A full-distributed pre-scheduled time UAV swarm cooperative control method

Abstract

The application relates to a full-distributed pre-scheduled time UAV cluster cooperative control method. The method comprises the following steps: a UAV double integrator dynamics model is established, positions and velocities are obtained, and a time-varying formation offset is introduced to generate a virtual cooperative state and a velocity tracking error; a time scale function which tends to infinity before a preset specified time and is a constant 1 after the preset specified time is constructed; a monotonically increasing edge and node adaptive gain is calculated by using a local error integral update law; an artificial potential field repulsive force gradient is calculated according to a neighboring distance; a virtual velocity instruction is generated by fusing the above parameters; a first stage control law containing feedforward compensation and disturbance estimation and a second stage control law with locked gain are respectively executed before and after the preset time based on a backstepping control model; and an actual control amount is output after being limited in amplitude. The application realizes full-distributed pre-scheduled time cooperative control, does not need global topological information, effectively avoids gain singularity and actuator saturation, and guarantees formation precision and obstacle avoidance safety.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) swarm collaborative control technology, and in particular to a fully distributed, pre-set time UAV swarm collaborative control method. Background Technology

[0002] Cooperative dynamic target tracking in multi-unmanned aerial vehicle (UAV) systems requires UAV swarms to continuously optimize their geometric encirclement configuration in real time while intercepting or monitoring maneuvering targets. Distributed time-varying optimization provides an effective mathematical framework for solving this type of multi-agent cooperative tracking problem. In practical tactical applications, the convergence speed of the system state is a crucial evaluation metric.

[0003] Traditional distributed optimization algorithms mostly rely on the principle of asymptotic convergence, which theoretically requires an infinitely long time to reach the global optimum, making it difficult to meet the needs of highly dynamic environments. While subsequent finite-time and fixed-time control techniques theoretically solved the problem of limited convergence speed, their mathematical upper bounds are often overly conservative and heavily dependent on complex global control parameters. To meet the stringent time requirements of practical engineering applications—specifically, forcing drone swarms to precisely reconfigure their formation before a manually specified countdown—current advanced methods are beginning to employ predetermined-time control techniques. The core idea is to introduce a timescale function containing time singularities into the feedback loop, using a function gain that gradually approaches infinity to force the system error to strictly decay to zero at a set time.

[0004] However, existing time-determined control technologies suffer from serious engineering flaws in practical applications. First, most existing time-determined protocols rely on the global topology information of the communication network. In bandwidth-constrained and decentralized real-world drone swarms, obtaining global information severely violates the core principle of fully distributed deployment. Second, real physical drones are constrained by dual-integrator dynamics models with deep position and velocity coupling characteristics. When time-determined control is directly applied to a dual-integrator system, the infinite gain phenomenon near the set time will drastically amplify high-frequency fluctuations through the dual-integrator chain, leading to extremely severe numerical singularities and actuator output saturation, ultimately causing the physical system to lose control or even crash. This engineering bottleneck significantly limits the practical expansion and application of time-determined control to complex dynamic systems.

[0005] Therefore, there is an urgent need in related technologies to solve the problems of predetermined time control relying on global topology information and the easy occurrence of gain singularity and actuator saturation in the dual integrator dynamics model. Summary of the Invention

[0006] Therefore, it is necessary to provide a fully distributed, pre-set timed drone swarm collaborative control method to address the aforementioned technical problems.

[0007] Firstly, this application provides a fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm cooperative control method. The method includes: Define the dynamic model and communication topology of the UAV dual integrator, obtain the current physical position and actual speed of each UAV, and introduce time-varying formation offset to generate virtual cooperative state and speed tracking error; Construct a time scale function, which tends to infinity with time before a preset specified time and becomes a constant 1 after the preset specified time; The edge adaptive gain and node adaptive gain are calculated using the monotonic integral update law of local position error and velocity error; The artificial potential repulsive gradient is calculated based on the physical distance between adjacent drones; Based on the time scale function, edge adaptive gain, node adaptive gain, artificial potential repulsion gradient, and adjacency matrix elements of the communication topology, a virtual speed control command is generated. Based on the backstepping control model, the first-stage control law is executed before the preset specified time, and the second-stage control law is executed after the preset specified time to obtain the actual control quantity; wherein the first-stage control law includes a virtual velocity feedforward compensation term, a gradient tracking term, and an upper limit estimate of the lumped unknown environmental disturbance; the second-stage control law locks the edge adaptive gain and the node adaptive gain to constants and stops the calculation of the artificial potential field repulsive gradient. The actual control quantity is limited to the maximum thrust saturation threshold and output to the UAV actuator.

[0008] Optionally, in one embodiment of this application, the calculation formula for the virtual cooperative state and velocity tracking error is as follows:

[0009] in, Indicates a virtual collaborative state. Indicates the current physical location. Indicates the time-varying formation offset. Indicates speed tracking error. Indicates actual speed. This represents the virtual speed command used as a reference by the lower-order speed loop actuator. This represents the first derivative of the formation offset vector.

[0010] Optionally, in one embodiment of this application, the expression of the time scale function before a preset specified time is:

[0011] in, This indicates a preset specified time. It is a scalar constant greater than 2. Indicates the initial time.

[0012] Optionally, in one embodiment of this application, the edge adaptive gain and node adaptive gain are calculated using the following monotonic integral update law:

[0013] in, The update law representing the edge adaptive gain. The update law representing the adaptive gain of the node. Indicates a virtual collaborative state. This represents the virtual cooperative state of neighbor node j. This represents the derivative of the function with respect to time at a preset time scale. Indicates speed tracking error. The learning rate is a constant. This represents the Euclidean second norm.

[0014] Optionally, in one embodiment of this application, the formula for calculating the repulsive gradient of the artificial potential field is:

[0015] in, For communication neighbor set, Let be the potential field function. This represents the sum of the gradients of the repulsive force from the artificial potential field acting on node i. The operator represents the gradient of the physical location pi of node i. and Let i and j represent the actual physical locations of node i and their communicating neighbor node j at time t, respectively.

[0016] Optionally, in one embodiment of this application, the virtual speed control command is represented as:

[0017] in, The gain is controlled by a normal number. These are the adjacency matrix elements of the topological graph. This is for edge-adaptive gain.

[0018] Optionally, in one embodiment of this application, in the first-stage control law, the feedforward compensation term of the virtual velocity is obtained by differentiating the virtual velocity command using the differential chain rule, the gradient tracking term is used to approximate the global optimal target, and the upper limit estimate of the lumped unknown environmental disturbance is used to offset the environmental disturbance.

[0019] Optionally, in one embodiment of this application, in the second-stage control law, the edge adaptive gain and node adaptive gain, which are locked as constants, are their instantaneous values ​​at the preset specified time, and the second-stage control law is a linear time-invariant control quantity controlled by proportional-derivative gain.

[0020] Optionally, in one embodiment of this application, limiting the actual control quantity within the maximum thrust saturation threshold includes: When the intensity of the actual control quantity exceeds the physical upper limit of the UAV actuator, the actual control quantity is truncated to the physical upper limit value.

[0021] The above-described fully distributed, pre-set timed UAV swarm collaborative control method has the following advantages compared with existing technologies: First, it is fully distributed and plug-and-play: Through fully distributed edge and node adaptive compensation, the integral gain monotonically increases within a predetermined time interval without requiring any prior knowledge of the global connectivity of the communication topology, thus achieving true plug-and-play characteristics in actual decentralized swarm networks.

[0022] Second, avoiding singularities and preventing thrust saturation: By using a segmented state switching architecture that balances the predetermined time accuracy requirements with physical tolerances, and backstepping feedforward analytical compensation, the gain singularity in the predetermined time control of the dual integrator system is completely eliminated, effectively avoiding the unavoidable huge control thrust mutation phenomenon in the prior art, and strictly constraining the maximum control input peak value within the saturation physical limit of the real UAV motor.

[0023] Third, a safe obstacle avoidance hard constraint guarantee: by integrating the underlying artificial potential field repulsion function into the outer ring of the position, the aircraft generate a strong reverse repulsion force when they approach each other, ensuring a collision-free hard constraint between the aircraft that is not lower than the set safe distance, thus avoiding the risk of crashes during dense swarm maneuvers. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating a fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method in one embodiment. Figure 2 This is a simulation diagram of the cooperative tracking trajectory of five drones in three-dimensional space in one embodiment; Figure 3 This is a graph showing the evolution of consistency error in a drone cluster in one embodiment. Figure 4 This is a graph showing the evolution of the speed tracking error of a drone swarm in one embodiment; Figure 5 This is a graph showing the evolution of control inputs of a UAV during cooperative tracking in one embodiment. Figure 6 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0026] In one embodiment, such as Figure 1 As shown, a fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm cooperative control method is provided, including the following steps: S101: Define the UAV dual integrator dynamics model and communication topology, obtain the current physical position and actual speed of each UAV, and introduce time-varying formation offset to generate virtual cooperative state and speed tracking error.

[0027] In this embodiment, the real physical UAV is modeled in three-dimensional space as a dual integrator system with lumped environmental interference to obtain the current physical position of the i-th UAV. and actual speed Subsequently, time-varying formation offsets are introduced to achieve specific geometric enclosing configurations. The error space is mapped to generate a virtual state space:

[0028] in, It represents a virtual collaborative state, transforming the traditional physical location consistency problem into a virtual state consistency problem, thereby giving the cluster the ability to change its formation over time; This is the first derivative of the formation offset vector. To decouple the dual integrator dynamics, this invention defines the velocity tracking error. ,in This indicates a virtual speed command that is referenced by the lower-order speed loop actuator.

[0029] S102: Construct a time scale function, which tends to infinity with time before a preset specified time and becomes a constant 1 after the preset specified time.

[0030] In this embodiment of the application, a time scale function is constructed. .definition , and when hour The scalar constant h>2 is used to avoid controlling singularity.

[0031] S103: Calculate edge adaptive gain and node adaptive gain using the monotonic integral update law of local position error and velocity error.

[0032] In this embodiment, to break the dependence on the eigenvalues ​​of the Laplacian matrix of the global communication topology graph, a monotonic integral update law is used to calculate the edge adaptive gain. and node adaptive gain :

[0033] in, This is the learning rate constant. This parameter monotonically increases with system runtime, using the position and velocity errors of local interactions to drive the gain to grow automatically, thus completely eliminating the dependence on the global communication topology connectivity.

[0034] S104: Calculate the repulsive gradient of the artificial potential field based on the physical distance between adjacent UAVs.

[0035] In this embodiment of the application, the repulsive gradient of the artificial potential field is calculated. The calculation formula is as follows: ,in For communication neighbor set, Let be the potential field function. This represents the sum of the gradients of the repulsive force from the artificial potential field acting on node i. The operator represents the gradient of the physical location pi of node i. and Let i and j represent the actual physical positions of node i and its communication neighbor node j at time t, respectively. When the physical distance between any two adjacent UAVs approaches the minimum safe distance, a very strong physical repulsive force is generated in the control law to achieve local obstacle avoidance.

[0036] S105: Based on the time scale function, edge adaptive gain, node adaptive gain, artificial potential repulsion gradient, and adjacency matrix elements of the communication topology, generate virtual speed control commands.

[0037] In one embodiment of this application, the virtual speed control command is represented as:

[0038] in, The gain is controlled by a normal number. These are the adjacency matrix elements of the topological graph. This is the edge-adaptive gain. The multiplier combination term approximates the value over time t. It will increase dramatically, generating a huge driving force that forces all drones to move closer to the target's relative position.

[0039] S106: Based on the backstepping control model, the first stage control law is executed before the preset specified time, and the second stage control law is executed after the preset specified time to obtain the actual control quantity; wherein the first stage control law includes a virtual velocity feedforward compensation term, a gradient tracking term, and an upper limit estimate of the lumped unknown environmental disturbance; the second stage control law locks the edge adaptive gain and the node adaptive gain to constants and stops the calculation of the artificial potential field repulsive gradient.

[0040] In this embodiment, state switching is performed based on a backstepping control model and a preset specified time to generate actual control quantities; to decouple deeply nonlinear relationships and avoid the infinite gain being drastically amplified by the double integrator chain, a preset specified time is used. The control law is switched in segments at all times: Before the preset specified time, that is At that time, finite-time and space formation assembly and obstacle avoidance are performed. First, the precise feedforward compensation term for the virtual velocity is calculated using the differential chain rule. To avoid spatial singularities; then, the first-stage actual control quantity is generated. :

[0041] in, To control the gain, For the gradient tracking term used to approximate the global optimum, This is an upper limit estimate of the lumped unknown environmental disturbances. The design directly cancels the rate of change of the virtual velocity in the inner loop, effectively preventing the accumulation and amplification of errors.

[0042] After a preset specified time, that is when At this time, asymptotically optimal steady-state tracking is performed. Adaptive gain is then applied. and It is forcibly locked to a constant, the infinite gain feedback chain is cut off, and a linear time-invariant tracking control quantity is generated. :

[0043] At this point, the system forcibly stops the repulsive potential field. and higher-order feedforward derivatives The calculation is smoothly transformed into a process subject to conventional proportional-differential gain. The control aims to achieve an asymptotically optimal tracking state to eliminate steady-state error.

[0044] S107: Limit the actual control quantity to within the maximum thrust saturation threshold and output it to the UAV actuator.

[0045] In this embodiment of the application, boundary constraints are executed. When the controller command The strength exceeds the physical limit of the drone actuator. When truncation protection is applied, the actual output is finally determined. The drive motor regulates the propeller speed. This outer constraint forces the commands to be tailored within the limits that the system can withstand, ensuring the physical safety of the aircraft.

[0046] The following specific embodiment illustrates the detailed implementation steps of the fully distributed, pre-defined timed UAV swarm cooperative control method of this application. Taking a swarm of five dual-integrator UAVs conducting a formation cooperative dynamic target tracking experiment as an example, the experimental field is a three-dimensional space, and the maneuvering target is set to exhibit a long-distance S-shaped curve motion. The UAVs in the swarm exchange local information via a wireless local area network, and the pre-defined convergence time of the task is set as follows: Considering the mechanical performance constraints of the physical motor, the maximum physical thrust limit of the actuator is set as follows: The specific implementation steps are as follows: Step 1: Establish the dynamic model and initial state of the dual integrator UAV: The actual motion state of a drone in three-dimensional space is affected by environmental wind field disturbances. The dynamic equations of the double integrator describe:

[0047] in, and Let i and j represent the position and velocity of the i-th UAV in three-dimensional space, respectively. The input acceleration is a bounded actual control input. In this experiment, five UAVs are initially randomly scattered in three-dimensional space, and the trajectory model of the maneuvering target is... , The specific initial positions and initial velocity parameters of each UAV and the moving target in the cluster are shown in Table 1: Table 1 Initial State Parameters of UAV Swarm and Maneuvering Target

[0048] Introducing time-varying formation offset By forming a pentagram formation, the state of each drone is mapped to a virtual collaborative state. and speed tracking error :

[0049] in, This indicates a virtual speed command that is referenced by the lower-order speed loop actuator.

[0050] Step 2: Based on local errors and the gradient of the artificial potential field, calculate the virtual velocity command and the fully distributed adaptive gain. The specific steps are as follows: s2.1 Setting obstacle avoidance safety distance When the physical distance between any two adjacent drones approaches this threshold, the artificial potential field function induces a repulsive gradient. To achieve collision avoidance.

[0051] s2.2 Constructing a time-scale function with time singularities To avoid second-order spatial singularities in backstepping differentiation, a time scaling exponent of h = 2.1 is chosen.

[0052] s2.3 Calculate edge adaptive gain using real-time state error With node adaptive gain Discarding global topology parameters, all gain and learning rate parameters are set to... The integral update law is expressed as:

[0053] s2.4, Set the position loop control parameter c1=0.6, and combine the above parameters to obtain the final virtual speed command:

[0054] Step 3: Based on the backstepping control model and a preset specified time, perform segmented state switching to generate actual control quantities: In step 3, the value approaching infinity is directly passed. This could lead to motor thrust overload or even loss of control, therefore the system is in a countdown... Time-triggered control law switching: s3.1, when At that time, the first phase of spatial formation assembly and obstacle avoidance is performed. The feedforward compensation term is analytically calculated using the differential chain rule. Set speed loop parameters Gradient tracking parameters and the upper limit of aggregate interference Generate actual control quantity :

[0055] s3.2 When t≥15s, perform the second stage of asymptotically optimal steady-state tracking. At this time, the gain is forcibly locked. and The constant value at the 15th second is used to cut off the infinite time feedback term, stop the calculation of the repulsive potential field, and transform it into a linear control quantity controlled by conventional proportional-differential gain:

[0056] Step 4: Apply saturation constraints based on the actuator's mechanical performance and obtain control results: Considering the upper limit of the actual motor's thrust, apply outer boundary constraints. When the controller command strength exceeds the actuator's upper limit... Cut-off protection is performed at that time:

[0057] Final output Drive drone swarms.

[0058] The effectiveness of the invention is verified through simulation experiments. Figure 2 As can be seen in the first phase of operation, the five drones quickly and safely approached the moving target and accurately completed the pentagram formation at the 15th second; then in the second phase, they smoothly followed the moving target with zero error. Figure 3 and Figure 4 The evolution trend of the error is shown. Under the control of the algorithm, the consistency error and velocity tracking error are forced to converge to zero at 15 seconds, and finally the asymptotic optimal tracking with zero steady-state error is achieved. Figure 5 This shows the changes in control inputs during system operation. Approaching the time singularity, the peak value of the input commands for each UAV is effectively limited to... Within the set range, motor thrust overload is avoided.

[0059] Table 2 shows the Monte Carlo random test data on the impact of different time scaling exponents h on system performance.

[0060] Table 2. Performance Comparison of Time Scaling Exponent under Maximum Physical Thrust Limits (Mean ± Standard Deviation)

[0061] As can be seen, using an excessively large time scaling exponent can cause control commands to exceed the actuator's physical limits for extended periods, disrupting the convergence loop. This embodiment uses a parameter of h=2.1, which further reduces control energy consumption and saturation rate while maintaining aggressive convergence, achieving... The meter-level terminal formation error accuracy demonstrates the advantages of the segmented control and feedforward compensation mechanism of this invention.

[0062] In summary, the fully distributed, pre-timed UAV swarm cooperative control method proposed in this application not only completely eliminates the dependence on global communication topology, achieving true distributed plug-and-play functionality, but also successfully circumvents the infinite gain singularity problem in dual integrator dynamics through a segmented state switching mechanism and backstepping feedforward compensation. Under the strict constraints of UAV physical thrust limitations and collision avoidance hard constraints, it safely and efficiently completes the high-precision cooperative tracking task at the pre-timed interval, significantly improving the control energy economy and engineering application reliability of UAV swarms in complex dynamic environments.

[0063] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0064] In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a fully distributed, time-determined, collaborative control method for a swarm of unmanned aerial vehicles (UAVs). The display screen can be an LCD screen or an e-ink display. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0065] Those skilled in the art will understand that Figure 6The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0066] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0067] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the steps in the above method embodiments.

[0068] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above method embodiments.

[0069] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties.

[0070] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0071] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0072] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm cooperative control method, characterized in that, The method includes: Define the dynamic model and communication topology of the UAV dual integrator, obtain the current physical position and actual speed of each UAV, and introduce time-varying formation offset to generate virtual cooperative state and speed tracking error; Construct a time scale function, which tends to infinity with time before a preset specified time and becomes a constant 1 after the preset specified time; The edge adaptive gain and node adaptive gain are calculated using the monotonic integral update law of local position error and velocity error; The artificial potential repulsive gradient is calculated based on the physical distance between adjacent drones; Based on the time scale function, edge adaptive gain, node adaptive gain, artificial potential repulsion gradient, and adjacency matrix elements of the communication topology, a virtual speed control command is generated. Based on the backstepping control model, the first-stage control law is executed before the preset specified time, and the second-stage control law is executed after the preset specified time to obtain the actual control quantity; wherein the first-stage control law includes a virtual velocity feedforward compensation term, a gradient tracking term, and an upper limit estimate of the lumped unknown environmental disturbance; the second-stage control law locks the edge adaptive gain and the node adaptive gain to constants and stops the calculation of the artificial potential field repulsive gradient. The actual control quantity is limited to the maximum thrust saturation threshold and output to the UAV actuator.

2. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The formulas for calculating the virtual cooperative state and velocity tracking error are as follows: in, Indicates a virtual collaborative state. Indicates the current physical location. Indicates the time-varying formation offset. Indicates speed tracking error. Indicates actual speed. This represents the virtual speed command used as a reference by the lower-order speed loop actuator. It represents the first derivative of the formation offset vector.

3. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The expression for the time scale function before a preset specified time is: in, This indicates a preset specified time. It is a scalar constant greater than 2. Indicates the initial time.

4. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The edge adaptive gain and node adaptive gain are calculated using the following monotonic integral update law: in, The update law representing the edge adaptive gain. The update law representing the adaptive gain of the node. Indicates a virtual collaborative state. This represents the virtual cooperative state of neighbor node j. This represents the derivative of the function with respect to time at a preset time scale. Indicates speed tracking error. The learning rate is a constant. This represents the Euclidean second norm.

5. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The formula for calculating the repulsive gradient of the artificial potential field is: in, For communication neighbor set, Let be the potential field function. This represents the sum of the gradients of the repulsive force from the artificial potential field acting on node i. The operator represents the gradient of the physical location pi of node i. and Let i and j represent the actual physical locations of node i and their communicating neighbor node j at time t, respectively.

6. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The virtual speed control command is represented as follows: in, The gain is controlled by a normal number. These are the adjacency matrix elements of the topological graph. This is for edge-adaptive gain.

7. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, In the first stage control law, the feedforward compensation term of the virtual velocity is obtained by differentiating the virtual velocity command using the differential chain rule, the gradient tracking term is used to approximate the global optimal target, and the upper limit estimate of the lumped unknown environmental disturbance is used to offset the environmental disturbance.

8. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, In the second-stage control law, the edge adaptive gain and node adaptive gain, which are locked as constants, are their instantaneous values ​​at the preset specified time. The second-stage control law is a linear time-invariant control quantity controlled by proportional-derivative gain.

9. The fully distributed, pre-set timed unmanned aerial vehicle (UAV) swarm collaborative control method according to claim 1, characterized in that, The step of limiting the actual control quantity within the maximum thrust saturation threshold includes: When the intensity of the actual control quantity exceeds the physical upper limit of the UAV actuator, the actual control quantity is truncated to the physical upper limit value.

Images