Low-altitude unmanned aerial vehicle formation cooperative control system and method

Abstract

The application discloses a low-altitude unmanned aerial vehicle formation cooperative control system and method, relates to the technical field of unmanned aerial vehicle flight control, and comprises a multi-source perception fusion module, a formation state cooperative calculation module, a flight path dynamic planning module, a formation obstacle avoidance cooperative control module, a multi-machine communication link management and control module, a formation posture cooperative stabilization module and a central cooperative control hub. Each module realizes data intercommunication and instruction cooperation through the central cooperative control hub, completes multi-source data fusion, formation state calculation, flight path planning, two-stage obstacle avoidance, communication link management and control, and flight posture stabilization control, and realizes full-process closed-loop cooperative control of low-altitude flight of the unmanned aerial vehicle formation. The application improves the cooperative control precision and environmental adaptation capability of the low-altitude unmanned aerial vehicle formation, guarantees the formation flight stability and communication transmission reliability, realizes low-altitude airspace regulation and control, and can adapt to formation flight task demands of multiple types of low-altitude unmanned aerial vehicles.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) flight control technology, and in particular to a low-altitude UAV formation cooperative control system and method. Background Technology

[0002] Low-altitude unmanned aerial vehicle (UAV) swarm flight technology has been widely applied in various fields such as land inspection, geographic surveying, emergency rescue, and scenario demonstrations. Swarm cooperative control is the core technology for ensuring the orderly execution of flight missions by multiple UAVs, directly determining the stability of swarm flight, the accuracy of mission execution, and the safety of the flight process. In low-altitude flight scenarios, environmental conditions are characterized by densely distributed obstacles, dynamically changing meteorological elements, and a complex and diverse electromagnetic environment. Simultaneously, low-altitude airspace control requirements are subject to dynamic adjustments, placing higher demands on the environmental adaptability, dynamic response capability, and global coordination capability of UAV swarm cooperative control systems. Existing swarm cooperative control systems mostly adopt centralized or distributed control modes with fixed architectures, exhibiting significant shortcomings in adaptability to complex low-altitude scenarios. The systems have limited multi-source perception fusion dimensions, and the perceived data is only transmitted as independent inputs to subsequent control modules, failing to establish a strong coupling relationship between perceived features and swarm state calculations. This makes it difficult to comprehensively capture the dynamic changes in the low-altitude environment and cannot provide comprehensive environmental compensation basis for swarm flight control, easily leading to problems such as increased swarm formation deviation and insufficient trajectory tracking accuracy. Existing formation obstacle avoidance control mostly adopts a single-aircraft independent obstacle avoidance mode, without establishing a collaborative obstacle avoidance mechanism between overall formation scheduling and individual aircraft execution. The disorderly execution of single-aircraft obstacle avoidance actions can easily lead to flight conflicts within the formation, making it difficult to simultaneously take into account both large-scale obstacle avoidance and the overall stability of the formation. At the same time, the existing system's trajectory planning stage does not fully incorporate the dynamic control constraints of low-altitude airspace, which can easily lead to conflicts between planned trajectories and controlled airspace.

[0003] Existing UAV swarm communication links mostly employ a fixed direct connection transmission mode. In scenarios where the swarm flight range expands and the spacing between UAVs increases, communication link interruptions and increased transmission latency are prone to occur, making it difficult to guarantee the real-time synchronous transmission of swarm control commands. Furthermore, in the complex electromagnetic environment of low altitudes, the communication links lack sufficient anti-interference capabilities, easily leading to command packet loss and data transmission anomalies, directly impacting the continuity and stability of swarm coordinated control. Current swarm coordination error assessments often use a fixed-weight, single-dimensional evaluation method, failing to dynamically adjust the weight allocation of the evaluation dimension according to the core control requirements of different mission types, and further failing to incorporate real-time sensing characteristics for dynamic weight correction. This makes it difficult to accurately quantify the swarm coordinated control accuracy under different mission scenarios, and cannot provide precise numerical basis for the dynamic correction of control commands. Simultaneously, the closed-loop control process of existing swarm coordinated control systems lacks a full-process hierarchical response mechanism. For different degrees of swarm coordination deviations, it cannot execute differentiated optimization adjustment strategies, easily leading to over-control or untimely adjustments, making it difficult to guarantee the stability of the swarm flight process and the continuity of mission execution, and failing to fully adapt to the actual application needs of routine UAV swarm flights in low-altitude scenarios. Summary of the Invention

[0004] This invention proposes a low-altitude unmanned aerial vehicle (UAV) formation cooperative control system and method to solve the problems mentioned in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a low-altitude unmanned aerial vehicle (UAV) formation cooperative control system, comprising the following modules: The multi-source perception fusion module is used to collect real-time position, heading, attitude, and flight speed data of each UAV in the formation, and simultaneously collect obstacle, weather, and airspace constraint data in the low-altitude environment. It completes the spatiotemporal registration and fusion of multi-source data and outputs wind field disturbance features, visual matching confidence features, and obstacle constraint features. The formation state collaborative solution module receives wind field disturbance features, visual matching confidence features, and obstacle constraint features output by the multi-source perception fusion module, forming a strong coupling relationship between perception features and formation state solution. Based on the fused UAV state data, the module dynamically and weights the relative position deviation, heading deviation, and attitude deviation of each UAV by combining the wind field disturbance features, visual matching confidence features, and obstacle constraint features corresponding to the coupling relationship, and constructs a global formation state matrix. The trajectory dynamic planning module is used to generate a global reference trajectory for the formation based on the formation mission objectives, airspace constraints, and wind field compensation parameters. The formation obstacle avoidance collaborative control module is used to perform two-level obstacle avoidance planning based on obstacle detection data: formation-level formation reconstruction and single-unit path fine-tuning. It synchronously coordinates the timing of obstacle avoidance actions of each UAV and completes flight conflict management within the formation. The multi-machine communication link management module is used to establish a formation-based distributed self-organizing communication network, support multi-hop relay transmission, monitor link quality in real time, and automatically switch communication frequency bands and optimal relay nodes. The formation attitude coordination stabilization module is used to adjust the power output and control surface angle of each UAV based on formation control commands and wind field compensation parameters. The central coordination and control hub is used to complete data communication and command coordination among various functional modules, and to complete closed-loop control and mission scheduling for the entire formation flight process.

[0006] Furthermore, it also includes a formation coordination error quantification and evaluation unit, the calculation of which is completed using the following formula: ; The total error of drone formation coordination. This is the position deviation weighting coefficient. For drones Real-time position vector, For formation reference position vector, Let be the Euclidean distance between the two. The maximum allowable threshold for formation position deviation. This is the heading deviation weighting coefficient. For drones Real-time heading angle For formation reference heading angle, This is the maximum permissible threshold for heading deviation. This is the attitude deviation weighting coefficient. For drones Real-time attitude angle vector, For formation reference attitude angle vector, The magnitude of the attitude deviation between the two is given. This represents the maximum permissible threshold for attitude deviation. It is a non-negative smoothing constant, and + + =1, , , All are positive real numbers, and the weighting coefficients are... , , In addition to adjusting according to the formation mission type, it also performs real-time dynamic correction by combining the wind field disturbance characteristics, visual matching confidence characteristics and obstacle constraint characteristics output by the multi-source perception fusion module.

[0007] Furthermore, it also includes an airspace constraint adaptation unit, which is used to interface with low-altitude airspace control data, extract the spatial boundaries and temporal constraints of no-fly zones, restricted flight zones, and temporary controlled airspaces, transform airspace constraint parameters into hard constraints for trajectory planning and obstacle avoidance control, automatically filter trajectory schemes that cross controlled airspaces, complete the risk management of illegal flights, and simultaneously match the dynamically updated airspace control instructions at low altitudes to complete the real-time adjustment of trajectory schemes. The airspace constraint features output by the airspace constraint adaptation unit are synchronously input into the multi-source perception fusion module to participate in the feature fusion process of multi-source data.

[0008] Furthermore, the multi-source perception fusion module is equipped with a binocular vision acquisition unit, a millimeter-wave radar detection unit, a BeiDou satellite positioning unit, an inertial navigation unit, and a meteorological sensing unit. The binocular vision acquisition unit and the millimeter-wave radar detection unit work together to detect the three-dimensional spatial position and motion state of low-altitude obstacles, and to simultaneously identify static obstacles and dynamic moving targets. The BeiDou satellite positioning unit and the inertial navigation unit use a loosely coupled fusion algorithm to output centimeter-level real-time position and attitude data of the UAV. The meteorological sensing unit collects real-time wind speed, wind direction, and air pressure data, calculates low-altitude wind field disturbance parameters, and converts them into flight control compensation quantities.

[0009] Furthermore, the trajectory dynamic planning module includes a global trajectory planning subunit and a local trajectory optimization subunit. The global trajectory planning subunit uses an improved A* algorithm to generate a conflict-free global reference trajectory for the formation from the takeoff point to the mission target point. At the same time, it sets safe distance constraints for trajectory nodes in combination with formation requirements. The local trajectory optimization subunit uses a model predictive control algorithm to perform rolling optimization of the sub-trajectories of each UAV based on real-time environmental changes, formation state deviations, and wind field compensation parameters, adjusting the spatial coordinates and arrival sequence of trajectory nodes.

[0010] Furthermore, the formation obstacle avoidance collaborative control module includes a formation-level obstacle avoidance scheduling subunit and a single-unit obstacle avoidance execution subunit. The formation-level obstacle avoidance scheduling subunit adjusts the overall flight formation and trajectory for static obstacles and large-scale airspace constraints, and completes large-scale obstacle avoidance through formation reconstruction, while simultaneously updating the sub-track nodes of each UAV. The single-unit obstacle avoidance execution subunit quickly plans the obstacle avoidance path for a single UAV within the formation constraints for dynamic obstacles and sudden risks, while simultaneously synchronizing the obstacle avoidance actions and speed adjustment timing of adjacent UAVs, and managing the risk of intra-formation flight collision caused by single-unit obstacle avoidance.

[0011] Furthermore, the multi-drone communication link management module adopts a distributed self-organizing network architecture and supports multi-hop relay transmission mode. When the distance between drones in the formation exceeds the direct communication range, the data is relayed and forwarded through an intermediate drone. At the same time, the signal strength, packet loss rate and transmission delay of each communication link are monitored in real time. The multi-drone communication link management module feeds back the real-time link quality characteristics to the central collaborative control center. The central collaborative control center inputs the link quality characteristics into the multi-source perception fusion module to participate in feature fusion, providing a basis for subsequent control parameter adjustments.

[0012] Furthermore, the low-altitude UAV formation cooperative control method includes the following steps: The multi-source data acquisition and fusion steps involve simultaneously acquiring flight status and low-altitude environment perception data through various types of sensing devices carried by the formation drones. This process completes multi-source data spatiotemporal reference registration, noise filtering, outlier removal, and feature fusion, generating wind field disturbance features, visual matching confidence features, and obstacle constraint features. The formation global state calculation steps directly input wind field disturbance features, visual matching confidence features, and obstacle constraint features into the calculation process, forming a strong coupling relationship between perception features and formation state calculation. Based on the fused formation state dataset, the relative position, heading, and attitude deviations of each UAV relative to the navigator and formation reference point are dynamically weighted and corrected by combining the wind field disturbance features, visual matching confidence features, and obstacle constraint features corresponding to the coupling relationship, and a global state matrix of the formation is constructed. The formation trajectory planning steps generate a globally conflict-free reference trajectory based on mission objectives, airspace constraints, environmental feature datasets, and wind field disturbance parameters. The formation-coordinated obstacle avoidance steps, based on the detection results of the spatial position and motion state of obstacles, execute formation-level formation reconstruction and single-machine-level path fine-tuning in a hierarchical manner, and coordinate the timing and speed adjustment strategies of obstacle avoidance actions. The formation flight cooperative control steps generate attitude control and power adjustment commands for each UAV based on the global reference track, real-time formation state deviation and wind field disturbance parameters, and adjust the power output and control surface deflection angle through a closed-loop control algorithm. The communication link management steps include establishing a distributed self-organizing communication network between UAVs and ground control terminals within the formation, completing low-latency two-way transmission of data and control commands, and monitoring and optimizing the quality of the communication link in real time. The closed-loop optimization scheduling process involves real-time collection of the actual flight status and mission execution of the formation, comparison with preset targets to complete the coordination error assessment, and dynamic optimization of the trajectory planning scheme and control parameters based on the assessment results.

[0013] Furthermore, in the global state calculation step of the formation, a multi-dimensional weighted fusion calculation method is adopted for the quantitative evaluation of formation coordination error. Differentiated weight coefficients are set for position deviation, heading deviation, and attitude deviation respectively. The weight coefficients can be dynamically adjusted according to the formation mission type. At the same time, they are corrected in real time by combining wind field disturbance characteristics, visual matching confidence characteristics, and obstacle constraint characteristics. The position deviation weight coefficient is increased for low-altitude inspection missions, the attitude and heading deviation weight coefficients are increased for formation light show missions, and the speed and heading deviation weight coefficients are increased for material transportation missions.

[0014] Furthermore, in the closed-loop optimization scheduling steps, a three-level hierarchical response mechanism is constructed. When the total coordination error is lower than the preset first-level safety threshold, the current control parameters and trajectory planning scheme are maintained in stable operation. When the total coordination error is between the first-level safety threshold and the second-level warning threshold, the local trajectory optimization and control parameter fine-tuning process is initiated. By adjusting the trajectory nodes and power output parameters on a small scale, the formation coordination deviation is quickly corrected. When the total coordination error exceeds the second-level warning threshold, the formation emergency scheduling mechanism is triggered, the current mission execution is suspended, the global trajectory is replanned and the formation formation is adjusted, and real-time warning information and status data are sent to the ground control terminal. The mission is resumed after the formation status returns to stability. The execution of each level of response mechanism is based on the real-time features output by the multi-source perception fusion module and the corrected status data output by the formation status coordination calculation module.

[0015] Compared with existing technologies, the beneficial effects of this invention are: This invention, through the design of a multi-source perception fusion module, realizes the synchronous acquisition and fusion processing of multi-dimensional data on UAV flight status and low-altitude environment. It can comprehensively capture the dynamic changes of the low-altitude environment, calculate the corresponding wind field compensation parameters and obstacle motion state parameters, provide comprehensive environmental data support for formation flight control, and improve the system's adaptability to complex low-altitude environments.

[0016] This invention establishes a strong coupling relationship between perception features and formation state calculation. The wind field disturbance features, visual matching confidence features, and obstacle constraint features output by the multi-source perception fusion module are directly input into the formation state collaborative calculation module. The relative position, heading, and attitude deviations of each UAV are dynamically corrected by combining the wind field disturbance features, visual matching confidence features, and obstacle constraint features. This breaks through the technical bottleneck of independent perception and calculation in traditional systems, improves the real-time performance and accuracy of formation state calculation, and solves the problem of increased formation deviation in complex low-altitude environments.

[0017] This invention achieves hierarchical control of large-scale obstacle avoidance and sudden risk handling through the design of a two-level collaborative obstacle avoidance mechanism at the formation and individual level. It can synchronously coordinate the timing of obstacle avoidance actions of each UAV, effectively control the flight conflicts within the formation that may occur during individual UAV obstacle avoidance, and take into account both the execution efficiency of formation obstacle avoidance and the overall stability of the formation.

[0018] This invention achieves real-time monitoring and dynamic optimization of communication link quality through the design of a distributed self-organizing communication network and a multi-hop relay transmission mode. It can automatically switch communication frequency bands and optimal relay nodes according to the link status, ensuring low-latency and stable transmission of formation control commands and improving the communication reliability of the system in complex low-altitude electromagnetic environments.

[0019] This invention utilizes a dynamically adjustable weighted formation coordination error quantification and evaluation system. This system can adjust the weight allocation of each evaluation dimension according to the control requirements of different task types. At the same time, it combines real-time perception features to dynamically correct the weights, accurately quantifying the formation coordination control precision under different scenarios. This provides a reliable numerical basis for the dynamic correction of formation control commands and improves the accuracy of formation coordination control.

[0020] This invention, through the design of a three-level hierarchical response mechanism, can execute differentiated optimization and adjustment strategies for different degrees of formation coordination deviation, realizing closed-loop coordinated control of the entire formation flight process, effectively maintaining the consistency of formation and the stability of the flight process, and ensuring the continuity of formation mission execution.

[0021] This invention, through the design of an airspace constraint adaptation unit, can transform low-altitude airspace control constraints into hard conditions for trajectory planning and obstacle avoidance control, ensuring that formation flight trajectories comply with low-altitude airspace control requirements, effectively managing the risk of illegal flight, and improving the system's adaptability to dynamic airspace control instructions.

[0022] This invention, through a modular system architecture design, enables data communication and command coordination among various functional modules, and constructs a closed-loop management and control system covering the entire process from environmental perception, state calculation, trajectory planning, obstacle avoidance control to command execution. This improves the overall operational stability of the UAV formation collaborative control system and can adapt to the needs of different types of formation flight missions in low-altitude scenarios. Attached Figure Description

[0023] Figure 1 This is a schematic block diagram of a low-altitude unmanned aerial vehicle (UAV) formation cooperative control system and method proposed in this invention; Figure 2 Flowchart of hierarchical collaborative obstacle avoidance and closed-loop optimization scheduling logic; Figure 3 Flowchart for multi-source sensing fusion and low-altitude wind field compensation processing; Figure 4 Flowchart for global and local flight path planning and airspace constraint verification for formation; Figure 5 This is a logic diagram for the dynamic management and control of distributed self-organizing communication networks and links. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0026] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0027] Reference Figures 1 to 5 A low-altitude unmanned aerial vehicle (UAV) formation cooperative control system includes the following modules: The multi-source perception fusion module is used to collect real-time position, heading, attitude, and flight speed data of each UAV in the formation, and simultaneously collect obstacle, weather, and airspace constraint data in the low-altitude environment. It completes the spatiotemporal registration and fusion of multi-source data and outputs wind field compensation parameters and obstacle motion state parameters. The formation state collaborative solution module and the features output by the multi-source perception fusion module form a strong coupling relationship between perception features and formation state solution. Based on the fused UAV state data, combined with the wind field disturbance features, visual matching confidence features and obstacle constraint features corresponding to the coupling relationship, the module dynamically and weightedly corrects the relative position deviation, heading deviation and attitude deviation of each UAV, thereby quantifying the formation collaborative control accuracy. The trajectory dynamic planning module is used to generate a global reference trajectory for the formation based on the formation mission objectives, airspace constraints and wind field compensation parameters, and to assign corresponding sub-track nodes and flight timing constraints to each UAV. The formation obstacle avoidance collaborative control module is used to perform two-level obstacle avoidance planning based on obstacle detection data: formation-level formation reconstruction and single-unit path fine-tuning. It synchronously coordinates the timing of obstacle avoidance actions of each UAV and completes flight conflict management within the formation. The multi-machine communication link management module is used to establish a formation-based distributed self-organizing communication network, support multi-hop relay transmission, monitor link quality in real time and automatically switch communication frequency bands and optimal relay nodes to achieve low-latency transmission of control commands. The formation attitude coordination stabilization module is used to adjust the power output and control surface angle of each UAV based on formation control commands and wind field compensation parameters, maintain flight attitude stability, and track preset track nodes. The central coordination and control hub is used to realize data communication and command coordination among various functional modules, and to complete the closed-loop control and mission scheduling of the entire formation flight process.

[0028] This invention also includes a formation coordination error quantification and evaluation unit, used to quantify and evaluate the position, heading, and attitude coordination deviations during formation flight. The calculation process is achieved through the following formula: ; The total error of drone formation coordination. This is the position deviation weighting coefficient. For drones Real-time position vector, For formation reference position vector, Let be the Euclidean distance between the two. The maximum allowable threshold for formation position deviation. This is the heading deviation weighting coefficient. For drones Real-time heading angle For formation reference heading angle, This is the maximum permissible threshold for heading deviation. This is the attitude deviation weighting coefficient. For drones Real-time attitude angle vector, For formation reference attitude angle vector, The magnitude of the attitude deviation between the two is given. This represents the maximum permissible threshold for attitude deviation. It is a non-negative smoothing constant, and + + =1, , , All values ​​are positive real numbers. Through weighted fusion calculation of multi-dimensional deviations, real-time quantitative evaluation of formation coordinated control accuracy is achieved. The weighting coefficients can be dynamically adjusted according to the formation mission type, providing accurate numerical basis for dynamic correction of formation control commands, improving the coordinated flight stability of the formation in complex low-altitude environments, and matching the formation control accuracy requirements under different mission scenarios. , , In addition to adjusting according to the formation mission type, it also performs real-time dynamic correction by combining the wind field disturbance characteristics, visual matching confidence characteristics and obstacle constraint characteristics output by the multi-source perception fusion module.

[0029] This invention also includes an airspace constraint adaptation unit, used to interface with low-altitude airspace control data, extract the spatial boundaries and temporal constraints of no-fly zones, restricted-fly zones, and temporary controlled airspaces, transform airspace constraint parameters into hard constraints for trajectory planning and obstacle avoidance control, automatically filter trajectory schemes that cross controlled airspaces, ensure that formation flight trajectories comply with low-altitude airspace control requirements, complete the risk management of illegal flights, and simultaneously match dynamically updated low-altitude airspace control instructions to complete real-time adjustments to trajectory schemes; the airspace constraint features output by the airspace constraint adaptation unit are synchronously input into the multi-source perception fusion module to participate in the feature fusion process of multi-source data.

[0030] In this invention, the multi-source perception fusion module is equipped with a binocular vision acquisition unit, a millimeter-wave radar detection unit, a BeiDou satellite positioning unit, an inertial navigation unit, and a meteorological sensing unit. The binocular vision acquisition unit and the millimeter-wave radar detection unit work together to detect the three-dimensional spatial position and motion state of low-altitude obstacles, achieving simultaneous identification of static obstacles and dynamic moving targets. The BeiDou satellite positioning unit and the inertial navigation unit use a loosely coupled fusion algorithm to output centimeter-level real-time position and attitude data of the UAV. The meteorological sensing unit collects real-time wind speed, wind direction, and air pressure data, calculates low-altitude wind field disturbance parameters, and converts them into flight control compensation quantities to compensate for the disturbance of complex low-altitude meteorological conditions on the stability of formation flight.

[0031] In this invention, the trajectory dynamic planning module includes a global trajectory planning subunit and a local trajectory optimization subunit. The global trajectory planning subunit uses an improved A* algorithm to generate a conflict-free global reference trajectory for the formation from the takeoff point to the mission target point. At the same time, it sets safe distance constraints for trajectory nodes in combination with formation requirements. The local trajectory optimization subunit uses a model predictive control algorithm to perform rolling optimization on the sub-trajectories of each UAV based on real-time environmental changes, formation state deviations, and wind field compensation parameters. It adjusts the spatial coordinates and arrival sequence of trajectory nodes, and sets the optimization window to 100ms, so that each UAV can accurately track the global reference trajectory while maintaining the consistency of the formation. The specific implementation of the improved A* algorithm is as follows: Algorithm inputs: takeoff point coordinates, mission target point coordinates, wind field disturbance characteristics output by the multi-source perception fusion module, airspace constraint characteristics output by the airspace constraint adaptation unit, formation safety distance parameters, and maximum flight speed and maximum turning radius parameters of the UAV; Cost function definition: The total cost function F(n) = G(n) + H(n) is used to evaluate the merits of each path node n, where: G(n) is the actual movement cost from the takeoff point to the current node n, which is calculated based on the UAV's flight distance and energy consumption. H(n) is the improved heuristic function, which is the core improvement of this algorithm. Its calculation formula is as follows: ; in: This is the distance cost weighting coefficient, with a value ranging from 0.4 to 0.6; The cost is the Euclidean distance from the current node n to the target node; This is the weighting coefficient for the spatial constraint cost, with a value ranging from 0.2 to 0.4. The cost of airspace constraints is infinity if node n is in a no-fly zone, the penalty value corresponding to the restricted flight level if it is in a restricted flight zone, and 0 if it is in free airspace. The weighting coefficient for the safety distance cost of formation is 0.1-0.25; The cost of safe spacing for formation is calculated based on the minimum distance between node n and the planned flight path nodes of adjacent UAVs. When the distance is less than the safe spacing, the cost increases exponentially as the distance decreases. This is the weighting coefficient for wind field disturbance costs, with a value ranging from 0.05 to 0.15. The cost of wind field disturbance is calculated based on the wind field disturbance intensity and wind direction at the current node n output by the multi-source sensing fusion module. The cost increases when flying against the wind and decreases when flying with the wind. All weight coefficients satisfy Furthermore, it can be dynamically adjusted according to the task type; Algorithm execution steps: Initialize the open and closed lists, add the takeoff point to the open list, and record its total cost F=0; Select the node with the smallest total cost F from the open list as the current node and move it to the closed list; Expand all neighboring nodes of the current node. The number of neighboring nodes and the step size are set according to the flight accuracy requirements of the UAV. The step size ranges from 5 to 20 meters. For each adjacent node, calculate its total cost F. If the node is in the closed list, skip it; if the node is not in the open list, add it to the open list and record its parent node; if the node is already in the open list and the newly calculated total cost is smaller, update its total cost and its parent node. Repeat steps 2-4 until the target node is added to the open list or the open list is empty; If the target node is found, the parent node is traced back from the target node to generate an initial global track; if the open list is empty and no target node is found, it is determined that there is no feasible track and a warning message is sent to the ground control terminal. The initial global trajectory is smoothed, and the trajectory nodes are fitted with cubic B-spline curves to ensure that the trajectory meets the maximum turning radius constraint of the UAV. Algorithm output: Smoothed global reference track for the formation, containing the spatial coordinates and estimated arrival times of a series of track nodes.

[0032] In this invention, the formation obstacle avoidance collaborative control module includes a formation-level obstacle avoidance scheduling subunit and a single-unit obstacle avoidance execution subunit. The formation-level obstacle avoidance scheduling subunit adjusts the overall flight formation and trajectory for static obstacles and large-scale airspace constraints, and completes large-scale obstacle avoidance through formation reconstruction, while simultaneously updating the sub-track nodes of each UAV. The single-unit obstacle avoidance execution subunit quickly plans the obstacle avoidance path for a single UAV within the formation constraints for dynamic obstacles and sudden risks, with a planning response time of no more than 200ms. At the same time, it synchronizes the obstacle avoidance actions and speed adjustment timing of adjacent UAVs, sets the action execution time difference to no less than 50ms, manages the risk of intra-formation flight collision caused by single-unit obstacle avoidance, and balances obstacle avoidance efficiency and formation coordination stability.

[0033] In this invention, the multi-drone communication link management module adopts a distributed self-organizing network architecture, supporting multi-hop relay transmission mode. When the distance between UAVs in the formation exceeds the direct communication range, data relay is completed through intermediate UAVs. At the same time, the signal strength, packet loss rate, and transmission delay of each communication link are monitored in real time. When the link quality is lower than a preset threshold, the communication frequency band and the optimal relay node are automatically switched to maintain the continuity and reliability of data transmission within the formation. The end-to-end transmission delay is controlled within 50ms, and the packet loss rate is less than 1%, achieving stable transmission of formation control commands in low-altitude complex electromagnetic environments. The multi-drone communication link management module feeds back the real-time link quality characteristics to the central collaborative control center. The central collaborative control center inputs the link quality characteristics into the multi-source perception fusion module to participate in feature fusion, providing a basis for subsequent control parameter adjustments.

[0034] In this invention, the low-altitude unmanned aerial vehicle (UAV) formation cooperative control method includes the following steps: The multi-source data acquisition and fusion steps involve simultaneously acquiring real-time flight status data and low-altitude environmental perception data from various types of sensing devices carried by UAVs within the formation. This process completes spatiotemporal reference registration, noise filtering, outlier removal, and feature fusion of the multi-source data. It also calculates low-altitude wind field disturbance parameters and obstacle motion state parameters, generating standardized formation status datasets and environmental feature datasets, as well as wind field disturbance features, visual matching confidence features, and obstacle constraint features. The formation global state calculation steps are based on the fused formation state dataset. The relative position, heading, and attitude deviation of each UAV in the formation relative to the navigator and the formation reference point are calculated to construct the formation global state matrix and complete the quantitative evaluation of the formation cooperative control accuracy. The formation trajectory planning steps, based on the formation mission objectives, low-altitude airspace constraints, environmental feature datasets and wind field disturbance parameters, generate a global conflict-free reference trajectory for the formation. Combined with the preset formation and UAV performance parameters, the steps assign corresponding sub-track nodes, flight speeds and timing constraints to each UAV in the formation. The formation-coordinated obstacle avoidance steps, based on the detection results of the spatial position and motion state of obstacles, execute obstacle avoidance path planning in a hierarchical manner, including formation-level reconstruction and individual-level path fine-tuning, coordinate the obstacle avoidance action timing and speed adjustment strategies of each UAV, complete static and dynamic obstacle avoidance, and manage flight conflicts within the formation. The formation flight cooperative control steps generate attitude control and power adjustment commands for each UAV based on the global reference track, real-time formation state deviation and wind field disturbance parameters. The power output and control surface deflection angle of the UAV are adjusted through a closed-loop control algorithm, so that the UAV can accurately track the preset track nodes and maintain the stability of the formation and the consistency of flight attitude. The communication link management steps include establishing a distributed self-organizing communication network between each UAV in the formation and the ground control terminal, completing low-latency two-way transmission of formation status data and collaborative control commands, monitoring and dynamically optimizing the communication link quality in real time, automatically switching communication frequency bands and optimal relay nodes, and maintaining the continuity and reliability of the formation's full-domain communication. The closed-loop optimization scheduling process involves real-time collection of actual flight status data and mission execution information, comparison with preset mission objectives to complete collaborative error assessment, and dynamic optimization of trajectory planning schemes and flight control parameters based on the assessment results, thereby achieving closed-loop collaborative control and mission scheduling throughout the entire formation flight process.

[0035] In this invention, during the formation global state calculation step, a multi-dimensional weighted fusion calculation method is adopted for the quantitative evaluation of formation coordination error. Differentiated weight coefficients are set for position deviation, heading deviation, and attitude deviation. The weight coefficients can be dynamically adjusted according to the formation mission type. For low-altitude inspection missions, the position deviation weight coefficient is increased; for formation light show missions, the attitude and heading deviation weight coefficients are increased; and for material transportation missions, the speed and heading deviation weight coefficients are increased. This ensures that the coordination error evaluation matches the core control requirements of the actual mission, providing a reliable numerical basis for the accurate correction of subsequent flight control commands.

[0036] In this invention, a three-level hierarchical response mechanism is constructed in the closed-loop optimization scheduling step. When the total coordination error is lower than the preset first-level safety threshold, the current control parameters and trajectory planning scheme are kept stable, and only routine status monitoring is performed. When the total coordination error is between the first-level safety threshold and the second-level warning threshold, the local trajectory optimization and control parameter fine-tuning process is initiated. By adjusting the trajectory nodes and power output parameters on a small scale, the formation coordination deviation is quickly corrected, and the formation consistency is restored. When the total coordination error exceeds the second-level warning threshold, the formation emergency scheduling mechanism is triggered, the current task execution is suspended, the global trajectory is replanned, and the formation is adjusted. At the same time, real-time warning information and status data are sent to the ground control terminal. The task is resumed after the formation status is restored to stability. The execution of each level of response mechanism is based on the real-time features output by the multi-source perception fusion module and the corrected status data output by the formation status coordination calculation module.

[0037] The following two examples further illustrate the specific implementation of this system: Example 1: Collaborative Control Implementation of Low-Altitude Transmission Line Inspection UAV Squadron This embodiment is applied to a low-altitude inspection scenario of a 220kV overhead transmission line corridor. The formation consists of 4 rotary-wing UAVs, with a maximum flight time of 45 minutes for a single UAV. The inspection task is to complete a full-section inspection of the corridor environment, tower bodies, and line hardware along a pre-set transmission line corridor. The total length of the inspected line is 25 kilometers. The line corridor traverses hilly areas and has obstacles such as dense towers, trees, and temporary construction facilities along the line. At the same time, there are clear airspace control constraints in the line protection zone, and it is necessary to deal with complex environmental conditions such as low-altitude wind disturbances and dynamic bird targets.

[0038] After the system starts, the central collaborative control center completes the initialization and data communication configuration of each functional module, and the multi-source perception fusion module starts the data acquisition process. Each UAV is equipped with a binocular vision acquisition unit and a millimeter-wave radar detection unit, which simultaneously collect obstacle data in the corridor at a sampling frequency of 20Hz, complete the three-dimensional spatial positioning and motion state recognition of static towers, trees and dynamic birds and construction machinery, and generate obstacle constraint features; the Beidou satellite positioning unit and the inertial navigation unit adopt a loosely coupled fusion algorithm to output centimeter-level real-time position, heading, attitude and flight speed data of each UAV; the meteorological sensing unit collects wind speed, wind direction and air pressure data along the inspection line in real time, generates visual matching confidence features, calculates low-altitude wind field disturbance parameters and generates wind field disturbance features, and simultaneously completes the spatiotemporal reference registration and feature fusion of all collected data to generate a standardized formation state dataset and environmental feature dataset.

[0039] The multi-source perception fusion module directly outputs the generated wind field disturbance features, visual matching confidence features, and obstacle constraint features to the formation state collaborative solution module, forming a strong coupling relationship between perception features and formation state solution. The formation state collaborative solution module, based on the fused formation state dataset, uses the lead UAV as a reference benchmark and dynamically weights and corrects the relative position deviation, heading deviation, and attitude deviation of each follower UAV relative to the lead UAV and the formation reference point by combining wind field disturbance characteristics, visual matching confidence characteristics, and obstacle constraint characteristics corresponding to the coupling relationship. This constructs a global formation state matrix. The collaborative error quantification evaluation unit then quantifies and evaluates the formation collaborative control accuracy, with the position deviation weight set to be the highest, matching the core track tracking requirements of the inspection mission. Simultaneously, the weight coefficients are dynamically adjusted based on real-time wind field disturbance intensity, visual matching confidence, and obstacle density: when the wind field disturbance intensity exceeds 3 m / s, the attitude deviation weight coefficient is increased by 15%; when the visual matching confidence is below 0.7, the position deviation weight coefficient is increased by 20%; and when the obstacle density exceeds 5 obstacles per 100m, the heading deviation weight coefficient is increased by 10%.

[0040] The global trajectory planning subunit of the trajectory dynamic planning module receives the corrected global state matrix of the formation. Combining the transmission line alignment, airspace constraints of the line protection zone, and no-fly zone boundary conditions, it uses an improved A* algorithm to generate a global conflict-free reference trajectory for the formation from takeoff point to inspection endpoint. The lateral safety distance between adjacent UAV trajectory nodes is set to be no less than 30 meters, and the longitudinal safety distance to be no less than 50 meters. The local trajectory optimization subunit, based on real-time wind field disturbance parameters and formation state deviation, uses a model predictive control algorithm to perform rolling optimization of each UAV's sub-trajectory. The optimization window is set to 100ms, and the spatial coordinates and arrival sequence of trajectory nodes are adjusted in real time to ensure that each UAV accurately tracks the global reference trajectory. The airspace constraint adaptation unit connects to low-altitude air traffic control data in real time, extracts the no-fly boundaries of the line protection zone, and transforms them into hard constraints for trajectory planning. It automatically filters trajectory schemes that cross controlled areas. Simultaneously, the airspace constraint features output by the airspace constraint adaptation unit are synchronously input into the multi-source perception fusion module to participate in the feature fusion process of multi-source data.

[0041] The formation obstacle avoidance collaborative control module executes two levels of obstacle avoidance planning. The formation-level obstacle avoidance scheduling subunit receives obstacle constraint features output by the multi-source perception fusion module. For large-scale static obstacles such as densely packed power lines and tower clusters, or large groves of trees, it adjusts the overall formation from a single-line inspection formation to a trapezoidal formation, and synchronously updates the sub-track nodes of each UAV to complete large-scale obstacle avoidance. The single-aircraft obstacle avoidance execution subunit receives corrected relative deviation data output by the formation state collaborative calculation module. For dynamic targets such as sudden bird attacks or temporary construction obstacles, it quickly plans the single-aircraft obstacle avoidance path within the formation constraints, with a planning response time of no more than 200ms. At the same time, it synchronizes the obstacle avoidance actions and speed adjustment timing of adjacent UAVs, setting the action execution time difference to no less than 50ms to control flight conflicts within the formation caused by single-aircraft obstacle avoidance.

[0042] The multi-drone communication link management module establishes a formation-distributed self-organizing communication network, adopting a multi-hop relay transmission mode. When the distance between the first and last drones in the formation exceeds the direct communication range, the intermediate drones relay and forward inspection data and control commands. The module monitors the signal strength, packet loss rate, and transmission delay of each communication link in real time. When the link quality is lower than a preset threshold, it automatically switches the communication frequency band and the optimal relay node. The end-to-end transmission delay is controlled within 50ms, and the packet loss rate is less than 1%, ensuring the stable transmission of control commands in the complex electromagnetic environment of hilly areas. At the same time, the multi-drone communication link management module feeds back the real-time link quality characteristics to the central collaborative control center. The central collaborative control center inputs the link quality characteristics into the multi-source perception fusion module to participate in feature fusion, providing a basis for subsequent control parameter adjustments.

[0043] The formation attitude cooperative stabilization module, based on formation control commands and wind field compensation parameters issued by the central cooperative control center, adjusts the power output and control surface deflection angles of each UAV in real time to compensate for the impact of low-altitude wind disturbances on flight attitude, maintain UAV flight attitude stability, and accurately track preset track nodes. In the closed-loop optimization scheduling process, the system constructs a three-level hierarchical response mechanism. The execution of each level of the response mechanism is based on the real-time features output by the multi-source perception fusion module and the corrected state data output by the formation state cooperative calculation module. Differential adjustment strategies are executed based on real-time cooperative errors to complete the closed-loop cooperative control and task scheduling of the entire formation flight process.

[0044] Table 1: Comparison of Formation Control Performance in Transmission Line Inspection Scenarios

[0045] Table 1 shows the data from the statistical analysis of five consecutive inspection missions conducted by the two systems in this embodiment under the same inspection route and environmental conditions. Traditional fixed-parameter formation control systems lack wind field compensation and a two-stage cooperative obstacle avoidance mechanism. In hilly, low-altitude wind-affected environments, their formation maintenance and track tracking accuracy are insufficient, obstacle avoidance response is delayed, and communication link stability is poor, making it difficult to complete detailed inspections of the entire line. The system of this invention improves formation flight stability through multi-source sensing fusion and wind field compensation. Its two-stage obstacle avoidance mechanism enables rapid obstacle response and avoidance, and its adaptive communication link ensures the continuity of command transmission, fully adapting to the mission requirements of low-altitude power transmission line inspections.

[0046] Example 2: Example of Cooperative Control of UAVs for Urban Low-Altitude Formation Light Show This embodiment is applied to a nighttime low-altitude light show in the core urban area. The formation consists of 30 small rotary-wing UAVs, each equipped with a programmable lighting module. The show mission is to complete the preset formation pattern changes, dynamic shape performances and synchronized output of lighting effects in the low-altitude airspace of 100m-150m in the core urban area. The surrounding airspace of the show includes urban high-rise buildings, restricted airspace and temporary control areas. It is necessary to cope with the complex electromagnetic environment of the urban low-altitude airspace and the disturbance of building wind fields, while meeting the strict airspace control requirements and the synchronization requirements of the formation shape.

[0047] After the system starts, the central collaborative control center completes the initialization and data communication configuration of each functional module, and the multi-source perception fusion module starts the full-process data acquisition. The binocular vision acquisition unit and millimeter-wave radar detection unit on each UAV collect the spatial position data of high-rise buildings and low-altitude obstacles around the exhibition airspace in real time, complete the synchronous identification of static obstacles and temporary low-altitude flying targets, and generate obstacle constraint features; the Beidou satellite positioning unit and inertial navigation unit adopt a loosely coupled fusion algorithm to output the centimeter-level real-time position, heading, and attitude data of each UAV, and generate visual matching confidence features; the meteorological sensing unit collects wind speed and direction data in the low airspace of the city, calculates the wind field disturbance parameters caused by building flow around the wind and generates wind field disturbance features, completes the spatiotemporal reference registration and feature fusion of all collected data, and generates a standardized formation state dataset and environmental feature dataset.

[0048] The multi-source perception fusion module directly outputs the generated wind field disturbance features, visual matching confidence features, and obstacle constraint features to the formation state collaborative solution module, forming a strong coupling relationship between perception features and formation state solution. The formation state collaborative solution module, based on the fused formation state dataset, uses the formation virtual center point as a reference benchmark and dynamically weights and corrects the relative position deviation, heading deviation, and attitude deviation of each UAV relative to the formation reference point by combining wind field disturbance characteristics, visual matching confidence characteristics, and obstacle constraint characteristics. This constructs a global formation state matrix. The collaborative error quantification evaluation unit then quantifies and evaluates the formation collaborative control accuracy. The weighting coefficients are set to have the highest weights for attitude deviation and heading deviation, matching the core requirement of formation synchronization for light show missions. Simultaneously, the weighting coefficients are dynamically adjusted based on real-time building wind field disturbance intensity, visual matching confidence, and the strictness of airspace constraints: when the building wind field disturbance intensity exceeds 2 m / s, the attitude deviation weighting coefficient is increased by 20%; when the visual matching confidence is below 0.8, the position deviation weighting coefficient is increased by 15%; and when entering the boundary of temporary controlled airspace, the heading deviation weighting coefficient is increased by 25%.

[0049] The global trajectory planning subunit of the trajectory dynamic planning module receives the corrected global formation state matrix and, combined with the constraints of the performance airspace boundary, urban no-fly zones and restricted flight zones, and the timing of performance pattern changes, uses an improved A* algorithm to generate a global reference trajectory for the formation. It assigns corresponding pattern change nodes, trajectory coordinates, and flight timing constraints to each UAV. The local trajectory optimization subunit, based on real-time wind field disturbance parameters and formation state deviations, uses a model predictive control algorithm to perform rolling optimization of each UAV's sub-trajectories. The optimization window is set to 100ms, and the spatial coordinates and arrival timing of trajectory nodes are adjusted in real time to ensure that each UAV arrives at the preset formation position accurately and synchronously. The airspace constraint adaptation unit connects to urban low-altitude airspace control data in real time, extracts the spatial boundaries and temporal constraints of no-fly zones and temporary controlled airspaces, transforms them into hard constraints for trajectory planning, and automatically filters trajectory schemes that cross controlled airspaces. Simultaneously, the airspace constraint features output by the airspace constraint adaptation unit are synchronously input into the multi-source perception fusion module to participate in the feature fusion process of multi-source data.

[0050] The formation obstacle avoidance collaborative control module executes two-level obstacle avoidance planning. The formation-level obstacle avoidance scheduling subunit receives obstacle constraint features output by the multi-source perception fusion module. For tall buildings and fixed airspace constraints around the exhibition airspace, it adjusts the trajectory and formation boundary of the overall formation shape change. Through overall formation reconstruction, it completes large-scale fixed obstacle avoidance and updates the sub-track nodes of each UAV simultaneously. The single-UAV obstacle avoidance execution subunit receives corrected relative deviation data output by the formation state collaborative solution module. For dynamic obstacles such as sudden low-altitude flying targets, it completes rapid planning of single-UAV obstacle avoidance paths within the formation shape constraints. The planning response time does not exceed 200ms. At the same time, it synchronizes the obstacle avoidance actions and speed adjustment timing of adjacent UAVs, setting the action execution time difference to be no less than 50ms to control the risk of intra-formation flight collision caused by single-UAV obstacle avoidance.

[0051] The multi-aircraft communication link management module establishes a formation-distributed self-organizing communication network, adopting a multi-hop relay transmission mode. When the formation range expands, causing the direct communication between the edge UAVs and the ground to be interrupted, the control commands and status data are relayed and forwarded through the UAV at the formation center. The module monitors the signal strength, packet loss rate, and transmission delay of each communication link in real time. When the complex electromagnetic environment of the city causes the link quality to fall below the preset threshold, it automatically switches the communication frequency band and the optimal relay node. The end-to-end transmission delay is controlled within 50ms, and the packet loss rate is less than 1%, ensuring the stable transmission of light synchronization commands and formation control commands during the performance. At the same time, the multi-aircraft communication link management module feeds back the real-time link quality characteristics to the central collaborative control center. The central collaborative control center inputs the link quality characteristics into the multi-source perception fusion module to participate in feature fusion, providing a basis for subsequent control parameter adjustments.

[0052] The formation attitude coordination and stabilization module, based on formation control commands and wind field compensation parameters issued by the central coordination control center, adjusts the power output and control surface deflection angle of each UAV in real time to compensate for the impact of building wind field disturbances on flight attitude, maintain the stability of UAV flight attitude and light orientation, and accurately track preset track nodes. In the closed-loop optimization scheduling process, the system constructs a three-level hierarchical response mechanism. The execution of each level of the response mechanism is based on the real-time features output by the multi-source perception fusion module and the corrected state data output by the formation state coordination calculation module. Based on the real-time coordination error, differentiated adjustment strategies are executed to complete the closed-loop coordinated control and task scheduling of the entire performance.

[0053] Table 2: Comparison of Formation Control Performance in Urban Light Show Scenes

[0054] Table 2 shows data from test results of three consecutive complete performances of the two systems in this embodiment, under the same performance airspace and the same formation scheme. Traditional performance formation control systems lack dynamic weighted collaborative error assessment and adaptive communication link mechanisms. In complex urban electromagnetic environments, this results in high packet loss rates in command transmission, insufficient synchronization of formation changes, and inadequate adaptation to dynamic airspace control constraints, posing airspace compliance risks. The system of this invention improves formation synchronization accuracy through multi-dimensional weighted adjustable collaborative error assessment, ensures stable command transmission in complex electromagnetic environments through an adaptive communication link, and achieves airspace compliance control throughout the performance using an airspace constraint adaptation unit, fully adapting to the mission requirements of urban low-altitude light shows.

[0055] refer to Figure 1 This diagram illustrates the closed-loop process of the system's entire lifecycle, from environmental perception to mission execution. The process begins with multi-source perception fusion, integrating visual, radar, satellite positioning, and meteorological data to output accurate environmental and wind field parameters. This is followed by state calculation and trajectory planning, generating reference trajectories based on global mission objectives and assigning sub-nodes to each individual drone. During flight, the obstacle avoidance coordination module and communication link management ensure the safety of the formation and the stability of data transmission in complex environments. Finally, the attitude coordination stabilization module drives the UAV's propulsion system, and the central coordination control center performs error assessment and dynamic trajectory correction based on real-time feedback, achieving precise control throughout the entire process.

[0056] Reference Figure 2This diagram details the core decision-making logic of the system when dealing with sudden obstacles and formation deviations. The cooperative obstacle avoidance logic is divided into two levels: the formation level is responsible for reconfiguring formations for large-scale obstacles, while the individual aircraft level is responsible for rapid path fine-tuning for sudden dynamic targets. Regarding closed-loop optimization scheduling, the system triggers a three-level response mechanism based on the quantified cooperative error: maintaining routine monitoring for low errors; initiating track and parameter fine-tuning for medium errors; and forcibly triggering emergency scheduling for high errors, suspending the mission and replanning. This mechanism ensures that the formation can maintain a high degree of flight safety and mission continuity even under complex low-altitude electromagnetic and weather interference.

[0057] Reference Figure 3 This diagram illustrates the perception and environmental compensation logic at the system's front end. The multi-source perception module integrates binocular vision, millimeter-wave radar, BeiDou satellite, and inertial navigation data. It achieves centimeter-level positioning through a loosely coupled fusion algorithm, simultaneously identifying both static and dynamic obstacles. Specifically, the meteorological sensing unit collects real-time data such as wind direction and speed, calculates wind field disturbance parameters, and converts them into compensation quantities for flight control. This counteracts the interference of complex low-altitude weather on formation attitude stability, ensuring the formation's path tracking accuracy under wind field disturbances.

[0058] Reference Figure 4 This diagram illustrates the two-tiered architecture of trajectory planning and airspace compliance checks. The global planning subunit uses an improved algorithm to generate conflict-free reference trajectories, while the local optimization subunit utilizes model predictive control to perform rolling optimization in 100ms windows, adjusting the flight sequence of each aircraft. An airspace constraint adaptation unit is embedded in the process to automatically extract no-fly and restricted-fly zone boundaries, ensuring that all trajectory schemes comply with air traffic control requirements. This combination of long-term planning and real-time optimization guarantees a high degree of coordination within the formation during mission execution.

[0059] refer to Figure 5 This diagram illustrates the communication architecture for maintaining the reliability of formation command transmission. The system constructs a distributed self-organizing network supporting multi-hop relay. The link management module monitors the signal strength and packet loss rate of each communication link in real time. Once the quality falls below a threshold, the system automatically switches communication frequency bands or reselects the optimal relay node. Through this dynamic switching mechanism, the system can control end-to-end transmission latency to an extremely low range, ensuring that commands from the central coordination control center can be delivered to each UAV in real time, supporting formation coordination in complex environments.

[0060] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A low-altitude unmanned aerial vehicle (UAV) formation cooperative control system, characterized in that, Includes the following modules: The multi-source perception fusion module is used to collect real-time position, heading, attitude, and flight speed data of each UAV in the formation, and simultaneously collect obstacle, weather, and airspace constraint data in the low-altitude environment. It completes the spatiotemporal registration and fusion of multi-source data and outputs wind field disturbance features, visual matching confidence features, and obstacle constraint features. The formation state collaborative solution module, based on the wind field disturbance features, visual matching confidence features, and obstacle constraint features output by the multi-source perception fusion module, forms a strong coupling relationship between the perception features and the formation state solution. It is used to solve the relative position deviation, heading deviation, and attitude deviation of each UAV in the formation based on the fused UAV state data. Combined with the wind field disturbance features, visual matching confidence features, and obstacle constraint features corresponding to the coupling relationship, the relative position deviation, heading deviation, and attitude deviation of each UAV are dynamically weighted and corrected to construct the global state matrix of the formation. The trajectory dynamic planning module is used to generate a global reference trajectory for the formation based on the formation mission objectives, airspace constraints, and wind field compensation parameters. The formation obstacle avoidance collaborative control module is used to perform two-level obstacle avoidance planning based on obstacle detection data: formation-level formation reconstruction and single-unit path fine-tuning. It synchronously coordinates the timing of obstacle avoidance actions of each UAV and completes flight conflict management within the formation. The multi-machine communication link management module is used to establish a formation-based distributed self-organizing communication network, support multi-hop relay transmission, monitor link quality in real time, and automatically switch communication frequency bands and optimal relay nodes. The formation attitude coordination stabilization module is used to adjust the power output and control surface angle of each UAV based on formation control commands and wind field compensation parameters. The central coordination and control hub is used to complete data communication and command coordination among various functional modules, and to complete closed-loop control and mission scheduling for the entire formation flight process.

2. The low-altitude unmanned aerial vehicle (UAV) formation cooperative control system according to claim 1, characterized in that, It also includes a formation coordination error quantification and evaluation unit, the calculation of which is completed using the following formula: ; The total error of drone formation coordination. This is the position deviation weighting coefficient. For drones Real-time position vector, For formation reference position vector, Let be the Euclidean distance between the two. The maximum allowable threshold for formation position deviation. This is the heading deviation weighting coefficient. For drones Real-time heading angle For formation reference heading angle, This is the maximum permissible threshold for heading deviation. This is the attitude deviation weighting coefficient. For drones Real-time attitude angle vector, For formation reference attitude angle vector, The magnitude of the attitude deviation between the two is given. This represents the maximum permissible threshold for attitude deviation. It is a non-negative smoothing constant, and + + =1, , , All are positive real numbers, and the weighting coefficients are... , , In addition to adjusting according to the formation mission type, it also performs real-time dynamic correction by combining the wind field disturbance characteristics, visual matching confidence characteristics and obstacle constraint characteristics output by the multi-source perception fusion module.

3. The low-altitude unmanned aerial vehicle (UAV) formation cooperative control system according to claim 1, characterized in that, It also includes an airspace constraint adaptation unit, which is used to interface with low-altitude airspace control data, extract the spatial boundaries and temporal constraints of no-fly zones, restricted flight zones, and temporary controlled airspaces, transform airspace constraint parameters into hard constraints for trajectory planning and obstacle avoidance control, automatically filter trajectory schemes that cross controlled airspaces, complete the risk management of illegal flights, and match the dynamically updated airspace control instructions at low altitudes to complete the real-time adjustment of trajectory schemes; the airspace constraint features output by the airspace constraint adaptation unit are synchronously input into the multi-source perception fusion module to participate in the feature fusion process of multi-source data.

4. The low-altitude UAV formation cooperative control system according to claim 1, characterized in that, The multi-source perception fusion module is equipped with a binocular vision acquisition unit, a millimeter-wave radar detection unit, a BeiDou satellite positioning unit, an inertial navigation unit, and a meteorological sensing unit. The binocular vision acquisition unit and the millimeter-wave radar detection unit work together to detect the three-dimensional spatial position and motion state of low-altitude obstacles, and to simultaneously identify static obstacles and dynamic moving targets. The BeiDou satellite positioning unit and the inertial navigation unit use a loosely coupled fusion algorithm to output centimeter-level real-time position and attitude data of the UAV. The meteorological sensing unit collects real-time wind speed, wind direction, and air pressure data, calculates low-altitude wind field disturbance parameters, and converts them into flight control compensation quantities.

5. The low-altitude unmanned aerial vehicle (UAV) formation cooperative control system according to claim 1, characterized in that, The trajectory dynamic planning module includes a global trajectory planning subunit and a local trajectory optimization subunit. The global trajectory planning subunit uses an improved A* algorithm to generate a conflict-free global reference trajectory for the formation from the takeoff point to the mission target point. At the same time, it sets safe distance constraints for trajectory nodes in combination with formation requirements. The local trajectory optimization subunit uses a model predictive control algorithm to perform rolling optimization on the sub-trajectories of each UAV based on real-time environmental changes, formation state deviations and wind field compensation parameters, and adjusts the spatial coordinates and arrival sequence of trajectory nodes.

6. The low-altitude unmanned aerial vehicle (UAV) formation cooperative control system according to claim 1, characterized in that, The formation obstacle avoidance collaborative control module includes a formation-level obstacle avoidance scheduling subunit and a single-unit obstacle avoidance execution subunit. The formation-level obstacle avoidance scheduling subunit adjusts the overall flight formation and trajectory for static obstacles and large-scale airspace constraints, and completes large-scale obstacle avoidance through formation reconstruction, while simultaneously updating the sub-track nodes of each UAV. The single-unit obstacle avoidance execution subunit quickly plans the obstacle avoidance path for a single UAV within the formation constraints for dynamic obstacles and sudden risks, while simultaneously synchronizing the obstacle avoidance actions and speed adjustment timing of adjacent UAVs, and managing the risk of intra-formation flight collision caused by single-unit obstacle avoidance.

7. The low-altitude unmanned aerial vehicle (UAV) formation cooperative control system according to claim 1, characterized in that, The multi-drone communication link management module adopts a distributed self-organizing network architecture and supports multi-hop relay transmission mode. When the distance between drones in the formation exceeds the direct communication range, data is relayed and forwarded through intermediate drones. At the same time, the signal strength, packet loss rate and transmission delay of each communication link are monitored in real time. The multi-drone communication link management module feeds back the real-time link quality characteristics to the central collaborative control center. The central collaborative control center inputs the link quality characteristics into the multi-source perception fusion module to participate in feature fusion, providing a basis for subsequent control parameter adjustment.

8. A method for cooperative control of low-altitude unmanned aerial vehicle (UAV) formations, applied to the cooperative control system for low-altitude UAV formations according to any one of claims 1-7, characterized in that, Includes the following steps: The multi-source data acquisition and fusion steps involve simultaneously acquiring flight status and low-altitude environment perception data through various types of sensing devices carried by the formation drones. This process completes multi-source data spatiotemporal reference registration, noise filtering, outlier removal, and feature fusion, generating wind field disturbance features, visual matching confidence features, and obstacle constraint features. The formation global state calculation step involves directly inputting the wind field disturbance features, visual matching confidence features, and obstacle constraint features into the calculation process, forming a strong coupling relationship between the perception features and the formation state calculation. Based on the fused formation state dataset, the relative position, heading, and attitude deviations of each UAV relative to the navigator and the formation reference point are dynamically weighted and corrected by combining the wind field disturbance features, visual matching confidence features, and obstacle constraint features corresponding to the coupling relationship, thereby constructing the formation global state matrix. The formation trajectory planning steps generate a globally conflict-free reference trajectory based on mission objectives, airspace constraints, environmental feature datasets, and wind field disturbance parameters. The formation-coordinated obstacle avoidance steps, based on the detection results of the spatial position and motion state of obstacles, execute formation-level formation reconstruction and single-machine-level path fine-tuning in a hierarchical manner, and coordinate the timing and speed adjustment strategies of obstacle avoidance actions. The formation flight cooperative control steps generate attitude control and power adjustment commands for each UAV based on the global reference track, real-time formation state deviation and wind field disturbance parameters, and adjust the power output and control surface deflection angle through a closed-loop control algorithm. The communication link management steps include establishing a distributed self-organizing communication network between UAVs and ground control terminals within the formation, completing low-latency two-way transmission of data and control commands, and monitoring and optimizing the quality of the communication link in real time. The closed-loop optimization scheduling process involves real-time collection of the actual flight status and mission execution of the formation, comparison with preset targets to complete the coordination error assessment, and dynamic optimization of the trajectory planning scheme and control parameters based on the assessment results.

9. A low-altitude unmanned aerial vehicle (UAV) formation cooperative control method according to claim 8, characterized in that, In the formation global state calculation step, a multi-dimensional weighted fusion calculation method is adopted for the quantitative evaluation of formation coordination error. Differentiated weight coefficients are set for position deviation, heading deviation, and attitude deviation. The weight coefficients can be dynamically adjusted according to the formation mission type. At the same time, real-time correction is made in combination with wind field disturbance characteristics, visual matching confidence characteristics, and obstacle constraint characteristics. The position deviation weight coefficient is increased for low-altitude inspection missions, the attitude and heading deviation weight coefficients are increased for formation light show missions, and the speed and heading deviation weight coefficients are increased for material transportation missions.

10. A low-altitude unmanned aerial vehicle (UAV) formation cooperative control method according to claim 8, characterized in that, In the closed-loop optimization scheduling process, a three-level hierarchical response mechanism is constructed. When the total coordination error is lower than the preset first-level safety threshold, the current control parameters and trajectory planning scheme are kept stable. When the total coordination error is between the first-level safety threshold and the second-level warning threshold, the local trajectory optimization and control parameter fine-tuning process is initiated. By adjusting the trajectory nodes and power output parameters in a small range, the formation coordination deviation is quickly corrected. When the total collaborative error exceeds the level 2 warning threshold, the formation emergency dispatch mechanism is triggered, the current mission execution is suspended, the global trajectory is replanned and the formation formation is adjusted, and real-time warning information and status data are sent to the ground control terminal. The mission will continue to be executed after the formation status returns to stability. The execution of each level of response mechanism is based on the real-time features output by the multi-source perception fusion module and the corrected status data output by the formation status collaborative solution module.

Images