A method for cooperative formation motion decision-making with or without people based on boundary traction strategy

Abstract

A method for manned-unmanned cooperative formation movement decision based on boundary traction strategy belongs to the field of large-scale manned-unmanned cooperative combat. The method comprises: obtaining current situation information and determining manned-unmanned cooperative formation objects; constructing a movement model of heterogeneous unmanned aerial vehicles (UAVs) based on the situation information; determining a dynamic cluster area according to a current task phase based on a set of manned-unmanned cooperative formation nodes and the movement model of the heterogeneous UAVs; performing regional allocation based on the dynamic cluster area and generating a virtual traction point; calculating a boundary traction force based on the virtual traction point; calculating a safety repulsion force based on the Euclidean distance between nodes; integrating control inputs; and updating the movement state of the heterogeneous UAVs based on the integrated control inputs. The proposed autonomous movement decision method for UAVs effectively reduces the frequency of movement control of manned aircrafts on UAVs while ensuring the task execution requirements.

Description

Technical Field

[0001] This invention belongs to the field of large-scale manned-unmanned cooperative combat, and proposes a manned-unmanned cooperative formation motion decision-making method based on boundary traction strategy. Background Technology

[0002] In the field of manned-unmanned cooperative swarm motion control, existing research mainly revolves around two levels: rigid formation control based on preset patterns and bio-inspired flexible swarm behavior modeling. Rigid formation control methods typically assign fixed relative positions or formation templates to UAVs, achieving high-precision formation maintenance in known, static environments. However, these methods heavily rely on precise global information and stable communication topologies. In dynamic adversarial and threat-emerging air combat scenarios, their reconfiguration capabilities are poor, fault tolerance is low, and they cannot adapt to rapidly changing mission requirements and battlefield situations. On the other hand, flexible swarm models based on Reynolds rules (velocity alignment, cohesion, separation) and their derivative methods achieve self-organized motion through local interactions, exhibiting good emergent characteristics in swarm aggregation, collision avoidance, and velocity synchronization. However, existing flexible swarm strategies often assume a fixed movement area or a non-threatening environment, with swarm boundaries typically being preset geometric shapes (such as circles or ellipses), lacking mechanisms for dynamic adjustment based on manned aircraft command intentions, UAV functional roles, and real-time threats. In highly dynamic manned-unmanned collaborative combat missions, this leads to either unmanned aerial vehicle (UAV) swarms becoming overly confined to fixed areas and losing tactical initiative, or being exposed to threats or falling outside the effective command range of manned aircraft due to rigid boundaries. Furthermore, most existing methods treat swarm movement and mission execution as separate processes, failing to integrate the reconnaissance and defense capabilities of UAVs into movement decision-making, resulting in a disconnect between "movement" and "mission," and failing to maximize the overall combat effectiveness of the swarm. Simultaneously, existing strategies often neglect the distribution quality of the swarm within a region; poor uniformity directly impacts search coverage or defense response efficiency. Regarding control burden, most methods still require high-frequency intervention or global planning by manned aircraft, failing to truly achieve high-order autonomous collaboration between humans and aircraft in the loop. Pilot cognitive load remains heavy, hindering the practical application of large-scale heterogeneous UAV formations. Summary of the Invention

[0003] The fundamental objective of this invention is to overcome a series of problems exposed by existing manned-unmanned cooperative swarm motion control in dynamic adversarial environments, such as insufficient adaptability, weak task coordination, and heavy human burden. It proposes a manned-unmanned cooperative formation motion decision-making method based on a boundary traction strategy. This method aims to construct a task-driven, boundary-flexible, uniformly distributed, and autonomously stable swarm motion system, thereby achieving efficient, robust, and intelligent coordination of heterogeneous UAV formations in complex and ever-changing air combat scenarios. The key technical challenges that need to be addressed include: Adaptive reconfiguration of swarm regions in dynamic environments: How to overcome the constraints of fixed regions or formations and design a mechanism that can automatically and in real-time adjust the swarm movement boundaries, number of regions, and spatial configuration based on the real-time flight status of manned aircraft, the number, location, and distance of threat targets, and the functional type of UAVs. This ensures that the swarm layout always matches the tactical phase, maximizing mission execution efficiency and survivability while maintaining the overall integrity of the formation. Autonomous guidance and uniform distribution of UAVs within a region: How to guide UAVs to autonomously and quickly enter and remain in the designated mission area within a dynamically defined, irregular swarm region, avoiding UAVs lingering at the boundary or excessively clustering. Integrated design and stability assurance of swarm movement rules under multiple constraints: How to integrate multiple requirements such as speed alignment, collision avoidance separation between individuals, and boundary maintenance towards the interior of the region, designing a reliable swarm movement rule with theoretical convergence guarantees. This rule must ensure that the UAV swarm can still operate stably as a whole under environments with dynamic traction, random disturbances, and external threats, achieving a multi-objective balance of speed convergence, safe spacing, and region maintenance, while significantly reducing reliance on high-frequency control commands from manned aircraft and improving system autonomy.

[0004] The technical solution of this invention to solve the key technical problem is: To achieve the above objectives, this invention provides a method for unmanned cooperative formation motion decision-making based on a boundary traction strategy. This method executes according to a time-flow sequence: "situational information acquisition and object determination, motion model construction, dynamic cluster area update, area allocation and virtual traction point generation, boundary traction force calculation, safety repulsion force calculation, velocity alignment force calculation, and comprehensive control input generation and motion state update." Subsequent steps all use the data, areas, traction points, or control variables obtained from previous steps as input, ensuring continuous, clear, and repeatable data processing. Its key features include the following steps: S1: Obtain current situation information and determine whether there are any unmanned cooperative formation targets; Based on the current mission phase and situational awareness data, a manned-unmanned cooperative formation is identified to participate in the coordinated motion decision-making. This manned-unmanned cooperative formation includes a master manned aircraft and multiple heterogeneous unmanned aerial vehicles (UAVs), wherein the master manned aircraft is denoted as […]. Heterogeneous drones include search-type and defense-type heterogeneous drones. Indicates the type number of heterogeneous drones. Indicates the search for heterogeneous drones. Indicates defense against heterogeneous drones; This indicates the serial number of a heterogeneous UAV of the same type. The unmanned cooperative formation participating in cooperative motion decision-making is represented as: in, Indicates time A collection of unmanned collaborative formation nodes; Indicates the first Search for heterogeneous drones; Indicates the first Defend against heterogeneous drones; Indicates the number of heterogeneous drones searched; This indicates the number of heterogeneous drones being defended against.

[0005] Based on the aforementioned set of manned and unmanned cooperative formation nodes, a communication topology centered on the manned main control aircraft is constructed: in, Indicates time The communication topology; Indicates time A set of communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment Communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment The Euclidean distance; Indicates that the main control unit is manned. At any moment The position vector; Heterogeneous drones At any moment The position vector; Indicates the maximum command and communication distance of the manned aircraft under main control; This represents the L2 norm.

[0006] The set of unmanned cooperative formation nodes obtained in step S1 Communication topology In addition to the position vectors of each node, the subsequent step S2 constructs a heterogeneous UAV motion model and provides state variables for subsequent dynamic cluster area updates and control input calculations.

[0007] S2: Construct a motion model for the heterogeneous UAV based on the situational information from step S1; Heterogeneous drones The position vector, velocity vector, and virtual integrated control input vector are respectively expressed as: in, , and These represent heterogeneous drones. The three coordinate components in the inertial coordinate system; Heterogeneous drones At any moment The velocity vector; , and These represent heterogeneous drones. The three velocity components in the inertial coordinate system; Heterogeneous drones At any moment The virtual integrated control input vector; , and These represent heterogeneous drones. Three virtual integrated control input components in the inertial coordinate system; superscript This represents the transpose of a vector.

[0008] The simplified motion model of the heterogeneous UAV is as follows: in, express The first derivative with respect to time; express The first derivative with respect to time.

[0009] When it is necessary to convert the virtual integrated control input into the actual input of the heterogeneous UAV, the lift, thrust, and roll angle of the heterogeneous UAV are obtained by the following formula: in, Heterogeneous drones At any moment Lift; Heterogeneous drones At any moment The thrust; Heterogeneous drones At any moment Roll angle; Heterogeneous drones The quality; Represents gravitational acceleration; Heterogeneous drones At any moment Flight path angle; Heterogeneous drones At any moment The heading angle; Heterogeneous drones At any moment Resistance.

[0010] The heterogeneous UAV motion model and state variables obtained in step S2 are followed by a dynamic cluster area update in step S3 based on the current mission stage and threat target information.

[0011] S3: Based on the set of unmanned cooperative formation nodes obtained in step S1 and the motion model of heterogeneous UAVs obtained in step S2, determine the dynamic cluster area according to the current mission stage.

[0012] Dynamic cluster regions are uniformly represented as ,in Indicates the type number of heterogeneous drones. This indicates the dynamic clustering region number corresponding to heterogeneous UAVs of the same type. A dynamic clustering region is defined as follows: in, Indicates time The Type 1 One dynamic cluster area; Represents dynamic cluster area Any position vector within; Represents dynamic cluster area The minimum radius; Represents dynamic cluster area The maximum radius; Representing vectors The direction angle relative to the horizontal axis of the inertial coordinate system; Represents dynamic cluster area The central direction angle; Represents dynamic cluster area The range of angles.

[0013] During the cruise phase, the search for heterogeneous drones corresponds to a search dynamic cluster area. Defending against heterogeneous drones corresponds to a dynamic cluster defense area. The search dynamic cluster area is symmetrically arranged along the flight path of the main manned aircraft; the parameters of the defense dynamic cluster area are updated according to the following formula: in, This indicates the maximum radius of the defense dynamic cluster area during the cruise phase. This represents the minimum radius of the defense dynamic cluster area during the cruise phase. This indicates the angular range of the defense dynamic cluster area during the cruise phase. Indicates the safe distance between heterogeneous drones; This indicates the maximum permissible angle for the defense dynamic cluster area during the cruise phase; , and These represent the maximum radius adjustment, minimum radius adjustment, and angle adjustment of the defense dynamic cluster area during the cruise phase, respectively.

[0014] During the threat target response phase, if detected For each threat target, the dynamic swarm area corresponding to the defense against heterogeneous drones is divided into: One defense dynamic cluster area. For the first One threat target Its corresponding defense dynamic cluster area The parameters are updated using the following formula: in, Indicates the threat target number, and ; Indicate threat target At any moment The position vector; Indicates that the main control unit is manned. With threat targets Euclidean distance; Indicates the defense dynamic cluster area The maximum radius; Indicates the defense dynamic cluster area The minimum radius; Indicates the defense dynamic cluster area Angle range; Indicates allocation to the dynamic defense cluster area The number of defensive heterogeneous drones; and These represent the defense dynamic cluster areas. The maximum radius adjustment amount and the minimum radius adjustment amount; Indicates the defense dynamic cluster area The central direction angle; and These represent the threat targets. The horizontal and vertical coordinates; and These respectively indicate that the main control is a manned aircraft. The horizontal and vertical coordinates; This represents the arctangent function in the four quadrants.

[0015] The dynamic cluster region obtained in step S3 In the subsequent step S4, the heterogeneous drones are assigned to the corresponding dynamic cluster areas, and virtual traction points are generated for each dynamic cluster area.

[0016] S4: Based on the dynamic cluster area in step S3, perform area allocation and generate virtual traction points; Based on the dynamic cluster region obtained in step S3 A region allocation strategy based on a consensus-based binding algorithm is adopted to assign heterogeneous drones to corresponding dynamic cluster regions. For any heterogeneous drone to be assigned... With any dynamic cluster region First, calculate the distance evaluation value from the virtual traction point, then calculate the regional bid value: in, Heterogeneous drones To the dynamic cluster area Corresponding virtual traction point Distance evaluation metric; Heterogeneous drones For dynamic cluster regions Regional bid value; Indicates the quantity demand weight; Indicates distance weight; Represents dynamic cluster area The required first Number of different types of heterogeneous drones; This indicates that the data has been assigned to a dynamic cluster area. The Number of different types of heterogeneous drones; This represents the function that takes the maximum value. Represents dynamic cluster area Corresponding virtual traction point The position vector.

[0017] Based on regional bid value The size of heterogeneous drones Assigned to the dynamic cluster region with the largest bid value, resulting in the region allocation: in, Indicates allocation to dynamic cluster area A collection of heterogeneous drones; This indicates that it is already in the dynamic cluster area. An internal collection of heterogeneous drones; This indicates that it has been allocated to the dynamic cluster area. However, the heterogeneous drone ensemble is still located outside this dynamic cluster area; This indicates the set union operation.

[0018] For the dynamic cluster area of ​​defense during the threat target response phase Its virtual traction point The position vector is: in, Indicates the defense dynamic cluster area Virtual traction point and main control manned machine Euclidean distance; Indicates the defense dynamic cluster area virtual traction point The position vector.

[0019] During the cruise phase, search dynamic cluster areas. Virtual traction points and defense dynamic cluster areas The virtual traction points are as follows: in, Indicates the search for dynamic cluster regions during the cruise phase. virtual traction point The position vector; Indicates the dynamic cluster area for defense during the cruise phase. virtual traction point The position vector; Indicates that the main control unit is manned. At any moment The heading angle.

[0020] The region allocation result obtained in step S4 and virtual traction points The subsequent step S5 calculates the boundary traction force for the heterogeneous UAVs that have been assigned but are still located outside the corresponding dynamic cluster area.

[0021] S5: Calculate the boundary traction force based on the virtual traction point in step S4; Based on the virtual traction point obtained in step S4 and regional allocation results heterogeneous drones Calculate the random boundary adjustment factor: in, Heterogeneous drones At any moment The random boundary adjustment factor; Indicates the interval A function that generates random numbers internally.

[0022] Based on the random boundary adjustment factor, the inner and outer decision distances within the dynamic cluster area are obtained: in, Heterogeneous drones Relative to dynamic cluster area The inner distance is determined; Heterogeneous drones Relative to dynamic cluster area The outer edge of the distance is determined.

[0023] Heterogeneous drones With virtual traction point The Euclidean distance between them is: in, Heterogeneous drones With virtual traction point At any moment The Euclidean distance.

[0024] The boundary traction force is expressed as: in, Heterogeneous drones At any moment The boundary traction force it is subjected to; Heterogeneous drones Non-zero boundary traction force experienced when located outside the corresponding dynamic cluster region or not satisfying the region constraints; This represents the boundary traction weight. Boundary traction is used to pull heterogeneous UAVs from outside the dynamic cluster area to inside the corresponding dynamic cluster area, and reduces the risk of heterogeneous UAVs stagnating near the area boundary through a random boundary adjustment factor.

[0025] The boundary traction force obtained in step S5 The subsequent step S6 further calculates the safety repulsion force between the heterogeneous drones.

[0026] S6: Calculate the safety repulsion force based on the Euclidean distance between nodes in step S1; Based on the position vectors of the heterogeneous UAVs obtained in step S1, calculate the position vectors of any two different heterogeneous UAVs. and Euclidean distance between them: in, Indicates difference from Another heterogeneous drone; Indicates the type number of another heterogeneous drone; This indicates the serial number of another heterogeneous drone within the same type; Heterogeneous drones heterogeneous drones At any moment The Euclidean distance.

[0027] A safety repulsion force is generated when the Euclidean distance between two heterogeneous drones is less than the safe distance; otherwise, the safety repulsion force is zero. The safety repulsion force between the two heterogeneous drones is: in, Heterogeneous drones heterogeneous drones The resulting safety repulsion force; This represents the weight of the safety repulsion force.

[0028] make In addition to the current heterogeneous drones All heterogeneous drones except those mentioned above. The total safety repulsion force is: in, Heterogeneous drones At any moment The total safety repulsion force experienced.

[0029] The total safety repulsion force obtained in step S6 The subsequent step S7 further calculates the velocity alignment force and generates the integrated control input.

[0030] S7: Generate integrated control input based on steps S5 and S6; Based on the velocity vectors of the heterogeneous UAV and the main manned aircraft obtained in step S2, the velocity vectors of the heterogeneous UAV are calculated. Speed ​​alignment force: in, Heterogeneous drones At any moment The velocity alignment force it receives; Indicates the velocity alignment force weight; Indicates that the main control unit is manned. At any moment The velocity vector; This represents a positive distance constant used to avoid a denominator of zero.

[0031] Based on the boundary traction force obtained in step S5, the total safety repulsion force obtained in step S6, and the velocity alignment force obtained in this step, a heterogeneous UAV is generated. Comprehensive control input: in, Heterogeneous drones At any moment The comprehensive control input.

[0032] The integrated control input obtained in step S7 The subsequent step S8 updates the motion state of the heterogeneous drones and feeds the updated state back to step S1 for collaborative motion decision-making at the next moment.

[0033] S8: Update the motion state of the heterogeneous UAV based on the integrated control input from step S7; Based on the integrated control input obtained in step S7 According to discrete time step Updating heterogeneous drones Velocity vector and position vector: in, Heterogeneous drones The velocity vector at the next moment; Heterogeneous drones The position vector at the next moment.

[0034] After the update is complete, determine whether the heterogeneous drones satisfy the following cooperative motion constraints: in, Indicates the speed alignment error threshold; This indicates the safe distance between the main manned aircraft and the heterogeneous unmanned aircraft; Indicates that the next moment the main control is a manned aircraft. heterogeneous drones The Euclidean distance between them; Indicates the next moment of heterogeneous drones heterogeneous drones The Euclidean distance between them; Indicates that the next moment the main control is a manned aircraft. The speed.

[0035] If all the above cooperative motion constraints are satisfied, the heterogeneous UAV motion decision result at the current moment is output; if any cooperative motion constraint is not satisfied, the process returns to step S3, updates the dynamic cluster area based on the updated master manned aircraft state, heterogeneous UAV state, and threat target state, and continues to execute steps S4 to S8. This achieves adaptive, continuous, and secure cluster motion decision-making for manned-UAV cooperative formations in cruise, single-threat-target, and multi-threat-target scenarios.

[0036] The beneficial effects of this invention are: The boundary-traction strategy-based manned-unmanned cooperative formation motion decision-making method described above has the following advantages compared with existing autonomous motion decision-making methods for heterogeneous UAVs in manned-unmanned cooperative formations: 1) Addressing the shortcomings of existing fixed formations or fixed-shape clusters in dynamic adversarial environments, this invention proposes a dynamic cluster area update mechanism. This mechanism can autonomously adjust the size, quantity, and spatial configuration of the cluster area based on the manned-unmanned status and real-time threat situation, achieving seamless and flexible switching between multiple mission phases, from cruising to multi-threat response, significantly improving the environmental adaptability and tactical flexibility of the formation; 2) Addressing the shortcomings of traditional cluster rules that rely on simple local interactions and struggle to achieve intelligent distribution and maintenance within the mission area, this invention designs a boundary-traction strategy centered on virtual traction points, replacing the traditional global aggregation force, guiding UAVs to autonomously enter and evenly distribute within the dynamic area. 3) Addressing the issues of heavy control burden and lack of stable theoretical guarantees in existing methods, this invention constructs a cluster motion model that integrates traction, alignment, and separation rules, and rigorously proves the convergence of the system in the dynamic region through Lyapunov stability analysis. While ensuring speed synchronization, formation maintenance, and collision avoidance, it significantly reduces the real-time control frequency and cognitive load of manned aircraft pilots. 4) Addressing the limitations of existing research in systematically evaluating the distribution quality of clusters, this invention proposes a multi-dimensional uniformity evaluation index that integrates Voronoi diagrams, nearest neighbor distances, grid statistics, and boundary adaptability. This index can quantitatively evaluate the distribution effect of cluster motion in different scenarios, providing a scientific basis for strategy optimization and performance analysis, and further enhancing the reliability and practicality of the method. Attached Figure Description

[0037] Figure 1 It is the result of dynamic swarming motion of searching for heterogeneous UAVs under the wave-like motion of the main controlled manned aircraft; Figure 2 It is the result of dynamic swarm movement of searching for and defending against heterogeneous drones under the circular motion of the main manned aircraft; Figure 3 It is a simulation result of the circular cluster motion under multiple threat targets. Detailed Implementation

[0038] The technical solution of the present invention will be further described below with reference to specific embodiments.

[0039] A method for decision-making in unmanned cooperative formation based on a boundary traction strategy is proposed. This method executes according to a time-series flow: "situational information acquisition and object determination, motion model construction, dynamic cluster area update, area allocation and virtual traction point generation, boundary traction force calculation, safety repulsion force calculation, velocity alignment force calculation, and comprehensive control input generation and motion state update." Subsequent steps use the data, areas, traction points, or control variables obtained from previous steps as input, ensuring continuous, clear, and repeatable data processing. Its key features include the following steps: S1: Obtain current situation information and determine whether there are any unmanned cooperative formation targets; Based on the current mission phase and situational awareness data, a manned-unmanned cooperative formation is identified to participate in the coordinated motion decision-making. This manned-unmanned cooperative formation includes a master manned aircraft and multiple heterogeneous unmanned aerial vehicles (UAVs), wherein the master manned aircraft is denoted as […]. Heterogeneous drones include search-type and defense-type heterogeneous drones. Indicates the type number of heterogeneous drones. Indicates the search for heterogeneous drones. Indicates defense against heterogeneous drones; This indicates the serial number of a heterogeneous UAV of the same type. The unmanned cooperative formation participating in cooperative motion decision-making is represented as: in, Indicates time A collection of unmanned collaborative formation nodes; Indicates the first Search for heterogeneous drones; Indicates the first Defend against heterogeneous drones; Indicates the number of heterogeneous drones searched; This indicates the number of heterogeneous drones being defended against.

[0040] Based on the aforementioned set of manned and unmanned cooperative formation nodes, a communication topology centered on the manned main control aircraft is constructed: in, Indicates time The communication topology; Indicates time A set of communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment Communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment The Euclidean distance; Indicates that the main control unit is manned. At any moment The position vector; Heterogeneous drones At any moment The position vector; Indicates the maximum command and communication distance of the manned aircraft under main control; This represents the L2 norm.

[0041] The set of unmanned cooperative formation nodes obtained in step S1 Communication topology In addition to the position vectors of each node, the subsequent step S2 constructs a heterogeneous UAV motion model and provides state variables for subsequent dynamic cluster area updates and control input calculations.

[0042] S2: Construct a motion model for the heterogeneous UAV based on the situational information from step S1; Heterogeneous drones The position vector, velocity vector, and virtual integrated control input vector are respectively expressed as: in, , and These represent heterogeneous drones. The three coordinate components in the inertial coordinate system; Heterogeneous drones At any moment The velocity vector; , and These represent heterogeneous drones. The three velocity components in the inertial coordinate system; Heterogeneous drones At any moment The virtual integrated control input vector; , and These represent heterogeneous drones. Three virtual integrated control input components in the inertial coordinate system; superscript This represents the transpose of a vector.

[0043] The simplified motion model of the heterogeneous UAV is as follows: in, express The first derivative with respect to time; express The first derivative with respect to time.

[0044] When it is necessary to convert the virtual integrated control input into the actual input of the heterogeneous UAV, the lift, thrust, and roll angle of the heterogeneous UAV are obtained by the following formula: in, Heterogeneous drones At any moment Lift; Heterogeneous drones At any moment The thrust; Heterogeneous drones At any moment Roll angle; Heterogeneous drones The quality; Represents gravitational acceleration; Heterogeneous drones At any moment Flight path angle; Heterogeneous drones At any moment The heading angle; Heterogeneous drones At any moment Resistance.

[0045] The heterogeneous UAV motion model and state variables obtained in step S2 are followed by a dynamic cluster area update in step S3 based on the current mission stage and threat target information.

[0046] S3: Based on the set of unmanned cooperative formation nodes obtained in step S1 and the motion model of heterogeneous UAVs obtained in step S2, determine the dynamic cluster area according to the current mission stage. Dynamic cluster regions are uniformly represented as ,in Indicates the type number of heterogeneous drones. This indicates the dynamic clustering region number corresponding to heterogeneous UAVs of the same type. A dynamic clustering region is defined as follows: in, Indicates time The Type 1 One dynamic cluster area; Represents dynamic cluster area Any position vector within; Represents dynamic cluster area The minimum radius; Represents dynamic cluster area The maximum radius; Representing vectors The direction angle relative to the horizontal axis of the inertial coordinate system; Represents dynamic cluster area The central direction angle; Represents dynamic cluster area The range of angles.

[0047] During the cruise phase, the search for heterogeneous drones corresponds to a search dynamic cluster area. Defending against heterogeneous drones corresponds to a dynamic cluster defense area. The search dynamic cluster area is symmetrically arranged along the flight path of the main manned aircraft; the parameters of the defense dynamic cluster area are updated according to the following formula: in, This indicates the maximum radius of the defense dynamic cluster area during the cruise phase. This represents the minimum radius of the defense dynamic cluster area during the cruise phase. This indicates the angular range of the defense dynamic cluster area during the cruise phase. Indicates the safe distance between heterogeneous drones; This indicates the maximum permissible angle for the defense dynamic cluster area during the cruise phase; , and These represent the maximum radius adjustment, minimum radius adjustment, and angle adjustment of the defense dynamic cluster area during the cruise phase, respectively.

[0048] During the threat target response phase, if detected For each threat target, the dynamic swarm area corresponding to the defense against heterogeneous drones is divided into: One defense dynamic cluster area. For the first One threat target Its corresponding defense dynamic cluster area The parameters are updated using the following formula: in, Indicates the threat target number, and ; Indicate threat target At any moment The position vector; Indicates that the main control unit is manned. With threat targets Euclidean distance; Indicates the defense dynamic cluster area The maximum radius; Indicates the defense dynamic cluster area The minimum radius; Indicates the defense dynamic cluster area Angle range; Indicates allocation to the dynamic defense cluster area The number of defensive heterogeneous drones; and These represent the defense dynamic cluster areas. The maximum radius adjustment amount and the minimum radius adjustment amount; Indicates the defense dynamic cluster area The central direction angle; and These represent the threat targets. The horizontal and vertical coordinates; and These respectively indicate that the main control is a manned aircraft. The horizontal and vertical coordinates; This represents the arctangent function in the four quadrants.

[0049] The dynamic cluster region obtained in step S3 In the subsequent step S4, the heterogeneous drones are assigned to the corresponding dynamic cluster areas, and virtual traction points are generated for each dynamic cluster area.

[0050] S4: Based on the dynamic cluster area in step S3, perform area allocation and generate virtual traction points; Based on the dynamic cluster region obtained in step S3 A region allocation strategy based on a consensus-based binding algorithm is adopted to assign heterogeneous drones to corresponding dynamic cluster regions. For any heterogeneous drone to be assigned... With any dynamic cluster region First, calculate the distance evaluation value from the virtual traction point, then calculate the regional bid value: in, Heterogeneous drones To the dynamic cluster area Corresponding virtual traction point Distance evaluation metric; Heterogeneous drones For dynamic cluster regions Regional bid value; Indicates the quantity demand weight; Indicates distance weight; Represents dynamic cluster area The required first Number of different types of heterogeneous drones; This indicates that the data has been assigned to a dynamic cluster area. The Number of different types of heterogeneous drones; This represents the function that takes the maximum value. Represents dynamic cluster area Corresponding virtual traction point The position vector.

[0051] Based on regional bid value The size of heterogeneous drones Assigned to the dynamic cluster region with the largest bid value, resulting in the region allocation: in, Indicates allocation to dynamic cluster area A collection of heterogeneous drones; This indicates that it is already in the dynamic cluster area. An internal collection of heterogeneous drones; This indicates that it has been allocated to the dynamic cluster area. However, the heterogeneous drone ensemble is still located outside this dynamic cluster area; This indicates the set union operation.

[0052] For the dynamic cluster area of ​​defense during the threat target response phase Its virtual traction point The position vector is: in, Indicates the defense dynamic cluster area Virtual traction point and main control manned machine Euclidean distance; Indicates the defense dynamic cluster area virtual traction point The position vector.

[0053] During the cruise phase, search dynamic cluster areas. Virtual traction points and defense dynamic cluster areas The virtual traction points are as follows: in, Indicates the search for dynamic cluster regions during the cruise phase. virtual traction point The position vector; Indicates the dynamic cluster area for defense during the cruise phase. virtual traction point The position vector; Indicates that the main control unit is manned. At any moment The heading angle.

[0054] The region allocation result obtained in step S4 and virtual traction points The subsequent step S5 calculates the boundary traction force for the heterogeneous UAVs that have been assigned but are still located outside the corresponding dynamic cluster area.

[0055] S5: Calculate the boundary traction force based on the virtual traction point in step S4; Based on the virtual traction point obtained in step S4 and regional allocation results heterogeneous drones Calculate the random boundary adjustment factor: in, Heterogeneous drones At any moment The random boundary adjustment factor; Indicates the interval A function that generates random numbers internally.

[0056] Based on the random boundary adjustment factor, the inner and outer decision distances within the dynamic cluster area are obtained: in, Heterogeneous drones Relative to dynamic cluster area The inner distance is determined; Heterogeneous drones Relative to dynamic cluster area The outer edge of the distance is determined.

[0057] Heterogeneous drones With virtual traction point The Euclidean distance between them is: in, Heterogeneous drones With virtual traction point At any moment The Euclidean distance.

[0058] The boundary traction force is expressed as: in, Heterogeneous drones At any moment The boundary traction force it is subjected to; Heterogeneous drones Non-zero boundary traction force experienced when located outside the corresponding dynamic cluster region or not satisfying the region constraints; This represents the boundary traction weight. Boundary traction is used to pull heterogeneous UAVs from outside the dynamic cluster area to inside the corresponding dynamic cluster area, and reduces the risk of heterogeneous UAVs stagnating near the area boundary through a random boundary adjustment factor.

[0059] The boundary traction force obtained in step S5 The subsequent step S6 further calculates the safety repulsion force between the heterogeneous drones.

[0060] S6: Calculate the safety repulsion force based on the Euclidean distance between nodes in step S1; Based on the position vectors of the heterogeneous UAVs obtained in step S1, calculate the position vectors of any two different heterogeneous UAVs. and Euclidean distance between them: in, Indicates difference from Another heterogeneous drone; Indicates the type number of another heterogeneous drone; This indicates the serial number of another heterogeneous drone within the same type; Heterogeneous drones heterogeneous drones At any moment The Euclidean distance.

[0061] A safety repulsion force is generated when the Euclidean distance between two heterogeneous drones is less than the safe distance; otherwise, the safety repulsion force is zero. The safety repulsion force between the two heterogeneous drones is: in, Heterogeneous drones heterogeneous drones The resulting safety repulsion force; This represents the weight of the safety repulsion force.

[0062] make In addition to the current heterogeneous drones All heterogeneous drones except those mentioned above. The total safety repulsion force is: in, Heterogeneous drones At any moment The total safety repulsion force experienced.

[0063] The total safety repulsion force obtained in step S6 The subsequent step S7 further calculates the velocity alignment force and generates the integrated control input.

[0064] S7: Generate integrated control input based on steps S5 and S6; Based on the velocity vectors of the heterogeneous UAV and the main manned aircraft obtained in step S2, the velocity vectors of the heterogeneous UAV are calculated. Speed ​​alignment force: in, Heterogeneous drones At any moment The velocity alignment force it receives; Indicates the velocity alignment force weight; Indicates that the main control unit is manned. At any moment The velocity vector; This represents a positive distance constant used to avoid a denominator of zero.

[0065] Based on the boundary traction force obtained in step S5, the total safety repulsion force obtained in step S6, and the velocity alignment force obtained in this step, a heterogeneous UAV is generated. Comprehensive control input: in, Heterogeneous drones At any moment The comprehensive control input.

[0066] The integrated control input obtained in step S7 The subsequent step S8 updates the motion state of the heterogeneous drones and feeds the updated state back to step S1 for collaborative motion decision-making at the next moment.

[0067] S8: Update the motion state of the heterogeneous UAV based on the integrated control input from step S7; Based on the integrated control input obtained in step S7 According to discrete time step Updating heterogeneous drones Velocity vector and position vector: in, Heterogeneous drones The velocity vector at the next moment; Heterogeneous drones The position vector at the next moment.

[0068] After the update is complete, determine whether the heterogeneous drones satisfy the following cooperative motion constraints: in, Indicates the speed alignment error threshold; This indicates the safe distance between the main manned aircraft and the heterogeneous unmanned aircraft; Indicates that the next moment the main control is a manned aircraft. heterogeneous drones The Euclidean distance between them; Indicates the next moment of heterogeneous drones heterogeneous drones The Euclidean distance between them; Indicates that the next moment the main control is a manned aircraft. The speed.

[0069] If all the above cooperative motion constraints are satisfied, the heterogeneous UAV motion decision result at the current moment is output; if any cooperative motion constraint is not satisfied, the process returns to step S3, updates the dynamic cluster area based on the updated master manned aircraft state, heterogeneous UAV state, and threat target state, and continues to execute steps S4 to S8. This achieves adaptive, continuous, and secure cluster motion decision-making for manned-UAV cooperative formations in cruise, single-threat-target, and multi-threat-target scenarios.

[0070] Simulation experiments verify: To further demonstrate that the manned-unmanned cooperative formation motion decision-making method based on boundary traction strategy described in this invention can be repeatedly implemented by those skilled in the art, and to verify its effectiveness in the cruise phase, single-threat target response phase, and multi-threat target response phase, this simulation experiment takes a master-controlled manned aircraft leading multiple heterogeneous unmanned aerial vehicles (UAVs) in cooperative motion as the object, and proceeds in the following order: "simulation parameter setting—dynamic cluster area update—virtual traction point generation—integrated control input calculation—motion state update—result evaluation". The purpose of the simulation experiment is to verify that the method can enable heterogeneous UAVs to enter and remain within a specified dynamic cluster area, satisfy speed alignment, area maintenance, and safety interval constraints during motion, and maintain good cluster distribution uniformity and motion stability when dynamic threats change.

[0071] 1. Simulation experiment scenario and parameter assignment: The simulation experiment sets up a two-dimensional flight mission area and uses approximate conditions at the same altitude to verify the motion of the manned-unmanned cooperative formation. The manned-unmanned cooperative formation includes a master manned aircraft, multiple search heterogeneous UAVs, and multiple defense heterogeneous UAVs. Among them, the search heterogeneous UAVs are used to maintain search coverage advantage in the forward area of ​​the master manned aircraft, and the defense heterogeneous UAVs are used to adjust the dynamic defense cluster area according to changes in the position, number, and distance of threat targets.

[0072] The main parameter settings for the simulation experiment are shown in the table below.

[0073] Among the parameters mentioned above, both the search and defense heterogeneous UAVs can acquire their own position, speed, and heading information; the controlling manned aircraft periodically collects the status information of the heterogeneous UAVs and sends its own status and dynamic threat situation to the heterogeneous UAVs. Based on these parameters, three simulation scenarios are constructed: a cruise phase, a single-threat target response phase, and a multi-threat target response phase.

[0074] 2. Simulation Experiment Constraints and Evaluation Indicators: To verify the feasibility of the method of this invention, the simulation experiment focused on checking whether the heterogeneous UAVs met the constraints of velocity alignment, area preservation, and safe spacing. These constraints can be expressed as: in, Heterogeneous drones The velocity vector at time t; This represents the velocity vector of the manned aircraft under main control at time t; Indicates the speed alignment error threshold; This represents the z-th type of heterogeneous UAV that is planned to be included in the dynamic clustering region at time t; This represents the h-th type and k-th dynamic cluster region at time t; This indicates the main control manned aircraft and heterogeneous unmanned aircraft. The Euclidean distance between them; Heterogeneous drones heterogeneous drones The Euclidean distance between them; This indicates the safe distance between the main manned aircraft and the heterogeneous unmanned aircraft; This indicates the safe distance between heterogeneous drones.

[0075] Furthermore, to quantify the distribution quality of heterogeneous UAVs within a dynamic swarm area, a comprehensive swarm uniformity evaluation index is introduced: in, The comprehensive cluster uniformity index represents time t; This represents a uniformity index based on the Voronoi diagram; Indicates a uniformity index based on proximity distance; Indicates the uniformity index based on grid partitioning; Indicates the boundary adaptability index; , , , Let each of the four categories of indicators represent their respective weights, and satisfy the following conditions: in, Used to characterize the consistency of the Voronoi region area corresponding to each heterogeneous UAV; Used to characterize the degree of balance in the nearest neighbor distances among heterogeneous drones; Used to characterize the volatility and distribution uncertainty of the number of grid cells within a dynamic cluster area; This index is used to characterize the degree to which the distribution of heterogeneous UAVs conforms to the boundary constraints of the dynamic swarm region. It is a comprehensive swarm uniformity index. The larger the value, the more evenly the heterogeneous drones are distributed within the corresponding dynamic cluster area, and the better the mission coverage and response quality.

[0076] 3. Simulation Experiment Procedure: 1) Within a 100km × 100km two-dimensional flight mission area, a master manned aircraft is deployed, and a circle with the master manned aircraft as the center and a radius of [missing information] is formed. The initial positions for searching and defending against heterogeneous drones are randomly generated within a circular area; 2) Based on the position, heading, and speed information of the main manned aircraft, as well as the position and number of threat targets, update the search dynamic cluster area and the defense dynamic cluster area according to step S3 in the instruction manual; 3) Based on the updated dynamic cluster area, assign the heterogeneous UAVs to the corresponding dynamic cluster area according to step S4 in the instruction manual, and generate virtual traction points for each dynamic cluster area; 4) For heterogeneous UAVs located outside the corresponding dynamic cluster area or that do not meet the area constraints, calculate the boundary traction force according to step S5 in the instruction manual; 5) Calculate the safety repulsion force based on the Euclidean distance between the heterogeneous drones, following step S6 in the instruction manual; 6) Based on the speed difference between the main manned aircraft and the heterogeneous UAV, calculate the speed alignment force according to step S7 in the instruction manual, and generate the comprehensive control input by combining the boundary traction force, safety repulsion force and speed alignment force; 7) Update the velocity vector and position vector of the heterogeneous UAV according to step S8 in the instruction manual, and determine whether the velocity alignment, region preservation and safety interval constraints are satisfied at each simulation moment; 8) Statistically analyze the overall cluster uniformity index at two typical times, t=50s and t=75s, and evaluate the effectiveness of the method of the present invention in combination with the motion trajectory diagram.

[0077] 4. Simulation results and discussion: During the cruise phase, a master manned aircraft and eight heterogeneous search drones are initially deployed. The master manned aircraft moves along a wavy trajectory, and the heterogeneous search drones are distributed within a forward-facing fan-shaped dynamic cluster area. Simulation results show that the heterogeneous search drones can rapidly enter the designated dynamic search cluster area from their initial random positions under the influence of boundary traction forces and maintain a stable cluster distribution along with the master manned aircraft. Since no threat targets are introduced in this scenario, the size of the dynamic search cluster area remains essentially constant, but its spatial position updates in real time with the movement of the master manned aircraft. Figure 1 As shown.

[0078] Furthermore, a master manned aircraft, eight search heterogeneous UAVs, and eight defense heterogeneous UAVs were configured, with the master manned aircraft moving along a circular trajectory. Simulation results show that the search heterogeneous UAVs can remain within the forward search dynamic cluster area, and the defense heterogeneous UAVs can remain within the close-range defense dynamic cluster area. Both types of heterogeneous UAVs can achieve smooth movement within their respective dynamic cluster areas, indicating that the method of this invention can handle cooperative formation movement decisions under heterogeneous task attributes. Figure 2 As shown.

[0079] In the multi-threat target response phase, two dynamic threat targets are set, and they are considered valid threat targets when the distance between them and the main manned aircraft reaches 50km. Simulation results show that when valid threat targets appear, the defense dynamic cluster area can automatically adjust the number and size of the area according to the number, orientation, and distance of the threat targets; the defense heterogeneous UAVs move towards the corresponding defense dynamic cluster area under the action of virtual traction points, and maintain smooth and safe cluster movement under the action of velocity alignment force and safety repulsion force. These results demonstrate that the method of this invention can achieve adaptive area reconstruction and autonomous response of heterogeneous UAVs under dynamic threat conditions. Figure 3 As shown.

[0080] 5. Overall uniformity index results: Based on the simulation results in the paper, two typical times, t=50s and t=75s, were used to statistically analyze the comprehensive cluster uniformity index of each dynamic cluster region under different adversarial scenarios. The results are shown in the table below.

[0081] As shown in the table above, during the cruise phase, the overall cluster uniformity index of the search dynamic cluster area increased from 0.865 to 0.905, and the overall cluster uniformity index of the defense dynamic cluster area increased from 0.833 to 0.865. This indicates that heterogeneous UAVs can gradually form a more uniform dynamic cluster distribution under the influence of boundary traction and velocity alignment forces. During the multi-threat target phase, although defending heterogeneous UAVs requires reclassification into two defense dynamic cluster areas based on the threat targets, the overall cluster uniformity index of each area is higher than 0.75, indicating that the method of this invention can still maintain good cluster distribution quality even when dynamic threats cause area reconstruction.

[0082] The simulation experiments described above demonstrate that the manned-unmanned cooperative formation motion decision-making method based on boundary traction strategy described in this invention can adaptively update the dynamic cluster area according to the status of the master manned aircraft and threat target information at different stages of confrontation. It generates integrated control inputs for heterogeneous UAVs through virtual traction points, boundary traction forces, safety repulsion forces, and velocity alignment forces, enabling heterogeneous UAVs to enter and remain within the corresponding dynamic cluster area. The simulation results further show that the method can maintain a stable distribution of search and defense heterogeneous UAVs during the cruise phase; adjust the defense dynamic cluster area according to the threat orientation during the single threat target response phase; and divide multiple defense dynamic cluster areas according to the number of threat targets during the multi-threat target response phase, while maintaining good motion smoothness, safety intervals, and regional uniformity. Therefore, the method is reproducible and can serve as an effective technical solution for autonomous motion decision-making in dynamic confrontation environments for manned-unmanned cooperative formations.

Claims

1. A method for decision-making on the movement of unmanned cooperative formations based on a boundary traction strategy, characterized in that, The steps are as follows: S1: Obtain current situation information and determine whether there are any unmanned cooperative formation targets; S2: Construct a motion model for the heterogeneous UAV based on the situational information from step S1; S3: Based on the set of unmanned cooperative formation nodes obtained in step S1 and the motion model of heterogeneous UAVs obtained in step S2, determine the dynamic cluster area according to the current mission stage. Dynamic cluster regions are uniformly represented as ,in Indicates the type number of heterogeneous drones. This indicates the dynamic clustering region number corresponding to heterogeneous UAVs of the same type; the dynamic clustering region is defined as: in, Indicates time The Type 1 One dynamic cluster area; Represents dynamic cluster area Any position vector within; Represents dynamic cluster area The minimum radius; Represents dynamic cluster area The maximum radius; Representing vectors The direction angle relative to the horizontal axis of the inertial coordinate system; Represents dynamic cluster area The central direction angle; Represents dynamic cluster area Angle range; During the cruise phase, the search for heterogeneous drones corresponds to a search dynamic cluster area. Defending against heterogeneous drones corresponds to a dynamic cluster defense area. The search dynamic cluster area is symmetrically arranged along the flight path of the main manned aircraft; the parameters of the defense dynamic cluster area are updated according to the following formula: in, This indicates the maximum radius of the defense dynamic cluster area during the cruise phase. This represents the minimum radius of the defense dynamic cluster area during the cruise phase. This indicates the angular range of the defense dynamic cluster area during the cruise phase. Indicates the safe distance between heterogeneous drones; This indicates the maximum permissible angle for the defense dynamic cluster area during the cruise phase; , and These represent the maximum radius adjustment, minimum radius adjustment, and angle adjustment of the defense dynamic cluster area during the cruise phase, respectively. During the threat target response phase, if detected For each threat target, the dynamic swarm area corresponding to the defense against heterogeneous drones is divided into: The first dynamic defense cluster area; for the first One threat target Its corresponding defense dynamic cluster area The parameters are updated using the following formula: in, Indicates the threat target number, and ; Indicate threat target At any moment The position vector; Indicates that the main control unit is manned. With threat targets Euclidean distance; Indicates the defense dynamic cluster area The maximum radius; Indicates the defense dynamic cluster area The minimum radius; Indicates the defense dynamic cluster area The range of angles; Indicates allocation to the dynamic defense cluster area The number of defensive heterogeneous drones; and These represent the defense dynamic cluster areas. The maximum radius adjustment amount and the minimum radius adjustment amount; Indicates the defense dynamic cluster area The central direction angle; and These represent the threat targets. The horizontal and vertical coordinates; and These respectively indicate that the main control is a manned aircraft. The horizontal and vertical coordinates; Represents the arctangent function in the four quadrants; S4: Based on the dynamic cluster area in step S3, perform area allocation and generate virtual traction points; Based on the dynamic cluster region obtained in step S3 A region allocation strategy based on a consistency binding algorithm is adopted to assign heterogeneous UAVs to corresponding dynamic cluster regions; for any heterogeneous UAV to be assigned... With any dynamic cluster region First, calculate the distance evaluation value from the virtual traction point, then calculate the regional bid value: in, Heterogeneous drones To the dynamic cluster area Corresponding virtual traction point Distance evaluation metric; Heterogeneous drones For dynamic cluster regions Regional bid value; Indicates the quantity demand weight; Indicates distance weight; Represents dynamic cluster area The required first Number of different types of heterogeneous drones; This indicates that the data has been assigned to a dynamic cluster area. The Number of different types of heterogeneous drones; This represents the function that takes the maximum value. Represents dynamic cluster area Corresponding virtual traction point The position vector; Based on regional bid value The size of heterogeneous drones Assigned to the dynamic cluster region with the largest bid value, resulting in the region allocation: in, Indicates allocation to dynamic cluster area A collection of heterogeneous drones; This indicates that it is already in the dynamic cluster area. An internal collection of heterogeneous drones; This indicates that it has been allocated to the dynamic cluster area. However, the heterogeneous drone ensemble is still located outside this dynamic cluster area; This represents the set union operation; For the dynamic cluster area of ​​defense during the threat response phase Its virtual traction point The position vector is: in, Indicates the defense dynamic cluster area Virtual traction point and main control manned machine Euclidean distance; Indicates the defense dynamic cluster area virtual traction point The position vector; During the cruise phase, search dynamic cluster areas. Virtual traction points and defense dynamic cluster areas The virtual traction points are as follows: in, Indicates the search for dynamic cluster regions during the cruise phase. virtual traction point The position vector; Indicates the dynamic cluster area for defense during the cruise phase. virtual traction point The position vector; Indicates that the main control unit is manned. At any moment The heading angle; S5: Calculate the boundary traction force based on the virtual traction point in step S4; S6: Calculate the safety repulsion force based on the Euclidean distance between nodes in step S1; S7: Generate integrated control input based on steps S5 and S6; S8: Update the motion state of the heterogeneous UAV based on the integrated control input from step S7.

2. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 1, characterized in that, The specific steps of S1 are as follows: Based on the current mission phase and situational awareness data, a manned-unmanned cooperative formation participating in the coordinated motion decision-making is identified; the manned-unmanned cooperative formation includes a master-controlled manned aircraft and multiple heterogeneous unmanned aerial vehicles (UAVs), wherein the master-controlled manned aircraft is denoted as […]. Heterogeneous drones include search-type heterogeneous drones and defense-type heterogeneous drones; Indicates the type number of heterogeneous drones. Indicates the search for heterogeneous drones. Indicates defense against heterogeneous drones; Indicates the serial number of a heterogeneous drone of the same type; The presence or absence of unmanned coordinated formations participating in coordinated motion decision-making is represented as follows: in, Indicates time A collection of unmanned collaborative formation nodes; Indicates the first Search for heterogeneous drones; Indicates the first Defend against heterogeneous drones; Indicates the number of heterogeneous drones searched; Indicates the number of heterogeneous drones used for defense; Based on the aforementioned set of manned and unmanned cooperative formation nodes, a communication topology centered on the manned main control aircraft is constructed: in, Indicates time The communication topology; Indicates time A set of communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment Communication links; Indicates that the main control unit is manned. heterogeneous drones At any moment The Euclidean distance; Indicates that the main control unit is manned. At any moment The position vector; Heterogeneous drones At any moment The position vector; Indicates the maximum command and communication distance of the manned aircraft under main control; This represents the L2 norm.

3. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 2, characterized in that, The specific steps of S2 are as follows: Heterogeneous drones The position vector, velocity vector, and virtual integrated control input vector are respectively expressed as: in, , and These represent heterogeneous drones. The three coordinate components in the inertial coordinate system; Heterogeneous drones At any moment The velocity vector; , and These represent heterogeneous drones. The three velocity components in the inertial coordinate system; Heterogeneous drones At any moment The virtual integrated control input vector; , and These represent heterogeneous drones. Three virtual integrated control input components in the inertial coordinate system; superscript Indicates vector transpose; The simplified motion model of the heterogeneous UAV is as follows: in, express The first derivative with respect to time; express The first derivative with respect to time; When it is necessary to convert the virtual integrated control input into the actual input of the heterogeneous UAV, the lift, thrust, and roll angle of the heterogeneous UAV are obtained by the following formula: in, Heterogeneous drones At any moment Lift; Heterogeneous drones At any moment The thrust; Heterogeneous drones At any moment Roll angle; Heterogeneous drones The quality; Represents gravitational acceleration; Heterogeneous drones At any moment Flight path angle; Heterogeneous drones At any moment The heading angle; Heterogeneous drones At any moment Resistance.

4. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 3, characterized in that, The specific steps of step S5 are as follows: Based on the virtual traction point obtained in step S4 and regional allocation results heterogeneous drones Calculate the random boundary adjustment factor: in, Heterogeneous drones At any moment The random boundary adjustment factor; Indicates the interval A function that generates random numbers internally; Based on the random boundary adjustment factor, the inner and outer decision distances within the dynamic cluster area are obtained: in, Heterogeneous drones Relative to dynamic cluster area The inner distance is determined; Heterogeneous drones Relative to dynamic cluster area The outer edge of the distance is used to determine the distance; Heterogeneous drones With virtual traction point The Euclidean distance between them is: in, Heterogeneous drones With virtual traction point At any moment The Euclidean distance; The boundary traction force is expressed as: in, Heterogeneous drones At any moment The boundary traction force it is subjected to; Heterogeneous drones Non-zero boundary traction force experienced when located outside the corresponding dynamic cluster region or not satisfying region constraints; This represents the boundary traction weight; the boundary traction is used to pull heterogeneous UAVs from outside the dynamic cluster area to inside the corresponding dynamic cluster area, and reduces the risk of heterogeneous UAVs stagnation near the area boundary through a random boundary adjustment factor.

5. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 4, characterized in that, The specific steps of step S6 are as follows: Based on the position vectors of the heterogeneous UAVs obtained in step S1, calculate the position vectors of any two different heterogeneous UAVs. and Euclidean distance between them: in, Indicates difference from Another heterogeneous drone; Indicates the type number of another heterogeneous drone; This indicates the serial number of another heterogeneous drone within the same type; Heterogeneous drones heterogeneous drones At any moment The Euclidean distance; A safety repulsion force is generated when the Euclidean distance between two heterogeneous drones is less than the safe distance; otherwise, the safety repulsion force is zero. The safety repulsion force between the two heterogeneous drones is: in, Heterogeneous drones heterogeneous drones The resulting safety repulsion force; Indicates the weight of the safety repulsion force; make In addition to the current heterogeneous drones All heterogeneous drones except those mentioned above; heterogeneous drones The total safety repulsion force is: in, Heterogeneous drones At any moment The total safety repulsion force experienced.

6. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 5, characterized in that, The specific steps of step S7 are as follows: Based on the velocity vectors of the heterogeneous UAV and the main manned aircraft obtained in step S2, the velocity vectors of the heterogeneous UAV are calculated. Speed ​​alignment force: in, Heterogeneous drones At any moment The velocity alignment force it is subjected to; Indicates the velocity alignment force weight; Indicates that the main control unit is manned. At any moment The velocity vector; This represents a positive distance constant used to avoid fractions with a denominator of zero. Based on the boundary traction force obtained in step S5, the total safety repulsion force obtained in step S6, and the velocity alignment force obtained in this step, a heterogeneous UAV is generated. Comprehensive control input: in, Heterogeneous drones At any moment The comprehensive control input.

7. The method for decision-making on manned / unmanned cooperative formation motion based on boundary traction strategy according to claim 6, characterized in that, The specific steps of S8 are as follows: Based on the integrated control input obtained in step S7 According to discrete time step Updating heterogeneous drones Velocity vector and position vector: in, Heterogeneous drones The velocity vector at the next moment; Heterogeneous drones The position vector at the next moment; After the update is complete, determine whether the heterogeneous drones satisfy the following cooperative motion constraints: in, Indicates the speed alignment error threshold; This indicates the safe distance between the main manned aircraft and the heterogeneous unmanned aircraft; Indicates that the next moment the main control is a manned aircraft. heterogeneous drones The Euclidean distance between them; Indicates the next moment of heterogeneous drones heterogeneous drones The Euclidean distance between them; Indicates that the next moment the main control is a manned aircraft. speed; If all the above cooperative motion constraints are satisfied, the heterogeneous UAV motion decision result at the current moment is output; if any cooperative motion constraint is not satisfied, return to step S3, update the dynamic cluster area again based on the updated master manned aircraft state, heterogeneous UAV state and threat target state, and continue to execute steps S4 to S8; thereby realizing adaptive, continuous and safe cluster motion decision of manned and unmanned cooperative formation in cruise, single threat target and multi-threat target scenarios.

Images