Data-driven hierarchical disturbance compliant consensus control method for unmanned aerial vehicle swarm

Abstract

The present application relates to the technical field of unmanned aerial vehicle cluster cooperative control, and discloses a data-driven unmanned aerial vehicle cluster hierarchical disturbance compliance consistency control method, comprising the following steps: S101: establishing a dynamic model of a multi-unmanned aerial vehicle cluster system under the action of multiple source disturbances; S102: designing a preset time distributed observer to estimate leader state information; S103: designing a projection-based parameter estimator to estimate unknown nonlinear terms in the system; S104: constructing a preset time nonlinear disturbance observer for unmeasurable disturbances; and S105: designing a data-driven disturbance compliance switching controller. The method realizes active disturbance control and reasonable utilization of disturbances, ensures high-precision consistency, significantly improves system safety and dynamic performance, effectively reduces system energy consumption, avoids actuator saturation and structural damage by suppressing unnecessary trajectory correction, and thus comprehensively improves the efficiency and economy of multi-unmanned aerial vehicle cluster cooperative operation.

Description

Technical Field

[0001] This invention belongs to the field of UAV swarm collaborative control technology, specifically involving a data-driven hierarchical disturbance control, compliant and consistent control method for UAV swarms. Background Technology

[0002] With the rapid development of intelligent computing and automatic control technologies, multi-UAV swarm systems are widely used in fields such as unmanned cooperative control, formation flight, and cooperative perception. Consistency control, as the fundamental theory of their group cooperative behavior, has become a research focus. Distributed control, which can achieve global coordination by relying only on local information interaction, has gradually become the mainstream research framework. However, existing distributed consistency control methods usually rely on accurate system model information, which is difficult to obtain in real-world scenarios with large swarm sizes or frequent changes in communication topology, severely restricting engineering applications.

[0003] To overcome the limitations of model dependence, data-driven control methods have gained attention. Model-free adaptive control methods rely solely on system input-output data to construct the control law. However, existing parameter estimators often employ constant weights and step size factors, making it difficult to balance transient estimation performance with steady-state estimation accuracy. Furthermore, existing data-driven consensus control methods do not adequately consider the optimality of control performance. Some optimal consensus control schemes are prone to algebraic loops and causality problems when solving for the optimal control law. Mitigating these issues often comes at the cost of reduced control accuracy or diminished engineering applicability.

[0004] Furthermore, multi-UAV swarm systems are inevitably affected by multi-source interference during actual operation. Existing disturbance rejection control methods lack the ability to actively detect and accurately estimate interference, and are mostly designed for single-source interference, making it difficult to cope with the actual working conditions of multi-source interference coupling. Existing control strategies generally lack compliance, and when tracking reference trajectories that exceed acceptable ranges, they are prone to actuator saturation and overload, leading to performance degradation or even structural damage. Moreover, compliant control methods based on virtual admittance models often use constant admittance parameters, making it difficult to adapt to the dynamic changes in mission requirements and interference environments. In addition, interference does not always have a negative impact on system performance. If interference can be actively managed and rationally utilized, unnecessary control inputs can be reduced and system energy consumption can be lowered. Therefore, there is an urgent need to design a cooperative control scheme that combines compliance and disturbance control capabilities. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a data-driven hierarchical disturbance mitigation, compliant, and consistent control method for UAV swarms. By introducing a hierarchical control structure and constructing a preset time-distributed observer, the control accuracy and dynamic response performance of the system are effectively improved. Simultaneously, by combining a projection operator parameter estimator, a preset time-nonlinear disturbance observer, a disturbance effectiveness evaluation function, and a compliant switching mechanism, the method achieves active mitigation and rational utilization of disturbances. This reduces system energy consumption and excessive trajectory correction while avoiding actuator saturation and potential structural damage risks, significantly improving the collaborative control performance and operational economy of multi-UAV swarm systems. This addresses the problems of insufficient system anti-disturbance capability and compliance performance, slow convergence speed, and high energy consumption under multi-source disturbances in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a data-driven hierarchical disturbance control, compliant, and consistent control method for UAV swarms, comprising the following steps: S101: Establish a dynamic model of a multi-UAV swarm system under multi-source interference, wherein the multi-source interference includes unmeasurable internal interference and measurable external interference, and the dynamic model encompasses the kinematic model of the leader UAV and the dynamic model of the follower UAV, and the control input of the follower UAV is subject to actuator saturation constraints. S102: Design a pre-defined time-distributed observer to estimate the leader's state information, realize the estimation of the leader's position and velocity state information, and this observer is suitable for scenarios where some followers cannot communicate directly with the leader; S103: Design a projection-based parameter estimator to estimate unknown nonlinear terms in a system. By constructing a criterion function and solving for the extrema, the online estimation of unknown nonlinear parameters of the system is completed. S104: Construct a preset time nonlinear interference observer for unmeasurable interference, and combine a preset time function and a nonlinear gain matrix to achieve accurate estimation of unmeasurable interference; S105: Design a data-driven disturbance control compliant switching controller, which includes a disturbance compensation controller and an admittance-based compliant controller. The controller achieves adaptive switching between two control modes based on the system tracking error state through a switching function.

[0007] Preferably, in step S101, the kinematic model expression of the leader drone is:

[0008] Among them, among them, Indicates the sampling interval. and Indicate their position and velocity respectively. and Represents a predefined constant matrix. , and Represent the real number field respectively; Under the influence of multi-source interference, the first The dynamic model expression for a follower drone system is:

[0009] in, , and These represent its position, velocity, and acceleration, respectively. Indicates its quality, This indicates unmeasurable internal disturbances. This indicates measurable external interference acquired through sensors. For controlling input.

[0010] Preferably, in step S102, the preset time distributed observer introduces a preset time function. The preset time function expression is:

[0011] in, and Indicates an adjustable parameter. Indicates the preset time; The expression for the preset time-distributed observer is:

[0012] in, and These represent estimates of the leader's position and speed, respectively. , and It is a diagonal gain matrix. , , , It is an adjustable parameter. and This is an item for exchanging information with neighboring nodes.

[0013] Preferably, the criterion function of the projection operator parameter estimator in step S103 is:

[0014] in, Indicates the time-varying weighting factor. These are estimates of unknown nonlinear parameters; The expression for the projection operator parameter estimator is:

[0015] in, This represents the time-varying step size factor.

[0016] Preferably, the expression for the preset time nonlinear disturbance observer in step S104 is:

[0017] in, and They represent and The estimated value, , , and For observation gain, and This is an adjustable coefficient. This indicates the output estimation error. This is a nonlinear gain matrix.

[0018] Preferably, the expression for the diagonal elements of the nonlinear gain matrix is:

[0019] in, express The One portion, , ,and All parameters are adjustable. It is a symbolic function.

[0020] Preferably, the expression for the interference compensation controller in step S105 is:

[0021] in, The time-varying step size factor is used to adjust tracking performance. Indicates the time-varying weighting factor; Provide an estimate of the leader's expected output; This is the estimated total disturbance.

[0022] Preferably, the expression for the adaptive gain virtual admittance model constructed in step S105 based on the admittance-compliant controller is:

[0023] The expression for the admittance-based compliant controller is:

[0024] in, , and These are the equivalent inertia, damping, and stiffness adaptive gain terms, respectively. For a reshaped, smooth trajectory.

[0025] Preferably, the expression for the switching function in step S105 is:

[0026] in, This represents a pre-set threshold, expressed as: , and All are bounded positive numbers; The control law of the disturbance control and compliant switching controller is: .

[0027] A multi-UAV swarm control system employs the aforementioned data-driven hierarchical disturbance control and compliant consistency control method for UAV swarms to achieve swarm collaborative consistency control under multi-source interference environments. The system includes multiple UAV nodes, a communication module, a sensor module, and a control module. The sensor module collects UAV position and speed information as well as measurable external interference. The communication module enables local information exchange between UAV nodes. The control module embeds a preset time-distributed observer, a projection operator parameter estimator, a preset time nonlinear interference observer, and a disturbance control and compliant switching controller to complete state estimation, parameter identification, interference observation, and control decision-making.

[0028] Compared with existing technologies, this invention provides a data-driven hierarchical disturbance control, compliant, and consistent control method for UAV swarms, which has the following advantages: The data-driven hierarchical disturbance mitigation compliant optimal consensus control method proposed in this invention does not rely on a precise mathematical model of the system. It constructs the control law solely through the system's input and output data, significantly improving its applicability and robustness under complex environments and model uncertainties. It effectively overcomes the problem of strong dependence on system models in existing distributed consensus control methods.

[0029] This invention introduces a hierarchical control structure, effectively decoupling consistent control from optimal control design. Structurally, it avoids the algebraic loops and causality problems easily encountered in traditional optimal control methods. While ensuring system stability, it significantly improves the system's control accuracy and dynamic response performance. Furthermore, the design of a pre-set time-distributed observer enables pre-set time estimation of the leader's state, enhancing the system's control effectiveness and operational safety in time-sensitive application scenarios.

[0030] This invention's preset time-nonlinear interference observer can accurately estimate multi-source interference online. Combined with an interference effectiveness evaluation function and a compliant switching mechanism, it achieves proactive control and rational utilization of interference. When the interference direction is consistent with the leader's movement direction, the interference reduces control input; when the interference direction is opposite, targeted compensation is performed. This reduces system energy consumption, minimizes excessive trajectory correction, and avoids actuator saturation and potential structural damage risks, significantly improving the collaborative control efficiency and operational economy of multi-UAV swarm systems.

[0031] The adaptive gain virtual admittance model of this invention can adjust the system compliance online according to the changes in task requirements and interference environment, so that the system can maintain good interactive safety and control performance in complex dynamic environment, and further expand the engineering application scenarios of the method. Attached Figure Description

[0032] Figure 1 This is a flowchart illustrating the data-driven hierarchical disturbance control, compliant, and optimal consistency control scheme for multi-UAV clusters according to the present invention. Figure 2 This is a block diagram illustrating the control principle of the data-driven multi-UAV cluster hierarchical disturbance mitigation, compliant, and optimal consistency control scheme of the present invention. Detailed Implementation

[0033] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0034] Example: See attached document Figures 1 to 2 A data-driven hierarchical perturbation-compliant consistency control method for UAV swarms includes the following steps: S101: Establish a dynamic model of a multi-UAV swarm system under multi-source interference, wherein the multi-source interference includes unmeasurable internal interference and measurable external interference, and the dynamic model encompasses the kinematic model of the leader UAV and the dynamic model of the follower UAV, and the control input of the follower UAV is subject to actuator saturation constraints. S102: Design a pre-defined time-distributed observer to estimate the leader's state information, realize the estimation of the leader's position and velocity state information, and this observer is suitable for scenarios where some followers cannot communicate directly with the leader; S103: Design a projection-based parameter estimator to estimate unknown nonlinear terms in a system. By constructing a criterion function and solving for the extrema, the online estimation of unknown nonlinear parameters of the system is completed. S104: Construct a preset time nonlinear interference observer for unmeasurable interference, and combine a preset time function and a nonlinear gain matrix to achieve accurate estimation of unmeasurable interference; S105: Design a data-driven disturbance control compliant switching controller, which includes a disturbance compensation controller and an admittance-based compliant controller. The controller achieves adaptive switching between two control modes based on the system tracking error state through a switching function.

[0035] Combination Figure 1 and Figure 2 The present invention specifically includes the following steps: S101, the kinematic model of the leader drone is established as follows: (1) in, Indicates the sampling interval. and Indicate their position and velocity respectively. and Represents a predefined constant matrix. , and Represent the real number field respectively. 3D real vector sum A 3D real matrix.

[0036] Under the influence of multi-source interference, the first The dynamic model of a follower drone system is represented as follows: (2) in, , and These represent its position, velocity, and acceleration, respectively. Indicates its quality, This indicates unmeasurable internal disturbances. This represents measurable external disturbances detected by sensors. (Control input) Subject to actuator saturation constraints, its first Each component satisfies: (3) in, and It is a bounded positive constant, representing its saturation limit.

[0037] The saturation error introduced by the above saturation constraint can be expressed as: (4) Next, for ease of subsequent analysis, the system output is defined as... (5) in, and These are adjustable parameters for position and velocity, respectively.

[0038] Therefore, the system output can be further expressed as an autoregressive model in the following form: (6) in, Indicates system input and output Related nonlinear functions, Indicates unmeasurable interference terms. This indicates a measurable interference term.

[0039] Based on the above model, the system output can be further dynamically linearized to obtain the following equivalent representation: (7) in, For unknown nonlinear parameters, satisfying , It is a bounded constant. , , , , .

[0040] In S102, a pre-defined time-distributed observer is designed to estimate the leader's state information. In practical multi-UAV swarm systems, due to limitations such as communication topology and bandwidth, some followers cannot establish direct communication connections with the leader. Therefore, this paper constructs a pre-defined time-distributed observer, enabling followers to estimate the leader's state without directly obtaining the leader's information. A pre-defined time function is defined. as follows: (8) in, and Indicates an adjustable parameter. Indicates the preset time.

[0041] Using the aforementioned preset time function, a novel preset time distributed observer is constructed, the expression of which is: (9) in, and These represent estimates of the leader's position and speed, respectively. , and It is a diagonal gain matrix. , , , It is an adjustable parameter. and Designed as follows: (10) Define augmented vectors , and Furthermore, the augmented matrix is ​​defined as follows: and .make The error dynamics expression can be obtained as follows: (11) in .

[0042] To analyze the system stability, a Lyapunov function is constructed. Combining the error dynamics in (11), we can obtain that its derivative satisfies: (12) Among them, the second and third terms satisfy

[0043]

[0044] (13) Therefore, if appropriate parameters are selected so that and Then we can further obtain: (14) in, .

[0045] It is noted that It appears in an incremental form. Therefore, there must exist a margin. , making Established. And because... Established, can be launched Therefore, when the gain matrix satisfy and At that time, we can conclude that: (15) Therefore, it can be concluded that there must exist a finite time. This allows the velocity estimation error to converge to a sufficiently small bounded range. Furthermore, when hour, Therefore, the system can still maintain its stability. Furthermore, because... and Since the design form is the same, and the same analysis process is used, we can obtain the following: if the gain matrix satisfies... and If this is achieved, the position estimation error can also converge within a predetermined time, thus completing the estimation of the leader's state.

[0046] In S103, a projection operator parameter estimator is designed to estimate the unknown nonlinear terms in the system. First, the following criterion function is defined: (16) in, Indicates the time-varying weighting factor. express The estimated value.

[0047] Solve The projection operator parameter estimator can be derived, and its expression is: (17) in, This represents the time-varying step size factor, used to adjust the estimation performance of the parameter estimator.

[0048] Substitute (17) into the error dynamics equation We can obtain: (18) in, express 3D identity matrix This indicates the error in the interference estimation.

[0049] If satisfied , ,and (19) in, If the norm is a positive constant, then the parameter estimation error is guaranteed to be bounded. Furthermore, using the equivalence relation of the spectral norm, the constraint in (19) can be rewritten as: (20) Combining (18)-(20) and This can be further deduced to mean: (twenty one) in, Let represent a bounded positive constant. Therefore, we can conclude that the parameter estimation error... It can converge to a sufficiently small bounded range.

[0050] In S104, a preset-time nonlinear disturbance observer is constructed for unmeasurable external disturbances. This is achieved using the aforementioned preset-time function. A novel pre-set time nonlinear disturbance observer is designed, the expression of which is: (twenty two) in, and They represent and The estimated value, , , and For observation gain, and This is an adjustable coefficient. Furthermore... This indicates the output estimation error. Let a diagonal matrix be represented, whose diagonal elements are given by the following nonlinear function: (twenty three) in, express The One portion, , ,and All parameters are adjustable.

[0051] Define augmented vectors Using (22), we can obtain (twenty four) in,

[0052] in, .

[0053] Note , and All are bounded. Therefore, and It is a bounded positive constant. Furthermore, define... Combining (24) and We can obtain: (25) For the first term in the above expression, we have (26) in, .

[0054] According to (26), if there exists a suitable parameter such that , and ,in , and If all are bounded positive constants, then we can further obtain: (27) At the same time, according to (27), it can be known that Established, and Let be a bounded positive constant. Substituting (27) into (24), we get: (28) make If satisfied and Then we can deduce that: (29) Therefore, It evolves in a contracting recursive form, thus ensuring stable convergence of the system. Based on the fundamental inequality... It can be deduced that Therefore, there must be a finite time. This allows the estimation error to converge to a sufficiently small bounded range. Furthermore, when hour, Similarly, it can guarantee the boundedness of its estimation error.

[0055] In S105, a data-driven disturbance-controlling compliant switching controller is designed. To achieve compliant control of a multi-UAV swarm system under multi-source disturbances, existing methods typically generate reshaped compliant trajectories using a fixed-gain virtual admittance model. However, this method struggles to adapt to dynamic changes in mission requirements and the disturbance environment. Therefore, this paper further constructs a novel adaptive gain virtual admittance model, capable of online adjustment of system compliance based on changes in mission requirements and the disturbance environment. Combined with the aforementioned pre-designed time-distributed observer, the expected output based on the leader state estimation information obtained by the observer can be expressed as: (30) in, and These are adjustable control parameters for position and speed, respectively.

[0056] To achieve online reshaping of compliant trajectories, an adaptive gain virtual admittance model was constructed, the expression of which is: (31) in, This represents an estimate of the total system disturbance. , , , , , and Let represent the adaptive gain terms used to adjust the equivalent inertia, damping, and stiffness characteristics of the system, respectively, and design them as follows:

[0057]

[0058] (32) in, , and Indicates the initial admittance gain. This is the interference effectiveness evaluation function, designed based on whether the interference direction is consistent with the leader's movement direction. Its specific expression is: (33) Next, to achieve optimal consistency control in a multi-UAV swarm system, a criterion function is constructed. as follows: (34) in, This represents the time-varying weighting factor.

[0059] By solving Using the observer-based expected output and estimates of total disturbance Furthermore, an interference compensation controller is constructed, the expression of which is: (35) in, This represents a time-varying step size factor used to adjust tracking performance.

[0060] Furthermore, based on the readjusted compliant trajectory The admittance-based compliant controller is constructed as follows: (36) In practical applications, multi-UAV swarm systems can track the desired system output through an interference compensation controller. However, when the system excessively pursues strict trajectory tracking beyond its permissible operating range, actuators may saturate or experience excessive mechanical stress, leading to actuator failure or even structural damage. To avoid this, a switching function for monitoring system performance is defined based on the system's tracking error, with the following expression: (37) in, This represents a pre-set threshold, designed as follows: (38) in, and They are all bounded positive numbers.

[0061] Using the defined switching function Furthermore, a data-driven disturbance control and compliant switching controller is constructed, whose control law can be expressed as: (39) Scenario 1: Interference Compensation Control Mode Under the disturbance compensation control mode, using the disturbance compensation controller in (35), the system error dynamics can be obtained as follows: (40) in,

[0062] (41) and , , .

[0063] Note and All are bounded increments. Based on the analysis in steps S102-S104, it can be seen that... , and All remain bounded. Therefore, if the parameter selection satisfies... , ,and (42) in, If it is a bounded constant, then Established, and It is a bounded constant. Furthermore, using the spectral norm equivalence relation, the constraint condition in (42) can be equivalent to: (43) Combining (41) and (43), we can further derive the following: (44) Therefore, under the interference compensation control mode, the system tracking error can converge to a sufficiently small bounded range, thereby ensuring the stability and tracking performance of the system.

[0064] Scenario 2: Admittance-based compliant control mode Using the admittance-based compliant controller in (36), the system error dynamics can be obtained as follows:

[0065] (45) in,

[0066] (46) in, , .

[0067] When satisfied , And when the condition in (43) is true, it can be known that Established and It is a bounded constant. Combining (45) and (46), we can further derive: (47) Therefore, it can be seen that under the admittance-based compliant control mode, the system's tracking error can converge to a sufficiently small bounded range. Furthermore, combining the above-mentioned disturbance compensation control mode and admittance-based compliant control mode, under the action of the designed compliant switching controller, the system can still maintain stability when switching between different control modes, thus ensuring that the system tracking error always converges to a sufficiently small bounded range.

[0068] This embodiment designs a data-driven hierarchical disturbance mitigation compliant optimal consensus control method, effectively overcoming the problem of strong dependence on system models in existing distributed consensus control methods. It also structurally avoids algebraic loops and causality issues, ensuring the operational safety and reliability of the multi-UAV swarm system under multi-source disturbance environments, and improving the system's control accuracy and dynamic response performance. Furthermore, by introducing a projection operator parameter estimator, a preset time nonlinear disturbance observer, a disturbance evaluation function, and a compliant switching mechanism, adaptive adjustment to multi-source disturbances is achieved. While actively mitigating and rationally utilizing disturbances, it avoids actuator saturation and potential structural damage risks, thereby significantly improving the collaborative control efficiency and operational economy of the multi-UAV swarm system.

[0069] The above description is only a preferred embodiment 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 data-driven hierarchical perturbation-controlling, compliant, and consistent control method for UAV swarms, characterized in that, Includes the following steps: S101: Establish a dynamic model of a multi-UAV swarm system under multi-source interference, wherein the multi-source interference includes unmeasurable internal interference and measurable external interference, and the dynamic model encompasses the kinematic model of the leader UAV and the dynamic model of the follower UAV, and the control input of the follower UAV is subject to actuator saturation constraints. S102: Design a pre-defined time-distributed observer to estimate the leader's state information, realize the estimation of the leader's position and velocity state information, and this observer is suitable for scenarios where some followers cannot communicate directly with the leader; S103: Design a projection-based parameter estimator to estimate unknown nonlinear terms in a system. By constructing a criterion function and solving for the extrema, the online estimation of unknown nonlinear parameters of the system is completed. S104: Construct a preset time nonlinear interference observer for unmeasurable interference, and combine a preset time function and a nonlinear gain matrix to achieve accurate estimation of unmeasurable interference; S105: Design a data-driven disturbance control compliant switching controller, which includes a disturbance compensation controller and an admittance-based compliant controller. The controller achieves adaptive switching between the two control modes based on the system tracking error state through a switching function. The expression for the adaptive gain virtual admittance model constructed based on the admittance compliant controller is as follows: The expression for the admittance-based compliant controller is: in, , and These are the equivalent inertia, damping, and stiffness adaptive gain terms, respectively. For a reshaped, compliant trajectory; The expression for the switching function is: in, This represents a pre-set threshold, expressed as: , and All are bounded positive numbers; The control law of the disturbance control and compliant switching controller is: 。。 2. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 1, characterized in that, In step S101, the kinematic model expression of the leader drone is: Among them, among them, Indicates the sampling interval. and Indicate their position and velocity respectively. and Represents a predefined constant matrix. , and Represent the real number field respectively; Under the influence of multi-source interference, the first The dynamic model expression for a follower drone system is: in, , and These represent its position, velocity, and acceleration, respectively. Indicates its quality, This indicates unmeasurable internal disturbances. This indicates measurable external interference acquired through sensors. For controlling input.

3. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 1, characterized in that, In step S102, the preset time distributed observer introduces a preset time function. The preset time function expression is: in, and Indicates an adjustable parameter. Indicates the preset time; The expression for the predefined time-distributed observer is: in, and These represent estimates of the leader's position and speed, respectively. , and It is a diagonal gain matrix. , , , It is an adjustable parameter. and This is an item for exchanging information with neighboring nodes.

4. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 1, characterized in that, The criterion function of the projection operator parameter estimator in step S103 is: in, Indicates the time-varying weighting factor. These are estimates of unknown nonlinear parameters; The expression for the projection operator parameter estimator is: in, This represents the time-varying step size factor.

5. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 1, characterized in that, The expression for the preset time nonlinear disturbance observer in step S104 is: in, and They represent and The estimated value, , , and For observation gain, and This is an adjustable coefficient. This indicates the output estimation error. This is a nonlinear gain matrix.

6. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 5, characterized in that, The expression for the diagonal elements of the nonlinear gain matrix is: in, express The One portion, , ,and All parameters are adjustable. It is a symbolic function.

7. The data-driven hierarchical perturbation-compliant consistency control method for UAV swarms according to claim 1, characterized in that, The expression for the interference compensation controller in step S105 is: in, The time-varying step size factor is used to adjust tracking performance. Indicates the time-varying weighting factor; Provide an estimate of the leader's expected output; This is the estimated total disturbance.

8. A multi-UAV swarm control system, characterized in that, The data-driven hierarchical disturbance handling compliance and consistency control method for UAV swarms described in any one of claims 1-7 is used to achieve swarm collaborative consistency control under multi-source interference environments. The system includes multiple UAV nodes, a communication module, a sensor module, and a control module. The sensor module collects UAV position and speed information and measurable external interference. The communication module realizes local information exchange between UAV nodes. The control module embeds a preset time-distributed observer, a projection operator parameter estimator, a preset time nonlinear interference observer, and a disturbance handling compliance switching controller to complete state estimation, parameter identification, interference observation, and control decision-making.

Images