Method for designing target interception configuration of unmanned aerial vehicle group imitating eagle group hunting behavior

Abstract

The invention discloses an eagle-group hunting-behavior-imitating unmanned aerial vehicle group target interception configuration design method. The method is characterized by comprising the following steps that 1, the states of unmanned aerial vehicles and a target are initialized; 2, determining an unmanned aerial vehicle task subgroup based on an eagle group energy model; step 3, an unmanned aerial vehicle task subgroup control structure simulating an eagle group hierarchical structure; 4, formation of an interception configuration imitating eagle group pursuing interception; 5, designing an interception guidance law; step 6, equivalent control instruction conversion; and 7, the unmanned aerial vehicle cluster interception process data is displayed. According to the method, strategies such as an energy model, a hierarchical structure and pursuing and intercepting in eagle-group cooperative hunting are used for reference, so that the problem of cooperative intercepting configuration design of a current unmanned aerial vehicle cluster on a moving target in a three-dimensional space is solved, and the dynamic adaptive capacity of the unmanned aerial vehicle cluster on the target in a complex environment is improved; the execution efficiency and adaptability of the collaborative interception task are further improved, and the interception configuration can be kept for a long time.

Description

Technical Field

[0001] This invention relates to a method for designing a drone swarm target interception configuration that mimics the hunting behavior of eagle flocks, belonging to the field of drone cooperative interception and control technology. Background Technology

[0002] Swarm cooperative interception refers to the construction of a swarm collaborative control method through autonomous coordination among individuals within a swarm, enabling rapid response and efficient interception of moving targets. Its core lies in the distributed structure of the swarm, the formation and maintenance of dynamic expected interception configurations, overcoming the limitations of traditional single-point interception. As an important branch of this technology, UAV swarm cooperative interception utilizes a swarm of multiple UAVs, combining cooperative detection, intelligent decision-making, and distributed strike capabilities to form an asymmetric combat advantage, thereby improving overall combat effectiveness and achieving effective interception of incoming targets. The multi-UAV cooperative interception approach has two significant advantages: first, UAVs share current status and battlefield information through communication, improving target interception capabilities; second, each UAV can restrict target movement within a certain azimuth, thus significantly increasing the interception rate of mobile targets and improving the probability of successfully capturing designated targets. With technological advancements and increasing demands, the development direction of UAV swarm cooperative interception technology focuses on the construction of intelligent collaborative decision-making systems. UAV swarm cooperative interception operations involve technologies from multiple fields, including cooperative detection, cooperative guidance, and formation flight control, among other key technologies. The design of cooperative guidance laws for multiple UAVs in the terminal guidance phase is one of the most important technologies in cooperative interception.

[0003] As a typical apex predator, eagles typically possess exceptional hunting abilities, exhibiting highly adaptable and cooperative hunting behaviors. After locking onto prey, eagles employ various methods, including high-speed swoops, to launch attacks and capture, demonstrating highly effective guidance. In situations where prey resources are scarce or the prey is large and highly agile, certain eagle species, such as the Harris Eagle, communicate through calls and flight postures, showcasing various cooperative hunting strategies, including leader rotation, encirclement and interception, and coordinated driving. These strategies allow eagle flocks to intercept targets within a smaller area before capturing them, significantly increasing the success rate. The collective intelligence of eagle flocks not only enhances hunting efficiency but also reduces individual risk, making them a model of survival in nature. The cooperative hunting behavior of eagle flocks provides important inspiration and technical reference for the configuration design of UAV swarm interception and control problems in three-dimensional space, especially in mission scenarios requiring efficient multi-aircraft collaboration, demonstrating significant research and application value. Summary of the Invention

[0004] This invention addresses the problem of designing a swarm-based cooperative interception configuration for aerial targets by unmanned aerial vehicle (UAV) swarms. It provides a method and implementation process for designing a UAV swarm target interception configuration that mimics the hunting behavior of eagle flocks. The proposed method can be applied to scenarios such as active territory defense and interception and destruction of enemy targets, aiming to solve the current problem of designing cooperative interception configurations for moving targets by UAV swarms in three-dimensional space. By drawing on energy models, hierarchical structures, and encirclement and interception strategies from cooperative hunting by eagle flocks, the proposed method improves the dynamic adaptability of UAV swarms to targets in complex environments, further enhancing the efficiency and adaptability of cooperative interception missions, and ensuring that the interception configuration can be maintained long-term.

[0005] A method for designing a drone swarm target interception configuration that mimics the hunting behavior of eagle flocks, the process of which is as follows: Figure 1 As shown, the specific implementation steps are as follows: Step 1: Initialize the UAV and target status Initialize the initial state values ​​of individuals in the UAV swarm, including type, position, velocity, angle, and mission status; initialize the initial state values ​​of the target, including type, position, velocity, and target maneuver strategy; initialize the desired configuration constraints, etc. Initialize the data recording module, including the UAV model, motion constraints, target motion characteristics, total simulation time, step size, etc., and initialize data storage. The initial mission status is "idle".

[0006] Step 2: Determining the UAV mission subgroup based on the Eagle Flock Energy Model In the early stages of cooperative hunting, eagle flocks spontaneously form hunting groups led by a alpha eagle. During the hunt, if other individuals join and the flock size exceeds a certain threshold, it will spontaneously split into smaller sub-flocks. The determination of drone mission sub-flocks based on an eagle flock energy model is as follows: Which individuals participate in mission sub-flocks targeting specific objectives are determined.

[0007] Based on the requirements of the interception mission, only certain types of drones in a swarm (such as those with large communication range and strong detection capabilities) can act as the lead drone, thereby achieving effective monitoring of the swarm and the target, playing a role similar to that of a hawk. When a detectable drone in the swarm detects a target, the lead drone closest to the swarm will bind to the target, ensuring the formation of subsequent sub-swarms and subsequent interception missions. The lead drone first determines the size of the task sub-swarm based on the target type through table lookup or command from the control center. The goal of task subgroup formation is to select individuals from a drone swarm to form a task subgroup of a specified size. This invention focuses on the problem of drone swarm interception of a target swarm. By determining task subgroups for specific targets, the swarm-to-swarm interception is decomposed into subgroup interception of a single target, thus simplifying the complex problem. Accordingly, this invention presents a task subgroup formation process based on the eagle swarm energy model.

[0008] The eagle flock energy model needs to consider the energy gain and cost of individuals, and the following is designed to represent the benefit value of an individual participating in the interception of the current target: (1) in, To obtain positive benefits from participating in the interception mission against this target, in relation to one's own type Target type Group size Energy can be obtained Expected level of participation Interception success rate Related; The cost of participating in an interception mission, and the relative distance Self-type Target type Expected resource expenditure Target Threat Level Related. and The optional calculation formulas are as follows: (2) in, , , For configurable coefficients, , , , The reward is obtained by defining a function or looking up a table, depending on the type of the target or the target itself. Long machines already bound to the target will continuously collect the participation rewards of individuals in the cluster whose task status is "idle" for that target. After a certain period of waiting and updating, the rewards will be sorted and obtained from the top performers. Each drone is an individual participant in the task subgroup. The lead drone then issues the task subgroup formation results, and the participating individuals change their task status to "assigned".

[0009] Specifically, for swarms composed of multiple types of drones, when a task subgroup requires drones of different types, the processing is based on type; that is, for each type... Size of the subgroup in demand The corresponding lead machine collects the profit value of this type of individual and selects the top performers. Frame (when the type is long machine, select the front) shelf).

[0010] Step 3: Control Structure for Unmanned Aerial Vehicle Mission Subgroups Based on Eagle Swarm Hierarchy Eagle flocks typically exhibit a linear, hierarchical group dominance hierarchy during their breeding cycle, defining the roles and dominance relationships of individuals within the flock during hunting. There are three status levels within an eagle flock: Alpha, Beta, and Gamma. Within each level, females are larger and have a degree of dominance over males of the same level. Drawing inspiration from this hierarchical structure of eagle flocks, the following task subgroup control structure for a drone swarm is designed.

[0011] S31. Subgroup Hierarchical Design Based on Eagle Flock Hierarchy. In the hierarchical relationship of an eagle flock, individual eagles are subordinate to individuals of higher rank but have leadership over members of lower rank. This hierarchical relationship is fixed within a breeding cycle, and individuals will only occupy one rank. Alpha and Beta rank members are usually adults, while Gamma rank members may be juveniles. A typical eagle flock consisting of four individuals usually includes an adult female Alpha, an adult male Alpha, an adult male Beta, and an immature male Gamma. Based on this characteristic of the eagle flock hierarchy, the following UAV swarm control hierarchy is designed.

[0012] To enhance the control center's management effectiveness over the swarm system, the control center is defined as the Alpha level. Entities at this level that can directly command the swarm, such as operators, ground control stations, and the master drone, are given the highest decision-making authority and direct control over individual drones. This ensures the system possesses multi-task collaborative execution capabilities and can more flexibly handle complex interception tasks. The Beta level consists of the lead drone, the core entity that coordinates the behavior of slave drones, assists in forming interception configurations, and determines interception requirements during the swarm's autonomous task execution. Specifically, task phases are defined based on the characteristics of the specific interception target, such as long-range tracking, configuration formation, and close-range interception phases. Referring to the unchanging nature of the eagle flock's hierarchical structure during its breeding cycle, different lead drones are configured for different task phases based on requirements and individual capabilities. Furthermore, to mitigate the impact of lead drone failure on the mission, several similar UAVs are also configured at the Beta level, but are controlled by the lead drone and have similar functions, capable of replacing the lead drone in case of damage or failure. Specifically, the lead aircraft maintains a hierarchy table to record the hierarchy of each individual in the current task subgroup, avoiding hierarchical management chaos. Drawing inspiration from the alpha eagle rotation mechanism in eagle hunting behavior, a transfer rule between the Gamma and Beta hierarchies is constructed, and an individual performance value is designed: (3) in, For the resources held, This represents the current tier. Only drones that meet the type and possess Beta-level capabilities can be transferred to a higher tier, and only those with high efficiency values ​​can be transferred, allowing the tier structure to be adjusted under specific circumstances. The remaining drones form the Gamma tier, which is responsible for interception missions. They possess relatively flexible maneuverability and can effectively intercept and threaten targets within a certain radius spherical area, thereby deterring targets. Multiple Gamma-level drones form an interception configuration that confines the target within this area, achieving coordinated interception. Specifically, when dealing with complex interception missions, the Gamma tier can be further divided into "male" and "female" drones, resulting in different interception effects on the target. Alternatively, "female" drones in the Gamma tier can achieve cross-tier identity conversion through certain rules, enhancing the overall swarm's resilience and mission execution capabilities.

[0013] S32. Subgroup Communication Topology Design Based on Henneberg Construction. After determining the hierarchical structure of the UAV swarm, a communication topology that meets connectivity and hierarchical requirements can be designed to reduce unnecessary communication links. Starting from the Alpha level and the lead UAV, the communication topology of the task subgroup is constructed using Henneberg's vertex addition method: for each Beta level individual added, two communication links are added: one with the Alpha level and one with the previous Beta level individual; for each Gamma level individual added, two communication links are added: one with the nearest Beta level individual and one with the previous Gamma level individual. Specifically, for the first added Gamma level individual, two communication links are added: one with the nearest Beta level individual and one with the Alpha level. For transitions between levels, individuals are renumbered according to their performance values, and then the communication topology is formed again using Henneberg's vertex addition method. The resulting structure guarantees connectivity.

[0014] Step 4: Formation of an interception configuration mimicking a flock of eagles encircling and blocking the enemy. After locking onto prey, eagles exhibit cooperative behavior, forming a hunting formation to encircle and intercept the target. This formation is often dynamic, centered on the prey's trajectory, restricting escape and effectively confining the prey's range within the formation. As the cooperative hunt progresses, the eagles gradually reduce the formation until it is limited to a small area advantageous to the flock. Drawing inspiration from this cooperative hunting behavior of eagles, the following process is designed to illustrate the formation of an interception configuration for a mobile target by a drone swarm.

[0015] S41. Design of Desired Orientation Constraints for Flock Encirclement and Interception of UAVs. Combining orientation rigidity theory and subgroup communication topology, orientation constraints between neighbors in a UAV subgroup are designed so that the subgroup configuration can be determined solely based on orientation. The orientation between UAVs is defined as: (4) in, , drones , position vector, Indicates drone The neighbor set. Furthermore, the desired orientation constraint is given in advance based on the interception requirements, and the target's position in the desired interception configuration can be determined using the lead aircraft's orientation constraint relative to the target. It means that the distance is determined by Confirmed. Specifically, This is related to the relative situational relationship between the lead aircraft and the target, as well as the interception and detection requirements. For example, to improve detection advantage, the elevation angle of the lead aircraft relative to the target is... azimuth angle is Thus, the lead aircraft consistently maintains a position above and behind the target, continuously aiming at it. At this time, ,in For the desired configuration constraints between the lead aircraft and the target aircraft in the standard desired configuration, and Since the subgroup communication topology constructed using the Henneberg construction method already satisfies the orientation rigidity requirement, as long as the orientations of neighbors within the subgroup satisfy the specified desired orientation constraints... The shape of the subgroup configuration can then be uniquely determined. Therefore, the goal of configuration formation is transformed into: (5) Specifically, considering the characteristics of eagle flocks' pursuit and interception behavior, directional constraints are expected. The resulting configuration is scalable, and the scaling factor is... The intercept configuration can be uniquely determined by the lead aircraft. In other words, the lead aircraft can adjust its distance from the target according to interception requirements, and the size of the intercept configuration is also adjusted by the lead aircraft in this process, thereby continuously forming the desired configuration, such as... Figure 2 As shown. Furthermore, combining this with the theory of azimuth rigidity, from... and The desired orientation constraint of an individual relative to a target can be obtained using vector addition. ,in This is a constraint on the expected orientation of an individual relative to the lead aircraft.

[0016] S42. Target State Cooperative Observation. Within the task subgroup, some individuals may lack target detection capabilities or have sensor malfunctions, preventing them from directly acquiring target information. Therefore, a target state cooperative observer is designed to assist subgroup members in acquiring target state information in real time through communication interaction. In this invention, different target state cooperative observer designs are used for individuals that have acquired target information. For individuals that can detect the target, the target state cooperative observer is designed as follows: (6) in, Indicates drone Individuals of the target can be detected among its neighbors. , For the observed target's position and velocity information, and These are the observations of the target state. For individuals where the target cannot be detected, the target state cooperative observer is designed as follows: (7) in, , , , , For observer parameters, the function ,and

[0017] S43, Desired Orientation Constraint Transformation. From S41, the desired orientation constraint of the individual relative to the target in the current situation can be obtained as follows: This describes the relative orientation information of an individual within the target's local coordinate system. Observations of the target's position can be obtained from S42. and Information from the homing aircraft can be directly obtained from communication interactions. and Therefore, the desired azimuth constraint can be transformed into a desired azimuth angle relative to the target in the line-of-sight coordinate system. Desired elevation angle and expected distance The specific formula is as follows: (8) in, This refers to the distance between the individual aircraft and the target when the distance between the lead aircraft and the target is one unit length, and when the desired configuration is achieved. , , These are the actual azimuth, actual elevation, and actual distance of the individual relative to the target.

[0018] At this point, the interception configuration of the task subgroup is transformed into the orientation and distance relationship of the individual relative to the target. It is only necessary to guide the UAV to satisfy these three relationships to achieve the interception configuration of the target. Similar to the encirclement and interception behavior of a flock of eagles, the scaling factor of the configuration is controlled by the lead aircraft, and the configuration change requirements can be controlled by giving relevant instructions to the lead aircraft.

[0019] Step 5: Design of the Interception Guidance Law The design goal of the interception guidance law is to ensure that: (9) At this point, numerous guidance laws are available. Based on the relative motion equations in three-dimensional space, a control law is designed to satisfy equation (9) to achieve guidance control, and the obtained control quantity is the desired three-axis acceleration value of the UAV. In particular, this invention does not limit the specific guidance control law, and this is not the focus of this invention.

[0020] Step Six: Equivalent Control Command Conversion Step five yields the desired three-axis acceleration values ​​for the individual UAV. Optionally, for system demonstration, a control command conversion function is used to convert acceleration-type control commands into equivalent overload-type control commands for the UAV. (10) in, It is an overload type control command, and , , For the tangential overload, normal overload and roll angle of the UAV.

[0021] After conversion, the overload control command can be directly used as input to the UAV motion model to obtain the next UAV state. If the termination condition is not met, jump to step two and continue execution; if the termination condition is met, jump to step seven. The termination condition is determined according to the requirements. Common termination conditions include reaching the final desired configuration and maintaining it for a specified time, and reaching the maximum running time.

[0022] Step 7: Data display of the drone swarm interception process During the operation of the method proposed in this invention, the data recording module continuously records necessary information such as the desired configuration result, control commands, UAV swarm status, and target status. Optionally, during demonstration, the data recording module can process the recorded data to display information such as changes in inter-UAV distance and actual interception configuration changes. After the proposed method finishes running, specific data from the entire process can be displayed, showing the UAV swarm interception control results and the configuration formation effect under the current settings.

[0023] This invention provides a method for designing a target interception configuration for a drone swarm that mimics the hunting behavior of eagle flocks. Its advantages and effects are as follows: by drawing on the energy model, hierarchical structure, and encirclement and interception strategies in cooperative hunting by eagle flocks, it aims to solve the problem of designing a cooperative interception configuration for moving targets in three-dimensional space for current drone swarms, improve the dynamic adaptability of drone swarms to targets in complex environments, and maintain the interception configuration for a long time. Attached Figure Description

[0024] Figure 1 A flowchart illustrating the configuration design for drone swarm target interception mimicking eagle hunting behavior.

[0025] Figure 2 This is a schematic diagram illustrating the configuration changes of a simulated encirclement and interception formation.

[0026] Figure 3 A schematic diagram of a subgroup structure design for an eagle-like flock.

[0027] Figure 4 Simulation diagram of the collaborative interception configuration.

[0028] Figure 5 This is a simulation diagram showing the change in distance between the UAV and the target during a coordinated interception process.

[0029] Figure 6 Simulation diagram of the coordinated interception trajectory.

[0030] The labels and symbols in the diagram are explained as follows: , , —The position of the UAV in the ground coordinate system Detailed Implementation

[0031] The effectiveness of the proposed method for designing a drone swarm target interception configuration that mimics eagle hunting behavior will be verified through specific examples below.

[0032] Step 1: Initialize the UAV and target status Initialize the initial state values ​​of individuals in the UAV swarm, including type, position, velocity, angle, and mission status; initialize the initial state values ​​of the target, including type, position, velocity, and target maneuver strategy; initialize the desired configuration constraints, etc. Initialize the data recording module, including the UAV model, motion constraints, target motion characteristics, total simulation time, step size, etc., and initialize data storage. The initial mission status is "idle". In this example, the UAV model adopts a three-freedom overload model, i.e.:

[0033] in, and Let U be the position and velocity of the UAV in the ground coordinate system, and The magnitude of the speed; For the UAV, the Euler angles are pitch, yaw and roll, respectively. It is the acceleration due to gravity; , This refers to tangential and normal overloads for the UAV. The overload control command is defined as follows: .

[0034] Motion constraints can be expressed as:

[0035] Among them, subscript , Describe the minimum and maximum allowable values ​​for the corresponding quantities. In this example, take... .

[0036] The target's maneuverability characteristics are set as follows:

[0037] The maximum simulation time is 140 seconds, with a step size of 0.02 seconds. In this example, consider a target with... Unmanned aerial vehicles (UAVs) are divided into two categories: one type has detection capabilities and is denoted as... One type can serve as the lead aircraft; the other type lacks detection capabilities and is designated as... .

[0038] Step 2: Determining the UAV mission subgroup based on the Eagle Flock Energy Model The task subgroup determination of UAVs based on the eagle flock energy model is presented. The process for determining which individuals participate in a task subgroup targeting a specific objective is described below, and the task subgroup formation process based on the eagle flock energy model is described. The size of the task subgroup acquired by the lead aircraft at this point is also described. Two drones of each type. (To be continued) The formula for calculating the benefit value of an individual participating in the interception of the current target is:

[0039] in, , , , , , At this point, the gains are mainly related to distance and drone type, with individuals closer to the drone receiving greater gains.

[0040] The lead drone, already bound to the target, collects the participation benefits of individuals in the cluster whose task status is "idle" for that target. After a certain period of waiting and updating, it sorts and selects drone individuals as participants in the task subgroup. The lead drone then issues the task subgroup formation results, and the participating individuals change their task status to "assigned".

[0041] Step 3: Control Structure for Unmanned Aerial Vehicle Mission Subgroups Based on Eagle Swarm Hierarchy Drawing inspiration from the hierarchical structure of eagle flocks, the following task subgroup control structure for UAV swarms is designed.

[0042] S31. Sub-swarm hierarchy design based on eagle-swarm hierarchical structure. Based on the characteristics of the eagle-swarm hierarchy, the following UAV swarm control hierarchy is designed. The operator, ground control station, and mother aircraft—control centers capable of directly commanding the swarm—are defined as the Alpha level. The Beta level consists of the lead aircraft and another UAV. The drone, Gamma level is The two drones. The lead drone maintains a hierarchy table to record the hierarchy of each individual in the current task subgroup, preventing hierarchy management chaos. Transfers between Gamma and Beta hierarchies are not allowed.

[0043] S32. Subgroup Communication Topology Design Based on Henneberg Construction. Starting from the Alpha level and the lead machine, the communication topology of the task subgroup is constructed using Henneberg's vertex addition method: For each Beta level individual added, two communication links are added: one with the Alpha level and one with the previous Beta level individual; for each Gamma level individual added, two communication links are added: one with the nearest Beta level individual and one with the previous Gamma level individual. Specifically, for the first added Gamma level individual, two communication links are added: one with the nearest Beta level individual and one with the Alpha level. The resulting communication topology is as follows: Figure 3 As shown.

[0044] Step 4: Formation of an interception configuration mimicking a flock of eagles encircling and blocking the enemy. Drawing inspiration from the cooperative hunting behavior of eagles encircling and intercepting targets, the following process is designed to form an interception configuration of a drone swarm against a mobile target.

[0045] S41. Design of desired orientation constraint for eagle-like encirclement and interception. In this example, the desired configuration is a regular tetrahedron, restricting the target to the center position, thus obtaining the standard desired configuration (target at the origin, lead aircraft at...). Desired configuration constraints at time ) The desired angle of elevation of the lead aircraft relative to the target is... azimuth angle is Thus, the lead aircraft tends to remain above and behind the target and continuously aim at it. Then it can be done and It is expressed. At this time, , As long as the orientations of neighbors in a subgroup satisfy the specified desired orientation constraints. The shape of the subgroup configuration can then be uniquely determined. Furthermore, combining this with the theory of orientation stiffness, from... and The desired orientation constraint of an individual relative to a target can be obtained using vector addition. ,in This is a constraint on the expected orientation of an individual relative to the lead aircraft.

[0046] S42. Target State Cooperative Observation. For individuals capable of detecting targets, the target state cooperative observer is designed as follows:

[0047] in, Indicates drone Individuals of the target can be detected among its neighbors. , For the observed target's position and velocity information, and These are the observations of the target state. For individuals where the target cannot be detected, the target state cooperative observer is designed as follows:

[0048] in, , , , , ,function ,and:

[0049] S43, Desired Orientation Constraint Transformation. From S41, the desired orientation constraint of the individual relative to the target in the current situation can be obtained as follows: This describes the relative orientation information of an individual within the target's local coordinate system. Observations of the target's position can be obtained from S42. and Information from the homing aircraft can be directly obtained from communication interactions. and Therefore, the desired azimuth constraint can be transformed into a desired azimuth angle relative to the target in the line-of-sight coordinate system. Desired elevation angle and expected distance The specific formula is as follows:

[0050] in, Let be the distance between the individual and the target when the distance between the lead aircraft and the target is 1 and the desired configuration is achieved. , , These are the actual azimuth, actual elevation, and actual distance of the individual relative to the target.

[0051] At this point, the interception configuration of the task subgroup is transformed into the orientation and distance relationship between the individual drones and the target. By guiding the drones to satisfy these three relationships, the interception configuration can be achieved, and the scaling factor of the configuration is controlled by the lead drone. In this example, the configuration size variation requirement is designed as follows:

[0052] Step 5: Design of the Interception Guidance Law The design goal of the interception guidance law is to ensure that:

[0053] In this invention, a guidance control law based on a sliding mode surface is used to obtain the desired three-axis acceleration values ​​of an individual UAV.

[0054] Step Six: Equivalent Control Command Conversion Step five yields the desired three-axis acceleration values ​​for the individual UAV. Optionally, for system demonstration, a control command conversion method is used to convert acceleration-type control commands into equivalent UAV overload-type control commands, i.e.:

[0055] in, It is an overload type control command, and , , For the tangential overload, normal overload and roll angle of the UAV.

[0056] After the conversion, the overload control command can be directly used as input to the UAV motion model to obtain the next UAV state. If the termination condition is not met, jump to step two and continue execution; if the termination condition is met, jump to step seven. In this example, two termination conditions are selected: reaching the final desired configuration and maintaining it for 5 seconds, and reaching the maximum simulation time. The process ends when either one is met.

[0057] Step 7: Data display of the drone swarm interception process During the operation of the method proposed in this invention, the data logging module continuously records necessary information such as the desired configuration result, control commands, UAV swarm status, and target status. Optionally, during demonstration, the data logging module can process the recorded data to display information such as changes in inter-UAV distance and actual interception configuration changes. After the proposed method finishes running, specific data from the entire process can be displayed, showing the UAV swarm interception control results and the configuration formation effect under the current settings. For this example, Figure 4 This shows the cooperative interception configuration at a certain moment. Figure 5 The process demonstrated the change in distance between the drone and the target. Figure 6 For coordinated interception trajectories. These results demonstrate the effectiveness of the proposed method, enabling the design, formation, and maintenance of configurations.

Claims

1. A method for designing a UAV swarm target interception configuration imitating the hunting behavior of an eagle swarm, characterized in that: The method comprises: Step 1: initializing the unmanned aerial vehicle and target state; Step 2: determining the unmanned aerial vehicle task subgroup based on the eagle swarm energy model; The determination of the unmanned aerial vehicle task subgroup based on the eagle swarm energy model determines which individuals participate in the task subgroup for a specific target according to the interception yield value of the eagle swarm energy model; Step 3: the unmanned aerial vehicle task subgroup control structure based on the eagle swarm hierarchical structure, comprising: S31, subgroup hierarchical design based on the eagle swarm hierarchical structure; S32, subgroup communication topology design based on Henneberg construction; Step 4: interception configuration formation of the eagle swarm pursuit and interception, comprising S41, expected direction constraint design of the eagle swarm pursuit and interception; S42, target state cooperative observation; S43, expected direction constraint conversion; Step 5: interception guidance law design, obtaining the three-axis acceleration value expected by the unmanned aerial vehicle individual; Step 6: equivalent control instruction conversion; acceleration type control instruction is converted into unmanned aerial vehicle overload type control instruction by using control instruction conversion; Step 7: unmanned aerial vehicle cluster interception process data display.

2. The method of claim 1, wherein: The specific process of step 2 is as follows: The determination of the unmanned aerial vehicle task subgroup based on the eagle swarm energy model determines which individuals participate in the task subgroup for a specific target according to the following method: When the individuals in the UAV cluster discover the target, the long-range aircraft closest to the cluster will be bound to the target, ensuring the subsequent formation of subgroups and subsequent interception tasks; the long-range aircraft first determines the task subgroup size according to the target type through table lookup or control center issuance ; the target of the task subgroup formation is to select individuals from the UAV cluster to form a task subgroup of a specified size; the task subgroup formation process based on the hawk group energy model; The individual energy acquisition and cost need to be considered in the eagle swarm energy model, and the following represents the individual's yield value participating in the interception of the current target:

3. Wherein, The positive benefit that can be obtained by participating in the interception task of this target is related to the type of the self , the type of the target , the group size , the obtainable energy , the expected participation level , and the interception success rate ; The cost required for participating in the interception task is related to the relative distance , the type of the self , the type of the target , the expected amount of resources to be paid , and the threat level of the target ; And The calculation formula is as follows: ； wherein, , , are configurable coefficients, , , , then according to the type of itself or target by defining function or look-up table to obtain; the long machine bound with the target will continuously collect the participation benefits of the individual with the task state of "idle" in the cluster to the target, and after waiting and updating, the front is obtained by sorting The individual of the aerial vehicle as the task subgroup participant; the long machine issues the task subgroup construction result, and the participating individual changes the task state to "has been allocated".

4. The method of claim 2, wherein: For the cluster composed of multiple types of UAVs, when the task sub-group has demand for different types of UAVs, then the processing is carried out according to the type, that is, for the type , the size of the sub-group having demand , the corresponding long plane collects the income value of the individuals of this type, and selects the top , when the type is a long plane, the top is selected.

5. The method of claim 1, wherein: The specific process of the subgroup hierarchical design based on the eagle swarm hierarchical structure in step S31 is as follows; The control center is defined as the Alpha level; the operator, the control ground station or the mother machine is at this level, which is the entity directly commanding the cluster, and has the highest decision level position and direct control authority over the individual; the Beta level is the long machine, which is the core individual of the cooperative behavior of the cluster in the process of autonomous task execution, and assists in forming the interception configuration and judging the interception demand; in order to weaken the influence of long machine damage on the task, a number of similar unmanned aerial vehicles are also configured in the Beta level, but are dominated by the long machine, and replace the long machine when the long machine is damaged or fails; the long machine maintains a level table to record the level of each individual in the current task subgroup, avoiding confusion in level management; the transfer rules between Gamma level and Beta level are constructed, and the individual performance value is designed: ； wherein, is the held resource, is the current level; only the drone individuals that meet the type, have the Beta level ability requirement and the high performance value can transfer the level, so that the level structure is adjusted and configured under certain conditions; the remaining drones are formed into the Gamma level, which is the performer of the interception task, effectively intercepts the target within the threat radius spherical area, and in response to complex interception tasks, the Gamma level is further divided into "male" and "female", which form different interception effects on the target, or the Gamma level "female" realizes cross-level identity conversion through rules, enhancing the anti-damage and task execution capability of the entire cluster.

6. The method of claim 4, wherein: According to the characteristics of the specific interception target, the task stage is formulated, including but not limited to the long-distance tracking stage, the configuration formation stage, and the near-distance interception stage; referring to the invariable characteristics of the eagle hierarchical structure in the breeding cycle, different long machines are configured for different stages according to the demand and individual ability difference in different task stages.

7. The method of claim 1, wherein: The specific process of the expected direction constraint design of the eagle swarm pursuit and interception in step S41 is as follows: The direction constraint between neighbors in the unmanned aerial vehicle subgroup is designed so that the configuration of the subgroup can be determined only according to the direction; the direction between unmanned aerial vehicles is defined as: ； where, , are the position vectors of the UAVs , , respectively, denotes the neighbor set of the UAV ; further, the desired orientation constraint is given in advance according to the interception requirement, and the position of the target in the desired interception configuration is denoted by the orientation constraint of the interceptor relative to the target , where the distance is determined by ; and the relative situation relationship between the interceptor and the target, the interception detection requirement; in order to improve the detection advantage, the elevation angle of the interceptor relative to the target is , and the azimuth angle is , so that the interceptor is always inclined to be above the target on the back side and constantly aligned with the target; at this time, , where, is the desired configuration constraint between the interceptor and the target in the standard desired configuration, and ; since the subgroup communication topology built by the Henneberg construction method has already met the azimuth rigidity requirement, as long as the azimuth between the neighbors in the subgroup meets the prescribed desired azimuth constraint , the shape of the subgroup configuration is determined; therefore, the target conversion of the configuration formation is: ； Desired bearing constraint of individual relative to target The configuration formed is scalable, and the scaling factor Is uniquely determined by the lead; from And The desired bearing constraint of individual relative to target is obtained according to vector addition Wherein, The desired bearing constraint of individual relative to lead.

8. The method of claim 1, wherein: The specific process of the target state cooperative observer design in step S42 is as follows: For individuals who can detect the target, the target state cooperative observer is designed as: ； wherein, represents a drone whose neighbor detects a target, , is position, velocity information of the observed target, and is the observation value of the target state; For individuals who cannot detect the target, the target state cooperative observer is designed as: ； wherein , , , , is an observer parameter, the function , and 。

Images