A method for mapping behaviors of a UAV cluster based on fusion of communication and perception information

Abstract

The application discloses a kind of based on communication and perception information fusion UAV cluster behavior mapping method, applied to the field of UAV cluster autonomous decision control, including: establishing the communication and perception model of UAV, and carry out fusion, obtain friendly party UAV information, obstacle information and ground target information, and based on this, design UAV basic behavior rule;Set the information weight of each cluster behavior, rule weight;To friendly party UAV information, obstacle information and ground target information are characterized and handled, obtain characterization information, obtain optimal cluster behavior in combination with the information weight of each cluster behavior, and based on the rule weight of optimal cluster behavior, the output of UAV basic behavior rule is weighted summation, obtain decision output, conversion obtains expected waypoint.The application realizes when UAV cluster communication is interfered, can be applicable to fixed-wing UAV that cannot effectively obtain perception information due to fast flight speed, long distance between aircraft.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned aerial vehicle cluster autonomous decision control, and particularly relates to a method for mapping behaviors of unmanned aerial vehicle cluster based on fusion of communication and perception information. BACKGROUND

[0002] Unmanned aerial vehicle cluster plays an important role in tasks such as express delivery, agricultural plant protection, emergency rescue, remote sensing observation, and light performance. In order to achieve higher task performance, it is necessary for the unmanned aerial vehicle cluster to make collaborative decisions.

[0003] The existing cluster decision algorithm has a high degree of dependence on data interaction between individuals. When a small and low-cost unmanned aerial vehicle cluster is expanded to a larger scale, the amount of data interaction between individuals will increase exponentially, which will bring great pressure to networking communication, resulting in a decline in communication quality and a reduction in cluster performance. In addition, in complex outdoor environments, the shielding of various complex obstacles and uncertain electromagnetic interference will also hinder or even completely disable the information transmission between individuals, at which time the collaborative ability between cluster individuals will be completely lost, greatly affecting the task efficiency.

[0004] Currently, the research on decision algorithms for unmanned aerial vehicle cluster communication interference is mainly focused on obtaining information through perception means as an alternative to communication. For example, in the paper “Swarm of micro flying robots in the wild”, Xin Zhou team developed a micro rotor unmanned aerial vehicle platform equipped with a depth camera. Without relying on external positioning and inter-machine communication, the platform only obtains information in the neighborhood through visual perception, and achieves tasks such as collaborative navigation, collaborative target positioning and tracking, obstacle avoidance in the region, and formation keeping in complex outdoor environments. In the paper “Bio-Inspired Compact Swarms of Unmanned Aerial Vehicles without Communication and External Localization”, Pavel Petracek team placed ultraviolet markers on the rotor unmanned aerial vehicle. The unmanned aerial vehicle determines the position and distance of friendly unmanned aerial vehicles by identifying the markers. This method requires maintaining a proper distance between unmanned aerial vehicles, and has a high requirement for the accuracy of the installation position of the markers.

[0005] However, at present, most of the researches on marker-based and vision-based UAV cluster take rotor-wing UAVs as test platforms. The rotor-wing UAVs have slow flight speed and stable flight state, which is conducive to ensuring the accuracy and stability of visual identification in the flight process. However, the fixed-wing UAVs have fast flight speed, and the distance between UAVs usually exceeds the available range of the sensors carried by the rotor-wing UAVs, so it is difficult to keep the friendly UAVs to be identified within the sensing range all the time, and the stability of the identification process is low. Therefore, the sensing means for supplementing communication information in the rotor-wing UAV cluster cannot achieve the corresponding effect when used in the fixed-wing UAV cluster.

[0006] Therefore, how to provide a UAV cluster behavior mapping method based on communication and sensing information fusion capable of autonomous decision-making when the communication of the fixed-wing UAV cluster is interfered is a problem to be solved by those skilled in the art. SUMMARY

[0007] Therefore, the present application provides a UAV cluster behavior mapping method based on communication and sensing information fusion. Taking a small-sized and low-cost fixed-wing UAV cluster as the research object, when the communication is interfered and causes the UAVs to be disconnected with part of the friendly individuals in the neighborhood, the sensing information is used as a supplement, the communication information and the sensing information are fused, the fused information obtained in real time is mapped to the decision-making behavior of the UAV individuals by using the mapping method, and the effect of the UAV cluster autonomously deciding to complete the task target is presented on the macro level.

[0008] To achieve the above object, the present application adopts the following technical solutions:

[0009] A UAV cluster behavior mapping method based on communication and sensing information fusion, comprising:

[0010] Step (1): establishing a communication model and a sensing model of the UAV, and fusing to obtain friendly UAV information, obstacle information and ground target information.

[0011] Step (2): based on the friendly UAV information, the obstacle information and the ground target information, designing a basic behavior rule of the UAV.

[0012] Step (3): setting the information weight and the rule weight of each cluster behavior; performing characterization processing on the friendly UAV information, the obstacle information and the ground target information to obtain characterization information, combining the information weight of each cluster behavior to obtain an optimal cluster behavior, and based on the rule weight of the optimal cluster behavior, performing weighted summation on the output of the basic behavior rule of the UAV to obtain a decision output, and converting to obtain an expected waypoint.

[0013] Optionally, in step (1), the communication model is as follows:

[0014]

[0015]

[0016]

[0017] in, For UAV U i state; p i and v i UAV U i Position and velocity in a two-dimensional Cartesian coordinate system; For UAV U i The neighborhood of the drone U i The composition of friendly drones within the communication range; U k For drones i Friendly drones within the communication range; U is the set of drones; p k For UAV U k Position in a two-dimensional Cartesian coordinate system; For UAV U i The communication range of UAV i A circle with a center of For UAV U i Obtained in UAV U i Friendly drones within communication range U k Complete status information of v k For UAV U k Velocity in two-dimensional Cartesian coordinates.

[0018] Optionally, in step (1), the perception model includes:

[0019] Friendly drone status information is as follows:

[0020]

[0021]

[0022] in, For UAV U i The set of friendly drones detected; U k For UAV U i Sensed friendly drones; p k For UAV U k Position in a two-dimensional Cartesian coordinate system; For UAV U i The perception range is the field of view angle θ and the perception radius d p The fan shape; For UAV Ui Sensed friendly drone U k Status information; α i，k For friendly drones U k In UAV U i Relative angle within the field of view; φ k For friendly drones U k The heading of is expressed as a vector;

[0023] Ground target status information is as follows:

[0024]

[0025]

[0026] in, For UAV U i The set of perceived ground targets; T j For UAV U i Perceived ground targets; T is the ground target set; p j For ground targets T j Position in a two-dimensional Cartesian coordinate system; For UAV U i Perceived ground target T j Status information; α i，j For ground targets T j In UAV U i Relative angle within the field of view;

[0027] Obstacle status information is as follows:

[0028]

[0029]

[0030] in, For UAV U i The set of perceived obstacles; O z For UAV U i Perceived obstacles; O is the set of obstacles; p z Obstacle O z Position in a two-dimensional Cartesian coordinate system; UAVU i Perceived obstacles O z Status information; α i，z Obstacle O z In UAV U i Relative angle within the field of view.

[0031] Optionally, in step (1), the friendly drone information is as follows:

[0032]

[0033]

[0034] wherein, is the friendly unmanned aerial vehicle U k in the neighborhood of the unmanned aerial vehicle U i distribution situation; is the motion trend of the friendly unmanned aerial vehicle U k

[0035] The ground target information is as follows:

[0036] p j = F(p i , a i，j , p k , a k，j );

[0037]

[0038]

[0039] wherein, a k，j is the relative angle of the ground target T j in the field of view of the unmanned aerial vehicle U k F represents the information fusion process; is the ground target set obtained by information fusion; is the ground target set perceived by the unmanned aerial vehicle U k χ T is the ground target information obtained by information fusion;

[0040] The obstacle information is as follows:

[0041] p z = F(p i , a i，z , p k , a k，z );

[0042]

[0043]

[0044] wherein, a k，z is the relative angle of the obstacle O z in the field of view of the unmanned aerial vehicle U k is the obstacle set obtained by information fusion; is the obstacle set perceived by the unmanned aerial vehicle U k ​​The set of perceived obstacles; χ O is the obstacle information obtained by information fusion.

[0045] Optionally, in step (2), the basic behavior rules of the drone include:

[0046] Formation rules: Make UAV U i With UAV i Friendly drones in the vicinity U k The flight directions tend to be consistent, as follows:

[0047]

[0048] Gathering rules: Make drones U i UAV i Friendly drones in the vicinity U k The cluster as a whole shows a trend of center aggregation, as shown below:

[0049]

[0050] Dispersion rules: making drones U i UAV i Friendly drones in the vicinity U k The geometric center of the cluster flies in the opposite direction, and the cluster as a whole shows a trend of rapid dispersion, as shown below:

[0051]

[0052] Collision avoidance rules: Make UAV i With UAV i Friendly drones in the vicinity U k Keep a certain distance, as follows:

[0053]

[0054]

[0055] Among them, d i，k For UAV U i With UAV i Friendly drones in the vicinity U k distance; c4 is a constant;

[0056] Obstacle avoidance rules: Make the drone U i Maintain a safe distance from perceived obstacles as follows:

[0057]

[0058] Approach target rule: Make the drone U iFly in the direction of the discovered ground target as follows:

[0059]

[0060] Keep away from target rule: Make the drone U i Fly in the opposite direction to the detected ground target as follows:

[0061]

[0062] Optionally, in step (3), set the information weight and rule weight of each cluster behavior as follows: A set of information weights Q s With a set of rule weights W s Indicates as follows:

[0063]

[0064] Assuming e types of cluster behaviors, the decision model q is as follows:

[0065]

[0066] Optionally, in step (3), the characterization information is as follows:

[0067]

[0068] Among them, b i，1 Used to represent drone U i The heading of the drone U i Friendly drones in the vicinity U k The size of the angle between the comprehensive headings is used to represent the UAV U i With UAV i Friendly drones in the vicinity U k Differences in movement trends; b i，2 Used to represent drone U i Friendly drones in the vicinity U k The distribution of the position is used to represent the UAV U i The compactness or arrangement of the surrounding friendly drones; b i，3 and b i，4 Indicates whether ground targets and obstacles are perceived.

[0069] Optionally, in step (3), the information weight of each clustering behavior is combined to obtain the optimal clustering behavior as follows:

[0070] Use the representation information and the information weight of each cluster behavior to perform weighted summation, and take the cluster behavior corresponding to the maximum value of the sum as the optimal cluster behavior as follows:

[0071]

[0072] Optionally, in step (3), the output of the basic behavior rule of the UAV is weighted and summed based on the rule weight of the optimal swarm behavior to obtain a decision output, as follows:

[0073]

[0074] wherein, is the decision output, and represents the desired heading of the UAV U i .

[0075] Optionally, in step (3), after obtaining the decision output, a desired waypoint is obtained by conversion, as follows:

[0076]

[0077] wherein, the meaning of the extend function is to extend a line from the current position p i of the UAV U i as a starting point and in the direction of the desired heading , and the intersection of the extended line and the boundary of the task area is the desired waypoint Wp i .

[0078] According to the technical solution described above, compared with the prior art, the application proposes a UAV swarm behavior mapping method based on communication and perception information fusion, which has the following advantages:

[0079] By fusing the communication information and the perception information, the fusion information obtained in real time is mapped into the decision behavior of the UAV individual by using the mapping method, and when the UAV swarm communication is interfered, the fixed-wing UAV which cannot effectively obtain the perception information due to the high flight speed and the long distance between UAVs can be applied.

[0080] The UAV swarm is completely autonomously controlled, when the swarm size is expanded or a new swarm behavior appears, only the weight parameter corresponding to the task target (new swarm behavior) needs to be set, and the swarm can autonomously emerge the swarm behavior meeting the task expectation.

[0081] The behavior mapping has extensibility, the basic behavior of the UAV in the behavior rule set can be added or deleted according to the demand such as the complexity of the task (at this time, the basic behavior rule of the UAV cannot well constitute the complex swarm behavior), so that the behavior mapping process according to the fusion information is more targeted and the task target is better completed. BRIEF DESCRIPTION OF DRAWINGS

[0082] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0083] Figure 1 Schematic diagram of the method of the present invention.

[0084] Figure 2 Schematic diagram of the communication model of the present invention.

[0085] Figure 3 Schematic diagram of the perception model of the present invention.

[0086] Figure 4 Schematic diagram of ground target information after the communication model and perception model of the present invention are integrated. DETAILED DESCRIPTION

[0087] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0088] Example 1:

[0089] Embodiment 1 of the present invention discloses a method for mapping the behavior of a drone cluster based on the fusion of communication and perception information, such as Figure 1 As shown, including:

[0090] Step (1): Establish the communication model and perception model of the drone as follows:

[0091] Communication models, such as Figure 2 As shown below:

[0092]

[0093]

[0094]

[0095] in, For UAV U i state; p i and v i UAV U i Position and velocity in a two-dimensional Cartesian coordinate system; For UAV U i The neighborhood of the drone U i The composition of friendly drones within the communication range; U k For drones i Friendly drones within the communication range; U is the set of drones; p k For UAV U k Position in a two-dimensional Cartesian coordinate system; For UAV U i The communication range of UAV i A circle with a center of For UAV U i Obtained in UAV U i Friendly drones within communication range U k Complete status information of v k For UAV U k Velocity in two-dimensional Cartesian coordinates.

[0096] Perception models, such as Figure 3 As shown, including:

[0097] Friendly drone status information is as follows:

[0098]

[0099]

[0100] in, For UAV U i The set of friendly drones detected; U k For UAV U i Sensed friendly drones; p k For UAV U k Position in a two-dimensional Cartesian coordinate system; For UAV U i The perception range is the field of view angle θ and the perception radius d p The fan shape; For UAV U i Sensed friendly drone U k Status information; α i，k For friendly drones U k In UAV U i Relative angle within the field of view; φ k For friendly drones U k The heading of is expressed as a vector;

[0101] Ground target status information is as follows:

[0102]

[0103]

[0104] wherein, is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U j is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U j is the set of ground targets T perceived by the UAV U j is the set of ground targets T perceived by the UAV U is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U j is the set of ground targets T perceived by the UAV U i，j is the set of ground targets T perceived by the UAV U j is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U

[0105] is the set of ground targets T perceived by the UAV U

[0106]

[0107]

[0108] wherein, is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U z is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U z is the set of ground targets T perceived by the UAV U z is the set of ground targets T perceived by the UAV U is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U z is the set of ground targets T perceived by the UAV U i，z is the set of ground targets T perceived by the UAV U z is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U

[0109] is the set of ground targets T perceived by the UAV U

[0110] is the set of ground targets T perceived by the UAV U

[0111]

[0112]

[0113] wherein, is the set of ground targets T perceived by the UAV U k is the set of ground targets T perceived by the UAV U i is the set of ground targets T perceived by the UAV U For friendly drones U k Movement trend. i The information obtained by communication takes precedence over the information obtained by perception. i Able to communicate and sense friendly U at the same time k When U i Unable to communicate but able to sense U k When using sensory information.

[0114] Ground target information, such as Figure 4 As shown, when the drone U i Perceive the target And there are friendly machines with which to communicate in the neighborhood Perceive the same target When U i According to its own position p i The relative position α of the target within its own perception range i，j , and friendly machine U k Position p k With the goal in U k Relative orientation within the sensing range α k，j , according to F(·) fusion to obtain the target T j Complete information as follows:

[0115] p j =F(p i , α i，j , p k , α k，j )

[0116]

[0117]

[0118] Among them, α k，j For ground targets T j In UAV U k The relative angle within the field of view; F represents the information fusion process; is the ground target set obtained by information fusion; For UAV U k The set of perceived ground targets; χ T It is the ground target information obtained by information fusion.

[0119] Obstacle information: A single drone can only obtain target and obstacle information within its sensing range through perception. i Sense obstacles And there are friendly machines with which to communicate in the neighborhood Sense the same obstacle When U i According to its own position p i The relative position α to the obstacle within its own perception range i，z , and friendly machine U k Position p k With obstacles in U k Relative orientation within the sensing range α k，z , fusion obtains obstacle O j Complete information as follows:

[0120] p z =F(p i , α i，z , p k , α k，z );

[0121]

[0122]

[0123] Among them, α k，z Obstacle O z In UAV U k Relative angle within the field of view; is the obstacle set obtained by information fusion; For UAV U k The set of perceived obstacles; χ O is the obstacle information obtained by information fusion.

[0124] Step (2): Based on the friendly UAV information, obstacle information, and ground target information, basic UAV behavior rules are designed. The basic UAV behavior rule set g designed in the present invention contains a total of 7 behavior rules. The 7 rules are related to friendly UAVs, obstacles, and ground targets, including:

[0125] Formation rules: Make UAV U i With UAV i Friendly drones in the vicinity U k The flight directions tend to be consistent, as follows:

[0126]

[0127] Gathering rules: Make drones U i UAV i Friendly drones in the vicinity U k The cluster as a whole shows a trend of center aggregation, as shown below:

[0128]

[0129] Dispersion rules: making drones U i UAV i Friendly drones in the vicinity U k The geometric center of the cluster flies in the opposite direction, and the cluster as a whole shows a trend of rapid dispersion, as shown below:

[0130]

[0131] Collision avoidance rules: Make UAV i With UAV i Friendly drones in the vicinity U k Keep a certain distance, as follows:

[0132]

[0133]

[0134] Among them, d i，k For UAV U i With UAV i Friendly drones in the vicinity U k distance; c4 is a constant.

[0135] Obstacle avoidance rules: Make the drone U i Maintain a safe distance from perceived obstacles as follows:

[0136]

[0137] Approach target rule: Make the drone U i Fly in the direction of the discovered ground target as follows:

[0138]

[0139] Keep away from target rule: Make the drone U i Fly in the opposite direction to the detected ground target as follows:

[0140]

[0141] After obtaining the output of each behavioral rule, using different behavioral weights to combine them can lead to the emergence of different cluster behaviors. Therefore, how individuals select the optimal set of behavioral weights at the current moment from multiple sets of weights is the key to the effectiveness of cluster behavior emergence.

[0142] Step (3): Set the information weight and rule weight of each cluster behavior as follows:

[0143] The sth type of clustering behavior A set of information weights Q s A set of rule weights W s As shown below:

[0144]

[0145] A set of e cluster behaviors, and a decision model q is as shown below:

[0146]

[0147] The information of friendly unmanned aerial vehicles, obstacles and ground targets is characterized and processed to obtain characterization information, as shown below:

[0148]

[0149] Wherein, b i，1 is used to represent the heading of the unmanned aerial vehicle U i and the angle between the heading of the unmanned aerial vehicle U i and the comprehensive heading of the friendly unmanned aerial vehicles U k in the neighborhood of the unmanned aerial vehicle U i , which is used to represent the difference in motion trend between the unmanned aerial vehicle U i and the friendly unmanned aerial vehicles U k in the neighborhood of the unmanned aerial vehicle U i，2 ; b i is used to represent the distribution of the positions of the friendly unmanned aerial vehicles U k in the neighborhood of the unmanned aerial vehicle U i , which is used to represent the compactness or combing degree of the distribution of the friendly unmanned aerial vehicles around the unmanned aerial vehicle U i，3 ; b i，4 and b s respectively represent whether the ground target and the obstacle are perceived.

[0150] The information weight Q i of each cluster behavior is used to select the optimal cluster behavior at the current moment according to the characterization information b The characterization information is respectively weighted and summed with the information weight of each cluster behavior, and the cluster behavior corresponding to the maximum value obtained by the summation is taken as the optimal cluster behavior As shown below:

[0151]

[0152] The rule weight based on the optimal cluster behavior is used to weight and sum the output of the basic behavior rule of the unmanned aerial vehicle, and the decision output is obtained, as shown below:

[0153]

[0154] Wherein, is the decision output, which represents the unmanned aerial vehicle Ui The desired heading needs to be converted into a more upper control quantity, i.e. the waypoint Wp of the UAV i .

[0155] The decision output is converted to obtain the desired waypoint as follows:

[0156]

[0157] Wherein, the meaning of the extend function is to make an extension line with the current position p of the UAV U i As the starting point, and the desired heading i As the direction, and the intersection of the extension line and the boundary of the task area is the desired waypoint Wp i .

[0158] The embodiment of the application discloses a UAV cluster behavior mapping method based on communication and perception information fusion. By fusing communication and perception information, abstract representation is carried out to obtain representation information, and the optimal cluster behavior of the maximum sum is obtained by weighted summation combined with the information weight of each cluster behavior; the output of the basic behavior rule of the UAV is weighted and summed based on the rule weight of the optimal cluster behavior to obtain a decision output, and the decision output is converted into a desired heading. The application not only realizes the complementation of advantages and disadvantages of communication and perception information, and is applicable to fixed-wing UAVs which cannot effectively obtain perception information due to high flight speed and long inter-aircraft distance when the UAV cluster communication is disturbed, but also realizes that when the cluster size is expanded, only the weight parameter corresponding to the task target needs to be set, and the cluster can autonomously emerge the cluster behavior conforming to the task expectation, i.e. the UAV cluster is completely autonomously controlled, and the behavior mapping has high expansibility, can increase or delete the basic behavior of the cluster UAV according to the task complexity and other needs, so that the behavior mapping process according to the fused information is more targeted, and the task target is better completed.

[0159] In the specification, each embodiment is described in a progressive manner, and each embodiment mainly explains the difference from other embodiments, and the same or similar parts of each embodiment can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the related parts can be referred to the method part.

[0160] ​The foregoing description of the disclosed embodiments enables a person skilled in the art to make or use the application. Modifications of these embodiments will occur to persons of skill in the art, and that the appended claims are intended to cover all such modifications that do not depart from the true spirit and scope of the application. Therefore, the application is not limited to the embodiments shown but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for mapping behaviors of a UAV swarm based on fusion of communication and perception information, characterized in that, The application relates to a decision-making model for unmanned aerial vehicles (UAVs) in a swarm mode, comprising the following steps: Step (1): establishing a communication model and a perception model of the UAV, and fusing the two models to obtain friendly UAV information, obstacle information and ground target information; Step (2): designing basic behavior rules of the UAV based on the friendly UAV information, the obstacle information and the ground target information; Step (3): setting information weights and rule weights of each swarm behavior; characterizing the friendly UAV information, the obstacle information and the ground target information to obtain characteristic information, combining the information weights of each swarm behavior to obtain an optimal swarm behavior, and performing weighted summation on the output of the basic behavior rules of the UAV based on the rule weight of the optimal swarm behavior to obtain a decision output, and finally converting the decision output into a desired waypoint. 2.The method of claim 1, wherein, In step (1), the communication model is as follows: wherein, is the state of the UAV U i ; p i and v i are the position and velocity of the UAV U i in a two-dimensional Cartesian coordinate system, respectively; is the neighborhood of the UAV U i , consisting of friendly UAVs within the communication range of the UAV U i ; U k is a friendly UAV within the communication range of the UAV U i ; U is a set of UAVs; p k is the position of the UAV U k in a two-dimensional Cartesian coordinate system; is the communication range of the UAV U i , which is a circle with the UAV U i as the center; is the complete state information of the friendly UAV U i within the communication range of the UAV U k obtained by the UAV U i ; v k is the velocity of the UAV U k in a two-dimensional Cartesian coordinate system. 3.The method of claim 2, wherein, In step (1), the perception model comprises: The friendly UAV state information is as follows: in, For UAV U i The set of friendly drones detected; U k For UAV U i Sensed friendly drones; p k For UAV U k Position in a two-dimensional Cartesian coordinate system; For UAV U i The perception range is the field of view angle θ and the perception radius d p The fan shape; For UAV U i Sensed friendly drone U k Status information; α i，k For friendly drones U k In UAV U i Relative angle within the field of view; ψ k For friendly drones U k The heading of is expressed as a vector; The ground target state information is as follows: wherein is the unmanned aerial vehicle U i a set of perceived ground targets T j is the unmanned aerial vehicle U i a perceived ground target T is a set of ground targets T j is the ground target T j a position in a two-dimensional Cartesian coordinate system; is the unmanned aerial vehicle U i a perceived ground target T j state information of the perceived ground target T i，j is the ground target T j a relative angle within a field of view of the unmanned aerial vehicle U i a relative angle within a field of view of the unmanned aerial vehicle U The obstacle state information is as follows: wherein is the UAV U i a set of perceived obstacles; O z is the UAV U i a perceived obstacle; O is a set of obstacles; p z is an obstacle O z a position in a two-dimensional Cartesian coordinate system; is the UAV U i a perceived obstacle O z state information; a i，z is an obstacle O z a relative angle within a field of view range of the UAV U i a relative angle within a field of view range of the UAV U 4.The method of claim 3, wherein, In step (1), the friendly UAV information is as follows: wherein, is a friendly drone U k In the drone U i distribution within the neighborhood; is a friendly drone U k motion trend; The ground target information is as follows: p j = F(p i , a i，j , p k , a k，j ); wherein, α k，j is the relative angle of the ground target T j within the field of view of the UAV U k ; F represents the information fusion process; is the set of ground targets obtained by information fusion; is the set of ground targets perceived by the UAV U k ; χ T is the ground target information obtained by information fusion; The obstacle information is as follows: P z = F(p i , a i，z , p k , a k，z ); wherein α k，z is the obstacle O z in the field of view of the UAV U k at the relative angle; is the set of obstacles obtained by information fusion; is the set of obstacles perceived by the UAV U k χ 0 is the obstacle information obtained by information fusion.

5. The method of claim 4, wherein, In step (2), the basic behavior rules of the UAV comprise: Formation rules: make the flight direction of the drones U i and the drones U i in the neighborhood of the friendly drones U k tend to be consistent, as follows: Gathering rules: Make drones U i UAV i Friendly drones in the vicinity U k The cluster as a whole shows a trend of center aggregation, as shown below: Dispersion rules: making drones U i UAV i Friendly drones in the vicinity U k The geometric center of the cluster flies in the opposite direction, and the cluster as a whole shows a trend of rapid dispersion, as shown below: Collision avoidance rules: cause the drone U i to keep a certain distance from the drone U i friendly drones U k in its neighborhood, as follows: Among them, d i，k For UAV U i With UAV i Friendly drones in the vicinity U k distance; c4 is a constant; Obstacle avoidance rules: cause the drone U i to maintain a safe distance from perceived obstacles, as follows: Proximity rule: cause the drone U i to fly towards the direction of the discovered ground target, as follows: Away from target rule: cause the drone U i to fly in the opposite direction of the discovered ground target, as follows: 6.The method of claim 5, wherein, In step (3), the information weights and rule weights of each swarm behavior are set as follows: The s-th cluster behavior By a set of information weights Q s With a set of rule weights W s Indicated as follows: If there are e kinds of swarm behaviors, the decision model q is as follows: 7.The method of claim 6, wherein, In step (3), the characteristic information is as follows: wherein b i，1 is used to represent the heading of the UAV U i and the angle between the heading of the UAV U i and the heading of the friendly UAVs U k in the neighborhood of the UAV U i is used to represent the difference between the motion trend of the UAV U i and the motion trend of the friendly UAVs U k in the neighborhood of the UAV U i，2 is used to represent the distribution of the friendly UAVs U i in the neighborhood of the UAV U k is used to represent the compactness or the combing degree of the friendly UAVs U i around the UAV U i，3 and b i，4 respectively represent whether the ground targets and obstacles are perceived. 8.The method of claim 7, wherein, In step (3), the optimal swarm behavior obtained by combining the information weights of each swarm behavior is as follows: The information weight of each cluster behavior is weighted and summed with the characterization information, and the cluster behavior corresponding to the maximum of the sum is taken as the optimal cluster behavior As follows: 9.The method of claim 7, wherein, In step (3), the decision output obtained by performing weighted summation on the output of the basic behavior rules of the UAV based on the rule weight of the optimal swarm behavior is as follows: wherein, is a decision output representing a desired heading of the UAV U i . 10.The method of claim 9, wherein, In step (3), after obtaining the decision output, the desired waypoint is converted as follows: Wherein, the meaning of extend function is to extend the current position p i of the unmanned aerial vehicle U i as the starting point, in the direction of the desired heading , and the intersection point of the extended line and the boundary of the task area is the desired waypoint Wp i .

Images