Multi-unmanned aerial vehicle hierarchical distributed control method based on preset synchronization time convergence

Abstract

The invention relates to a multi-unmanned aerial vehicle hierarchical distributed control method based on preset synchronization time convergence, belongs to the field of flight control, and is used for processing the control requirements of consistency planning and accurate tracking in a multi-unmanned aerial vehicle formation. The method comprises the following steps: constructing a hierarchical cooperative control framework based on a reference instruction of a consistency rule and robust adaptive tracking control, obtaining control input by using the hierarchical cooperative control framework so as to realize sliding mode adaptive tracking control of preset synchronization time, and combining the control input and an attitude dynamic model of multiple unmanned aerial vehicles so as to realize sliding mode adaptive tracking control of the multiple unmanned aerial vehicles. And synchronization time convergence and consistent cooperative control of multi-machine attitudes are realized. Compared with the prior art, the method has the advantages that the coupling problem of cooperative consistency and model uncertainty of a multi-unmanned aerial vehicle system is considered, hierarchical control based on consensus and estimation is designed, the reliability of tracking control is high, the uncertainty estimation precision is enhanced, the effective maintenance of the flight formation is ensured, and a new technical approach is provided for improving the flight performance.

Description

Technical Field

[0001] The present invention belongs to the field of flight control and relates to an aircraft control method, and in particular to a multi-UAV hierarchical distributed control method based on preset synchronization time convergence. Background Art

[0002] With the rapid development of drone technology and its widespread application in military, agriculture, logistics, and rescue operations, multi-UAV collaboration has become a key development direction for future intelligent systems, leveraging team capabilities, expanding mission envelopes, and improving mission completion rates. In multi-UAV collaborative operations, drones require precise attitude control to achieve formation stability, task division, and efficient collaboration. This is particularly true in scenarios such as target search, disaster relief, and precision operations. Collaborative attitude control can improve the overall efficiency and reliability of the system. Furthermore, real-time information sharing and attitude coordination between different drones can effectively avoid collisions and mission conflicts, enhancing system flexibility and robustness. Therefore, research on multi-UAV collaborative attitude control methods is crucial for advancing multi-UAV technology.

[0003] In multi-UAV collaborative operations, the core requirement of attitude control is to achieve multi-machine synchronization, including attitude consistency and time synchronization. However, in practical applications, UAV systems are often faced with the uncertainty of dynamic models and the influence of external interference. These problems can significantly reduce the effectiveness of attitude response and the overall performance of the system. In the traditional distributed consistency control framework, the multi-machine synchronization problem is tightly coupled with the single-machine interference suppression problem. This coupling relationship restricts the convergence performance of consistency control, especially in the case of strong external interference or large differences in the dynamic characteristics of UAVs, the consistency effect is significantly reduced. Therefore, how to effectively separate the multi-machine synchronization and single-machine interference suppression problems and design a distributed consistency control method with strong robustness and excellent dynamic performance to improve the attitude response quality of the multi-UAV system has become an important research topic in the current field of multi-UAV collaborative control.

[0004] The paper "A hierarchical design framework for distributed control of multi-agent systems" (Xiangyu Wang, Automatica, 2024) addresses the multi-agent control problem and proposes a distributed control law based on a hierarchical design framework. This decouples agent cooperation from individual rules and designs the distributed control law as a reference signal generator design layer and a tracking controller design layer. The controller is combined with the generator instructions generated by the virtual nodes to achieve collaborative tracking. However, this prior art primarily focuses on the hierarchical framework itself, ignoring the impact of individual dynamic uncertainty and external interference on collaborative control performance. Furthermore, the design fails to consider state response effects, making it difficult to achieve high-precision control in complex flight environments. Summary of the Invention

[0005] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a multi-UAV hierarchical distributed control method based on preset synchronization time convergence and suitable for complex environments to meet the multi-UAV collaborative attitude control requirements.

[0006] The purpose of the present invention can be achieved by the following technical solutions:

[0007] A hierarchical distributed control method for multiple UAVs based on preset synchronization time convergence, the method comprising: constructing a hierarchical cooperative control framework based on reference instructions and robust adaptive tracking control based on consistency rules; utilizing the hierarchical cooperative control framework to obtain control inputs to implement sliding mode adaptive tracking control with preset synchronization time; and combining the control inputs with attitude dynamics models of multiple UAVs to achieve synchronized time convergence and consistent cooperative control of the attitudes of the multiple UAVs;

[0008] Among them, the hierarchical collaborative control framework includes a reference attitude instruction generation layer and an attitude tracking control layer. In the reference attitude instruction generation layer, the sliding mode state observation technology is combined with the consistency of multiple UAVs to establish a reference attitude instruction guidance law, obtain reference attitude instructions, and realize the estimation of the navigator state quantity; the attitude tracking control layer obtains the dynamic uncertainty estimation value and disturbance observation value respectively through the neural network uncertainty estimation and disturbance observer based on parallel estimation, and generates control input in combination with the reference attitude instruction.

[0009] Furthermore, the attitude dynamics model of the multi-UAV is:

[0010]

[0011] in, and x i,2 =[p i ,qi ,r i ] T Represents the flight state of the i-th UAV, which are the roll angle Pitch angle θ, yaw angle ψ, roll angular velocity p, pitch angular velocity q, yaw angular velocity r; u i =[δ ai ,δ ei ,δ ri ] T Represents the control input of the i-th UAV, which are the aileron δ a , elevator δ e , rudder δ r ;d i represents the external disturbance of the i-th UAV; f i,2 represents the dynamic uncertainty function of the i-th UAV; g i,1 、g i,2 represents the coefficient matrix.

[0012] Furthermore, the established reference attitude command guidance law is expressed as:

[0013]

[0014] Among them, ξ i,1 ,ξ i,2 They are respectively expressed as the estimated value of the leader's state quantity, ξ i,3 Indicates the extended state quantity; κ1>0, κ2>0 and κ3>0 represent control parameters, T c Indicates the preset convergence time; sig n m (ε i,k ) represents the control function, k=1,2,3;ε i,k Represents the coordination error of multiple UAVs.

[0015] Furthermore, the expression of the control function is:

[0016] sig n m (ε i,k )=||ε i,k || m sig n (ε i,k )

[0017]

[0018] Furthermore, the expression of the multi-UAV coordination error is:

[0019]

[0020] Among them, a ij is the connection coefficient in the multi-UAV communication topology graph theory, the j-th UAV is the neighbor node of the i-th UAV, is the set of neighbors that are topologically connected to the i-th drone, i and j both represent nodes in the communication topology graph; the communication between the leader and follower i is a one-way communication from the leader to the follower i, b i represents the communication edge weight; x 0,1 and x 0,2 represents the flight state of the leader, and u0 represents the control input of the leader.

[0021] Furthermore, the expression of the kinetic uncertainty estimate is:

[0022]

[0023] in, represents the kinetic uncertainty estimate; represents the estimated value of the optimal weight of the neural network, Represents the basis functions of the neural network.

[0024] Furthermore, the update law of the neural network weight estimate is:

[0025]

[0026] Among them, Γ i >0,k xi >0 and k σi >0 indicates control parameters; s i represents the preset time sliding surface; represents the state estimation error; represents the perturbed observation.

[0027] Furthermore, the state estimation error represents the deviation between the constructed parallel estimation model and the system dynamics model x i,2 、 Represents the flight state quantity of the i-th UAV and the corresponding estimated state;

[0028] The parallel estimation model constructed according to the kinetic model structure is:

[0029]

[0030] Among them, κ5>0 represents the control parameter, represents the kinetic uncertainty estimate, represents the perturbed observation value; T c Indicates the preset convergence time; represents the control function, g i,2 represents the coefficient matrix; u i represents the control input of the i-th UAV.

[0031] Furthermore, the disturbance observation value is obtained based on the integral sliding mode disturbance observer and is expressed as

[0032]

[0033] Among them, χ i,1 and χ i,2 represents the auxiliary signal; κ6>0 represents the control parameter; represents the estimated value of kinetic uncertainty; T c Indicates the preset convergence time; sig n m (χ i,2 -x i,2 ) represents the control function, g i,2 represents the coefficient matrix; u i represents the control input of the i-th UAV.

[0034] Furthermore, the tracking control law for generating the preset synchronization time of the control input is:

[0035]

[0036] Among them, κ5>0 represents the control parameter, represents the kinetic uncertainty estimate, represents the perturbed observation value; T c Indicates the preset convergence time; sig n m (s i ) represents the control function, s i represents the preset time sliding surface, s i =e i,2 +s si , e i,2 represents the tracking error; g i,2 represents the coefficient matrix; u i represents the control input of the i-th UAV.

[0037] The present invention also provides a computer-readable storage medium, comprising one or more programs for execution by one or more processors of an electronic device, wherein the one or more programs include instructions for executing the multi-UAV hierarchical distributed control method based on preset synchronization time convergence as described above.

[0038] Compared with the prior art, the present invention has the following beneficial effects:

[0039] 1. The method of the present invention is oriented to the multi-UAV attitude dynamics model, designs a hierarchical collaborative control framework of reference instructions based on consistency rules and robust adaptive tracking control, decouples the influence of dynamic uncertainty interference on the attitude synchronization of multiple UAVs; introduces the synchronization time convergence performance into the instruction and controller design respectively, adopts a neural network estimator based on parallel estimation to generate accurate feedforward information to eliminate the influence of uncertainty on tracking response, realizes the collaborative control of multi-UAV attitude in complex environments, improves the robustness and adaptability of the control system, enhances the uncertainty estimation accuracy to ensure the effective maintenance of the flight formation, and provides a new technical approach to improve flight performance.

[0040] 2. The present invention designs a hierarchical design framework for distributed command generation and adaptive posture tracking control, which realizes universal and scalable distributed control without changing the original individual controllers.

[0041] 3. The present invention designs an attitude reference instruction generator based on integral sliding mode observation in combination with cooperative tracking error, which realizes the synchronous estimation of the leader state within the preset time, and the attitude instructions of multiple UAV systems are consistent.

[0042] 4. The present invention uses a neural network learning technology based on parallel estimation to estimate dynamic uncertainty, uses an integral sliding mode observer with a preset synchronization time to estimate external disturbances, and designs a high-precision feedforward sliding mode adaptive control law with a preset synchronization time to improve tracking control reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a schematic diagram of the principle of the present invention. DETAILED DESCRIPTION

[0044] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0045] This embodiment provides a multi-UAV hierarchical distributed control method based on preset synchronization time convergence, which is applied to a hypersonic aircraft system, specifically a multi-UAV system, including: constructing a hierarchical cooperative control framework based on reference instructions and robust adaptive tracking control based on consistency rules, such as Figure 1As shown, the hierarchical collaborative control framework is used to obtain control inputs to implement sliding mode adaptive tracking control with a preset synchronization time. Combining the control inputs with the attitude dynamics models of multiple UAVs, synchronized time convergence and consistent collaborative control of the multi-UAV attitudes are achieved. The hierarchical collaborative control framework includes a reference attitude instruction generation layer and an attitude tracking control layer. The reference attitude instruction generation layer utilizes sliding mode state observation technology combined with multi-UAV consistency to establish a reference attitude instruction guidance law, obtain reference attitude instructions, and estimate the state of the leader. The attitude tracking control layer obtains dynamic uncertainty estimates and disturbance observations through a parallel estimation-based neural network uncertainty estimation and disturbance observer, respectively, and generates control inputs in combination with the reference attitude instructions.

[0046] This method constructs a hierarchical design framework for reference command and tracking control. It introduces a collaborative error design for a reference command estimation law with a preset synchronization time convergence to ensure the coordinated consistency of commands for multiple UAVs in a leader-follower structure. A neural network is used to estimate attitude dynamics uncertainty, and an integral sliding mode observer is used to estimate external disturbances. A parallel estimation model is established by combining the dynamic and disturbance estimates. The neural network weight update law is designed by combining the tracking error and state estimation error. Finally, a sliding mode surface with a preset synchronization time is used to design UAV control inputs based on precise estimation feedforward. This method considers the coupling between the coordinated consistency of the multi-UAV system and model uncertainty, and designs a consensus-based and estimation-based hierarchical control. This enhances the uncertainty estimation accuracy and ensures the effective maintenance of the flight formation, providing a new technical approach for improving flight performance.

[0047] The key points of the above method are described as follows:

[0048] (a) The UAV attitude dynamics model is:

[0049]

[0050] in, and x i,2 =[p i ,q i ,r i ] T Represents the flight state of the i-th UAV, which are the roll angle Pitch angle θ, yaw angle ψ, roll angular velocity p, pitch angular velocity q, yaw angular velocity r; u i =[δ ai ,δ ei ,δ ri ] T Represents the control input of the i-th UAV, which are the aileron δ a , elevator δ e , rudder δ r ;di =[d 1i ,d 2i ,d 3i ] T represents the external disturbance of the i-th UAV;

[0051]

[0052]

[0053] Where J represents the moment of inertia, Indicates dynamic pressure, S wi , B wi , C wi are the reference wing area, span and average aerodynamic chord length of the i-th UAV respectively; C Li , C Mi , C Ni , represents the aerodynamic torque coefficient of the i-th UAV.

[0054] In this embodiment, S wi =334.7m 2 , B wi =18.29m, C wi =24.38m; M a Represents the flight Mach number of the drone, i=1,…,5

[0055]

[0056] (b) Guidance based on consistency of reference attitude commands with preset synchronization time

[0057] The leader model in a multi-UAV system is:

[0058]

[0059] in, and x 0,2 =[p0,q0,r0] T Indicates the flight state of the navigator, u0=[δ a0 ,δ e0 ,δ r0 ] T Represents the control input of the navigator.

[0060] The collaborative error of multiple UAVs is defined as:

[0061]

[0062] Among them, ξi,1 ,ξ i,2 Respectively expressed as x 0,1 and x 0,2 The estimated value of ξ i,3 Indicates the extended state quantity; is the connection coefficient in the multi-UAV communication topology graph theory, the j-th UAV is the neighbor node of the i-th UAV, is the set of neighbors that are topologically connected to the i-th UAV, i and j both represent nodes in the communication topology graph, 0 represents the leader node, 1,…,N represents the followers; the communication between the leader and follower i is a one-way communication from the leader to the follower i, b i Indicates the communication edge weight. In this embodiment,

[0063] The sliding mode state observation technology is used to estimate the state of the leader, and the reference attitude command guidance law is established by combining the consistency of multiple UAVs:

[0064]

[0065] in, κ1>0, κ2>0 and κ3>0 represent the control parameters to be designed. In this embodiment, κ1=2, κ2=2 and κ3=2, T c =4 indicates the preset convergence time,

[0066]

[0067] in, k=1,2,3.

[0068] (c) Sliding mode adaptive tracking control with preset synchronization time

[0069] Combined with the reference posture command, the tracking error is defined as:

[0070]

[0071] The preset time sliding surface is established as:

[0072]

[0073] Wherein, κ4>0 represents the control parameter to be designed. In this embodiment, κ4=2,

[0074]

[0075] Among them, ∈>0 represents a small constant.

[0076] The tracking control law for the preset synchronization time is designed as follows:

[0077]

[0078] Wherein, κ5>0 represents the control parameter to be designed. In this embodiment, κ5=2. represents the kinetic uncertainty estimate, represents the perturbed observation value; The specific expression of is similar to formula (11).

[0079] (d) Neural Network Uncertainty Estimation and Disturbance Observer Based on Parallel Estimation

[0080] Using neural networks to estimate the dynamic uncertainty function f in the individual posture model i,2 , and its estimated value expression is:

[0081]

[0082] in, represents the estimated value of the optimal weight of the neural network, represents the basis function of the neural network, X i =[-1, 1]×[-1, 1]×[-1, 1]×[-1, 1]×[-1, 1]×[-1, 1]×[-1, 1].

[0083] The parallel estimation model is constructed according to the kinetic model structure:

[0084]

[0085] in, represents the state estimation error; The specific expression of is similar to formula (11).

[0086] The update law for the estimated weights of the neural network is:

[0087]

[0088] Among them, Γ i >0,k xi >0 and k σi >0 represents the control parameter to be designed. In this embodiment, Γ i =0.2, k xi =30 and k σi =0.1.

[0089] The following disturbance observer is designed to obtain the auxiliary signal χ i,1 and χ i,2 , whose expression is:

[0090]

[0091] in, represents the disturbance observation value, κ6=2 represents the control parameter to be designed; The specific expression of is similar to formula (11).

[0092] (e) According to the obtained control input u i , returning to the attitude dynamics model of multiple UAVs, the synchronous time convergence and consistent coordination of the multi-drone attitudes are achieved.

[0093] If the above method is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0094] Parts of the present invention that are not described in detail belong to common knowledge among those skilled in the art.

[0095] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A multi-UAV hierarchical distributed control method based on preset synchronization time convergence, characterized in that: The method includes: constructing a hierarchical cooperative control framework based on consistency rule reference instructions and robust adaptive tracking control, using the hierarchical cooperative control framework to obtain control input to implement sliding mode adaptive tracking control with preset synchronization time, combining the control input with the attitude dynamics model of multiple UAVs to achieve synchronous time convergence and consistent cooperative control of the multi-UAV attitudes; Among them, the hierarchical collaborative control framework includes a reference attitude instruction generation layer and an attitude tracking control layer. In the reference attitude instruction generation layer, the sliding mode state observation technology is combined with the consistency of multiple UAVs to establish a reference attitude instruction guidance law, obtain reference attitude instructions, and realize the estimation of the navigator state quantity; the attitude tracking control layer obtains the dynamic uncertainty estimation value and disturbance observation value respectively through the neural network uncertainty estimation and disturbance observer based on parallel estimation, and generates control input in combination with the reference attitude instruction.

2. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 1 is characterized in that: The attitude dynamics model of the multi-UAV is: in, and x i,2 =[p i ,q i ,r i ] T Represents the flight state of the i-th UAV, which are the roll angle Pitch angle θ, yaw angle ψ, roll angular velocity p, pitch angular velocity q, yaw angular velocity r; u i =[δ ai ,δ ei ,δ ri ] T Represents the control input of the i-th UAV, which are the aileron δ a , elevator δ e , rudder δ r ;d i represents the external disturbance of the i-th UAV; f i,2 represents the dynamic uncertainty function of the i-th UAV; g i,1 、g i,2 represents the coefficient matrix.

3. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 1 is characterized in that: The reference attitude command guidance law established is expressed as: Among them, ξ i,1 ,ξ i,2 They are respectively expressed as the estimated value of the leader's state quantity, ξ i,3 Indicates the extended state quantity; κ1>0, κ2>0 and κ3>0 represent control parameters, T c Indicates the preset convergence time; sig n m (ε i,k ) represents the control function, ε i,k Represents the coordination error of multiple UAVs.

4. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 3 is characterized in that: The expression of the control function is: sig n m (e i,k )=||e i,k || m sig n (e i,k ) 5. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 3 is characterized in that: The expression of the multi-UAV coordination error is: Among them, a ij is the connection coefficient in the multi-UAV communication topology graph theory, the j-th UAV is the neighbor node of the i-th UAV, is the set of neighbors that are topologically connected to the i-th drone, i and j both represent nodes in the communication topology graph; the communication between the leader and follower i is a one-way communication from the leader to the follower i, b i represents the communication edge weight; x 0,1 and x 0,2 represents the flight state of the leader, and u0 represents the control input of the leader.

6. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 1 is characterized in that: The expression for the kinetic uncertainty estimate is: in, represents the kinetic uncertainty estimate; represents the estimated value of the optimal weight of the neural network, θ i Represents the basis functions of the neural network.

7. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 6 is characterized in that: The update law of the neural network weight estimate is: Among them, Γ i >0,k xi >0 and k σi >0 indicates control parameters; s i represents the preset time sliding surface; represents the state estimation error; represents the perturbed observation.

8. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 7 is characterized in that: The state estimation error represents the deviation between the constructed parallel estimation model and the system dynamics model. x i,2 、 Represents the flight state quantity of the i-th UAV and the corresponding estimated state; The parallel estimation model constructed according to the kinetic model structure is: Among them, κ5>0 represents the control parameter, represents the kinetic uncertainty estimate, represents the perturbed observation value; T c Indicates the preset convergence time; represents the control function, g i,2 represents the coefficient matrix; u i represents the control input of the i-th UAV.

9. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 6 is characterized in that: The disturbance observation value is obtained based on the integral sliding mode disturbance observer and is expressed as Among them, χ i,1 and χ i,2 represents the auxiliary signal; κ6>0 represents the control parameter; represents the estimated value of kinetic uncertainty; T c Indicates the preset convergence time; sig n m (χ i,2 -x i,2 ) represents the control function, g i,2 represents the coefficient matrix; u i represents the control input of the i-th UAV.

10. The multi-UAV hierarchical distributed control method based on preset synchronization time convergence according to claim 1, characterized in that: The tracking control law for generating the preset synchronization time of the control input is: Among them, κ5>0 represents the control parameter, represents the kinetic uncertainty estimate, represents the perturbed observation value; T c Indicates the preset convergence time; sig n m (s i ) represents the control function, s i represents the preset time sliding surface, s i =e i,2 +s si , e i,2 represents the tracking error; g i,2 represents the coefficient matrix; u i Represents the control input of the i-th UAV.

Images