Unmanned aerial vehicle formation obstacle avoidance control method based on bird survival pressure mechanism

Abstract

The invention discloses an unmanned aerial vehicle formation obstacle avoidance control method based on a bird survival pressure simulating mechanism. The method comprises the following steps: step 1, constructing an unmanned aerial vehicle and virtual leader motion control model; 2, designing an unmanned aerial vehicle formation controller; 3, designing an obstacle avoidance controller based on a bird survival pressure mechanism; 4, designing an unmanned aerial vehicle formation obstacle avoidance controller; 5, designing a virtual leader controller; and 6, updating the states of the unmanned aerial vehicle and the virtual leader model. The method has the advantages that 1) the problem that the unmanned aerial vehicle cannot bypass an obstacle when directly facing the obstacle in a traditional artificial potential field method is solved; (2) a consistent formation method with a virtual leader is adopted, in practical application, a ground station can serve as the virtual leader, the relative position and formation between the unmanned aerial vehicles can be effectively kept, and rapid and stable switching of the formation can be achieved; and 3) on the premise of ensuring that the formation of the unmanned aerial vehicles is basically kept, efficient obstacle avoidance of the unmanned aerial vehicles in the formation state can be realized.

Description

Technical Field

[0001] The present invention is a method for obstacle avoidance control of UAV formation based on the survival pressure mechanism of birds, belonging to the field of UAV autonomous control. Background Art

[0002] With the rapid development of UAV technology, its applications in military, agriculture, environmental monitoring, logistics and other fields are becoming increasingly widespread. Especially in the aspect of UAV formation flight, it can effectively improve the operation efficiency and data acquisition ability. However, the complexity of formation flight and obstacles in the dynamic environment pose severe challenges to safe flight. For this reason, researchers have proposed various formation and obstacle avoidance control methods to solve the collision risks that UAVs may encounter during formation flight.

[0003] The existing UAV formation control methods mainly include behavior-based formation strategies, leader methods, artificial potential field methods, graph theory methods, and reinforcement learning methods, etc. The behavior-based formation method usually simulates the movement patterns of biological groups, such as bird flocks or fish schools, and realizes the cooperative flight of UAVs by designing the interaction rules between individuals. This method has a fast response ability in a dynamic environment and can quickly adjust the flight strategy according to environmental changes. However, in the case of complex obstacles, the stability and coordination of the formation may be affected, resulting in an increased collision risk.

[0004] The leader method sets a certain UAV in the formation as the "leader", and other UAVs maintain the formation shape by adjusting their relative positions. The advantage of this method is that it is easy to implement and calculate and is suitable for large-scale formations. However, when encountering obstacles, the relative position adjustment between the formation members and the leader may cause the overall formation to lose coordination and affect the obstacle avoidance effect. At the same time, the traditional artificial potential field method simulates the interaction between UAVs and obstacles by constructing a potential field model, driving the UAVs to avoid obstacles and move towards the target. Although this method is relatively simple to implement, it still faces the problem of local minima in a dynamic environment, which may cause the UAVs to be "stuck" between obstacles and unable to avoid obstacles smoothly.

[0005] The graph theory method models UAVs and obstacles as nodes and edges of a graph, and uses graph algorithms for path planning. This method performs well in a static environment, but when dealing with a dynamic environment, it is still a challenge to update the graph structure and calculate the optimal path in real time. In addition, with the progress of deep learning and reinforcement learning, learning-based strategies have gradually become a research hotspot. These methods automatically learn how to achieve formation flight and obstacle avoidance in a complex environment by training models, and have good dynamic adaptability and flexibility. However, the training process of reinforcement learning usually requires a large amount of data and computing resources, and may not be applicable to application scenarios with high real-time requirements.

[0006] The survival state of birds in nature is comprehensively affected by various external factors, including predation by natural enemies, pollution and destruction of the living environment, etc. When facing these survival pressures, birds secrete a hormone called corticosterone to regulate their physiological and behavioral responses. As a stress hormone, the amount of corticosterone secreted directly reflects the degree of environmental pressure borne by birds. Relevant research shows that an increase in the concentration of corticosterone in the body can not only affect the physical condition of birds, but also reflect their long-term living state through the sparseness of feathers. This biological mechanism based on stress response reflects the rapid perception and adaptive ability of birds to environmental changes. Birds secrete corticosterone by perceiving the threat of natural enemies and make stress responses when the content reaches a certain threshold. In the obstacle avoidance control of UAV formations, it is also necessary to achieve flexible responses to complex environments. By mapping the corticosterone secretion and regulation mechanism of birds in a stress environment to the control strategy of UAV formation obstacle avoidance, simulating the stress accumulation and attenuation mechanism of birds when perceiving obstacle threats, precise obstacle avoidance and efficient cooperation can be achieved.

[0007] In summary, how to simulate the stress accumulation and attenuation mechanism of birds when perceiving obstacle threats to solve the problem that traditional artificial potential field methods cannot avoid obstacles when UAVs face obstacles head-on is an urgent problem to be solved in the present invention. Summary of the Invention

[0008] The present invention proposes a UAV formation obstacle avoidance control method based on the bird survival stress mechanism, adopting a consensus formation control method with a virtual leader. This virtual leader is not a real UAV individual in the air, but a virtual UAV model running in the ground station. It transmits data in real time with real UAVs, guides them to achieve cooperative formation, and does not need to perceive and avoid obstacles. In addition, the artificial potential field obstacle avoidance strategy is improved through the bird survival stress mechanism to enhance the flexibility and autonomy of UAV formations in avoiding obstacles in complex environments.

[0009] The present invention proposes a UAV formation obstacle avoidance control method based on the bird survival stress mechanism, as Figure 1 shown, and its specific implementation steps are as follows:

[0010] Step 1: Construction of the motion control model for UAVs and the virtual leader

[0011] S11. Build a UAV motion control model

[0012] The UAV motion control model adopts an autopilot holding model, including speed holding, heading holding, and altitude holding, as shown in Equation (1).

[0013]

[0014] Among them, x i 、yi and h i are the components of the position of UAV i along the three axes in the ground coordinate system, V i , ψ i and λ i are the horizontal velocity, heading angle and altitude change rate of UAV i respectively, τ V , τ ψ and (τ λ , τ h ) are the time constants of the speed-keeping autopilot, heading-keeping autopilot and altitude-keeping autopilot respectively, V i c , and are the control inputs of the three autopilots of UAV i respectively, N = {1, 2,..., n} is the set of UAVs, and n is the number of UAVs.

[0015] The constraints on the UAV motion control model are shown in Equation (2).

[0016] Among them, V min and V max are the minimum horizontal speed and maximum horizontal speed of the UAV respectively, and are the minimum heading angular velocity and maximum heading angular velocity of the UAV respectively, λ min and λ max are the minimum altitude change rate and maximum altitude change rate of the UAV respectively.

[0017] The state transition of the UAV model is shown in Equation (3).

[0018]

[0019] Among them, p i ∈R 3 is the three-dimensional position vector of UAV i, v i ∈R 3 is the three-dimensional velocity vector of UAV i, and R 3 is the three-dimensional Euclidean space.

[0020] The communication radius of the UAV is r c , and data transmission can only be carried out between UAVs within the communication range. The neighbor set of UAV i is N i , and its definition is shown in Equation (4).

[0021] N i = {j ∈ N|||p j - p i || ≤ r c , j ≠ i} (4)

[0022] where p j ∈R 3 is the three - dimensional position vector of UAV j, and ||·|| is the Euclidean distance of the vector.

[0023] S12. Build a virtual leader motion control model

[0024] All UAVs are guided by a virtual leader to achieve cooperative formation flight. The virtual leader can be deployed on the ground station and adopt the same motion control model, state transition and constraints as the UAVs. In addition, the virtual leader can communicate with all UAVs and exchange data without the limitation of the communication radius.

[0025] Step 2: Design of UAV formation controller

[0026] S21. Design of UAV horizontal plane formation controller

[0027] The controller of the UAV in the horizontal plane is shown in Equation (5).

[0028]

[0029] where are the control quantities of UAV i on two axes in the horizontal plane, is the horizontal position vector of UAV i, is the horizontal velocity vector of UAV i, is the horizontal position vector of the virtual leader, is the horizontal velocity vector of the virtual leader, is the expected horizontal position spacing between UAV i and the virtual leader, and are the horizontal formation controller gains, and R 2 is the two - dimensional Euclidean space.

[0030] S22. Design of UAV vertical plane formation controller

[0031] The controller of the UAV in the vertical plane is shown in Equation (6).

[0032]

[0033] where is the vertical control quantity of UAV i, is the height of UAV i, is the height change rate of UAV i, is the height of the virtual leader, and r i 3 ∈R 1 is the expected height spacing between UAV i and the virtual leader, and is the gain of the high formation controller, R 1 is a one-dimensional Euclidean space.

[0034] Step 3: Design of the obstacle avoidance controller based on the bird survival pressure mechanism

[0035] S31. Build an obstacle model

[0036] Design a three-dimensional spherical obstacle, and define the obstacle set as N o ={1, 2,..., n o}, n o is the number of obstacles, and the center position of the o-th obstacle is p o ∈R 3 and the radius is d o . The virtual leader does not detect obstacles and does not need to avoid obstacles.

[0037] S32. Design an artificial potential field obstacle avoidance controller imitating the bird survival pressure mechanism

[0038] Each obstacle has a repulsive potential field on the UAV. The potential field of obstacle o on UAV i is shown in Equation (7).

[0039]

[0040] Among them, d io =||p i -p o || is the three-dimensional distance between UAV i and obstacle o, d δ =d o +d safe +d pre is the effective action distance of the potential field, (d o +d safe ) is the minimum allowable distance between the UAV and the obstacle, d safe is the protection distance to ensure that the UAV maintains a safe isolation from the obstacle during obstacle avoidance, d pre is the warning distance, which gives the UAV the ability to react in advance and provides a necessary buffer, d obs is the sensing distance of the UAV to the obstacle, k b is the repulsive potential field strength coefficient, the larger k b , the stronger the repulsive force, k c is the attenuation coefficient, the larger k c , the slower the repulsive force decays.

[0041] Take the negative gradient of Equation (6) with respect to d io to obtain the repulsive force generated by the obstacles within the sensing range of UAV i on it as shown in Equation (8).

[0042]

[0043] Introduce a mechanism that imitates the survival pressure of birds for the repulsive force to improve the obstacle avoidance effect, as shown in Equation (9).

[0044]

[0045] Among them, k r is the adjustment coefficient, and H is the content of corticosterone in birds. When birds face long-term survival pressure, they will secrete a hormone called corticosterone by themselves. The more its content in the body, the more threats the birds face and the greater the survival pressure (as shown in the appendix Figure 2 ), and at the same time, its concentration will also slowly decay over time. Establish its mathematical model, as shown in Equation (10).

[0046]

[0047] Among them, C(t) is the accumulation of the survival pressure of birds at time t, and λ H is the decay gain, Δt is the sampling time, and the survival pressure of birds 1 3 is a column vector with all elements being 1.

[0048] Step 4: Design of the obstacle avoidance controller for the UAV formation

[0049] S41. Integrate the UAV formation and the obstacle avoidance controller

[0050] Integrate the UAV formation controller and the obstacle avoidance controller to obtain the formation obstacle avoidance controller as shown in Equation (11).

[0051]

[0052] Among them, is the total output of the formation obstacle avoidance controller of UAV i, is the output of the formation controller of UAV i, is the output of the obstacle avoidance controller of UAV i, and k α and k o are the controller gains.

[0053] S42. Design the control quantity converter for the UAV formation obstacle avoidance

[0054] For the UAV formation obstacle avoidance controller, design a control quantity converter to convert its total output into the control inputs of three autopilots, as shown in Equation (12).

[0055]

[0056] Step 5: Design of the virtual leader controller

[0057] S51. Design the control input of the virtual leader autopilot

[0058] The desired control input of the virtual leader autopilot is given as shown in Equation (13).

[0059]

[0060] Wherein, and are the control inputs of the three autopilots of the virtual leader respectively. V expect , ψ expect and h expect are the desired horizontal speed, desired heading angle and desired altitude respectively. The virtual leader tracks the given desired state, and then guides the UAV to perform state transformation.

[0061] Step 6: Update the model states of the UAV and the virtual leader

[0062] S61. Update the states of the UAV and the virtual leader

[0063] The model adopts a discretized update mechanism. Define the sampling time Δt. Taking the UAV as an example, the update mechanism is shown in Equation (14).

[0064]

[0065] The state update methods of the virtual leader and the UAV are the same and will not be elaborated here.

[0066] The present invention proposes a UAV formation obstacle avoidance control method based on the bird survival pressure mechanism, and its advantages are as follows: 1) On the basis of the traditional artificial potential field obstacle avoidance strategy, the corticosterone secretion and regulation mechanism of birds in a stress environment are mapped into the stress obstacle avoidance strategy of UAVs, solving the problem that the UAV cannot bypass the obstacle when facing the obstacle directly in the traditional artificial potential field method; 2) Adopt the consensus formation method with a virtual leader. In practical applications, the ground station can be used as the virtual leader, which can effectively maintain the relative positions and formations of UAVs and can realize the fast and stable switching of formations; 3) By combining the formation controller and the obstacle avoidance controller, on the premise of ensuring that the formation of UAVs is basically maintained, the UAV can achieve efficient obstacle avoidance in the formation state. Description of the drawings

[0067] Figure 1 is a flow chart of a UAV formation obstacle avoidance control method based on the bird survival pressure mechanism.

[0068] Figure 2 is a schematic diagram of bird survival pressure.

[0069] Figure 3 It is the three - dimensional trajectory diagram of the UAV in the simulation experiment.

[0070] Figure 4 It is the two - dimensional trajectory diagram of the UAV in the simulation experiment.

[0071] Figure 5 It is the change diagram of the UAV's heading angle with time in the simulation experiment.

[0072] Figure 6 It is the change diagram of the UAV's horizontal speed with time in the simulation experiment.

[0073] Figure 7 It is the change diagram of the UAV's altitude with time in the simulation experiment. Specific implementation mode

[0074] Next, the effectiveness of the UAV cooperative formation obstacle - avoidance control method proposed in the present invention based on the bird survival pressure mechanism is verified through simulation experiments. In this simulation experiment, the number of UAVs is set to 4, there is 1 virtual leader, the sampling time Δt is 0.05 s, and the total simulation time is 180 s. The software used in this simulation experiment is MATLAB 2023b version.

[0075] For the UAV cooperative formation obstacle - avoidance control method based on the bird survival pressure mechanism, the specific practice steps of its simulation experiment are as follows:

[0076] Step 1: Construction of the UAV and virtual leader motion control models

[0077] S11. Build the UAV motion control model

[0078] Set the time constant τ of the UAV autopilot V to 3 s, τ ψ to 0.75 s, τ λ to 0.3 s, τ h to 1 s. Set the initial positions of the 4 UAVs to be (0, 0, 0) m, (- 10, 20, 0) m, (- 15, - 15, 0) m, (- 25, - 30, 0) m respectively, the initial horizontal speed V i (0) is 5 m / s for all, the initial heading angle ψ i (0) is 45° for all, and the initial altitude change rate λ i (0) is 0 m / s for all. Set the minimum and maximum horizontal speeds of the UAV to be 2 m / s and 10 m / s respectively, the minimum and maximum heading angular velocities to be - 0.2 rad / s and 0.2 rad / s respectively, and the minimum and maximum altitude change rates to be - 3 m / s and 3 m / s respectively. Set the communication radius of the UAV to be 200 m.

[0079] S12. Build a virtual leader motion control model

[0080] The autopilot time constant of the virtual leader takes the same value as that of the UAV. Set the initial position of the virtual leader to (0, 0, 0) m, the initial horizontal speed to 5 m / s, the initial heading angle to 45°, and the initial altitude change rate to 3 m / s. The constraints on the horizontal speed, heading angular velocity, and altitude change rate of the virtual leader are the same as those of the UAV.

[0081] Step 2: Design the UAV formation controller

[0082] First, design two UAV formation shapes. The expected positions of the virtual leader and the first UAV coincide. When 0 ≤ t < 95 s, a diamond formation shape is adopted; when t ≥ 95 s, a single-file formation shape is adopted. The expected position spacings (in meters) between the UAVs and the virtual leader are given respectively. For the diamond formation shape

[0083]

[0084] For the single-file formation shape

[0085]

[0086] Determine its neighbors according to the communication radius of each UAV and its own position at time t. The virtual leader is always a neighbor of all UAVs.

[0087] S21. Design the UAV horizontal formation controller

[0088] Set the gains of the UAV horizontal formation controller to 0.77666, to 5.01476, to 4.50190, to 15.7241. According to the horizontal position vectors and horizontal velocity vectors of UAV i, its neighbors, and the virtual leader at time t, as well as the expected horizontal position spacings, calculate the output of the UAV i horizontal formation controller through Equation (5)

[0089] S22. Design the UAV vertical formation controller

[0090] Set the gains of the UAV altitude formation controller to 16.71015, to 2.12736, to 6.18283, is 4.13172. According to the altitude positions and altitude change rates of UAV i, its neighbors, and the virtual leader at time t, as well as the desired vertical position spacing, the output of the formation controller for UAV i in the vertical plane is calculated through Equation (6).

[0091] Step 3: Design of the UAV obstacle avoidance controller

[0092] S31. Build the obstacle model

[0093] Set the number of obstacles n o to 5. The central positions of the obstacles are (44.96, 45, 61) m, (112, 130, 67) m, (237, 208, 70) m, (237, 178, 70) m, and (235, -17, 70) m respectively, and the radius of each obstacle is 8 m.

[0094] S32. Design an artificial potential field obstacle avoidance controller that imitates the survival pressure mechanism of birds

[0095] Set the protection distance d safe to 3 m. Then the minimum allowable distance (d o + d safe ) between the UAV and the obstacle is 11 m. Set the warning distance d pre to 4 m. Then the effective action distance d δ of the potential field is 15 m. Set the sensing distance d obs of the UAV to the obstacle to 30 m. The repulsive potential field strength coefficient k b of the obstacle is 600, the attenuation coefficient k c is 6.2, the adjustment coefficient k r is 0.1, and the attenuation gain λ H is 1. According to the sensing distance of UAV i and its own position at time t, determine the obstacles that need to be avoided, and calculate the output of the obstacle avoidance controller for UAV i through Equation (9).

[0096] Step 4: Design of the UAV formation obstacle avoidance controller

[0097] S41. Integrate the UAV formation and obstacle avoidance controllers

[0098] Set the gain k α of the formation obstacle avoidance controller to 1, and k o to 80. Calculate the output u i (t) of the formation obstacle avoidance controller for UAV i according to Equation (11), that is

[0099] S42. Design a converter for the UAV formation obstacle avoidance control quantity

[0100] According to the calculated u i (t), and the horizontal velocity V i (t), heading angle ψ i (t), altitude position h i (t) and altitude change rate λ i (t) of UAV i at time t, the control inputs of the three autopilots of UAV i are calculated by Equation (12) as V i c (t), and

[0101] Step Five: Design of Virtual Leader Controller

[0102] S51. Design the control input of the virtual leader autopilot

[0103] Set the expected value V expect (t) of the virtual leader to be always 5 m / s, h expect (t) to be always 70 m, and ψ expect (t) to be

[0104]

[0105] Assign it to the control input of the virtual leader autopilot through Equation (13).

[0106] Step Six: Update the Model States of UAVs and Virtual Leaders

[0107] S61. Update the states of UAVs and virtual leaders

[0108] According to V i (t), ψ i (t), h i (t), λ i (t), V i c (t), and of UAV i at time t, calculate and by Equation (1). Then, update the states through Equation (14) combined with the constraints given by Equation (2) to obtain x i (t + Δt), y i (t + Δt), h i (t + Δt), V i (t + Δt), ψ i (t + Δt) and λ i (t + Δt), and perform state conversion through Equation (3) to obtain p i (t + Δt) and v i(t + Δt). The state of the virtual leader is updated in the same way. Repeat the above steps until the set simulation time is reached.

[0109] Figure 3 is the three - dimensional trajectory diagram of the UAV for the simulation experiment, Figure 4 is the two - dimensional trajectory diagram of the UAV for the simulation experiment, Figure 5 is the variation diagram of the heading angles of 4 UAVs over time, Figure 6 is the variation diagram of the horizontal speeds of 4 UAVs over time, Figure 7 is the variation diagram of the altitudes of 4 UAVs over time. Figure 3 and Figure 4 The obstacle radius shown in o + d safe ).

[0110] Because the initial climb rate of the virtual leader is set to 3 m / s and the desired altitude is set to 70 m, the UAVs start from the initial position and follow the virtual leader to climb to 70 m, and avoid the first obstacle by reducing the altitude during the climbing process.

[0111] After crossing the first obstacle, the UAVs climb to 70 m, start to fly horizontally, basically form the preset diamond formation, and continue to follow the virtual leader. During this process, they avoid the second obstacle by increasing the altitude.

[0112] At 52.5 s, the desired heading angle of the virtual leader changes from 45° to 0°. The UAVs respond in time, change the heading angle, and avoid the third and fourth obstacles by increasing the altitude, and the formed diamond formation basically does not change significantly.

[0113] At 52.5 s, the desired heading angle of the virtual leader changes from 0° to - 90°, and the heading angles of the UAVs also change to - 90°. At 95 s, they receive the desired formation change instruction and quickly change from the diamond formation to the single - file formation.

[0114] At 122.5 s, the UAVs follow the virtual leader to change the heading angle from - 90° to - 180°, and bypass the fifth obstacle from the upper left in the single - file formation. The formation is basically not damaged, and the altitude quickly returns to 70 m.

Claims

1. A UAV formation obstacle avoidance control method based on bird survival pressure mechanism, characterized by: The method comprises the following steps: Step 1: Construction of UAV and virtual leader motion control model The virtual leader is deployed on the ground station and uses the same motion control model, state transition and constraints as the drones. It communicates with all drones and exchanges data without any communication radius limitation. Step 2: Design of UAV formation controller, including the design of UAV horizontal plane formation controller and UAV vertical plane formation controller; Step 3: Design of obstacle avoidance controller based on bird survival stress mechanism Build an obstacle model and design an artificial potential field obstacle avoidance controller that simulates the survival pressure mechanism of birds; Among them, the virtual leader does not detect obstacles and does not need to avoid them; Step 4: Design of drone formation obstacle avoidance controller, including integration of drone formation controller and obstacle avoidance controller, and design of drone formation obstacle avoidance control quantity converter Step 5: Design the virtual leader controller and specifically design the virtual leader autopilot control input.

2. The method according to claim 1, characterized in that: The specific process of designing the artificial potential field obstacle avoidance controller simulating the bird survival pressure mechanism is as follows: Each obstacle has a repulsive potential field on the UAV. The potential field of obstacle o on UAV i is shown in formula (7); Among them, d io =||p i -p o || is the three-dimensional distance between drone i and obstacle o, d δ =d o +d safe +d pre is the effective action distance of the potential field, (d o +d safe ) is the minimum allowable distance between the drone and the obstacle, d safe To protect the distance and ensure that the drone is safely separated from obstacles when avoiding them, d pre To provide the necessary buffer for early warning distance, the drone is given the ability to react in advance. obs is the distance that the drone perceives obstacles, k b is the repulsive potential field strength coefficient, k b The larger the k, the stronger the repulsion. c is the attenuation coefficient, k c The larger it is, the slower the repulsive force decays; For formula (6), io Calculate the negative gradient to obtain the repulsive force exerted on drone i by obstacles within its sensing range As shown in formula (8); The bird-like survival pressure mechanism is introduced into the repulsive force to improve the obstacle avoidance effect, as shown in formula (9); in, k r is the regulation coefficient, H is the corticosterone content in the bird's body, which will be secreted by the bird when it faces survival pressure for a long time.

3. The method according to claim 2, characterized in that: The corticosterone content in the bird body, the mathematical model is shown in formula (10); Among them, C(t) is the accumulated amount of bird survival pressure at time t, λ H is the attenuation gain, Δt is the sampling time, and the survival pressure of birds 13 is a column vector whose elements are all 1.

4. The method according to claim 1, characterized in that: The specific process of integrating the UAV formation controller and the obstacle avoidance controller is as follows: S41, integrated drone formation and obstacle avoidance controller The UAV formation controller and obstacle avoidance controller are integrated to obtain the formation obstacle avoidance controller as shown in formula (11); in, is the total output of the obstacle avoidance controller of the UAV formation i, is the output of the UAV i formation controller, is the output of the obstacle avoidance controller of UAV i, k α and k o is the controller gain.

5. The method according to claim 1, characterized in that: The specific process of step five is as follows: Designing Virtual Leader Autopilot Control Inputs Given the virtual leader’s desired autopilot control input, as shown in Equation (13); in, and are the control inputs of the three autopilots of the virtual leader, V expect , expect and h expect They are the expected horizontal speed, expected heading angle and expected altitude respectively. The virtual leader tracks the given expected state and guides the UAV to change its state.

6. The method according to claim 1, characterized in that: The method also includes the following steps: Step 6: The states of the UAV and virtual leader models are updated, and both models adopt a discrete update mechanism.

Images