Unmanned aerial vehicle formation flight control method and device, unmanned aerial vehicle and storage medium

Abstract

This invention discloses a method, apparatus, UAV, and storage medium for UAV formation flight control. The method includes: acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs within the formation; determining a control mode based on the flight mission command data, wherein the control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode; based on the determination result of the control mode, using an L1 guidance algorithm based on adaptive trajectory tracking, integrating UAV spacing control logic and / or formation control logic for dynamic adjustment of the formation, obtaining control commands for the desired trajectory and / or attitude, and executing the control commands; cyclically executing the above steps to ensure the continuity and dynamic adaptability of UAV and its formation flight. The above solution adopts a fully decentralized distributed control architecture with a fault-tolerant mechanism, significantly improving the reliability and resilience of the formation system.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, specifically to a UAV formation flight control method, device, UAV, and storage medium. Background Technology

[0002] Vertical takeoff and landing (VTOL) fixed-wing unmanned aerial vehicles (UAVs) combine the site adaptability of multi-rotor VTOL with the advantages of fixed-wing long endurance and high speed, making them core operational equipment in fields such as power line inspection, geographic mapping, and emergency search and rescue. In large-scale, high-precision mission scenarios (such as cross-regional power line inspection and large-area terrain mapping), multi-UAV formation operations can significantly improve mission efficiency—for example, a 6-UAV formation can cover 4-6 times more area than a single UAV, and can reduce operational errors through data complementarity among multiple UAVs. Therefore, formation cooperative control technology has become a key direction for the large-scale application of VTOL fixed-wing UAVs.

[0003] Existing vertical takeoff and landing (VTOL) fixed-wing UAV formation control technologies have significant shortcomings: centralized control schemes rely excessively on central nodes such as ground stations or lead aircraft. If the central node fails (e.g., GPS interruption or communication failure), the entire formation is prone to disintegration, and command relay is prone to delays during the switching between vertical takeoff and landing and horizontal cruise modes, leading to configuration deviations. A few distributed control schemes either have complex algorithms that require high-performance hardware support, making deployment difficult and costly, or they are not adapted to cross-flight mode requirements and can only maintain formation in a single mode, making configuration disarrays easy to occur during switching. At the same time, in terms of communication, star-shaped networking limits the "decentralized" advantages of distributed systems. Traditional data transmission radios have high latency (≥100ms) and narrow bandwidth (≤0.8Mbps), which cannot meet the real-time interaction requirements of distributed control for the position and attitude information of neighboring nodes. Summary of the Invention

[0004] In view of the above problems, the present invention is proposed to provide a method, apparatus, drone, and storage medium for drone formation flight control that overcomes or at least partially solves the above problems.

[0005] According to one aspect of the present invention, a method for controlling unmanned aerial vehicle (UAV) formation flight is provided, wherein each UAV within the formation is executed separately, the method comprising: Acquire or update formation parameters, flight mission command data, and flight status data of other UAVs in the formation; The control mode is determined based on the flight mission command data. The control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode. Based on the judgment result of the control mode, the L1 guidance algorithm based on adaptive trajectory tracking is used to integrate the UAV spacing control logic and / or formation control logic for dynamic adjustment of formation to obtain the control command for the desired trajectory and / or attitude, and the control command is executed. Repeat the above steps to ensure the continuity and dynamic adaptability of the drone and its formation flight.

[0006] In some implementations, acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs within the formation includes: The system uses a Kalman filter algorithm to fuse GPS data and IMU sensor data from drones to remove noise. At the same time, a prediction and compensation algorithm based on historical trajectories is used to estimate the status data of other drones in the formation that have been lost.

[0007] In some implementations, under normal control enabled and conventional flight mode, the L1 guidance algorithm based on adaptive trajectory tracking includes: Select a reference point with a distance L1 on the desired path, wherein the distance L1 of the reference point is selected from one of the following: minimum L1 distance, set as a specific multiple of the UAV's minimum turning radius to ensure a smooth trajectory during low-speed take-off and landing; maximum L1 distance, set according to the mission cruise speed to adapt to the response requirements when the UAV flies at a speed higher than a preset threshold; normal L1, determined by the product of the current UAV's flight speed and gain coefficient; and / or, Correcting when a UAV veers off the target flight path includes: calculating the lateral deviation distance of the UAV relative to the target flight path using a vector cross product algorithm; when the lateral deviation distance is less than or equal to a preset lateral deviation threshold, approaching the target flight path using a first gain rate based on the Sigmoid function; when the lateral deviation distance exceeds the preset lateral deviation threshold, approaching the target flight path using a second gain rate based on the Sigmoid function, wherein the first gain rate is less than the second gain rate; and / or, Planning the turning trajectory of the UAV includes: continuously detecting the angle between flight segments, and determining the target flight segment as a turning segment when the angle exceeds a preset angle threshold; in the turning segment, determining the pre-turn trigger distance based on the current flight speed of the UAV, and generating a transition trajectory between the two flight segments using circular interpolation, with the interpolation radius not less than a preset multiple of the minimum turning radius of the UAV; and realizing the turn based on the aforementioned turning trigger distance and transition trajectory.

[0008] In some implementations, the UAV spacing control logic and / or formation control logic for dynamic formation adjustment includes: The drone formation switching logic includes: upon receiving a ground station command or when formation switching conditions are triggered, performing a formation switch based on formation parameters in a pre-built formation database; during formation switching, using a linear interpolation algorithm to set a reasonable configuration transition duration; and / or, The UAV longitudinal distance control logic includes: calculating the desired longitudinal distance with neighboring UAVs based on the wingman's sequence number in the formation; obtaining the actual longitudinal distance by performing a dot product of the neighboring UAV's real-time position vector and its current heading; determining the distance adjustment amount based on the product of the difference and a proportional coefficient when the difference between the actual and desired longitudinal distance exceeds a preset dead zone value; simultaneously setting a safe flight speed range for the UAV to ensure it flies within that range; and / or, The multi-drone speed coordination logic includes: taking the standard flight speed of the formation as the benchmark, accumulating the speed adjustment amounts of all neighboring drones to obtain the total adjustment command, thereby ensuring the precise spacing between the wingman and the lead drone while avoiding adjustment conflicts between multiple drones.

[0009] In some embodiments, the method further includes: By integrating artificial potential field method and speed obstacle method, a multi-level collision avoidance mechanism is constructed to achieve inter-machine collision avoidance in emergency situations, while also taking into account formation configuration maintenance. The design logic of the repulsive field in the artificial potential field method includes: treating each UAV as a repulsive source, forming a repulsive field that dynamically changes with distance around it; when the distance between two UAVs is less than a preset safe distance, the repulsive force increases non-linearly as the distance decreases; the direction of the repulsive force is along the line connecting the two UAVs, pointing away from each other, and is superimposed on the formation control commands to form a collision avoidance correction; a direction weighting coefficient is introduced to appropriately reduce the repulsive force gain for adjacent UAVs flying in the same direction, and increase the repulsive force gain for adjacent UAVs flying in opposite directions; and / or, The dynamic collision zone prediction logic of the speed obstacle method includes: real-time calculation of the range of possible collision directions within several control cycles based on the position, speed, and heading of neighboring aircraft, with the local aircraft as the origin, according to the relative speed vector of the neighboring aircraft and a preset collision radius, thereby constructing a speed obstacle zone; selecting feasible speeds outside the speed obstacle zone through a speed planning algorithm to ensure that the relative motion trajectories of the local aircraft and neighboring aircraft always maintain a safe distance; and / or, The logic of the collision avoidance and formation coordination mechanism includes: setting collision avoidance priority higher than formation control; when the detection distance is less than the safety threshold or when entering the speed obstacle area, suspending the normal formation configuration control and executing the collision avoidance command first; after the collision avoidance is completed, gradually restoring to the original formation position through a smooth transition algorithm to avoid the collision avoidance action from severely impacting the overall formation.

[0010] In some implementations, the calculation mode includes: calculating the target position and control variables but not outputting them to the UAV's flight control system; The pause-and-circle mode includes: recording the current position as the circling center point, fixing the target waypoint as the circling center point and reducing the airspeed to the minimum flight speed to form a stable circling trajectory; The disbanding mode includes: when the disbanding formation is triggered, each drone disperses along a preset safe course, increasing the distance between them to a safe range.

[0011] In some embodiments, the method further includes: External disturbances are estimated using wind field observers, and compensation terms are added to the control commands; robust control theory is used to design the controller to reduce the impact of parameter variations and model uncertainties on the system; and / or, Fault Tolerance Mechanism: Establish a multi-layered fault tolerance system, including sensor fault detection, communication fault handling, and controller fault switching, to ensure that a single point of failure does not affect the safe operation of the entire formation system. Combined with Lyapunov stability theory analysis, ensure stable operation under various scenarios; and / or, Transmission frequencies are set according to data priority, with position or speed information transmitted at high frequencies and static information such as formation configuration transmitted at low frequencies. Differential coding technology is used to compress data volume to adapt to the bandwidth requirements of mesh networking; and / or, Each UAV within the formation operates an independent control node, using the ROS2 framework and data distribution service to complete status interactions; and / or, Behavior tree nodes and cluster formation nodes are constructed within the UAV. The behavior tree nodes transmit mission phase status and formation control start / stop information to the cluster formation nodes, which in turn issue control commands to the flight control node, thereby achieving information interaction and closed-loop management of formation control; and / or, When communication is interrupted, timeout detection and state prediction are performed; when sensors malfunction, automatic switching to stand-alone safety mode is initiated; and formation is gradually restored after collision avoidance. Combined with adaptive parameter adjustment and interference compensation, stable operation in complex environments is ensured; and / or, Create a timer and trigger the control logic based on the timer.

[0012] According to another aspect of the present invention, a drone formation flight control device is provided, which is executed by each drone in the formation, the device comprising: The acquisition module is suitable for acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs in the formation. The judgment module is adapted to determine the control mode based on the flight mission command data. The control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode. The tracking module is adapted to obtain control commands for the desired trajectory and / or attitude based on the judgment result of the control mode, using the L1 guidance algorithm for adaptive trajectory tracking, and integrating the UAV spacing control logic and / or formation control logic for dynamic adjustment of formation, and to execute the control commands. The loop module is suitable for repeatedly executing the above steps to ensure the continuity and dynamic adaptability of UAVs and their formation flights.

[0013] According to another aspect of the present invention, a drone is provided, comprising: a processor and a memory arranged to store computer-executable instructions, which, when executed, cause the processor to perform a drone formation flight control method according to any one of the above embodiments.

[0014] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores one or more programs, which, when executed by a processor, implement the unmanned aerial vehicle (UAV) formation flight control method according to any one of the above.

[0015] As can be seen from the above, the UAV formation flight control method disclosed in this invention constructs a fully decentralized distributed control architecture of "distributed architecture + precise control + dynamic coordination + active collision avoidance + full-scenario management" and is equipped with a fault tolerance mechanism, which significantly improves the reliability and resilience of the formation system.

[0016] Furthermore, the above technical solution is based on the ROS2 communication framework and combines the decentralized characteristics of mesh networking to achieve autonomous and stable flight of multiple aircraft and emergency collision avoidance.

[0017] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and in order to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A flowchart illustrating a method for controlling unmanned aerial vehicle (UAV) formation flight according to an embodiment of the present invention is shown. Figure 2 A schematic diagram illustrating the process of communication between a ground station and a drone according to an embodiment of the present invention is shown; Figure 3A flowchart illustrating the start-up control process of a single unmanned aerial vehicle (UAV) according to an embodiment of the present invention is shown. Figure 4 A schematic diagram of a drone formation flight control device according to an embodiment of the present invention is shown; Figure 5 A schematic diagram of the structure of a drone (controller, flight controller) according to an embodiment of the present invention is shown. Detailed Implementation

[0019] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0020] Terminology Explanation: L1 guidance algorithm: The "intelligent guide" for drones. By selecting a reference point on the desired path, it helps the drone calculate how to turn and accelerate, thereby accurately tracking the trajectory. The steps include: 1. Selecting a reference point: Select a point at a distance of L1 on the desired path (such as a straight line or an arc) (L1 is the algorithm parameter, i.e., the "safe distance"); 2. Calculating lateral acceleration.

[0021] ROS2: Robot Operating System 2: An open-source framework for robot development. In this invention, it serves as the foundation for distributed control algorithms, providing functions such as node management and communication middleware interfaces to support standardized interaction between multiple UAV control nodes.

[0022] DDS: Data Distribution Service: A high-performance, real-time distributed data communication middleware protocol used in this invention for transmitting flight status (position, speed, etc.), formation configuration, and control commands between UAV nodes under the ROS2 framework, ensuring low latency and high reliability of distributed interaction.

[0023] GPS: Global Positioning System: A satellite positioning technology used in this invention to obtain real-time location information of UAVs, providing basic positioning data for trajectory tracking (such as L1 guidance algorithm), formation spacing calculation, and collision avoidance judgment; when combined with RTK (Real-Time Kinematic) technology, it can improve positioning accuracy and meet the position accuracy requirements of formation control.

[0024] IMU: Inertial Measurement Unit: A device that integrates sensors such as accelerometers and gyroscopes. In this invention, it is used to measure the attitude (pitch angle, roll angle, yaw angle) and motion acceleration of a UAV. By fusing Kalman filtering with GPS data, measurement noise is filtered out, providing reliable motion state input for the control algorithm.

[0025] SDK: Software Development Kit: In this invention, it refers to a standardized set of algorithm interfaces (such as waypoint correction and collision avoidance logic interfaces) provided to developers. It includes code examples, function descriptions, etc., and supports developers in adjusting control parameters (such as safety distance and correction gain) according to the performance of different models of vertical take-off and landing fixed-wing UAVs, thereby reducing the difficulty of technical adaptation.

[0026] frame (control enable flag): In this invention, it is used to identify the output enable status of formation control commands: frame=1 indicates control is enabled, and the target trajectory / attitude commands calculated by the system can be sent to the flight controller; frame=0 indicates control is disabled, and only the target position and control quantity are calculated, without outputting to the flight controller, which is used for pre-planning or monitoring scenarios.

[0027] command (flight mode instruction): In this invention, the flight mode of the UAV formation is defined as follows: command=0 / 2 corresponds to the normal flight mode (such as cruise, operation); command=1 corresponds to the pause and hover mode (the UAV hovers around the current position at the minimum flight speed); command=3 corresponds to the disband formation mode (each UAV disperses according to the preset safe heading). It is the core instruction parameter for all-scenario flight mode management.

[0028] IEEE 802.11s: A mesh network communication protocol based on the IEEE 802.11 standard. In this invention, it is one of the optional mesh networking protocols, supporting the dynamic construction of communication links between multiple UAV nodes, realizing decentralized data interaction, and providing communication support for distributed control (cooperatively adapted with ROS2 / DDS).

[0029] LoRa mesh: Long Range mesh network: A mesh networking scheme based on LoRa (long-range radio) technology. In this invention, it is used as a supplementary option to general mesh networking. It has the characteristics of low power consumption and long-distance transmission. It is suitable for convoy scenarios with high requirements for communication distance (such as cross-regional inspection) but slightly lower requirements for real-time performance. It can adapt to the communication needs of different operating environments.

[0030] Figure 1A flowchart illustrating a drone formation flight control method according to an embodiment of the present invention is shown. This method is used on drones, particularly vertical take-off and landing fixed-wing drones. The method includes the following steps: Step S110: Obtain or update formation parameters, flight mission command data, and flight status data of other UAVs in the formation; Step S120: Determine the control mode based on the flight mission command data. The control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode. Step S130: Based on the judgment result of the control mode, the L1 guidance algorithm based on adaptive trajectory tracking is used to integrate the UAV spacing control logic and / or formation control logic for dynamic adjustment of formation to obtain the control command for the desired trajectory and / or attitude, and the control command is executed. Step S140: Repeat the above steps to ensure the continuity and dynamic adaptability of the UAV and its formation flight.

[0031] Based on the UAV formation flight control method in this embodiment, a fully decentralized distributed control architecture of "distributed architecture + precise control + dynamic coordination + active collision avoidance + full-scenario management" is constructed, and a fault tolerance mechanism is provided to significantly improve the reliability and resilience of the formation system.

[0032] Specifically, the formation system adopts a completely decentralized deployment. Each vertical take-off and landing fixed-wing UAV is equipped with an independent computing unit and runs a complete control algorithm node. The nodes establish a standardized interactive network through ROS2's DDS communication middleware.

[0033] 1. Communication Protocol and Message Definition: Define a unified message format that covers flight status (position, speed, heading, updated at a frequency that adapts to the needs of formation coordination), formation configuration (formation type, relative offset of each aircraft), and control commands (speed, heading, altitude adjustment values) to ensure consistency and compatibility of information exchange between different nodes.

[0034] Combination Figure 2 As shown, the DDS2ROS node at the ground station transmits the PS11 / PS12 / PS13 formation parameters to the cluster formation node at the airborne computing terminal. Within the airborne computing terminal, the behavior tree node transmits the O6 mission phase status and BC4 formation control start / stop to the cluster formation node. The flight control node feeds back the BI3 flight control status to the cluster formation node, and the cluster formation node then issues the BC5 Offboard control command to the flight control node, thereby realizing information interaction and closed-loop management of formation control.

[0035] 2. Fault-tolerant mechanism design: Communication timeout detection is achieved through a high-precision timer that meets the requirements of real-time control. When the signal interruption of the neighboring machine is detected to exceed the preset threshold, the safety mode is automatically triggered. At the same time, a polynomial prediction compensation technology based on historical trajectory is adopted. Combined with the position change trend of the UAV in the previous few control cycles, the lost neighboring machine status data is estimated in real time to avoid formation instability caused by information loss.

[0036] 3. Task scheduling strategy: Within a single control cycle on the UAV, non-blocking task scheduling is executed in the order of "data update → target calculation → speed planning → command issuance" to ensure that there are no timing conflicts in each link and to guarantee the real-time performance of the system.

[0037] Furthermore, for each drone within the formation system, combined with Figure 3 As shown, in the initialization and data preparation phase, the "Vehicle node initialization" is executed first to provide a running platform for the multi-UAV formation control algorithm; then, "subscribing to various topics" is executed to obtain distributed interactive data such as multi-UAV flight status (position, speed, heading, etc.) and formation configuration parameters under the ROS2 / DDS communication framework, providing an input basis for the subsequent control algorithm; then, "creating a timer" is executed to match the precise period required for the real-time coordination of the formation and trigger the subsequent control logic.

[0038] During the data update and preprocessing stage, after the timer triggers "pause_callback", "data_update" is executed: GPS and IMU sensor data are fused through the Kalman filter algorithm to filter out noise. At the same time, based on the polynomial prediction compensation technology of historical trajectory, the lost state data of neighboring machines is estimated in real time (reflecting the fault tolerance mechanism of distributed control), providing reliable multi-machine state input for control decisions.

[0039] During the control mode branch decision phase, the branch logic of "judging the control mode" is entered (corresponding to the state machine design of full-scenario flight mode management): If the control mode is frame=1&command=0 / 2 (normal control enabled and in normal flight mode): "caculateTarget" is executed, which integrates the L1 guidance algorithm of adaptive trajectory tracking (including adaptive look-ahead distance and sideslip correction) and the inter-aircraft spacing / formation control logic of dynamic formation coordination to obtain the UAV's desired flight target; then, the desired trajectory, attitude and other instructions are sent to the flight control system through "Publish_cmd" to drive the UAV to perform formation actions.

[0040] If the control mode is frame=0&command=0 / 2 (calculation mode only): "caculateTarget" is executed, but the output is "Publish_waypoint" which is used for formation waypoint pre-planning or ground station monitoring, and does not directly output control commands to the flight controller.

[0041] If the control mode is frame=1&command=1 (pause hovering mode): after executing “caculateTarget”, “pauseControl” is triggered. The drone generates a stable hovering trajectory centered on the current position and at the minimum flight speed to ensure flight safety in emergency scenarios.

[0042] During the cyclic execution phase, after each of the above control mode branches has been executed, it returns to the "continue execution" stage, so that the control logic iterates with a timer cycle, ensuring the continuity and dynamic adaptability of formation control, and running through the core technical logic of distributed control architecture, adaptive trajectory tracking, and full-scenario mode management.

[0043] In some optional embodiments, acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs within the formation includes: By fusing GPS data and IMU sensor data from UAVs using the Kalman filter algorithm to filter out noise, and simultaneously using a multinomial prediction compensation algorithm based on historical trajectories to estimate the state data of other lost UAVs in the formation in real time, the fault-tolerant mechanism of distributed control is demonstrated, while providing reliable multi-aircraft state input for control decisions.

[0044] In some optional embodiments, under normal control enabled and conventional flight mode, the L1 guidance algorithm based on adaptive trajectory tracking includes: Based on the L1 guidance algorithm, the design is optimized for the characteristics of vertical take-off and landing fixed wings across flight modes (vertical take-off and landing, horizontal cruise, and mode switching) to ensure trajectory tracking accuracy.

[0045] Select a reference point with a distance of L1 on the desired path. The distance L1 of the reference point is selected from one of the following: minimum L1 distance, which is set as a specific multiple of the minimum turning radius of the UAV to ensure smooth trajectory during low-speed take-off and landing; maximum L1 distance, which is set according to the mission cruise speed to adapt to the response requirements when the UAV flies at a speed higher than a preset speed threshold; and normal L1, which is determined by the product of the current flight speed of the UAV and the gain coefficient.

[0046] Specifically, adaptive look-ahead distance calculation: The L1 distance is taken as a reasonable value among the following three: minimum L1 distance (set as a specific multiple of the drone's minimum turning radius to ensure smooth trajectory during low-speed take-off and landing), maximum L1 distance (set according to the mission cruise speed to adapt to the response requirements during high-speed flight), and L1 gain coefficient × current airspeed, to balance low-speed stability and high-speed responsiveness.

[0047] Correcting when a UAV veers off the target flight path includes: calculating the lateral deviation distance of the UAV relative to the target flight path using a vector cross product algorithm; when the lateral deviation distance is less than or equal to a preset lateral deviation threshold, approaching the target flight path using a first gain rate based on the Sigmoid function; when the lateral deviation distance exceeds the preset lateral deviation threshold, approaching the target flight path using a second gain rate based on the Sigmoid function, wherein the first gain rate is less than the second gain rate.

[0048] Specifically, the lateral offset of the UAV relative to the flight path (such as the Y-axis coordinate in the local coordinate system) is calculated by cross product operation of vectors in the local coordinate system. The Sigmoid function is introduced to realize adaptive adjustment of the correction gain. When the lateral offset is less than the preset minimum value, a low gain coefficient is used to avoid frequent adjustments. When the lateral offset exceeds the preset threshold, the gain coefficient is increased to quickly converge the deviation. The correction vector is vertically superimposed on the original waypoint to generate a new target position, ensuring the trajectory fit.

[0049] Planning the turning trajectory of the UAV includes: continuously detecting the angle between flight segments, and determining the target flight segment as a turning segment when the angle exceeds a preset angle threshold; in the turning segment, determining the pre-turn trigger distance based on the current flight speed of the UAV, and generating a transition trajectory between the two flight segments using circular interpolation, with the interpolation radius not less than a preset multiple of the minimum turning radius of the UAV; and realizing the turn based on the aforementioned turning trigger distance and transition trajectory.

[0050] Specifically, the angle detection of continuous flight segments (when the angle exceeds the preset angle threshold, it is determined to be a turning segment) can be used to calculate the pre-turn trigger distance in advance based on "current airspeed × specific coefficient"; the circular interpolation method is used to generate a transition trajectory between straight flight segments, and the interpolation radius is not less than a specific multiple of the minimum turning radius of the UAV to avoid abrupt trajectory changes when switching modes.

[0051] The control logic for waypoint correction using two parameters—forward and backward distance and lateral offset—is as follows: For the relative position deviation of UAVs during formation flight, the actual forward and backward distance (projected distance along the flight path) between the UAV and its neighboring aircraft (or the formation leader) is first calculated and compared with the desired forward and backward distance to obtain the distance deviation. Then, the lateral offset of the UAV relative to the target flight path is calculated using the cross product of vectors in the local coordinate system to determine the lateral deviation. Finally, the distance deviation and lateral offset deviation are converted into waypoint correction values, superimposed on the original target waypoint, and a corrected flight waypoint is generated, achieving precise control of the formation position. Simultaneously, a dynamic adjustment strategy for deviation correction is protected: the airspeed adjustment coefficient is dynamically adjusted based on the magnitude of the forward and backward distance deviation, and the lateral correction gain is adaptively adjusted using the Sigmoid function based on the magnitude of the lateral offset deviation, avoiding control oscillations caused by small deviations and ensuring the smoothness and accuracy of waypoint correction.

[0052] In some optional embodiments, the UAV spacing control logic and / or formation control logic for dynamic formation adjustment includes: The drone formation switching logic includes: when receiving a command from the ground station or when the formation switching condition is triggered, performing formation switching according to the formation parameters in the pre-built formation database; during formation switching, a reasonable configuration transition time is set using a linear interpolation algorithm.

[0053] Among them, a standardized formation database is established, which can store the relative position offsets of commonly used formations such as wedge, rhombus, and straight line (with the formation leader aircraft as the origin, the X / Y / Z axis offset parameters of each wingman are defined); it supports formation switching through ground station commands or mission triggers, and the switching process adopts a linear interpolation algorithm and sets a reasonable configuration transition time (dynamically adjusted according to the formation size) to ensure that the wingman positions change gradually and avoid instability caused by sudden attitude changes.

[0054] The UAV longitudinal distance control logic includes: calculating the expected longitudinal distance with neighboring UAVs based on the wingman's sequence number in the formation; obtaining the actual longitudinal distance by performing a dot product operation between the neighboring UAV's real-time position vector and the current heading; determining the distance adjustment amount based on the product of the difference and a proportional coefficient when the difference between the actual longitudinal distance and the expected distance exceeds a preset dead zone value; and setting the safe flight speed range for the UAV to ensure that the UAV flies within the safe flight speed range.

[0055] The process involves calculating the desired longitudinal distance between the wingman and neighboring aircraft based on the wingman's position in the formation (different formations correspond to different distance parameters, such as distance along the flight path in a straight formation and decreasing distance in a wedge formation). The actual longitudinal distance is obtained by multiplying the real-time position vector of the neighboring aircraft with the current heading. Error control employs a proportional controller with a dead zone. When the difference between the actual distance and the desired distance exceeds the preset dead zone value, the adjustment amount is calculated according to "airspeed increment = proportional coefficient × error value" (the proportional coefficient is dynamically adjusted according to airspeed). At the same time, a safe airspeed range is set (not lower than a specific multiple of the UAV's stall speed and not exceeding a specific proportion of the cruise speed) to ensure safe and reliable speed adjustment.

[0056] The multi-drone speed coordination logic includes: taking the standard flight speed of the formation as the benchmark, accumulating the speed adjustment amounts of all neighboring drones to obtain the total adjustment command, thereby ensuring the precise spacing between the wingman and the lead drone while avoiding adjustment conflicts between multiple drones.

[0057] Based on the standard flight speed of the formation, the total adjustment command is obtained by accumulating the speed adjustments of all neighboring aircraft, which ensures the precise spacing between the single wingman and the lead aircraft while avoiding adjustment conflicts between multiple aircraft.

[0058] In some embodiments, the method further includes: By integrating artificial potential field method and speed obstacle method, a multi-level collision avoidance mechanism is constructed to achieve inter-machine collision avoidance in emergency situations, while also taking into account formation configuration maintenance. The design logic of the repulsive field in the artificial potential field method includes: treating each UAV as a repulsive source, forming a repulsive field that dynamically changes with distance around it; when the distance between two UAVs is less than the preset safe distance, the repulsive force increases nonlinearly as the distance decreases; the direction of the repulsive force is along the line connecting the two UAVs and points away from each other, and is superimposed on the formation control command to form a collision avoidance correction; a direction weighting coefficient is introduced to appropriately reduce the repulsive force gain for adjacent UAVs flying in the same direction, and increase the repulsive force gain for adjacent UAVs flying in opposite directions. The dynamic collision zone prediction logic of the speed obstacle method includes: real-time calculation of the speed direction range of possible collisions within several control cycles based on the position, speed and heading of the neighboring aircraft, with the local aircraft as the origin, according to the relative speed vector of the neighboring aircraft and the preset collision radius, thereby constructing the speed obstacle zone; and selecting feasible speeds outside the speed obstacle zone through a speed planning algorithm to ensure that the relative motion trajectories of the local aircraft and the neighboring aircraft always maintain a safe distance. The logic of the collision avoidance and formation coordination mechanism includes: setting collision avoidance priority higher than formation control; when the detection distance is less than the safety threshold or when entering the speed obstacle area, suspending the normal formation configuration control and executing the collision avoidance command first; after the collision avoidance is completed, gradually restoring to the original formation position through a smooth transition algorithm to avoid the collision avoidance action from severely impacting the overall formation.

[0059] In some embodiments, the calculation mode includes: calculating the target position and control variables but not outputting them to the flight control system of the UAV; The pause-and-circle mode includes: recording the current position as the circling center point, fixing the target waypoint as the circling center point and reducing the airspeed to the minimum flight speed to form a stable circling trajectory; The disbanding mode includes: when the disbanding formation is triggered, each drone disperses along a preset safe course, increasing the distance between them to a safe range.

[0060] In some embodiments, the method further includes: External disturbances are estimated using wind field observers, and compensation terms are added to the control commands. Robust control theory is used to design the controller to reduce the impact of parameter variations and model uncertainties on the system. Fault tolerance mechanism: Establish a multi-level fault tolerance system, including sensor fault detection, communication fault handling, and controller fault switching, to ensure that a single point of failure does not affect the safe operation of the entire formation system. Combined with Lyapunov stability theory analysis, it ensures stable operation in various scenarios. The transmission frequency is set according to data priority, with position or speed information transmitted at high frequency and static information of formation configuration transmitted at low frequency. Differential coding technology is used to compress the data volume to adapt to the bandwidth requirements of mesh networking. Each UAV in the formation operates an independent control node and completes status interaction with the help of the ROS2 framework and data distribution service; Behavior tree nodes and cluster formation nodes are constructed within the UAV. The behavior tree nodes transmit the mission stage status and formation control start / stop to the cluster formation nodes, while the cluster formation nodes issue control commands to the flight control node, thereby realizing information interaction and closed-loop management of formation control. When communication is interrupted, timeout detection and state prediction are performed; when sensors malfunction, the system automatically switches to stand-alone safety mode and gradually restores formation after collision avoidance; combined with adaptive parameter adjustment and interference compensation, it ensures stable operation in complex environments. Create a timer and trigger the control logic based on the timer.

[0061] The above embodiments will be described in detail below through a specific example: The technical solution in this embodiment addresses the collaborative needs of vertical take-off and landing fixed-wing UAV formations by constructing a complete technical system of "distributed architecture + precise control + dynamic coordination + active collision avoidance + full-scenario management." The core technical modules and details are as follows: I. Distributed Control Architecture Design The system adopts a completely decentralized deployment. Each vertical take-off and landing fixed-wing UAV is equipped with an independent computing unit and runs a complete control algorithm node. The nodes establish a standardized interactive network through ROS2's DDS communication middleware.

[0062] 1. Communication Protocol and Message Definition: Define a unified message format that covers flight status (position, speed, heading, updated at a frequency that adapts to the needs of formation coordination), formation configuration (formation type, relative offset of each aircraft), and control commands (speed, heading, altitude adjustment values) to ensure consistency and compatibility of information exchange between different nodes.

[0063] 2. Fault-tolerant mechanism design: Communication timeout detection is achieved through a high-precision timer that meets the requirements of real-time control. When the signal interruption of the neighboring machine is detected to exceed the preset threshold, the safety mode is automatically triggered. At the same time, a polynomial prediction compensation technology based on historical trajectory is adopted. Combined with the position change trend of the UAV in the previous few control cycles, the lost state data of the neighboring machine is estimated in real time to avoid formation instability caused by information loss.

[0064] 3. Task scheduling strategy: Within a single control cycle, non-blocking task scheduling is executed in the order of "data update → target calculation → speed planning → instruction issuance" to ensure that there are no timing conflicts in each stage and to guarantee the real-time performance of the system.

[0065] II. Adaptive Tracking Technology Based on the L1 guidance algorithm, the design is optimized for the characteristics of vertical take-off and landing fixed wings across flight modes (vertical take-off and landing, horizontal cruise, and mode switching) to ensure trajectory tracking accuracy.

[0066] 1. Adaptive look-ahead distance calculation: The L1 distance is a reasonable value among the following three: minimum L1 distance (set as a specific multiple of the UAV’s minimum turning radius to ensure smooth trajectory during low-speed take-off and landing), maximum L1 distance (set according to the mission cruise speed to adapt to the response requirements during high-speed flight), and L1 gain coefficient × current airspeed, to balance low-speed stability and high-speed responsiveness.

[0067] 2. Side deviation correction mechanism: The side deviation distance of the UAV relative to the flight path is calculated by cross product operation of vectors in the local coordinate system (e.g., the Y-axis coordinate in the local coordinate system); the Sigmoid function is introduced to realize adaptive adjustment of correction gain. When the side deviation distance is less than the preset minimum value, a low gain coefficient is used to avoid frequent adjustments; when the side deviation distance exceeds the preset threshold, the gain coefficient is increased to quickly converge the deviation. The correction vector is vertically superimposed on the original waypoint to generate a new target position, ensuring trajectory fit.

[0068] 3. Pre-turn trajectory planning: By detecting the angle between consecutive flight segments (when the angle exceeds the preset angle threshold, it is determined to be a turning segment), the pre-turn trigger distance is calculated in advance based on "current airspeed × specific coefficient"; the circular interpolation method is used to generate a transition trajectory between straight flight segments, and the interpolation radius is not less than a specific multiple of the minimum turning radius of the UAV to avoid abrupt trajectory changes when switching modes.

[0069] III. Dynamic Formation Coordination Technology Focusing on "configuration maintenance + spacing control", the system achieves stability and flexibility in multi-aircraft cooperative flight.

[0070] 1. Formation Management and Switching: Establish a standardized formation database to store the relative position offsets of commonly used formations such as wedges, rhombuses, and straight lines (with the formation leader as the origin, defining the X / Y / Z axis offset parameters of each wingman); support formation switching via ground station commands or mission triggers. The switching process uses a linear interpolation algorithm and sets a reasonable configuration transition time (dynamically adjusted according to the formation size) to ensure gradual changes in wingman positions and avoid instability caused by sudden attitude changes.

[0071] 2. Precise longitudinal distance control: Based on the wingman's sequence number in the formation, the desired longitudinal distance with neighboring aircraft is calculated (different formations correspond to different distance parameters, such as distance along the flight path in a straight formation and decreasing distance in a wedge formation according to levels); the actual longitudinal distance is obtained by the dot product of the neighboring aircraft's real-time position vector and the current heading; error control adopts a proportional controller with dead zone. When the difference between the actual distance and the desired distance exceeds the preset dead zone value, the adjustment amount is calculated according to "airspeed increment = proportional coefficient × error value" (the proportional coefficient is dynamically adjusted according to airspeed). At the same time, a safe airspeed range is set (not lower than a specific multiple of the UAV's stall speed and not exceeding a specific proportion of the cruise speed) to ensure safe and reliable speed adjustment.

[0072] 3. Multi-aircraft speed coordination: Based on the standard formation flight speed, the total adjustment command is obtained by accumulating the corresponding speed adjustments of all neighboring aircraft, which ensures the accurate spacing between the wingman and the lead aircraft and avoids adjustment conflicts between multiple aircraft.

[0073] IV. Inter-machine collision avoidance technology By integrating artificial potential field method and speed obstacle method, a multi-level collision avoidance mechanism is constructed to achieve inter-machine collision avoidance in emergency situations, while also taking into account formation configuration maintenance.

[0074] 1. Repulsive Field Design Using Artificial Potential Field Method: Each UAV is considered a "repulsive source," forming a dynamically changing repulsive field around it that varies with distance. When the distance between two UAVs is less than a preset safe distance, the repulsive force increases non-linearly as the distance decreases (using the inverse square relationship of distance to enhance close-range collision avoidance response). The direction of the repulsive force points away from the other UAV along the line connecting the two UAVs, and is superimposed on the original formation control commands to form a collision avoidance correction. Simultaneously, a "direction weighting coefficient" is introduced to appropriately reduce the repulsive force gain for adjacent UAVs flying in the same direction (to avoid excessive collision avoidance disrupting formation), and to increase the repulsive force gain for adjacent UAVs flying in opposite directions (to enhance emergency response).

[0075] 2. Dynamic collision zone prediction using the speed obstacle method: Based on the position, speed, and heading of neighboring aircraft in real time, a "speed obstacle zone" is constructed. Taking the local aircraft as the origin, the range of speed directions in which a collision may occur within several future control cycles is calculated according to the relative speed vector of the neighboring aircraft and the preset collision radius. A feasible speed is selected outside the obstacle zone through a speed planning algorithm (such as sampling-evaluation-selection) to ensure that the relative motion trajectory between the local aircraft and the neighboring aircraft always maintains a safe distance.

[0076] 3. Collision avoidance and formation coordination mechanism: Collision avoidance is set to take priority over formation control. When a collision risk is detected (distance is less than the safety threshold or the speed obstacle zone is entered), the normal formation configuration control is suspended and the collision avoidance command is executed first. After the collision avoidance is completed (distance is restored to more than 1.5 times the safety distance), the formation is gradually restored to its original position through a smooth transition algorithm to avoid the collision avoidance action from having a severe impact on the overall formation.

[0077] V. All-Scenario Flight Mode Management State machine logic enables multi-scenario coverage, ensuring flight safety and controllability at different mission stages.

[0078] 1. Core State Definition and Transition: Four core states are defined: normal flight, calculation only, paused circling, and disbanded formation. Deterministic transitions are achieved based on "control enable flag (frame) + flight mode command (command)"—when frame=0, the system only calculates the target position and control variables but does not output them to the flight controller; when frame=1, normal control is executed; when command=1 triggers paused circling, the current position is recorded as the circling center, the target waypoint is fixed as the center point, and the airspeed is reduced to the minimum flight speed to form a stable circling trajectory; when command=3 triggers disbanded formation, each aircraft automatically disperses according to the preset safe heading, increasing the spacing to a safe range.

[0079] 2. Safety mode fallback mechanism: Real-time monitoring of communication links, sensor data (GPS positioning accuracy, IMU attitude stability) and the rationality of control commands. When GPS signal loss is detected (positioning accuracy exceeds the preset threshold) or there is no neighboring aircraft data for several consecutive control cycles, the formation coordination function is automatically stopped and switched to single-aircraft track tracking mode to maintain basic flight capability. After the anomaly is resolved, the formation function is gradually restarted through a progressive recovery strategy to avoid impacting system stability.

[0080] VI. Data Processing and Interface Adaptation Enhance system usability and compatibility, and ensure data validity and multi-device collaboration capabilities.

[0081] 1. Sensor data fusion: Kalman filtering algorithm is used to process GPS and IMU data to filter out measurement noise; a data validity verification mechanism is established to ensure the data quality input to the control algorithm through data range check, rate of change check, and consistency check.

[0082] 2. Communication optimization design: The transmission frequency is set according to data priority (critical information such as position / speed is transmitted at high frequency, and static information such as formation configuration is transmitted at low frequency). Differential coding technology is used to compress the data volume and adapt to the bandwidth requirements of mesh networking.

[0083] 3. Multi-interface compatibility: The interface design is compatible with mainstream flight control systems and supports standardized conversion of control commands (including coordinate system conversion, unit conversion, and command limiting). The ground station interface provides graphical mission planning, real-time status monitoring, and command issuance functions. At the same time, through modular architecture and plug-in interfaces, expansion space is reserved, which can flexibly integrate new formation algorithms or adapt to different models of vertical take-off and landing fixed-wing UAVs.

[0084] VII. Optimization of Control Accuracy and Stability Improve system performance through multiple technical means to meet the needs of complex operation scenarios.

[0085] 1. Parameter adaptive adjustment: Based on the flight environment (such as wind field) and mission requirements, the control parameters (including PID parameters, L1 gain coefficient, and speed limit value) are dynamically adjusted to ensure control accuracy under different operating conditions.

[0086] 2. Disturbance compensation and robust design: External disturbances (such as crosswinds) are estimated by wind field observers and compensation terms are added to the control commands; robust control theory is used to design the controller to reduce the impact of parameter changes and model uncertainties on the system.

[0087] 3. Fault Tolerance Mechanism: Establish a multi-level fault tolerance system covering sensor fault detection, communication fault handling, and controller fault switching to ensure that a single point of failure does not affect the safe operation of the entire formation system. Combined with Lyapunov stability theory analysis, ensure the stable operation of the system in various scenarios.

[0088] It should be noted that the core differences between the embodiments of the present invention and the prior art solutions are mainly reflected in the following four aspects, focusing on solving the bottlenecks of the prior art in terms of practicality, synergy, and reliability: I. Advantages of Distributed Control in "Lightweight Deployment" and "Cross-Mode Adaptation" Existing distributed formation control schemes are either algorithmically complex (requiring the solution of multi-machine coupled dynamics models), rely on high-performance hardware, and are difficult to adapt to small and medium-sized UAVs; or, after simplifying the algorithm, they only support a single flight mode and cannot cover the complete mission process.

[0089] This embodiment achieves lightweighting through "standardized nodes + simplified logic": each UAV only needs to run an independent control node and complete state interaction with the help of ROS2 / DDS, without the need for complex coupled calculations; at the same time, an adaptive trajectory tracking and mode switching collaborative strategy is designed for the entire formation process, which not only reduces the hardware threshold, but also solves the problems of "difficult deployment" or "poor adaptability" of existing technologies.

[0090] II. Deep Integration of Formation Coordination and Inter-Aircraft Collision Avoidance In existing technologies, formation control and collision avoidance are often designed independently: either the collision avoidance mechanism is simple (relying only on a preset safety distance), and the dynamic conflict response is lagging; or the formation completely falls apart after collision avoidance, requiring time-consuming reorganization.

[0091] This invention integrates the artificial potential field method and the velocity obstacle method to construct a "graded collision avoidance + smooth recovery" mechanism: at close range, it relies on the repulsive force field for rapid response, and in dynamic scenarios, it relies on the velocity obstacle to predict and avoid risks. After collision avoidance, the original formation is gradually restored through an interpolation algorithm, which not only solves the emergency collision avoidance problem, but also avoids the efficiency loss caused by "collision avoidance and formation separation".

[0092] III. Cooperative Optimization Design of Communication and Control In existing technologies, communication networking and formation control are mostly "loosely coupled": star networking limits the advantages of distributed systems, and traditional data transmission radios suffer from high latency and insufficient bandwidth, which can easily lead to untimely status interaction.

[0093] This invention combines the "decentralized, dynamic topology" characteristics of mesh networking with deep collaboration with distributed control: it uses DDS to distinguish high / low priority messages (location / speed is transmitted at high frequency, configuration information is transmitted at low frequency), and uses differential coding to compress data to adapt to mesh bandwidth; at the same time, it designs communication timeout prediction compensation (based on historical trajectory to estimate neighbor machine status) to solve the synchronization deviation problem caused by the "poor communication-control adaptation" in the existing technology.

[0094] IV. More comprehensive robustness protection Existing robust designs are mostly designed for single faults (such as communication interruption) and have simple recovery mechanisms (such as directly disbanding the formation).

[0095] This invention constructs a multi-level fault-tolerant system: "timeout detection + state prediction" as a double safety net when communication is interrupted, automatic switching to single-machine safety mode when sensors are abnormal, and gradual restoration of formation after collision avoidance; combined with adaptive parameter adjustment and interference compensation, it ensures stable operation in complex environments and avoids "single fault equals mission failure".

[0096] In summary, the key points and beneficial effects of the embodiments of the present invention include: 1. Decentralized distributed queue control architecture and fault tolerance mechanism A fully decentralized deployment scheme based on ROS2 / DDS: Each vertical take-off and landing fixed-wing UAV is equipped with an independent computing unit and a complete control node, and achieves decentralized interaction through standardized communication middleware; at the same time, it protects the distributed fault-tolerant method: combining communication timeout detection and historical trajectory multinomial prediction compensation technology, it can still estimate the state when the neighboring UAV signal is interrupted, ensuring the stability of the formation.

[0097] 2. Inter-machine collision avoidance technology and collaborative mechanism An inter-aircraft collision avoidance method integrating artificial potential field method and speed obstacle method: constructing a repulsive field that dynamically changes with distance using artificial potential field method (including direction weight coefficient to adapt to formation flight direction) and predicting dynamic collision area using speed obstacle method; at the same time, protecting the collaborative control strategy of collision avoidance and formation: setting collision avoidance priority and restoring formation configuration through smooth transition algorithm after collision avoidance to avoid collision avoidance action to prevent formation from being destroyed.

[0098] 3. Design of Formation Control Algorithm Based on Waypoint Correction Using Forward / Backward Distance and Lateral Deviation The control logic for waypoint correction using two parameters—forward and backward distance and lateral offset—is as follows: For the relative position deviation of UAVs during formation flight, the actual forward and backward distance (projected distance along the flight path) between the UAV and its neighboring aircraft (or the formation leader) is first calculated and compared with the desired forward and backward distance to obtain the distance deviation. Then, the lateral offset of the UAV relative to the target flight path is calculated using the cross product of vectors in the local coordinate system to determine the lateral deviation. Finally, the distance deviation and lateral offset deviation are converted into waypoint correction values, superimposed on the original target waypoint, and a corrected flight waypoint is generated, achieving precise control of the formation position. Simultaneously, a dynamic adjustment strategy for deviation correction is protected: the airspeed adjustment coefficient is dynamically adjusted based on the magnitude of the forward and backward distance deviation, and the lateral correction gain is adaptively adjusted using the Sigmoid function based on the magnitude of the lateral offset deviation, avoiding control oscillations caused by small deviations and ensuring the smoothness and accuracy of waypoint correction.

[0099] This invention uses a compound-wing (vertical take-off and landing fixed-wing) UAV as the core execution carrier, requiring a complete landing chain consisting of server-side, client-side, and development-side products. The core execution terminal, the compound-wing UAV, needs to integrate an independent computing unit (supporting ROS2 operation), a mesh communication module, and a GPS / RTK high-precision positioning module. It should be equipped with distributed control, inter-aircraft collision avoidance, and waypoint correction algorithms to meet the requirements of formation flight and collaborative operations under both vertical take-off and landing and horizontal cruise modes. The server is a formation control ground station system that can configure formation parameters (formation type, safety distance), monitor UAV position / speed and formation configuration in real time, issue task commands through an adapted communication protocol, and provide fault warning functions such as communication interruption and formation disarray. The client includes a field handheld terminal (tablet / industrial PC, supporting quick start / stop of formation, emergency disbandment, etc.) and a remote monitoring client (PC / cloud platform, for viewing cross-regional operation data), adapting to the operational needs of different scenarios. The development team provides an algorithm SDK and modular hardware adapters to facilitate rapid parameter adjustments (such as gain correction and collision avoidance thresholds) and hardware integration for different drone models. Multiple devices interact via communication links, forming a closed loop of "development and adaptation - central scheduling - operation execution - human-machine monitoring," enabling the practical application of the technology.

[0100] Combination Figure 4 As shown, the present invention also protects a drone formation flight control device 400, the drone formation flight control device 400 comprising: The acquisition module 410 is suitable for acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs in the formation. The judgment module 420 is adapted to determine the control mode based on the flight mission command data. The control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode. The tracking module 430 is adapted to obtain control commands for the desired trajectory and / or attitude based on the judgment result of the control mode, the L1 guidance algorithm for adaptive trajectory tracking, the UAV spacing control logic for dynamic formation adjustment and / or the formation control logic, and execute the control commands. The loop module 440 is suitable for repeatedly executing the above steps to ensure the continuity and dynamic adaptability of the UAV and its formation flight.

[0101] Based on the UAV formation flight control device in this embodiment, a fully decentralized distributed control architecture of "distributed architecture + precise control + dynamic coordination + active collision avoidance + full-scenario management" is constructed, and a fault tolerance mechanism is provided to significantly improve the reliability and resilience of the formation system.

[0102] In some embodiments, the acquisition module 410 acquires or updates formation parameters, flight mission command data, and flight status data of other UAVs within the formation, including: The system uses a Kalman filter algorithm to fuse GPS data and IMU sensor data from drones to remove noise. At the same time, a prediction and compensation algorithm based on historical trajectories is used to estimate the status data of other drones in the formation that have been lost.

[0103] In some embodiments, in the normal control enabled and conventional flight mode of the tracking module 430, the L1 guidance algorithm based on adaptive trajectory tracking includes: Select a reference point with a distance of L1 on the desired path. The distance L1 of the reference point is selected from one of the following: minimum L1 distance, which is set as a specific multiple of the minimum turning radius of the UAV to ensure a smooth trajectory during low-speed take-off and landing; maximum L1 distance, which is set according to the mission cruise speed to adapt to the response requirements when the UAV flies at a speed higher than a preset threshold; or normal L1 distance, which is determined by the product of the current flight speed of the UAV and the gain coefficient. Correcting when a UAV veers off the target flight path includes: calculating the lateral deviation distance of the UAV relative to the target flight path using a vector cross product algorithm; when the lateral deviation distance is less than or equal to a preset lateral deviation threshold, approaching the target flight path using a first gain rate based on the Sigmoid function; when the lateral deviation distance exceeds the preset lateral deviation threshold, approaching the target flight path using a second gain rate based on the Sigmoid function, wherein the first gain rate is less than the second gain rate. Planning the turning trajectory of the UAV includes: continuously detecting the angle between flight segments, and determining the target flight segment as a turning segment when the angle exceeds a preset angle threshold; in the turning segment, determining the pre-turn trigger distance based on the current flight speed of the UAV, and generating a transition trajectory between the two flight segments using circular interpolation, with the interpolation radius not less than a preset multiple of the minimum turning radius of the UAV; and realizing the turn based on the aforementioned turning trigger distance and transition trajectory.

[0104] In some embodiments, the UAV spacing control logic and / or formation control logic for dynamic formation adjustment in the tracking module 430 includes: The drone formation switching logic includes: when receiving a command from the ground station or when the formation switching condition is triggered, performing formation switching according to the formation parameters in the pre-built formation database; during formation switching, using a linear interpolation algorithm to set a reasonable configuration transition time; The UAV longitudinal distance control logic includes: calculating the desired longitudinal distance with neighboring UAVs based on the wingman's sequence number in the formation; obtaining the actual longitudinal distance by performing a dot product operation between the neighboring UAV's real-time position vector and the current heading; determining the distance adjustment amount based on the product of the difference and a proportional coefficient when the difference between the actual longitudinal distance and the desired distance exceeds a preset dead zone value; and setting the safe flight speed range for the UAV to ensure that the UAV flies within the safe flight speed range. The multi-drone speed coordination logic includes: taking the standard flight speed of the formation as the benchmark, accumulating the speed adjustment amounts of all neighboring drones to obtain the total adjustment command, thereby ensuring the precise spacing between the wingman and the lead drone while avoiding adjustment conflicts between multiple drones.

[0105] In some embodiments, the device is further adapted to: By integrating artificial potential field method and speed obstacle method, a multi-level collision avoidance mechanism is constructed to achieve inter-machine collision avoidance in emergency situations, while also taking into account formation configuration maintenance. The design logic of the repulsive field in the artificial potential field method includes: treating each UAV as a repulsive source, forming a repulsive field that dynamically changes with distance around it; when the distance between two UAVs is less than the preset safe distance, the repulsive force increases nonlinearly as the distance decreases; the direction of the repulsive force is along the line connecting the two UAVs and points away from each other, and is superimposed on the formation control command to form a collision avoidance correction; a direction weighting coefficient is introduced to appropriately reduce the repulsive force gain for adjacent UAVs flying in the same direction, and increase the repulsive force gain for adjacent UAVs flying in opposite directions. The dynamic collision zone prediction logic of the speed obstacle method includes: real-time calculation of the speed direction range of possible collisions within several control cycles based on the position, speed and heading of the neighboring aircraft, with the local aircraft as the origin, according to the relative speed vector of the neighboring aircraft and the preset collision radius, thereby constructing the speed obstacle zone; and selecting feasible speeds outside the speed obstacle zone through a speed planning algorithm to ensure that the relative motion trajectories of the local aircraft and the neighboring aircraft always maintain a safe distance. The logic of the collision avoidance and formation coordination mechanism includes: setting collision avoidance priority higher than formation control; when the detection distance is less than the safety threshold or when entering the speed obstacle area, suspending the normal formation configuration control and executing the collision avoidance command first; after the collision avoidance is completed, gradually restoring to the original formation position through a smooth transition algorithm to avoid the collision avoidance action from severely impacting the overall formation.

[0106] In some embodiments, the calculation mode includes: calculating the target position and control variables but not outputting them to the flight control system of the UAV; The pause-and-circle mode includes: recording the current position as the circling center point, fixing the target waypoint as the circling center point and reducing the airspeed to the minimum flight speed to form a stable circling trajectory; The disbanding mode includes: when the disbanding formation is triggered, each drone disperses along a preset safe course, increasing the distance between them to a safe range.

[0107] In some embodiments, the device is further adapted to: External disturbances are estimated using wind field observers, and compensation terms are added to the control commands. Robust control theory is used to design the controller to reduce the impact of parameter variations and model uncertainties on the system. Fault tolerance mechanism: Establish a multi-level fault tolerance system, including sensor fault detection, communication fault handling, and controller fault switching, to ensure that a single point of failure does not affect the safe operation of the entire formation system. Combined with Lyapunov stability theory analysis, it ensures stable operation in various scenarios. The transmission frequency is set according to data priority, with position or speed information transmitted at high frequency and static information of formation configuration transmitted at low frequency. Differential coding technology is used to compress the data volume to adapt to the bandwidth requirements of mesh networking. Each UAV in the formation operates an independent control node and completes status interaction with the help of the ROS2 framework and data distribution service; Behavior tree nodes and cluster formation nodes are constructed within the UAV. The behavior tree nodes transmit the mission stage status and formation control start / stop to the cluster formation nodes, while the cluster formation nodes issue control commands to the flight control node, thereby realizing information interaction and closed-loop management of formation control. When communication is interrupted, timeout detection and state prediction are performed; when sensors malfunction, the system automatically switches to stand-alone safety mode and gradually restores formation after collision avoidance; combined with adaptive parameter adjustment and interference compensation, it ensures stable operation in complex environments. Create a timer and trigger the control logic based on the timer.

[0108] It should be noted that the specific implementation methods of the above-mentioned device embodiments can be referred to the specific implementation methods of the corresponding methods described above, and will not be repeated here.

[0109] It should be noted that: The algorithms and displays provided herein are not inherently related to any particular computer, virtual device, or other equipment. Various general-purpose devices can also be used in conjunction with the teachings herein. The required structure for constructing such devices is apparent from the above description. Furthermore, this invention is not directed to any particular programming language. It should be understood that the contents of the invention described herein can be implemented using various programming languages, and the above description of specific languages ​​is for the purpose of disclosing the best mode of implementation of the invention.

[0110] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0111] Similarly, it should be understood that, in order to streamline the invention and aid in understanding one or more of the various aspects of the invention, features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof in the above description of exemplary embodiments of the invention. However, this disclosure should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim.

[0112] Those skilled in the art will understand that modules in the device of the embodiments can be adaptively changed and placed in one or more devices different from that embodiment. Modules, units, or components in the embodiments can be combined into a single module, unit, or component, and further, they can be divided into multiple sub-modules, sub-units, or sub-components. Except where at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or device so disclosed. Unless expressly stated otherwise, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0113] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features included in other embodiments but not others, combinations of features from different embodiments are meant to be within the scope of the invention and form different embodiments.

[0114] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some or all of the components in the UAV formation flight control device according to embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0115] This invention provides a non-volatile computer storage medium storing at least one executable instruction that can execute the above-described UAV formation flight control method in any of the above method embodiments.

[0116] Figure 5 The diagram shows a structural schematic of an embodiment of the UAV of the present invention. The UAV may be a vertical take-off and landing fixed-wing UAV. The specific embodiments of the present invention do not limit the specific structure of the UAV.

[0117] like Figure 5 As shown, the UAV (including a controller, control node, or flight controller) may include: a processor 502, a communications interface 504, a memory 506, and a communications bus 508.

[0118] The processor 502, communication interface 504, and memory 506 communicate with each other via communication bus 508. Communication interface 504 is used to communicate with other network elements such as clients or other servers. Processor 502 executes program 510, specifically performing the relevant steps in the above-described embodiment of the UAV formation flight control method for UAVs.

[0119] Specifically, program 510 may include program code that includes computer operation instructions.

[0120] Processor 502 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention. The one or more processors included in the drone may be processors of the same type, such as one or more CPUs; or they may be processors of different types, such as one or more CPUs and one or more ASICs.

[0121] Memory 506 is used to store program 510. Memory 506 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0122] Specifically, program 510 can be used to cause processor 502 to perform the operations corresponding to the above-described UAV formation flight control method embodiment.

[0123] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

Claims

1. A method for controlling formation flight of vertical take-off fixed-wing unmanned aerial vehicles, which is respectively executed by each unmanned aerial vehicle in the formation, characterized in that, The method includes: Acquire or update formation parameters, flight mission command data, and flight status data of other UAVs in the formation; The control mode is determined based on the flight mission command data. The control modes include: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, and disband mode. Based on the judgment result of the control mode, the L1 guidance algorithm based on adaptive trajectory tracking is used to integrate the UAV spacing control logic and / or formation control logic for dynamic adjustment of formation to obtain the control command for the desired trajectory and / or attitude, and the control command is executed. Repeat the above steps to ensure the continuity and dynamic adaptability of the drone and its formation flight; In normal control enabled and conventional flight mode, the L1 guidance algorithm based on adaptive trajectory tracking includes: Select a reference point with a distance of L1 on the desired path. The distance L1 of the reference point is selected from one of the following: minimum L1 distance, which is set as a specific multiple of the minimum turning radius of the UAV to ensure a smooth trajectory during low-speed take-off and landing; maximum L1 distance, which is set according to the mission cruise speed to adapt to the response requirements when the UAV flies at a speed higher than a preset threshold; or normal L1 distance, which is determined by the product of the current flight speed of the UAV and the gain coefficient. Correcting when a UAV veers off the target flight path includes: calculating the lateral deviation distance of the UAV relative to the target flight path using a vector cross product algorithm; when the lateral deviation distance is less than or equal to a preset lateral deviation threshold, approaching the target flight path using a first gain rate based on the Sigmoid function; when the lateral deviation distance exceeds the preset lateral deviation threshold, approaching the target flight path using a second gain rate based on the Sigmoid function, wherein the first gain rate is less than the second gain rate. Planning the turning trajectory of the UAV includes: continuously detecting the angle between flight segments, and determining the target flight segment as a turning segment when the angle exceeds a preset angle threshold; in the turning segment, determining the pre-turn trigger distance based on the current flight speed of the UAV, and generating a transition trajectory between the two flight segments using circular interpolation, with the interpolation radius not less than a preset multiple of the minimum turning radius of the UAV; and realizing the turn based on the pre-turn trigger distance and the transition trajectory.

2. The method of claim 1, wherein, Acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs within the formation includes: The system uses a Kalman filter algorithm to fuse GPS data and IMU sensor data from drones to remove noise. At the same time, a prediction and compensation algorithm based on historical trajectories is used to estimate the status data of other drones in the formation that have been lost.

3. The method according to claim 1 or 2, characterized in that, The drone spacing control logic and / or formation control logic for dynamic formation adjustment include: The drone formation switching logic includes: upon receiving a ground station command or when formation switching conditions are triggered, performing a formation switch based on formation parameters in a pre-built formation database; during formation switching, using a linear interpolation algorithm to set a reasonable configuration transition duration; and / or, The UAV longitudinal distance control logic includes: calculating the desired longitudinal distance with neighboring UAVs based on the wingman's sequence number in the formation; obtaining the actual longitudinal distance by performing a dot product of the neighboring UAV's real-time position vector and its current heading; determining the distance adjustment amount based on the product of the difference and a proportional coefficient when the difference between the actual and desired longitudinal distance exceeds a preset dead zone value; simultaneously setting a safe flight speed range for the UAV to ensure it flies within that range; and / or, The multi-drone speed coordination logic includes: taking the standard flight speed of the formation as the benchmark, accumulating the speed adjustment amounts of all neighboring drones to obtain the total adjustment command, thereby ensuring the precise spacing between the wingman and the lead drone while avoiding adjustment conflicts between multiple drones.

4. The method according to claim 1 or 2, characterized in that, The method further includes: By integrating artificial potential field method and speed obstacle method, a multi-level collision avoidance mechanism is constructed to achieve inter-machine collision avoidance in emergency situations, while also taking into account formation configuration maintenance. The design logic of the repulsive field in the artificial potential field method includes: treating each UAV as a repulsive source, forming a repulsive field that dynamically changes with distance around it; when the distance between two UAVs is less than a preset safe distance, the repulsive force increases non-linearly as the distance decreases; the direction of the repulsive force is along the line connecting the two UAVs, pointing away from each other, and is superimposed on the formation control commands to form a collision avoidance correction; a direction weighting coefficient is introduced to appropriately reduce the repulsive force gain for adjacent UAVs flying in the same direction, and increase the repulsive force gain for adjacent UAVs flying in opposite directions; and / or, The dynamic collision zone prediction logic of the speed obstacle method includes: real-time calculation of the range of possible collision directions within several control cycles based on the position, speed, and heading of neighboring aircraft, with the local aircraft as the origin, according to the relative speed vector of the neighboring aircraft and a preset collision radius, thereby constructing a speed obstacle zone; selecting feasible speeds outside the speed obstacle zone through a speed planning algorithm to ensure that the relative motion trajectories of the local aircraft and neighboring aircraft always maintain a safe distance; and / or, The logic of the collision avoidance and formation coordination mechanism includes: setting collision avoidance priority higher than formation control; when the detection distance is less than the safety threshold or when entering the speed obstacle area, suspending the normal formation configuration control and executing the collision avoidance command first; after the collision avoidance is completed, gradually restoring to the original formation position through a smooth transition algorithm to avoid the collision avoidance action from severely impacting the overall formation.

5. The method according to claim 1 or 2, characterized in that, The calculation modes include: calculating the target position and control variables but not outputting them to the UAV's flight control system; The pause-and-circle mode includes: recording the current position as the circling center point, fixing the target waypoint as the circling center point and reducing the airspeed to the minimum flight speed to form a stable circling trajectory; The disbanding mode includes: when the disbanding formation is triggered, each drone disperses along a preset safe course, increasing the distance between them to a safe range.

6. The method of claim 1 or 2, wherein, The method further includes: External disturbances are estimated using wind field observers, and compensation terms are added to the control commands; robust control theory is used to design the controller to reduce the impact of parameter variations and model uncertainties on the system; and / or, Fault Tolerance Mechanism: Establish a multi-layered fault tolerance system, including sensor fault detection, communication fault handling, and controller fault switching, to ensure that a single point of failure does not affect the safe operation of the entire formation system. Combined with Lyapunov stability theory analysis, ensure stable operation under various scenarios; and / or, Transmission frequencies are set according to data priority, with position or speed information transmitted at high frequencies and static information such as formation configuration transmitted at low frequencies. Differential coding technology is used to compress data volume to adapt to the bandwidth requirements of mesh networking; and / or, Each UAV within the formation operates an independent control node, using the ROS2 framework and data distribution service to complete status interactions; and / or, Behavior tree nodes and cluster formation nodes are constructed within the UAV. The behavior tree nodes transmit mission phase status and formation control start / stop information to the cluster formation nodes, which in turn issue control commands to the flight control node, thereby achieving information interaction and closed-loop management of formation control; and / or, When communication is interrupted, timeout detection and state prediction are performed; when sensors malfunction, automatic switching to stand-alone safety mode is initiated; and formation is gradually restored after collision avoidance. Combined with adaptive parameter adjustment and interference compensation, stable operation in complex environments is ensured; and / or, Create a timer and trigger the control logic based on the timer.

7. A vertical take-off and landing fixed-wing unmanned aerial vehicle (UAV) formation flight control device, wherein each UAV in the formation is executed separately, characterized in that, The device includes: The acquisition module is suitable for acquiring or updating formation parameters, flight mission command data, and flight status data of other UAVs in the formation. The judgment module is adapted to determine the control mode based on the flight mission command data. The control mode includes at least one of the following: normal control enabled and conventional flight mode, calculation mode, pause and hover mode, or disband mode. The tracking module is adapted to obtain control commands for the desired trajectory and / or attitude based on the judgment result of the control mode, using the L1 guidance algorithm for adaptive trajectory tracking, and integrating the UAV spacing control logic and / or formation control logic for dynamic adjustment of formation, and to execute the control commands. The loop module is suitable for repeatedly executing the above steps to ensure the continuity and dynamic adaptability of the UAV and its formation flight; In normal control enabled and conventional flight mode, the L1 guidance algorithm based on adaptive trajectory tracking includes: Select a reference point with a distance of L1 on the desired path. The distance L1 of the reference point is selected from one of the following: minimum L1 distance, which is set as a specific multiple of the minimum turning radius of the UAV to ensure a smooth trajectory during low-speed take-off and landing; maximum L1 distance, which is set according to the mission cruise speed to adapt to the response requirements when the UAV flies at a speed higher than a preset threshold; or normal L1 distance, which is determined by the product of the current flight speed of the UAV and the gain coefficient. Correcting when a UAV veers off the target flight path includes: calculating the lateral deviation distance of the UAV relative to the target flight path using a vector cross product algorithm; when the lateral deviation distance is less than or equal to a preset lateral deviation threshold, approaching the target flight path using a first gain rate based on the Sigmoid function; when the lateral deviation distance exceeds the preset lateral deviation threshold, approaching the target flight path using a second gain rate based on the Sigmoid function, wherein the first gain rate is less than the second gain rate. Planning the turning trajectory of the UAV includes: continuously detecting the angle between flight segments, and determining the target flight segment as a turning segment when the angle exceeds a preset angle threshold; in the turning segment, determining the pre-turn trigger distance based on the current flight speed of the UAV, and generating a transition trajectory between the two flight segments using circular interpolation, with the interpolation radius not less than a preset multiple of the minimum turning radius of the UAV; and realizing the turn based on the aforementioned turning trigger distance and transition trajectory.

8. A vertical take-off and landing fixed-wing unmanned aerial vehicle, characterized in that, It includes a processor and a memory arranged to store computer-executable instructions, which, when executed, cause the processor to perform the vertical take-off and landing fixed-wing unmanned aerial vehicle formation flight control method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores one or more programs that, when executed by a processor, implement the vertical take-off and landing fixed-wing unmanned aerial vehicle (UAV) formation flight control method according to any one of claims 1-6.

Images