CN112925345A - Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors - Google Patents

Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors Download PDF

Info

Publication number
CN112925345A
CN112925345A CN202110103753.5A CN202110103753A CN112925345A CN 112925345 A CN112925345 A CN 112925345A CN 202110103753 A CN202110103753 A CN 202110103753A CN 112925345 A CN112925345 A CN 112925345A
Authority
CN
China
Prior art keywords
unmanned aerial
wolf
aerial vehicle
representing
target
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110103753.5A
Other languages
Chinese (zh)
Other versions
CN112925345B (en
Inventor
段海滨
张岱峰
魏晨
邓亦敏
吴江
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beihang University
Original Assignee
Beihang University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beihang University filed Critical Beihang University
Priority to CN202110103753.5A priority Critical patent/CN112925345B/en
Publication of CN112925345A publication Critical patent/CN112925345A/en
Application granted granted Critical
Publication of CN112925345B publication Critical patent/CN112925345B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
    • G05D1/10—Simultaneous control of position or course in three dimensions
    • G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
    • G05D1/104—Simultaneous control of position or course in three dimensions specially adapted for aircraft involving a plurality of aircrafts, e.g. formation flying

Landscapes

  • Engineering & Computer Science (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention discloses a cluster enclosure control method of an unmanned combat aircraft imitating wolf flock hunting behaviors, which is characterized by comprising the following steps of: the method comprises the following steps: the method comprises the following steps: initializing an unmanned aerial vehicle motion model; step two: determining a wolf group hunting role observation level; step three: calculating a wolf group leader-following target state observer; step four: designing a head wolf-slave wolf interaction mechanism surrounding control law; step five: equivalent control instruction conversion; step six: and outputting the unmanned aerial vehicle cluster enclosure control track. The method provided by the invention uses the leader-following interaction relationship in the wolf pack opportunistic hunting strategy for reference, and maintains the cluster enclosure control connectivity and stability under the sensing limitation by enhancing the interaction of heterogeneous nodes under the condition of limited detection and communication range.

Description

Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors
Technical Field
The invention discloses an unmanned combat aircraft cluster enclosure control method imitating wolf flock hunting behaviors, which is used for unmanned combat aircraft cluster enclosure control under the condition of sensing limitation and belongs to the field of unmanned aerial vehicle cluster intelligent control and decision making.
Background
The unmanned combat aircraft cluster is a cluster system consisting of a group of unmanned combat aircraft which are gathered in a high-scale mode and mutually communicated in a complementary mode, can bear the task level which cannot be related to a single combat aircraft in the traditional sense, and has lower manufacturing cost and higher reliability while improving the success rate of a single task. Formation and enclosure is a typical form of cooperative operation of unmanned aircraft clusters, where enemy objects are enclosed in a circular area to interfere, direct or block the objects, thereby reducing their threat potential. And the enclosure control is used for ensuring that the unmanned combat aircraft enters the corresponding convex hull and is matched with other unmanned aerial vehicles to complete the target enclosure task.
The enclosure control problem has been widely studied in the field of consistency control of unmanned cluster systems, and the traditional method designs a corresponding control law based on a fixed or connected topological structure so that the cluster systems form stable enclosure formation. For a consistency enclosure control method based on fixed topology, a unidirectional or bidirectional communication topology is often adopted to combine with a virtual leader to realize cluster enclosure control, and an enclosure target is taken as the virtual leader. However, due to the influence of sensing range and performance, it is difficult for the unmanned combat aircraft cluster to ensure that each unmanned aerial vehicle can effectively detect the target and always maintain a fixed spatial topology. Especially, in a relatively-rejected countermeasure environment, signal interference existing at any moment easily causes suppression or even failure of perception capability of some unmanned aerial vehicles, and the cluster communication topology cannot guarantee lasting stability and connectivity. The enclosure control law based on the connected topological structure maintains the connectivity of the communication topology of the cluster system by adding connected keeping constraint in a potential field or a consistency control protocol. However, forcing topological connectivity into cluster control by connectivity maintenance constraints can lead to difficulties in forming a closed formation; moreover, the requirement on the calculation amount of the connectivity estimation is high, and high requirements are also put forward on the calculation capacity of the unmanned aerial vehicle.
According to related biological studies, carnivores such as wolves and wild dogs can adopt an opportunistic predation strategy in the hunting process to capture prey through long-term, continuous and cooperative pursuit. The synergy in the opportunistic predation strategy is embodied in the harbourne-differentiation from the wolf role of the predator during hunting, i.e. the strong individual conducts public pursuit on the prey as a first opinion, which entails major energy loss during hunting; the wolf uses the straight line approach to wrap the prey so as to obtain better chance to catch the prey. Therefore, the wolf colony role differentiation cooperation mechanism provides an effective technical approach for solving the problem of cluster enclosure control of unmanned fighter aircrafts.
Disclosure of Invention
The invention provides a cluster enclosure control method of an unmanned combat aircraft imitating wolves of wolfs, which aims to solve the problem of cluster enclosure control stability of the unmanned combat aircraft under the condition of sensing limitation and improve the cluster enclosure stability under the condition of sensing limitation by enhancing interaction of heterogeneous nodes under the condition of limited detection and communication range.
Aiming at the application requirements of cluster enclosure control of unmanned aerial vehicles under the condition of limited perception, the invention designs a wolf hunting behavior-imitating unmanned aerial vehicle cluster enclosure control method by using a role differentiation interaction mechanism in a wolf chance definition hunting strategy for reference, improves the cluster enclosure control stability of unmanned aerial vehicles by enhancing the interaction of heterogeneous nodes, and has the advantages of reliable communication maintenance characteristic, low requirement on airborne perception capability and the like. The method comprises the following concrete implementation steps:
the method comprises the following steps: initializing unmanned aerial vehicle motion models
Step two: determining a wolf group hunting character observation level, specifically comprising:
s21, initializing an observation level: initializing a hunting observation level of each unmanned aerial vehicle to be infinite;
s22, constructing a neighbor set: each unmanned aerial vehicle generates an unmanned aerial vehicle set adjacent to the unmanned aerial vehicle;
s23, constructing a neighbor hunting observation level set: each unmanned aerial vehicle calculates a hunting observation level set C of adjacent unmanned aerial vehicles;
s24, determining the self hunting observation level: each unmanned aerial vehicle judges whether the unmanned aerial vehicle has effective detection capability, if so, the observation level is 1; otherwise, the observation level is minC + 1;
step three: computing wolf group leader-following target state observer
S31, generating a spatial topological adjacency matrix A ═ aij]；
S32, calculating a target observation dependency coefficient of the adjacent unmanned aerial vehicle;
s33, calculating and executing a wolf pack leader-following target state observer, and providing target guidance for detecting the disabled unmanned aerial vehicle;
step four: design head wolf-slave wolf interaction mechanism enclosure control law
S41, designing a wolf head group potential function: for any detection unmanned aerial vehicle, calculating an interaction potential field of the wolf head group;
s42, designing a potential-pairing function from a wolf group: for any failed unmanned aerial vehicle, calculating an interaction potential field of a slave wolf group;
s43, design head wolf-interaction potential field from wolf: for any detection unmanned aerial vehicle, calculating the head wolf-slave wolf interaction potential field; for any failed unmanned aerial vehicle, calculating a slave wolf-head wolf interaction potential field;
s44, designing a wolf pack-prey interaction potential field: for any unmanned aerial vehicle, calculating a wolf pack-prey interaction potential field;
s45, execute wolf-slave interaction mechanism control law: for any unmanned aerial vehicle, calculating a horizontal bounding control law, wherein observed values of a target horizontal position, a target speed and a target acceleration are obtained by a leader-following target observer;
s46, executing consistency height control law: for any unmanned aerial vehicle, calculating a consistency height control law, wherein target height, vertical speed and acceleration observed values are obtained by a leader-following target observer;
step five: equivalent control instruction conversion
And C, converting the control command of the step four into corresponding airspeed, yaw and track angle commands.
Step six: outputting unmanned plane cluster enclosure control track
And the control command converter obtains the airspeed, yaw and track angle commands of the unmanned aerial vehicle, acts on the unmanned aerial vehicle motion model obtained in the step one, and outputs the unmanned aerial vehicle cluster surrounding flight track.
Further, the second step of determining the observation level of the wolf hunting character includes the following specific processes:
s21, initializing observation level
The unmanned aerial vehicle is equivalent to a wolf pack individual with different role types, and each unmanned aerial vehicle initializes the state observation level thereof
Figure BDA0002916567610000041
To infinity, i.e. to the ith unmanned aerial vehicle, hi＝∞；
S22, constructing a neighbor set
Each unmanned aerial vehicle generates an unmanned aerial vehicle set adjacent to the unmanned aerial vehicle set according to the following formula:
Ni(p)＝{j|||pi-pj||≤rc} (2)
wherein N isiA set of neighboring drones representing drone i; p is a radical ofiAnd pjRespectively representing the three-dimensional position vectors of the ith unmanned aerial vehicle and the jth unmanned aerial vehicle; p is an unmanned aerial vehicle cluster three-dimensional position stacking vector; r iscRepresents its communication radius;
s23, constructing a neighbor hunting observation hierarchy set
Each unmanned aerial vehicle calculates the hunting observation level set of the adjacent unmanned aerial vehicles to the target state according to the following method:
Figure BDA0002916567610000042
wherein, Ci(t) a set of adjacent drone state observation levels representing drone i; h isjRepresenting a jth unmanned aerial vehicle state observation level; t is t0Is the starting time;
s24, determining self hunting observation level
Whether each unmanned aerial vehicle judges whether the unmanned aerial vehicle isHas effective detection capability, if yes, the state observation level h isi1, namely, the unmanned aerial vehicle is the wolf individual; otherwise, its state observation level is hi＝minCi+1, the higher the hunting observation level, the stronger the slave wolf attribute of the drone, and the weaker the detection capability.
Further, step S32 is to calculate a target observation dependency coefficient of the neighboring unmanned aerial vehicle, and the specific process is as follows: each unmanned aerial vehicle calculates the observation state dependency of the unmanned aerial vehicle on the adjacent unmanned aerial vehicle according to the following formula:
Figure BDA0002916567610000051
wherein, betaij(t) represents a target observation dependency coefficient for drone i with respect to neighboring drone j; beta is a fixed gain coefficient; h isiRepresenting the state observation level attribute of the unmanned aerial vehicle i in the spatial topology; λ is a fixed coefficient.
Further, step S33 executes a wolf pack leader-following target state observer, which includes the following specific processes: adopting the following leader-following target state observer to provide target guidance for detecting the failed unmanned aerial vehicle:
Figure BDA0002916567610000052
wherein, XPThe real state of an observation target is represented, and elements such as a three-dimensional position, a speed and an acceleration can be represented; p is a target index;
Figure BDA0002916567610000053
and
Figure BDA0002916567610000054
target observation states of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle are respectively set; rhoiRepresenting the radial distance of the ith unmanned aerial vehicle relative to the target; r isdThe detection distance of the unmanned aerial vehicle relative to the target; omegaLRepresenting a set of detecting drones or a set of wolfs.
Further, the step S41 is to design the wolf pack potential-alignment function, and the specific process is as follows: for any detection unmanned plane i epsilon omegaLAnd if the head wolf is equivalent to the head wolf individual, designing the following adversity functions:
Figure BDA0002916567610000055
wherein the content of the first and second substances,
Figure BDA0002916567610000061
representing a wolf head group potential function; dijRepresents the radial distance between drones i and j; dLRepresents the desired spacing of the detecting drones from each other, which satisfies:
Figure BDA0002916567610000062
wherein, κLRepresenting a head wolf space configuration setting parameter; n is a radical ofLFor detecting the number of unmanned aerial vehicles; n is a radical ofFThe number of disabled unmanned aerial vehicles;
θ (x) is a communication quality piecewise function satisfying:
Figure BDA0002916567610000063
where ω ∈ (0,1) denotes a communication distance weight coefficient.
Further, the step S42 is to design a potential alignment function from the wolf-wolf group, and the specific process is as follows: let omegaFRepresenting the detection of a set of failed drones or a set of slave wolves, i ∈ Ω for any failed droneFThe following hedging functions were designed:
Figure BDA0002916567610000064
wherein the content of the first and second substances,
Figure BDA0002916567610000065
representing a potential function from a wolf group; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (8); dFRepresenting a desired spacing of the failed drones from each other, which satisfies:
Figure BDA0002916567610000066
wherein, κFRepresenting the parameters set from the wolf space configuration.
Further, the step S43 designs a wolf-wolf interaction potential field, and the specific process is as follows: for any detection unmanned plane i epsilon omegaLThe following hedging functions were designed:
Figure BDA0002916567610000067
wherein the content of the first and second substances,
Figure BDA0002916567610000071
representing a leading wolf-trailing wolf interaction adversary function of the detecting drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,LRepresents the expected spacing of the detecting drone relative to the failed drone, which satisfies:
dLF,L＝κLF,Lrc (13)
wherein, κLF,LRepresenting a leading wolf-trailing wolf interaction space configuration setting parameter;
similarly, for any invalid unmanned aerial vehicle i belongs to omegaFThe following hedging functions were designed:
Figure BDA0002916567610000072
wherein the content of the first and second substances,
Figure BDA0002916567610000073
a slave wolf-head wolf interaction advection function representing a failed drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,FRepresenting an expected spacing of a failed drone relative to a probe drone, which satisfies:
dLF,F＝κLF,Frc (15)
wherein, κLF,FRepresenting the configuration setting parameters from the wolf-wolf interaction space.
Further, the step S44 is to design a wolf pack-prey interaction potential field, which includes the following steps: for any drone i e {1,. and N }, it needs to move immediately after the prey and keep a certain desired enclosure distance, so the following adversarial function is designed:
Figure BDA0002916567610000074
wherein the content of the first and second substances,
Figure BDA0002916567610000075
representing a wolf-animal interaction adversity function;
Figure BDA0002916567610000076
representing a hedonic function from a wolf-prey interaction; dipRepresenting the radial distance between drone i and the target; r ise LAnd re FRespectively show that survey and unmanned aerial vehicle that becomes invalid expect to surround the interval, it satisfies respectively:
Figure BDA0002916567610000077
wherein alpha isLEpsilon (0,1) represents the attack coefficient of the wolf relative to the prey; alpha is alphaFE (0,1) represents the attack coefficient from wolf to game.
Further, the step S45 executes the wolf-wolf interaction mechanism control law, which includes the following specific procedures: for any drone i e { 1., N }, the following horizontal bounding control laws are implemented after combining the above head wolf-slave wolf interaction potential fields:
Figure BDA0002916567610000081
wherein the content of the first and second substances,
Figure BDA0002916567610000082
represents the horizontal direction control vector u of the equivalent second-order system of the ith unmanned aerial vehiclei,xAnd ui,yRespectively are control components of an equivalent second-order system in the directions of an x axis and a y axis;
Figure BDA0002916567610000083
is the gradient operator;
Figure BDA0002916567610000084
and
Figure BDA0002916567610000085
respectively representing horizontal position vectors of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle;
Figure BDA0002916567610000086
Figure BDA0002916567610000087
respectively representing the horizontal position, speed and acceleration observed values of the target by the ith unmanned aerial vehicle, and obtaining the observed values by a leader-following target observer; k3Representing the gain constant.
The invention provides a cluster enclosure control method of an unmanned combat aircraft imitating wolf hunting behaviors, and aims to solve the problem of cluster enclosure control stability of the unmanned aerial vehicle under the condition of sensing limitation. The invention uses the leader-following interaction relationship in the wolf colony opportunistic hunting strategy for reference, and maintains the cluster enclosure control connectivity and stability under the sensing limitation by enhancing the interaction of heterogeneous nodes under the condition of limited detection and communication range.
Drawings
FIG. 1 is a flow chart of an unmanned aerial vehicle cluster enclosure control algorithm of the present invention
FIG. 2 is a schematic diagram of a wolf hunting observation hierarchy and an interactive potential field
FIG. 3 unmanned aerial vehicle cluster initial time position distribution
FIG. 4 three-dimensional flight trajectory of unmanned aerial vehicle cluster
FIG. 5 illustrates a communication topology connectivity curve of an UAV cluster
FIG. 6 illustrates a variation curve of relative target distance of unmanned aerial vehicle cluster
FIG. 7 minimum spacing covariance curve for UAV cluster
The reference numbers and symbols in the figures are as follows:
λ2secondary small eigenvalue (topology connectivity) of unmanned plane cluster communication topology Laplace matrix
t simulation time
tsSimulation step length
i unmanned aerial vehicle sequence number
ΩFFailure unmanned plane set
CiNeighbor hunting observation hierarchy set of unmanned aerial vehicle i
hiHunting observation hierarchy of unmanned aerial vehicle i
Number of unmanned aerial vehicles
TmaxMaximum simulation duration
m units: rice and its production process
s unit: second of
Detailed Description
The effectiveness of the enclosure control method provided by the invention is verified through a specific example. In the example, 20 unmanned aerial vehicles are selected to form a cluster, the initial airspeed and the course of each unmanned aerial vehicle are randomly selected in the intervals (10,20) m/s and (0,2 pi), and the pitch angle is zero. Setting a target motion rule to meet the following maneuvering command of accelerating the screw
Figure BDA0002916567610000091
Wherein u isP＝(uP,x,uP,y,uP,z)TControlling a vector for a target three-dimensional acceleration; v. ofP＝(vP,x,vP,y,vP,z)TA target three-dimensional velocity vector is obtained; t ismax200s is the maximum simulation duration. The unmanned aerial vehicle cluster forms a circumference enclosure to the target under the condition that the sensing distance is limited, and the detection radius of the unmanned aerial vehicle cluster is rd110m, communication radius rc40 m. Unmanned aerial vehicle cluster is divided into N L10 detecting unmanned aerial vehicle and N F10 disabled drones. The starting positions of the respective machines and the target are shown in fig. 3, wherein the dotted line area represents the effective detection area of the target, and the chain line represents the communication range of the detecting unmanned aerial vehicle. The simulation environment of this example is configured as an intel i7-4790 processor, 3.60Ghz master frequency, 4G memory, with software as MATLAB 2010a version.
As shown in fig. 1, the method of the embodiment of the present invention comprises the following specific steps:
the method comprises the following steps: initializing unmanned aerial vehicle motion models
Initializing following unmanned aerial vehicle motion model
Figure BDA0002916567610000101
Wherein N is the number of unmanned aerial vehicles; { xi,yi,ziThe position of the ith unmanned aerial vehicle is shown in the specification; { Vi,χi,γiThe airspeed, the yaw angle and the track inclination angle of the ith unmanned aerial vehicle are set; { Vmin,Vmax,nmin,nmax,γmin,γmaxThe airspeed, transverse overload and track dip angle limit ranges are respectively set; g is the acceleration of gravity;
Figure BDA0002916567610000102
is an airspeed, yaw and track control command; { tauV,τχ,τγAnd the airspeed, yaw and track control response time of the autopilot are respectively.
Including N in a cluster systemLFrame-detectable nobodyMachine and NFFrame detection failure unmanned aerial vehicle. For detectable drones, with a detection distance rd(ii) a For a failed unmanned aerial vehicle, the unmanned aerial vehicle has no detection capability; both types of unmanned aerial vehicles have a finite detection radius rc。
The following parameters are used to configure and set the unmanned aerial vehicle motion model in the embodiment:
τV＝τχ＝τγ＝3s，Vmin＝10m/s，Vmax＝90m/s，nmax＝-nmin＝6，γmax＝-γmin＝π/4，g＝9.8m/s2。
step two: determining a wolf flock hunting character observation hierarchy
(1) Observation level initialization
As shown in fig. 2, the drone is equivalent to a wolf pack of individuals with different role types, and since the detection capabilities of the drone are different, the acquisition of the target observation information is different, and the hunting role observation level needs to be calculated. Each unmanned aerial vehicle initializes its state observation hierarchy
Figure BDA0002916567610000111
To infinity, i.e. to the ith unmanned aerial vehicle, hi＝∞。
(2) Building neighbor sets
Each unmanned aerial vehicle generates an unmanned aerial vehicle set adjacent to the unmanned aerial vehicle according to the following formula
Ni(p)＝{j|||pi-pj||≤rc} (2)
Wherein N isiA set of neighboring drones representing drone i; p is a radical ofiAnd pjRespectively representing the three-dimensional position vectors of the ith unmanned aerial vehicle and the jth unmanned aerial vehicle; p is an unmanned aerial vehicle cluster three-dimensional position stacking vector; r iscIndicating its communication radius.
(3) Constructing neighbor hunting observation hierarchy set
Each unmanned aerial vehicle calculates the hunting observation level set of the adjacent unmanned aerial vehicles to the target state according to the following method
Figure BDA0002916567610000112
Wherein, Ci(t) a set of adjacent drone state observation levels representing drone i; h isjRepresenting a jth unmanned aerial vehicle state observation level; t is t0Is the starting time.
(4) Determining self hunting observation hierarchy
Each unmanned aerial vehicle judges whether the unmanned aerial vehicle has effective detection capability or not, and if yes, the state observation level h of each unmanned aerial vehicle i1, namely, the unmanned aerial vehicle is the wolf individual; otherwise, its state observation level is hi＝minCi+1, the higher the hunting observation level, the stronger the slave wolf attribute of the drone, and the weaker the detection capability.
Step three: computing wolf group leader-following target state observer
(1) Generating spatial topological adjacency matrices
Calculating a spatial topological adjacency matrix a ═ a according to the following formulaij]
Figure BDA0002916567610000121
Wherein, aijI rows and j columns of elements representing the spatial topology adjacency matrix a at the current time.
(2) Calculating target observation dependency coefficients
Each unmanned aerial vehicle calculates the dependency of the unmanned aerial vehicle on the observation state of the adjacent unmanned aerial vehicle according to the following formula
Figure BDA0002916567610000122
Wherein, betaij(t) represents a target observation dependency coefficient for drone i with respect to neighboring drone j; beta is a fixed gain coefficient; h isiRepresenting the state observation level attribute of the unmanned aerial vehicle i in the spatial topology; λ is a fixed coefficient. In this embodiment, β is 5 and λ is 5.
(3) Executing wolf group leader-following target state observer
The following leading-following target state observer is adopted to provide target guidance for detecting the failed unmanned aerial vehicle
Figure BDA0002916567610000123
Wherein, XPThe real state of an observation target is represented, and elements such as a three-dimensional position, a speed and an acceleration can be represented; p is a target index;
Figure BDA0002916567610000124
and
Figure BDA0002916567610000125
target observation states of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle are respectively set; rhoiRepresenting the radial distance of the ith unmanned aerial vehicle relative to the target; r isdThe detection distance of the unmanned aerial vehicle relative to the target; omegaLRepresenting a set of detecting drones or a set of wolfs.
Step four: design head wolf-slave wolf interaction mechanism enclosure control law
(1) Designing head wolf group potential function
As shown in fig. 2, for any probe drone i e ΩLThe equivalent of the root of wolf is to design the following adversity function
Figure BDA0002916567610000131
Wherein the content of the first and second substances,
Figure BDA0002916567610000132
representing a wolf head group potential function; dijRepresents the radial distance between drones i and j; dLRepresenting a desired spacing of the detecting drones from each other, which satisfies
Figure BDA0002916567610000133
Wherein the content of the first and second substances,κLrepresenting a head wolf space configuration setting parameter; n is a radical ofLFor detecting the number of unmanned aerial vehicles; n is a radical ofFNumber of disabled drones.
Theta (x) is a communication quality piecewise function satisfying
Figure BDA0002916567610000134
Where ω ∈ (0,1) denotes a communication distance weight coefficient.
In this example,. kappa.L＝0.4，ω＝0.5。
(2) Designing a pennisetum group potential-pairing function
Let omegaFRepresenting probing a set of failed drones or a set of slave wolves, i e Ω, for any failed drone, as shown in fig. 2FDesigning the following advection function
Figure BDA0002916567610000135
Wherein the content of the first and second substances,
Figure BDA0002916567610000136
representing a potential function from a wolf group; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (8); dFIndicating a desired spacing of the failed drones from each other, which satisfies
Figure BDA0002916567610000137
Wherein, κFRepresenting the parameters set from the wolf space configuration.
In this example,. kappa.F＝1，ω＝0.5。
(3) Designing a leading wolf-trailing wolf interaction potential field
For any detection unmanned plane i epsilon omegaLAs shown in FIG. 2, the following logarithmic potential functions are designed
Figure BDA0002916567610000141
Wherein the content of the first and second substances,
Figure BDA0002916567610000142
representing a leading wolf-trailing wolf interaction adversary function of the detecting drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,LRepresents an expected spacing of a detecting drone relative to a failed drone, which satisfies
dLF,L＝κLF,Lrc (13)
Wherein, κLF,LHead wolf-slave wolf interaction space configuration setting parameters are represented.
Similarly, for any invalid unmanned aerial vehicle i belongs to omegaFDesigning the following advection function
Figure BDA0002916567610000143
Wherein the content of the first and second substances,
Figure BDA0002916567610000144
a slave wolf-head wolf interaction advection function representing a failed drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,FIndicating an expected spacing of failed drones relative to a detecting drone, which satisfies
dLF,F＝κLF,Frc (15)
Wherein, κLF,FRepresenting the configuration setting parameters from the wolf-wolf interaction space.
In this example,. kappa.LF,F＝0.5，ω＝0.5。
(4) Designing wolf colony-prey interaction potential field
For any drone i e { 1.,. N }, as shown in fig. 2, it needs to move immediately after the prey and keep a certain desired enclosure distance, so the following advection function is designed
Figure BDA0002916567610000145
Wherein the content of the first and second substances,
Figure BDA0002916567610000151
representing a wolf-animal interaction adversity function;
Figure BDA0002916567610000152
representing a hedonic function from a wolf-prey interaction; dipRepresenting the radial distance between drone i and the target; r ise LAnd re FRespectively represent the expected enclosure distance of the detection unmanned aerial vehicle and the failure unmanned aerial vehicle, which respectively satisfy
Figure BDA0002916567610000153
Wherein alpha isLEpsilon (0,1) represents the attack coefficient of the wolf relative to the prey; alpha is alphaFE (0,1) represents the attack coefficient from wolf to game.
In this example, αL＝0.3，αF＝0.48。
(5) Execution of wolf-head-slave wolf interaction mechanism control law
For any drone i e { 1.,. N }, the following horizontal bounding control laws are executed after combining the above head wolf-slave wolf-animal interaction potential fields
Figure BDA0002916567610000154
Wherein the content of the first and second substances,
Figure BDA0002916567610000155
represents the horizontal direction control vector u of the equivalent second-order system of the ith unmanned aerial vehiclei,xAnd ui,yRespectively are control components of an equivalent second-order system in the directions of an x axis and a y axis;
Figure BDA0002916567610000156
is the gradient operator;
Figure BDA0002916567610000157
and
Figure BDA0002916567610000158
respectively representing horizontal position vectors of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle;
Figure BDA0002916567610000159
Figure BDA00029165676100001510
respectively representing the horizontal position, speed and acceleration observed values of the target by the ith unmanned aerial vehicle, and obtaining the observed values by a leader-following target observer; k3Representing the gain constant. In this example, K3＝2。
(6) Execution of a consistent altitude control law
For any drone i e {1,. and N }, the following consistent altitude control law is implemented, keeping the same altitude as the target
Figure BDA0002916567610000161
Wherein u isi,zAn equivalent second-order system height control component of the ith unmanned aerial vehicle;
Figure BDA0002916567610000162
and
Figure BDA0002916567610000163
respectively representing the observed values of the unmanned aerial vehicle i on the target height, the vertical speed and the acceleration, and obtaining the observed values through a leading-following target observer; k1,K2And > 0 is height control gain. In this example, K1＝K2＝2。
Step five: equivalent control instruction conversion
The control command converter is adopted to convert the equivalent second-order system control commands of the unmanned aerial vehicle with the formulas (13) and (14) into corresponding airspeed, yaw and track angle commands
Figure BDA0002916567610000164
Wherein u isi＝(ui,x,ui,y,ui,z)TAn equivalent second-order system control vector of the ith unmanned aerial vehicle; giA conversion matrix for representing the control vector of the unmanned aerial vehicle motion model and the equivalent second-order model thereof, and having
Figure BDA0002916567610000165
Step six: outputting unmanned plane cluster enclosure control track
The control command converter of the formula (16) obtains the airspeed, yaw and track angle commands of the unmanned aerial vehicle, acts on the unmanned aerial vehicle motion model of the formula (1), and outputs the flight trajectory surrounded by the unmanned aerial vehicle cluster.
Fig. 4 shows three-dimensional flight trajectories of the unmanned aerial vehicle cluster encircling task of the present example, wherein the dotted line represents the circle that the unmanned aerial vehicle expects to encircle, and the position distribution of the clusters at 50s, 100s, 150s and 200s is shown. Can reachd, the unmanned aerial vehicle cluster has formed double formation fast, surveys unmanned aerial vehicle at the inner ring, and inefficacy unmanned aerial vehicle is at the outer loop. The invalid unmanned aerial vehicle stably tracks the target by following the adjacent detection unmanned aerial vehicle under the condition of no detection capability. Fig. 5 shows the connection situation of the space topology of 20 unmanned planes at each time in this example, where the sub-small eigenvalue λ of the cluster topology Laplace matrix is used2Approximately represented. Can obtain the lambda under the effect of the simulated wolf colony interaction potential field2And 1, a better level is kept, namely, the unmanned plane cluster keeps better connectivity. The results of fig. 6 and 7 show that under the action of the simulated wolf colony interaction potential field, the unmanned aerial vehicle cluster maintains a stable enclosure formation, the distances between the unmanned aerial vehicle clusters are uniform, and the relative target distance approaches the expected enclosure radius. Despite the lack of failure of unmanned aerial vehicleThe target detection function still keeps relatively accurate relative target expected distance under the combined action of the leader-follower interaction potential field and the target observer. Therefore, the unmanned aerial vehicle cluster enclosure control stability under the perception limited condition is well maintained by enhancing the interaction of heterogeneous nodes.

Claims (9)

1. The utility model provides a control method is enclosed to unmanned combat aircraft cluster of imitative wolf crowd hunting action which characterized in that: the method comprises the following steps:
the method comprises the following steps: initializing unmanned aerial vehicle motion models
Step two: determining a wolf group hunting character observation level, specifically comprising:
s21, initializing an observation level: initializing a hunting observation level of each unmanned aerial vehicle to be infinite;
s22, constructing a neighbor set: each unmanned aerial vehicle generates an unmanned aerial vehicle set adjacent to the unmanned aerial vehicle;
s23, constructing a neighbor hunting observation level set: each unmanned aerial vehicle calculates a hunting observation level set C of adjacent unmanned aerial vehicles;
s24, determining the self hunting observation level: each unmanned aerial vehicle judges whether the unmanned aerial vehicle has effective detection capability, if so, the observation level is 1; otherwise, the observation level is minC + 1;
step three: computing wolf group leader-following target state observer
S31, generating a spatial topological adjacency matrix A ═ aij]；
S32, calculating a target observation dependency coefficient of the adjacent unmanned aerial vehicle;
s33, calculating and executing a wolf pack leader-following target state observer, and providing target guidance for detecting the disabled unmanned aerial vehicle;
step four: design head wolf-slave wolf interaction mechanism enclosure control law
S41, designing a wolf head group potential function: for any detection unmanned aerial vehicle, calculating an interaction potential field of the wolf head group;
s42, designing a potential-pairing function from a wolf group: for any failed unmanned aerial vehicle, calculating an interaction potential field of a slave wolf group;
s43, design head wolf-interaction potential field from wolf: for any detection unmanned aerial vehicle, calculating the head wolf-slave wolf interaction potential field; for any failed unmanned aerial vehicle, calculating a slave wolf-head wolf interaction potential field;
s44, designing a wolf pack-prey interaction potential field: for any unmanned aerial vehicle, calculating a wolf pack-prey interaction potential field;
s45, execute wolf-slave interaction mechanism control law: for any unmanned aerial vehicle, calculating a horizontal bounding control law, wherein observed values of a target horizontal position, a target speed and a target acceleration are obtained by a leader-following target observer;
s46, executing consistency height control law: for any unmanned aerial vehicle, calculating a consistency height control law, wherein target height, vertical speed and acceleration observed values are obtained by a leader-following target observer;
step five: equivalent control instruction conversion
Converting the control instruction in the step four into corresponding airspeed, yaw and track angle instructions;
step six: outputting unmanned plane cluster enclosure control track
And the control command converter obtains the airspeed, yaw and track angle commands of the unmanned aerial vehicle, acts on the unmanned aerial vehicle motion model obtained in the step one, and outputs the unmanned aerial vehicle cluster surrounding flight track.
2. The method as claimed in claim 1, wherein the method comprises the steps of: determining the observation level of the wolf hunting character in the second step, wherein the specific process is as follows:
s21, initializing observation level
The unmanned aerial vehicle is equivalent to a wolf pack individual with different role types, and each unmanned aerial vehicle initializes the state observation level thereof
Figure FDA0002916567600000021
To infinity, i.e. to the ith unmanned aerial vehicle, hi＝∞；
S22, constructing a neighbor set
Each unmanned aerial vehicle generates an unmanned aerial vehicle set adjacent to the unmanned aerial vehicle set according to the following formula:
Ni(p)＝{j|||pi-pj||≤rc} (2)
wherein N isiA set of neighboring drones representing drone i; p is a radical ofiAnd pjRespectively representing the three-dimensional position vectors of the ith unmanned aerial vehicle and the jth unmanned aerial vehicle; p is an unmanned aerial vehicle cluster three-dimensional position stacking vector; r iscRepresents its communication radius;
s23, constructing a neighbor hunting observation hierarchy set
Each unmanned aerial vehicle calculates the hunting observation level set of the adjacent unmanned aerial vehicles to the target state according to the following method:
Figure FDA0002916567600000031
wherein, Ci(t) a set of adjacent drone state observation levels representing drone i; h isjRepresenting a jth unmanned aerial vehicle state observation level; t is t0Is the starting time;
s24, determining self hunting observation level
Each unmanned aerial vehicle judges whether the unmanned aerial vehicle has effective detection capability or not, and if yes, the state observation level h of each unmanned aerial vehiclei1, namely, the unmanned aerial vehicle is the wolf individual; otherwise, its state observation level is hi＝minCi+1, the higher the hunting observation level, the stronger the slave wolf attribute of the drone, and the weaker the detection capability.
3. The method as claimed in claim 1, wherein the method comprises the steps of: step S32 is to calculate the target observation dependency coefficient of the neighboring drone, and the specific process is as follows: each unmanned aerial vehicle calculates the observation state dependency of the unmanned aerial vehicle on the adjacent unmanned aerial vehicle according to the following formula:
Figure FDA0002916567600000032
wherein, betaij(t) represents a target observation dependency coefficient for drone i with respect to neighboring drone j; beta is a fixed gain coefficient; h isiRepresenting the state observation level attribute of the unmanned aerial vehicle i in the spatial topology; λ is a fixed coefficient.
4. The method as claimed in claim 1, wherein the method comprises the steps of: step S33 is executed to execute a wolf pack leader-following target state observer, and the specific process is as follows:
adopting the following leader-following target state observer to provide target guidance for detecting the failed unmanned aerial vehicle:
Figure FDA0002916567600000041
wherein, XPThe real state of an observation target is represented, and elements such as a three-dimensional position, a speed and an acceleration can be represented; p is a target index;
Figure FDA0002916567600000042
and
Figure FDA0002916567600000043
target observation states of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle are respectively set; rhoiRepresenting the radial distance of the ith unmanned aerial vehicle relative to the target; r isdThe detection distance of the unmanned aerial vehicle relative to the target; omegaLRepresenting a set of detecting drones or a set of wolfs.
5. The method as claimed in claim 1, wherein the method comprises the steps of: the step S41 designs the wolf pack potential-alignment function, and the specific process is as follows:
for any detection unmanned plane i epsilon omegaLAnd if the head wolf is equivalent to the head wolf individual, designing the following adversity functions:
Figure FDA0002916567600000044
wherein the content of the first and second substances,
Figure FDA0002916567600000045
representing a wolf head group potential function; dijRepresents the radial distance between drones i and j; dLRepresents the desired spacing of the detecting drones from each other, which satisfies:
Figure FDA0002916567600000046
wherein, κLRepresenting a head wolf space configuration setting parameter; n is a radical ofLFor detecting the number of unmanned aerial vehicles; n is a radical ofFThe number of disabled unmanned aerial vehicles;
θ (x) is a communication quality piecewise function satisfying:
Figure FDA0002916567600000047
where ω ∈ (0,1) denotes a communication distance weight coefficient.
6. The method as claimed in claim 1, wherein the method comprises the steps of: the step S42 is to design a pennisetum potential-alignment function, and the specific process is as follows:
let omegaFRepresenting the detection of a set of failed drones or a set of slave wolves, i ∈ Ω for any failed droneFThe following hedging functions were designed:
Figure FDA0002916567600000051
wherein the content of the first and second substances,
Figure FDA0002916567600000052
representing a potential function from a wolf group; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (8); dFRepresenting a desired spacing of the failed drones from each other, which satisfies:
Figure FDA0002916567600000053
wherein, κFRepresenting the parameters set from the wolf space configuration.
7. The method as claimed in claim 1, wherein the method comprises the steps of: the step S43 designs a wolf-wolf interaction potential field, and the specific process is as follows: for any detection unmanned plane i epsilon omegaLThe following hedging functions were designed:
Figure FDA0002916567600000054
wherein the content of the first and second substances,
Figure FDA0002916567600000055
representing a leading wolf-trailing wolf interaction adversary function of the detecting drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,LRepresents the expected spacing of the detecting drone relative to the failed drone, which satisfies:
dLF,L＝κLF,Lrc (13)
wherein, κLF,LRepresenting a leading wolf-trailing wolf interaction space configuration setting parameter;
similarly, for any invalid unmanned aerial vehicle i belongs to omegaFThe following hedging functions were designed:
Figure FDA0002916567600000056
wherein the content of the first and second substances,
Figure FDA0002916567600000057
a slave wolf-head wolf interaction advection function representing a failed drone; dijRepresents the radial distance between drones i and j; θ (x) is a piecewise function of the same form as formula (9); dLF,FRepresenting an expected spacing of a failed drone relative to a probe drone, which satisfies:
dLF,F＝κLF,Frc (15)
wherein, κLF,FRepresenting the configuration setting parameters from the wolf-wolf interaction space.
8. The method as claimed in claim 1, wherein the method comprises the steps of: the step S44 is to design a wolf pack-prey interaction potential field, and the specific process is as follows: for any drone i e {1,. and N }, it needs to move immediately after the prey and keep a certain desired enclosure distance, so the following adversarial function is designed:
Figure FDA0002916567600000061
wherein the content of the first and second substances,
Figure FDA0002916567600000062
representing a wolf-animal interaction adversity function;
Figure FDA0002916567600000063
representing a hedonic function from a wolf-prey interaction; dipRepresenting the radial distance between drone i and the target;
Figure FDA0002916567600000064
and
Figure FDA0002916567600000065
respectively show that survey and unmanned aerial vehicle that becomes invalid expect to surround the interval, it satisfies respectively:
Figure FDA0002916567600000066
wherein alpha isLEpsilon (0,1) represents the attack coefficient of the wolf relative to the prey; alpha is alphaFE (0,1) represents the attack coefficient from wolf to game.
9. The method as claimed in claim 1, wherein the method comprises the steps of: the step S45 executes the wolf-wolf interaction mechanism control law, which includes the following specific procedures: for any drone i e { 1., N }, the following horizontal bounding control laws are implemented after combining the above head wolf-slave wolf interaction potential fields:
Figure FDA0002916567600000067
wherein the content of the first and second substances,
Figure FDA0002916567600000071
represents the horizontal direction control vector u of the equivalent second-order system of the ith unmanned aerial vehiclei,xAnd ui,yRespectively are control components of an equivalent second-order system in the directions of an x axis and a y axis;
Figure FDA0002916567600000072
is the gradient operator;
Figure FDA0002916567600000073
and
Figure FDA0002916567600000074
respectively representing horizontal position vectors of an ith unmanned aerial vehicle and a jth unmanned aerial vehicle;
Figure FDA0002916567600000075
Figure FDA0002916567600000076
respectively representing the horizontal position, speed and acceleration observed values of the target by the ith unmanned aerial vehicle, and obtaining the observed values by a leader-following target observer; k3Representing the gain constant.
CN202110103753.5A 2021-01-26 2021-01-26 Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors Active CN112925345B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110103753.5A CN112925345B (en) 2021-01-26 2021-01-26 Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110103753.5A CN112925345B (en) 2021-01-26 2021-01-26 Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors

Publications (2)

Publication Number Publication Date
CN112925345A true CN112925345A (en) 2021-06-08
CN112925345B CN112925345B (en) 2022-01-28

Family

ID=76166280

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110103753.5A Active CN112925345B (en) 2021-01-26 2021-01-26 Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors

Country Status (1)

Country Link
CN (1) CN112925345B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113627082A (en) * 2021-08-03 2021-11-09 北京航空航天大学 Unmanned intelligent cluster confrontation control method based on wolf hunting behavior bionics

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108376282A (en) * 2018-01-18 2018-08-07 东莞理工学院 A kind of computational intelligence optimization method of simulation wolf pack behavior
CN109902796A (en) * 2019-03-14 2019-06-18 重庆邮电大学 A kind of improved wolf pack algorithm based on fireworks explosion
CN110069076A (en) * 2019-04-23 2019-07-30 北京航空航天大学 A kind of unmanned plane cluster air battle method for surrounding and seize behavior based on violent wolf
CN110554709A (en) * 2019-09-06 2019-12-10 哈尔滨工业大学（深圳） Distributed bionic multi-agent autonomous cluster control method
CN110608743A (en) * 2019-10-18 2019-12-24 南京航空航天大学 Multi-unmanned aerial vehicle collaborative route planning method based on multi-population chaotic grayling algorithm
CN111443728A (en) * 2020-03-25 2020-07-24 北京航空航天大学 Chaos wolf optimization-based unmanned aerial vehicle formation control method
CN111722643A (en) * 2020-06-12 2020-09-29 北京航空航天大学 Unmanned aerial vehicle cluster dynamic task allocation method imitating wolf colony cooperative hunting mechanism

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108376282A (en) * 2018-01-18 2018-08-07 东莞理工学院 A kind of computational intelligence optimization method of simulation wolf pack behavior
CN109902796A (en) * 2019-03-14 2019-06-18 重庆邮电大学 A kind of improved wolf pack algorithm based on fireworks explosion
CN110069076A (en) * 2019-04-23 2019-07-30 北京航空航天大学 A kind of unmanned plane cluster air battle method for surrounding and seize behavior based on violent wolf
CN110554709A (en) * 2019-09-06 2019-12-10 哈尔滨工业大学（深圳） Distributed bionic multi-agent autonomous cluster control method
CN110608743A (en) * 2019-10-18 2019-12-24 南京航空航天大学 Multi-unmanned aerial vehicle collaborative route planning method based on multi-population chaotic grayling algorithm
CN111443728A (en) * 2020-03-25 2020-07-24 北京航空航天大学 Chaos wolf optimization-based unmanned aerial vehicle formation control method
CN111722643A (en) * 2020-06-12 2020-09-29 北京航空航天大学 Unmanned aerial vehicle cluster dynamic task allocation method imitating wolf colony cooperative hunting mechanism

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
CHENGZHI QU: "A novel reinforcement learning based grey wolf optimizer algorithm for unmanned aerial vehicles (UAVs) path planning", 《APPLIED SOFT COMPUTING JOURNAL》 *
HAOHAN CHEN: "Multiple Unmanned Aerial Vehicle Autonomous Formation via Wolf", 《2016 IEEE/CSAA INTERNATIONAL CONFERENCE ON AIRCRAFT UTILITY SYSTEMS (AUS)》 *
刘森琪: "基于信息素启发狼群算法的UAV集群火力分配", 《北京航空航天大学学报》 *
张岱峰: "仿狼群狩猎空间交互机制的无人机集群合围控制", 《中国科学: 技术科学》 *
朱磊: "航空集群协同探测最优编队构型研究", 《火力与指挥控制》 *
段海滨: "从狼群智能到无人机集群协同决策", 《中国科学：信息科学》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113627082A (en) * 2021-08-03 2021-11-09 北京航空航天大学 Unmanned intelligent cluster confrontation control method based on wolf hunting behavior bionics
CN113627082B (en) * 2021-08-03 2024-02-13 北京航空航天大学 Unmanned intelligent cluster countermeasure control method based on wolf group hunting behavior bionical

Also Published As

Publication number Publication date
CN112925345B (en) 2022-01-28

Similar Documents

Publication Publication Date Title
CN113589842B (en) Unmanned cluster task cooperation method based on multi-agent reinforcement learning
CN108845590B (en) Multi-unmanned aerial vehicle cooperative formation control method in time delay environment
Khan et al. Cooperative robots to observe moving targets
CN108549407B (en) Control algorithm for multi-unmanned aerial vehicle cooperative formation obstacle avoidance
CN112880688B (en) Unmanned aerial vehicle three-dimensional track planning method based on chaotic self-adaptive sparrow search algorithm
CN110069076B (en) Unmanned aerial vehicle cluster air combat method based on wolf warns trapping behavior
CN108897312B (en) Method for planning continuous monitoring path of multiple unmanned aerial vehicles to large-scale environment
CN110608743A (en) Multi-unmanned aerial vehicle collaborative route planning method based on multi-population chaotic grayling algorithm
CN112180967B (en) Multi-unmanned aerial vehicle cooperative countermeasure decision-making method based on evaluation-execution architecture
Quintero et al. Robust UAV coordination for target tracking using output-feedback model predictive control with moving horizon estimation
CN113791634A (en) Multi-aircraft air combat decision method based on multi-agent reinforcement learning
CN112198892B (en) Multi-unmanned aerial vehicle intelligent cooperative penetration countermeasure method
Soria et al. The influence of limited visual sensing on the reynolds flocking algorithm
CN111077909B (en) Novel unmanned aerial vehicle self-group self-consistent optimization control method based on visual information
Stegagno et al. Relative localization and identification in a heterogeneous multi-robot system
CN114397911A (en) Unmanned aerial vehicle cluster confrontation decision-making method based on multiple intelligent agents
CN112925345B (en) Cluster enclosure control method of unmanned combat aircraft imitating wolf hunting behaviors
CN113139331A (en) Air-to-air missile situation perception and decision method based on Bayesian network
Fan et al. Path planning for a reconnaissance UAV in uncertain environment
CN113434966A (en) Elastic force field-based bee colony collision avoidance control method
CN112000070B (en) Heterogeneous cluster cooperative motion planning method
CN114372603A (en) Pigeon-group-imitated multi-learning-intelligence unmanned target drone collaborative route dynamic planning method
CN110162084B (en) Formation control method of flying missile cluster system based on consistency theory
CN113507339B (en) Unmanned aerial vehicle swarm countering method based on swarm behavior characteristics
Iovino et al. Implementation of a distributed flocking algorithm with obstacle avoidance capability for UAV swarming

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant