Aircraft cluster crossing method and system based on bionic visual projection field and identity balance

Abstract

The invention discloses an aircraft cluster control method and system based on a bionic visual projection field and identity balance, and belongs to the technical field of unmanned aerial vehicle clusters. Aiming at the problems that a traditional method depends on global planning and is poor in robustness in a communication limited and dynamic complex scene, two innovations are provided: one is a bionic visual projection field, and a dynamic environment model is generated through local sensing, so that an unmanned aerial vehicle autonomously avoids obstacles and cooperatively moves; and 2, a dynamic identity balancing strategy is adopted, leader or follower roles are allocated to cluster members according to tasks and positions, role adaptive switching is supported, and task fault tolerance is enhanced. The system is integrated with a multi-module cooperation mechanism, efficient cluster crossing is achieved in complex obstacle environments such as ruins and forests through local information interaction and rule simplification, and motion consistency and safety are guaranteed. The method is low in calculation requirement, high in adaptability and suitable for communication limited scenes such as post-disaster rescue, and the autonomous cooperation capability of the cluster is remarkably improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of drone cluster control, and specifically relates to a drone cluster traversal method and system based on bionic visual projection field and identity trade-off, which is particularly suitable for mission scenarios with communication denial, unknown environment or dynamic complex. Background Art

[0002] UAV swarm systems play an important role in the development of modern science and technology, and their application areas cover military and civilian aspects. Swarm crossing, as a key link in UAV swarm missions, is of great significance in exploring crossing paths in unknown environments. However, existing technical solutions have the following shortcomings. First, there is a strong dependence on communication. Traditional methods rely on real-time communication to achieve collaboration, but they cannot work in communication-denied environments. Second, there is poor environmental adaptability. It is difficult to plan paths in real time in unknown or dynamic environments, and they rely on global information. Third, there are computing resource limitations. Environmental modeling methods such as SLAM have high requirements for onboard computing power and are difficult to deploy on micro-UAVs. Fourth, there is insufficient robustness. Existing bionic swarm models focus on behavior reproduction and lack targeted design for complex tasks (such as swarm crossing).

[0003] Although biological swarming behaviors (such as bird migration and fish schooling) offer insights into self-organization and robustness for drone swarms, existing biomimetic approaches focus on simulating simple scenarios and fail to effectively address swarm traversal in complex environments. For example, methods based on the Boids rule or the Vicsek model struggle to maintain swarm coordination during communication interruptions, while artificial potential field methods are prone to the risk of swarm fragmentation due to topology switching. Therefore, a robust swarm traversal method that adapts to communication-denied environments, reduces computational overhead, and is highly desirable is urgently needed. Summary of the Invention

[0004] To address the aforementioned technical issues, the present invention provides a method and system for swarm traversal of unmanned aircraft based on a bionic visual projection field and identity trade-off. This method, combining bionics with swarm intelligent control technology, effectively addresses the shortcomings of existing technologies in unmanned swarm traversal missions in complex environments. This method, in particular, ensures that drone swarms can safely and efficiently execute missions in complex environments, such as those facing unknown or communication-denied environments.

[0005] By mimicking the visual interaction mechanisms of natural organisms, this invention designs a bionic visual projection field that uses a visual perception system to enable real-time monitoring and judgment of the swarm environment. Traditional unmanned swarm control methods typically rely on global path planning or preset paths. However, the bionic visual projection field of this invention can adjust the trajectory of each aircraft in real time based on local visual information. This allows drones to effectively coordinate their swarms, especially in communication-denied or unknown environments, relying on their local perception and visual projection.

[0006] Each aircraft collects visual information of its surroundings through sensors such as cameras and lidar, and updates the projection field data in real time. Each aircraft interacts locally with other aircraft around it based on visual information.

[0007] The aircraft in the cluster use the bionic visual projection field algorithm to judge their relative position with neighboring aircraft and obstacles in the environment, and automatically plan trajectories to ensure the orderly movement of the cluster.

[0008] Another technical solution of this invention is to dynamically adjust the collaboration mode of swarm members through an identity trade-off strategy. This strategy not only adjusts the roles of swarm members in real time based on information such as the spatial position, speed, and mission priority of each aircraft, but also the current state of the swarm, changes in the environment, and interactions between aircraft.

[0009] The identity trade-off strategy involves prioritizing each aircraft, dynamically determining its identity within the swarm (e.g., leader, follower, or collaborator) based on its current mission priority, location, payload, and other information. For example, an aircraft performing a critical mission is assigned the leader role, while other aircraft follow or collaborate based on their missions.

[0010] Identity adjustment: During mission execution, the roles of aircraft may be dynamically adjusted. For example, when a certain aircraft is in a dangerous area or the mission is completed, its identity may change from leader to collaborator. Other aircraft will assume the new leadership role to ensure that the swarm mission is not affected.

[0011] The identity trade-off strategy improves the swarm's autonomy and flexibility, allowing it to quickly adjust to unexpected events or environmental changes. By balancing each aircraft's mission priority with environmental information, it ensures efficient allocation of swarm resources.

[0012] This paper proposes a swarm traversal method based on local information. Unlike existing global path planning methods, this approach emphasizes real-time path adjustment based on local perception in dynamic environments. The aircraft not only adjusts its trajectory based on surrounding obstacles and the positions of other aircraft, but also ensures the uniformity and stability of the swarm's motion through a coordinated control algorithm within the swarm.

[0013] Path planning process:

[0014] Local perception: Each aircraft acquires visual information about its surroundings through a bionic visual projection field, while also sensing the relative positions and velocities of other aircraft. Based on this information, the aircraft plans a coordinated path with other aircraft according to pre-set collaborative control rules and identity trade-off strategies.

[0015] Dynamic Adjustment: During flight, the swarm dynamically adjusts its path based on environmental changes, such as the movement of obstacles or communication interruptions. Swarm members exchange local information and, through simplified control rules, ensure the consistency and safety of the swarm's overall path.

[0016] Traversing obstacles: When the swarm encounters complex obstacles (such as ruins or forests), each aircraft will adjust its trajectory based on local perception information while considering the overall motion state of the swarm to avoid cluster splitting or mission failure.

[0017] Key technological innovations:

[0018] Dynamic path adjustment based on local information avoids dependence on global path planning.

[0019] The collaborative control algorithm of cluster members ensures the coherence of path planning and the overall consistency of the cluster.

[0020] This invention also provides an aircraft swarm traversal system based on biomimetic visual projection fields and identity tradeoffs. The system comprises multiple unmanned aerial vehicle platforms, each equipped with a visual perception module, an identity tradeoff decision module, a local path planning module, and a swarm collaborative control module. These modules utilize local information exchange and biomimetic algorithms to ensure efficient coordination and autonomous mission execution within the swarm in complex environments.

[0021] Visual perception module: Each aircraft is equipped with a visual perception system to obtain real-time data of the surrounding environment (such as obstacles, the location of other aircraft, etc.).

[0022] Identity trade-off decision module: Dynamically adjusts the role of the aircraft in the cluster (leader, follower, or collaborator) based on the aircraft's mission, location, and environmental information.

[0023] Local path planning module: plans the aircraft's trajectory in real time based on the aircraft's local information, bionic visual projection field, and mission requirements.

[0024] Cluster collaborative control module: Through simplified cluster collaborative control rules, it ensures the coordination and consistency between each aircraft and guarantees the stability and safety of cluster movement.

[0025] Environmental perception and information collection

[0026] Each aircraft obtains information about its surrounding environment in real time through the visual perception module, including the position, speed and obstacle distribution of neighboring aircraft.

[0027] Identity trade-offs and role allocation

[0028] Based on mission priority, environmental information, and interactions between aircraft, the identity trade-off module assigns an appropriate role to each aircraft, determines whether it is a leader or a collaborator, and dynamically adjusts it.

[0029] Path planning and collaborative control

[0030] Each aircraft plans its flight path based on local perception information and cluster control rules. The aircraft will exchange information with neighboring aircraft to ensure the consistency and stability of the entire cluster.

[0031] Cluster traversal and task execution

[0032] During the mission, the aircraft continuously updates its position and status based on the bionic visual projection field, dynamically adjusting its path as it traverses the environment. The swarm collaborates efficiently to complete the mission in a complex environment.

[0033] The method, system, and device of the present invention offer the following benefits: By combining a bionic visual projection field with an identity trade-off strategy, the present invention provides a highly adaptable and robust method for swarming aircraft, enabling the safe and efficient completion of swarming missions in unknown or communication-restricted environments. This method not only represents a technological breakthrough in traditional path planning and swarm collaboration models, but also offers greater autonomy and flexibility in practical applications, promising broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 This is a schematic diagram of a typical unmanned swarm system R&D project.

[0035] Figure 2 This is a schematic diagram of a typical application scenario of cluster traversal.

[0036] Figure 3 This is a schematic diagram of bionic algorithm verification.

[0037] Figure 4 It is a schematic diagram of social gregarious organisms and typical cluster behavior phenomena in nature.

[0038] Figure 5 It is a schematic diagram of cluster interaction types.

[0039] Figure 6 It is a schematic diagram of the motion mode of the Vicsek model.

[0040] Figure 7 This is a schematic diagram of the perception rules and movement modes of the Couzin model.

[0041] Figure 8 is the phase diagram of the potential field force model.

[0042] Figure 9It is the overall framework of cluster crossing based on bionic visual projection field in communication denied environment.

[0043] Figure 10 It is a schematic diagram of the bionic visual projection mechanism.

[0044] Figure 11 This is a schematic diagram of the impact of the bionic visual projection field on the magnitude and direction of speed.

[0045] Figure 12 This is a schematic diagram of the cluster crossing mission scenario.

[0046] Figure 13 Schematic diagram of the parameter optimization framework based on genetic methods.

[0047] Figure 14 This is a schematic diagram of the cluster crossing method based on the visual projection field.

[0048] Figure 15 It is the movement trajectory and quantitative index of drone clusters in different canyons.

[0049] Figure 16 This is a schematic diagram of the indoor test system for drone clusters.

[0050] Figure 17 This is a real picture of the semi-physical test system.

[0051] Figure 18 It is a cluster crossing test scenario based on the visual projection field.

[0052] Figure 19 This is a schematic diagram of the fusion test scenario.

[0053] Figure 20 This is a schematic diagram of the field deployment of the cluster traversal method in the fusion scenario. DETAILED DESCRIPTION

[0054] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The step numbers in the following embodiments are provided for ease of description only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted based on the understanding of those skilled in the art.

[0055] Implementation of cluster crossing method based on visual projection field

[0056] Initialization phase: reference Figure 9The "Overall Framework for Swarm Traversal in Communication-Denied Environments Based on Bionic Visual Projection Fields" determines the size of the drone swarm based on actual mission requirements and initializes the position and velocity of each drone. Based on the optimization results of previous experiments, key system parameters such as the repulsion and attraction ranges are set, and the ratio of informational and non-informational individuals is determined to complete the initial configuration of the swarm.

[0057] Visual projection field construction and calculation stage: During the flight, based on Figure 10 "Schematic diagram of the bionic visual projection mechanism." The drone uses its onboard visual sensors to collect environmental information. Using an internal proprietary algorithm (core technology protected and not yet publicly available), it constructs a visual projection field and identifies the area within the field of view (repulsion or attraction). Using specific calculation rules (protected content), it derives the projection field function value, providing a basis for subsequent decision-making.

[0058] Speed alignment and task guidance calculation phase: With the help of Figure 10 and Figure 11 "Schematic diagram of the impact of the bionic visual projection field on speed magnitude and direction." Drones use the visual projection field to determine the speed of their neighbors through a unique speed estimation strategy (protection algorithm). A speed alignment mechanism (core protection algorithm) calculates speed alignment to ensure consistent speed within the cluster. Information individuals determine their mission guidance based on pre-set mission information (such as preferred flight direction) using a specific calculation method (protection content). Non-information individuals calculate their self-propulsion according to established rules (protection content) to ultimately determine the desired speed.

[0059] During the status update phase, the drone updates its position and speed based on the desired speed, kinematic principles, and the control logic of this invention (partial core logic protection). The drone monitors the flight process to see if it has reached its destination. If it has, the drone marks its status. This cycle continues until the flight is complete.

[0060] Cluster traversal method and system function test implementation

[0061] Adaptability test implementation: refer to Figure 12 "Schematic diagram of cluster crossing mission scenario", select narrow canyon and obstacle canyon scenarios for testing. Prepare a certain number of drones to form a cluster, and randomly determine the information individuals according to the rules. Use the Optitrack optical motion capture system (combined with Figure 16 The team used a drone swarm indoor test system (see "Drone Swarm Indoor Test System Diagram") to record movement trajectories and use data analysis software to calculate metrics such as velocity correlation and minimum distance. The swarm's behavior was observed as it traversed the scene, and the data was analyzed to assess the adaptability of the method.

[0062] Robustness test implementation: Figure 12In the narrow canyon scenario shown, different drone loss scenarios were set (e.g., randomly losing different numbers of drones at specific times). Each scenario was tested multiple times, with metrics such as mission completion time and speed correlation recorded. The data was analyzed to investigate the swarm's obstacle avoidance and cohesion performance when some members were lost, and to evaluate the robustness of the method.

[0063] Scalability testing was conducted in narrow canyon and obstacle canyon scenarios, testing clusters of varying sizes and individual information ratios. Each parameter set was randomly tested multiple times, and metrics such as task completion time were recorded. The data was analyzed to determine the impact of cluster size and individual information ratio on performance, and to determine the optimal configuration strategy.

[0064] Implementation of swarm-traveling aircraft system

[0065] Implementation of open source UAV swarm flight verification test platform: When building the platform, ensure Figure 16 The network communication subsystem (workstation and router), drone swarm subsystem (such as the Tello drone), and integrated positioning subsystem (Optitrack optical motion capture system) are functioning normally. The motion capture system collects data and transmits it to the workstation via the SDK. The swarm algorithm calculates control commands on the workstation and sends them to the drone via the Tello SDK, achieving precise control.

[0066] Implementation of the hardware-in-the-loop test system: When building the system, select an appropriate computer and an onboard processor including the Pixhawk flight controller and the Nvidia NX host computer (reference Figure 17 (Image of the "Hard-in-the-Loop Test System"). CopterSim, the communication core, transmits data to the onboard module via the MAVLink protocol, simulates flight scenarios, and transmits the data to the 3D engine for visualization. The drone is equipped with a specific depth camera and an appropriate field of view to achieve all-round perception.

[0067] Adaptability and scalability task capability test implementation: Figure 18 Flight tests were conducted in narrow canyons and obstacle canyons using the "Visual Projection Field-Based Swarm Crossing Test Scenario," with varying numbers of individual information units. Movement trajectories and mission completion times were recorded, and the data was compared and analyzed to verify the adaptability and scalability of the method.

[0068] Mission capability robustness testing implementation: Simulate drone failures during flight testing (e.g., a single drone stops functioning at a specific time). Use monitoring equipment to observe the swarm's trajectory, recording metrics such as mission completion time and collisions, to verify the robustness and fault tolerance of the method.

[0069] Method Perception range verification test implementation: Set a lower visual perception range (such as adjusting the rejection and field of view perception range) in Figure 18Test in a scenario. Record metrics such as movement trajectory and task completion time, and analyze the data to verify the performance of the method in a low-perception range.

[0070] Cluster traversal method and system deployment implementation in fusion scenarios: Building Figure 19 The fusion scenario of the "fusion test scenario diagram" expands the scale of the drone cluster and reasonably sets the number of information individuals. Use multiple monitoring devices to observe the cluster movement and record Figure 20 The quantitative indicator change curve in the "Diagram of Field Deployment of Cluster Traversal Method in Fusion Scenario" verifies the effectiveness of the method in complex environments.

[0071] This invention demonstrates how a bionic visual projection field and identity trade-off strategy work together to improve the adaptability and robustness of swarm traversal missions. The above examples describe in detail the application of this invention in complex environments and post-disaster rescue missions. The accompanying figures further illustrate how each step is combined with environmental perception, identity trade-offs, and swarm collaborative control logic to ensure the efficiency and safety of drone swarm mission execution in dynamic and complex environments.

[0072] The contents of the above method embodiments are all applicable to the present system embodiments. The functions specifically implemented by the present system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0073] The contents of the above method embodiments are all applicable to the present device embodiments. The functions specifically implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0074] The above is a specific description of the preferred implementation of the present invention, but the invention is not limited to the embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for aircraft cluster crossing based on bionic visual projection field and identity trade-off, characterized in that: The steps include: Constructing an implicit heterogeneous swarming framework: A swarm of drones is divided into information individuals and non-information individuals, which share common attributes but employ distinct behavioral decision-making methods. Information individuals are responsible for collecting and transmitting mission-critical information, while non-information individuals maintain a consistent speed with the swarm through self-propulsion. When swarm members interact, they do not distinguish between neighboring individuals; there are no explicit leaders or followers. Instead, members autonomously adjust their own motion based solely on the real-time motion status of their neighbors. Constructing a visual projection field: Utilizing the drone's onboard visual sensors, a visual projection field is constructed to reflect information about obstacles and other drones in the environment. The drone's field of view is divided into multiple equal parts, and the visual projection mechanism is abstracted as a piecewise function of line of sight, with exclusion and attraction regions defined. When the line of sight in the attraction region is blocked, the visual projection function outputs 1. 0 if not blocked; -1 if the exclusion area's view is blocked. Achieving velocity alignment: A drone uses its visual projection field to determine the angle of a neighboring drone in its field of view, estimates its distance to the neighbor, and converts the neighbor's position into its own coordinate system to estimate its velocity. Based on this, a velocity alignment mechanism reduces speed differences between drones, avoids oscillation, and ensures stable flight in the swarm. Perform mission guidance and self-propulsion: Information individuals guide the cluster flight according to the preferred flight direction pre-set by the operator before takeoff, and their mission guidance function is calculated by a specific formula; non-information individuals rely on the self-propulsion mechanism to autonomously maintain the current speed direction and maintain the desired speed of the cluster, and the self-propulsion function is also calculated by the corresponding formula. Update the drone status: Based on the results calculated in the above steps, update the drone's position and speed status at the next moment, and continue the cycle until the crossing mission is completed.

2. The method for aircraft cluster crossing based on bionic visual projection field and identity trade-off according to claim 1 is characterized in that: When constructing the visual projection field, the exclusion area is a circular area with a radius defined by a parameter, and the attraction area is a circular area from to ; objects appearing in the exclusion area are regarded as threats and the drone needs to avoid them, while obstacles in the attraction area do not need to be processed.

3. The method for aircraft cluster crossing based on bionic visual projection field and identity trade-off according to claim 1 is characterized in that: In the speed alignment mechanism, the gain coefficient of the speed alignment effect is, the number of neighbors distinguished by the drone based on the visual projection field is, and the speed of the neighbors estimated based on the visual projection field is. Speed alignment is achieved through the formula.

4. The method for aircraft cluster crossing based on bionic visual projection field and identity trade-off according to claim 1, characterized in that: The formula for the task guidance effect of the information individual is, where is the cluster speed that the UAV attempts to maintain and is the unit vector representing the task information; the formula for the self-driving effect of the non-information individual is.

5. An aircraft cluster crossing system based on bionic visual projection field and identity trade-off, characterized by: include: UAV swarm subsystem: consists of multiple UAVs, each of which is equipped with a swarm controller, speed controller, and attitude controller. The speed controller uses the classic PID control method, and the attitude controller relies on the DJI flight control method on the Tello Edu drone. Combined positioning subsystem: uses an external auxiliary positioning system, such as the Optitrack optical motion capture system, to sense the reflective markers on the drone through multiple infrared cameras to obtain the drone's position and attitude information in real time. Network communication and cloud control subsystem: It consists of workstations and routers and is responsible for communication between drones and data exchange between drones and motion capture systems. Semi-physical test subsystem: used to verify large-scale UAV swarm crossing missions, it virtualizes the basic functions, some special functions and actual operating environment of the swarm system, and physicalizes functions closely related to swarm control; it consists of multiple computers and airborne processors, realizes data transmission and control closed loop through a specific communication architecture, and simulates outdoor flight scenarios.

6. The aircraft swarm crossing system based on bionic visual projection field and identity trade-off according to claim 5 is characterized in that: The UAV is equipped with a visual sensor for constructing a visual projection field and realizing various functions of the cluster crossing method based on the visual projection field.

7. The aircraft swarm crossing system based on bionic visual projection field and identity trade-off according to claim 5, characterized in that: In the semi-physical test subsystem, each drone is equipped with multiple depth cameras with different field of view angles, which are used to implement a perception module based on a bionic vision mechanism.

8. The application of visual projection field in aircraft cluster crossing is characterized by: The visual sensor is used to generate a visual projection field, through which the repulsion, attraction and velocity alignment of the aircraft cluster are realized, thereby realizing the cluster's crossing and obstacle avoidance functions in a communication denial environment.

9. Application of implicit heterogeneous cluster control framework in aircraft cluster crossing, characterized by: Based on the division of information individuals and non-information individuals, the autonomous crossing mission of aircraft clusters in complex environments is realized through the implicit identity UAV control method.

10. The application of bionic visual perception and motion alignment mechanism in aircraft cluster crossing is characterized by: The aircraft's visual perception is used to achieve speed alignment and motion decision-making, especially in communication denial environments. By simulating the visual perception and flocking behavior of birds, exclusion and attraction areas are defined to ensure that the aircraft can safely avoid obstacles and maintain cluster coordinated flight in complex environments.

Images