angular rotation rate relative to the UAV's line of sight; y q is the angle of view pitch. d For the desired line-of-sight pitch angle, a tn This represents the normal acceleration of the target in the line-of-sight frame.

[0078] According to a preferred embodiment of the present invention, each UAV in the cluster is set with a different desired line-of-sight pitch angle, so that the UAVs in the cluster can fly toward the target from different angles and different paths.

[0079] In a preferred embodiment, the radial acceleration command a cx Represented as:

[0080]

[0081] Among them, a tx The acceleration component of the target along the line of sight in the line-of-sight frame is represented by t; k represents the convergence control coefficient. d The expected attack time is represented by t, and the flight time of the drone is represented by t. s This indicates the switching time of the tangential control command.

[0082] In a preferred embodiment, the control command a is represented as:

[0083]

[0084] in, The coordinate rotation matrix from the line-of-sight frame to the inertial frame is represented as:

[0085]

[0086] According to this invention, when the distance between the drone and the target is less than a preset value, i.e., the minimum miss distance, at the expected moment of impact, it indicates that the encirclement and capture is successful.

[0087] Furthermore, according to the present invention, the angle of the UAV at the final moment is equal to the desired attack angle, that is, it satisfies the mission requirement of striking at the desired angle at the desired moment.

[0088] Example

[0089] Example 1

[0090] A simulation experiment was conducted, in which two drones were used to coordinate and control the target for capture. The initial positions and initial velocities of the drones are shown in Table 1.

[0091] Table 1

[0092]

[0093] The speed limit for all drones is set to 20m / s.

[0094] The initial position of the target is (x t0 ,y t0 ,z t0 ) = (200, 300, 10) m, initial velocity is (V xt0 V yt0 ) = (3,0,1) m / s, performing three-dimensional maneuvering motion in the air, with the acceleration in the inertial frame set as:

[0095] The collaborative control method includes the following steps:

[0096] S1. Each drone in the cluster independently acquires the target's motion status information;

[0097] S2, the drones in the cluster obtain a unified expected attack time;

[0098] S3. Each drone in the cluster determines whether the expected attack time it has acquired is reasonable. If the expected attack time of any drone is unreasonable, a new expected attack time is acquired again until all drones determine that the expected attack time they have acquired is reasonable.

[0099] S4. Based on the expected attack time, each drone in the cluster receives its own control commands to control its own flight status and achieve the goal of hitting the target at the expected time.

[0100] In S2, the expected attack time t dSet to 20 seconds.

[0101] In S3, determine whether the expected attack time obtained in S2 is reasonable according to the following formula:

[0102]

[0103]

[0104]

[0105] Where k is set to 1, the expected attack time can be determined by the above formula.

[0106] In S4, the control commands are represented as follows:

[0107]

[0108]

[0109]

[0110]

[0111] Figure 2-5 The simulation process and results are shown. Figure 2 The movement trajectories of different drones and targets are shown. Figure 3 The relative distance between different drones and targets changes over time. Figure 4 The speed variation curves of different drones are shown. Figure 5 The curves showing the line-of-sight angle variation and the terminal line-of-sight angle values ​​for different UAVs are shown.

[0112] from Figure 2 , 3 As can be seen, for two unmanned aerial vehicles with different initial positions and speeds, both can make their relative distance from the target approach zero at the same moment, that is, hit the maneuvering target at the expected attack time.

[0113] from Figure 4 It can be seen that this method can predict the speed extreme value and determine whether the selection of the expected approach time is reasonable; it can meet its own speed limit and at the same time meet the mission target of simultaneous attack.

[0114] from Figure 5 It can be seen that this method can achieve all-round and multi-angle capture of the target by satisfying the time constraint, speed constraint and end angle constraint.

[0115] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front," and "rear," etc., indicate the orientation or positional relationship based on the orientation or positional relationship in the working state of this invention, and are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. Furthermore, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0116] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0117] The present invention has been described above with reference to preferred embodiments; however, these embodiments are merely exemplary and illustrative. Various substitutions and modifications can be made to the present invention based on these embodiments, all of which fall within the scope of protection of the present invention.

Claims

1. A method for collaborative control of a drone swarm mimicking wolf pack hunting, characterized in that, The method includes the following steps: S1. Each drone in the cluster independently acquires the target's motion status information; S2, the drones in the cluster obtain a unified expected attack time; S3. Each drone in the cluster determines whether the expected attack time it has acquired is reasonable. If the expected attack time of any drone is unreasonable, a new expected attack time is acquired again until all drones determine that the expected attack time they have acquired is reasonable. S4. Based on the expected attack time, each drone in the cluster receives its own control commands to control its own flight status and achieve the goal of hitting the target at the expected time. In S3, "reasonable" means that the expected attack time obtained meets the following conditions: in, This represents the target's initial velocity relative to the drone. Indicates the switching time of the tangential control command. To obtain the expected attack time, This represents the maximum speed limit of the drone itself, i.e., the speed ceiling. To control the convergence coefficient, The tangential control command switching time Obtained through the following formula: in, This represents the initial state variables of the drone. This represents the relative distance between the drone and the target at the initial moment; In S4, the control commands include normal acceleration commands and radial acceleration commands, wherein the radial acceleration commands are set based on the desired attack time; The normal acceleration command includes the Y-axis acceleration command in the line-of-sight frame. Acceleration command along the Z-axis in the line-of-sight frame , is represented as: in, This represents the projection of the target's velocity relative to the drone along the line of sight. This indicates the angular rate of rotation of the target relative to the line of sight of the drone; The angle of view (tilt / slope). To the desired angle of view, This represents the normal acceleration of the target in the line-of-sight frame. The radial acceleration command Represented as: in, This represents the acceleration component of the target along the line of sight in the line-of-sight frame. Indicates the control convergence coefficient; Indicates the expected attack time. This indicates the flight time of the drone. This indicates the switching time of the tangential control command.

2. The drone swarm cooperative control method for wolf pack encirclement as described in claim 1, characterized in that, In S1, the target's motion status information is obtained in real time through the electro-optical pod carried on the UAV.

3. The drone swarm cooperative control method for wolf pack encirclement as described in claim 1, characterized in that, In step S1, the motion state information of the target includes: The line-of-sight angle of the target relative to the pitch direction of the mission UAV, the line-of-sight angle of the target relative to the yaw direction of the mission UAV, and the distance of the target relative to the mission UAV.

4. The drone swarm cooperative control method for wolf pack encirclement as described in claim 1, characterized in that, Is it reasonable to determine the desired attack time based on the target's motion status information, the drone's own speed limit, and the relative distance between the target and the drone? 5. The drone swarm cooperative control method for wolf pack encirclement as described in claim 1, characterized in that, Set angle constraints in the normal acceleration command.

6. The drone swarm cooperative control method for wolf pack encirclement as described in claim 1, characterized in that, The control command Represented as: in, The coordinate rotation matrix from the line-of-sight frame to the inertial frame is represented as: 。

Images