A heterogeneous multi-unmanned system formation tracking control method based on dynamic topology

Abstract

The application discloses a heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure and belongs to the field of multi-agent formation and heterogeneous unmanned system control, and specifically proposes a consistent formation control method by introducing a virtual artificial potential field; for a UAV-UAVC heterogeneous system under air-ground cooperation, a unified mathematical model is established for consistent control, a dynamic topology communication structure is established based on state vectors and relative positions of the UAVCs, an artificial potential field method is added, and it is ensured that rapid obstacle avoidance and collision avoidance can be performed in the process of formation flight; experimental simulation results show that the method provides complete collision avoidance, obstacle avoidance and formation strategies for the unmanned heterogeneous system, can effectively perform obstacle avoidance, the obstacle avoidance trajectory is smooth, and the safety and stability of the unmanned system are ensured.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of multi-agent formation and heterogeneous unmanned system control, and particularly relates to a heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure. BACKGROUND

[0002] Compared with a single unmanned aerial vehicle, a multi-unmanned aerial vehicle system can effectively improve the information perception amount of the unmanned aerial vehicle system, improve the task completion rate, shorten the task completion time, and improve the risk resistance. At present, the multi-unmanned aerial vehicle formation system has been widely applied, such as multi-angle detection reconnaissance, multi-target tracking attack, heavy load transportation, formation flight performance, etc. In the future, the multi-unmanned aerial vehicle cooperative control technology will focus on complex and dynamically changing environments, and will no longer be multiple unmanned aerial vehicles of the same type performing tasks, but multiple unmanned aerial vehicles of different types performing heterogeneous formation flight.

[0003] Heterogeneous multi-agent formation refers to multiple types of agents completing a specified cluster task through information interaction, resource sharing, and division and cooperation, which has high application potential in scientific and engineering fields. The heterogeneous multi-agent system can complete formation tasks in a distributed manner and improve execution efficiency; can integrate the advantages of each agent to expand the functional range of the cluster; and can quickly and fully obtain environmental information for cluster system environmental perception. At present, the application value of the heterogeneous multi-agent system is increasingly prominent, especially under the background of the current land, sea, air and space joint three-dimensional combat concept, the heterogeneous multi-agent system formation has significant advantages.

[0004] The multi-unmanned aerial vehicle-multi-unmanned vehicle formation system is a typical heterogeneous multi-agent system, in which the unmanned aerial vehicle has a wide field of view, high mobility and flexibility, and the unmanned vehicle has stronger load and endurance capacity compared with the unmanned aerial vehicle. Through the complementary advantages of the two, the operation efficiency is doubled through air-ground cooperation. At present, the research on the formation control of the heterogeneous multi-unmanned system composed of unmanned aerial vehicles and unmanned vehicles is still in its infancy, and the research on the heterogeneous time-varying formation tracking control is also relatively lacking.

[0005] In the process of time-varying formation tracking of the unmanned aerial vehicle-unmanned vehicle system, obstacle avoidance and collision avoidance between agents are of great significance to improve the safety of the formation. Although the consensus algorithm can also play a role in separating the unmanned vehicles when the relative distance between the unmanned vehicles is less than a given formation shape, the corresponding input is generally linearly related to the relative displacement, and the collision avoidance process is relatively slow, which is difficult to deal with the rapid collision avoidance problem between the unmanned vehicles. Therefore, how to control the heterogeneous unmanned aerial vehicle-unmanned vehicle system in a consistent manner and achieve the requirements of obstacle avoidance and collision avoidance to ensure the safety of the formation has important research significance. SUMMARY

[0006] The application provides a heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure, and solves the problem that the prior art is difficult to cope with rapid collision avoidance between unmanned vehicles.

[0007] To achieve the above object, the application adopts the following technical scheme:

[0008] A heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure comprises the following steps:

[0009] Firstly, the kinematics and dynamics models of a quadrotor unmanned aerial vehicle and a three-wheeled unmanned vehicle are analyzed, the unmanned aerial vehicle system is decoupled and controlled, the height channel and the XOY plane position channel are controlled respectively, and the heterogeneous system mathematical model is unified; further, based on a "leader-follower" mode, a two-layer architecture is established, a distributed observer is used to estimate the state of a virtual leader, and a consensus protocol is proposed to realize that the heterogeneous system tracks a time-varying trajectory;

[0010] Secondly, based on the problem that the communication capability between unmanned vehicles is affected by distance, a communication neighborhood concept is abstracted, a state-dependent dynamic joint communication topology structure is established, and within the communication range, the connection strength between each intelligent agent and the neighbor changes linearly with the change of the relative position;

[0011] Finally, the safety area, the obstacle avoidance area and the attractive area of the unmanned vehicle are defined, a virtual artificial potential field is added, a sigma norm is introduced, a smooth and continuous attractive field potential energy function and a repulsive field potential energy function are respectively constructed, and the virtual potential field force received by each unmanned vehicle is obtained, so that the obstacle avoidance and collision avoidance control of the unmanned aerial vehicle and unmanned vehicle during time-varying tracking formation is realized.

[0012] Beneficial effects: The application provides a heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure, establishes an unmanned aerial vehicle-unmanned vehicle heterogeneous system time-varying formation tracking control model, designs a dynamic joint topology structure based on relative position aiming at the influence of the relative position change of the unmanned system on the communication structure, the model of the unmanned system is more in line with the actual situation, and the tracking control method has better and more stable effects in actual application; a consensus formation control protocol based on the dynamic structure is proposed for the unmanned system formation problem; the consensus algorithm and the artificial potential field method are combined, the repulsive field potential energy function between the unmanned vehicle and the obstacle is designed for the obstacle avoidance problem between the unmanned vehicle and the obstacle, the potential field function between the unmanned vehicles is designed for the collision avoidance problem between the unmanned vehicles and the communication loss problem when the relative position is too far, and the sigma norm is introduced to smooth the potential field function, so that the obstacle avoidance and collision avoidance in the formation tracking process are realized quickly and effectively, and simulation experiment results show that the application can provide complete collision avoidance, obstacle avoidance and formation strategy for the unmanned heterogeneous system, can effectively avoid obstacles, the obstacle avoidance trajectory is smooth, and the safety and stability of the unmanned system are ensured. BRIEF DESCRIPTION OF DRAWINGS

[0013] Figure 1 For the control method flowchart in the embodiment of the application;

[0014] Figure 2 For the definition of the communication-related area based on the ith unmanned vehicle obstacle avoidance in the embodiment of the application;

[0015] Figure 3 For the potential field function in the embodiment of the application, wherein a) the potential field function between the unmanned vehicle and the obstacle, b) the collision avoidance potential field function between the unmanned vehicles;

[0016] Figure 4 For the unmanned vehicle obstacle avoidance trajectory in the embodiment of the application;

[0017] Figure 5 For the unmanned vehicle-unmanned vehicle formation tracking trajectory in the embodiment of the application;

[0018] Figure 6 For the position consistency error time response curve in the embodiment of the application;

[0019] Figure 7 For the speed consistency error time response curve in the embodiment of the application;

[0020] Figure 8 For the basic principle diagram of the artificial potential field in the embodiment of the application;

[0021] Figure 9 For the force diagram of the unmanned vehicle in the artificial potential field in the embodiment of the application. DETAILED DESCRIPTION

[0022] The application will be described in detail below with reference to the accompanying drawings and embodiments:

[0023] As Figure 1 shown, a heterogeneous multi-unmanned system formation tracking control method based on a dynamic topology structure specifically includes the following steps:

[0024] Unmanned vehicle-unmanned aerial vehicle heterogeneous system model establishment

[0025] This embodiment considers a "leader-follower" architecture, assumes that there is one leader and M unmanned aerial vehicles and N-M unmanned vehicles as followers, and uses the set F = {1, 2,..., N} as the index of the followers in the formation.

[0026] The leader state space equation is:

[0027]

[0028] wherein, and are constant matrices; and represent the state and output of the leader, respectively;

[0029] Decouple the dynamics of the UAVs i∈F A ={1,2,…,M} into two parts, one is the formation control in XOY plane, the other is the trajectory tracking control in Z axis direction;

[0030] The dynamics of the UAVs in XOY plane is:

[0031]

[0032] The dynamics of the UAVs in Z axis direction is:

[0033]

[0034] where [p ix ,p iy ,p iz ] represents the position of the i-th UAV in space, [v ix ,v iy ,v iz ] represents the velocity of the UAV in space, θ i ,ψ i represent the roll angle, pitch angle and yaw angle of the UAV, respectively; g is the gravity acceleration; U1, U2, U3, U4 represent the control inputs of the UAV, which are defined as:

[0035]

[0036] where b is the lift coefficient of the UAV, d is the anti-torque coefficient of the motor, l is the distance between the center of mass of the UAV and the rotating shaft of the four-rotor, ω1, ω2, ω3, ω4 are the rotating speeds of the four groups of rotors of the UAV, U1 is the total lift T of the four propellers perpendicular to the fuselage direction, U2 represents the rotating moment τ x that affects the pitch motion of the aircraft, U3 represents the rotating moment τ y that affects the roll motion of the aircraft, and U4 represents the torque τ z that affects the yaw motion of the aircraft, which can be used as the pitch angle, roll angle and yaw angle control variables, respectively. In the actual flight process of the UAV, the motion control of the UAV is realized by controlling the rotating speed of the motor;

[0037] Let the center of mass of the UAV M g be at a distance l g from the front wheel, and the wheel diameter be 2r, and the distance between the rear wheels be 2R;

[0038] The dynamics of the UAVs i∈F GThe dynamic equations for {M+1,…,N} in the XOY plane are:

[0039]

[0040] In the formula x i =[p ix ,p iy ,θ] T Let [p] be the state vector of the i-th autonomous vehicle in the ground coordinate system. ix ,p iy Let θ represent the position of the unmanned vehicle in the ground coordinate system, and let θ represent the angle between the unmanned vehicle's direction of motion and the x-axis of the ground coordinate system, called the vehicle's attitude angle. Differentiating these angles yields the linear velocity and angular velocity in the ground coordinate system. v g Let ω be the linear velocity of the autonomous vehicle. g Indicates the center of mass M g The angular velocity of a point. ω L and ω R These are the angular velocities of the left and right drive wheels, respectively.

[0041] The dynamic equations of the unmanned vehicle-unmanned aerial vehicle heterogeneous system located in the XOY plane can be uniformly represented by the following linear state-space equations:

[0042]

[0043] in, These represent the position status, control input, and system output of the i-th drone or unmanned vehicle, respectively.

[0044] Regarding the leader matrices S and F, and the follower matrix A in this system... i B i C i It meets the following conditions:

[0045] 1) All eigenvalues ​​of matrix S have non-negative real parts;

[0046] 2) For All have (A) i B i It can be calmed, and matrix B i Reversible, (C i A i (F,S) can be detected, and matrix B is defined. p =[0 1] T , (S,B p Controllable;

[0047] 3) For the tracker matrix (A) in heterogeneous unmanned systems i B i ) and C i, i ∈ {1,..., N}, and leader matrices S and F, there exists a matrix pair (Π i , Γ i ) such that the following regulator equations hold:

[0048]

[0049] 4) Digraph represents the corresponding communication topology of the unmanned heterogeneous system, digraph has a spanning tree with the leader as the root node, and the communication between followers is undirected. At least one follower can accept the information of the leader.

[0050] Consensus algorithm based on dynamic topology

[0051] For each follower x i Design a distributed virtual leader observer:

[0052]

[0053] The state of the observer, i.e. the state of the follower system x i estimates the state of the leader q0, and γ is a constant to be designed, K p is the control gain matrix to be designed, B P = [0 1] T is a constant parameter matrix.

[0054] To achieve the formation control of the unmanned system, the following consensus control algorithm is proposed for i ∈ F:

[0055]

[0056] where K 1i is the feedback coefficient matrix, which can make A i + B i K 1i Hurwitz matrix, K 2i = Γ i - K 1i Π i , v i (t) is the input compensation signal of the time-varying formation, h xi represents the state deviation vector relative to the state of the follower x i (t);

[0057] Let q0= [q p0 q v0 ] T Then the above formula can be written as

[0058]

[0059] In the above formula:

[0060]

[0061] Assuming that each unmanned vehicle has limited communication capability, only communicating with other unmanned vehicles in its neighborhood, the communication neighborhood of an unmanned vehicle is defined as:

[0062]

[0063] where r c is called the aggregation radius, when ||x-p i ||2≤r c , the communication between this unmanned vehicle and the ith unmanned vehicle is good, at this time the weighted communication topology parameter between them is dynamically adjusted to w ij =1, r a is called the communication awareness radius, when ||x-p i ||2>r a , the communication between this unmanned vehicle and the ith unmanned vehicle is lost, at this time the weighted communication topology parameter between them is dynamically adjusted to w ij =0, when r c <||x-p i ||2≤r a , the communication between this unmanned vehicle and the ith unmanned vehicle changes linearly with their relative distance, at this time the weighted communication topology parameter between them is dynamically adjusted based on the state;

[0064] In summary, the element a ij in the adjacency matrix will be redefined as w ij which is dynamically adjusted in real time depending on the state quantity, the mathematical expression is as follows:

[0065]

[0066] where w ij is the weighted communication topology parameter between the ith and jth unmanned vehicles, p i and p j are symmetric, i.e. w ij = w ji , and the topology graph without considering time-varying communication is defined as

[0067] The state-dependent dynamic communication topology graph is defined as The initial state communication structure of the system is The neighbor set of each intelligent agent is N i ={j|||x-p i ||2≤r a , j=1, 2, …, N, j≠i}.

[0068] Consistent formation control algorithm with artificial potential field

[0069] The basic idea of artificial potential field method is that the unmanned vehicle, target point and obstacle in two-dimensional space are regarded as particles. In order to ensure the safety of formation and avoid obstacles and collision, repulsive force F rep will be applied to the unmanned vehicle by the obstacle and neighbor which is too close to cause collision; in order to ensure the stability of unmanned system formation and keep a certain formation, attractive force F att will be generated between unmanned vehicles which are relatively far away from each other, and the resultant force F d of attractive force and repulsive force is introduced into the formation consistency algorithm control system, so that the unmanned vehicle can keep relative formation and track the virtual leader to form formation under the conditions of obstacle avoidance and collision avoidance.

[0070] 1) Definition of potential field region

[0071] Taking the i-th unmanned vehicle as an example, the safety region of the unmanned vehicle is defined as follows:

[0072]

[0073] wherein r d is the minimum safety radius of the unmanned vehicle, if r , it means that the unmanned vehicle formation has collided;

[0074] The obstacle avoidance region of the unmanned vehicle is defined as follows:

[0075]

[0076] wherein r rep is the maximum obstacle avoidance radius of the unmanned vehicle, when the obstacle avoidance region of the unmanned vehicle appears obstacle or other unmanned vehicle, the obstacle avoidance and collision avoidance mode will be triggered, and the reverse potential field force will push the unmanned vehicle away to ensure the safety of the formation, and the range of each potential field region is shown in Figure 2 .

[0077] 2) Design of potential field function

[0078] Suppose the spatial position of unmanned vehicle i is p , the spatial position of unmanned vehicle j is p j = [p jx p jy ] T , the relative distance between the two unmanned vehicles is ||p ij || = ||p i -p j ||, and the potential field between the two unmanned vehicles is defined as ψ ij , then the potential field function ψ ij (||p ij ||) has the characteristics of continuity, always positive and unbounded.

[0079] Unmanned vehicle i The potential field force of the unmanned vehicle is: j

[0080]

[0081] where, represents the gradient of ψ ij (||p ij ||) at point p i ;

[0082] The total potential field of the unmanned vehicle i is represented as:

[0083]

[0084] where, N i ={j|||p i -p j ||≤r a}, represents the intelligent agents within the communication neighborhood of the unmanned vehicle i, r a is the communication radius of the unmanned vehicle;

[0085] When the relative distance is small, i.e., ||p ij ||→0, the potential field force between the unmanned vehicles is repulsive, and when the relative distance is large, i.e., ||p ij ||→r a , the unmanned vehicles are about to leave the communication range of each other, the potential field force between the unmanned vehicles is converted to attractive force to maintain the formation configuration; therefore, the attractive potential field and the repulsive potential field are designed respectively when designing the potential field of the unmanned vehicle, and the total potential field is as follows:

[0086]

[0087] The resultant force on the unmanned vehicle is:

[0088]

[0089] To ensure that the constructed potential field function is smooth and derivable, the σ norm is introduced, which is defined as follows:

[0090] Definition (σ norm) ||·|| σ :R n →R + is defined as:

[0091]

[0092] where, c a >0 is a constant.

[0093] a) Potential field function between unmanned vehicle and obstacle ​

[0094] In the obstacle avoidance area of the unmanned vehicle When the obstacle is detected, the obstacle generates a repulsive force on the unmanned vehicle, and the size of the repulsive force is negatively related to the distance between the obstacle and the unmanned vehicle, that is, the closer the unmanned vehicle is to the obstacle, the greater the repulsive force of obstacle avoidance.

[0095] The obstacle repulsive potential field function is selected as follows:

[0096]

[0097] Wherein, k o is a parameter to be designed, z=p io ||p i -p o -r o ||, p o is the position of the obstacle, r d is the minimum safety radius of the unmanned vehicle, r o is the radius of the obstacle, and r rep is the detection radius of the unmanned vehicle to the obstacle, when the distance between the unmanned vehicle and the obstacle is ||x-p i -r o || is less than r d , the potential field generates a repulsive force to ensure the safety of the formation; when the distance between the unmanned vehicle and the obstacle is greater than the detection radius, the potential field force is 0, assuming that the minimum safety radius is 1m, the maximum detection radius r rep is 5m, and the obstacle avoidance potential field function of the unmanned vehicle is shown in Figure 3 (a).

[0098] Then the corresponding potential field force between the unmanned vehicle and the obstacle is:

[0099]

[0100] b) Potential field function between unmanned vehicles

[0101] The inter-vehicle collision avoidance ensures that the unmanned vehicles maintain a safe distance when moving, so as not to collide, and the safety radius of the unmanned vehicle is given as r d When the relative distance between the unmanned vehicles is less than 2r d during the formation process, the two unmanned vehicles will generate a repulsive force in the same direction as the relative position vector, which is equivalent to that the unmanned vehicle i regards other unmanned vehicles in the formation as dynamic obstacles, and the collision avoidance neighborhood is

[0102] That is, the repulsive potential field function between the unmanned vehicles is as follows:

[0103]

[0104] Wherein, k r>0 is a parameter to be designed.

[0105] Definition of the attractive region of the unmanned vehicle wherein ra is the communication radius. When the relative distance between the unmanned vehicles is too large to exceed ra , the unmanned vehicles will lose communication. In order to maintain communication as much as possible, by applying virtual gravity to the unmanned vehicles, the unmanned vehicles with a relatively large distance are accelerated to approach each other until the equilibrium state is achieved.

[0106]

[0107] wherein, k a >0 is a parameter to be designed.

[0108] Through the design of the above potential field function, the repulsive force between the unmanned vehicles is

[0109]

[0110] The attractive force between the unmanned vehicles is

[0111]

[0112] When the relative distance is 0≤z<2r d , the unmanned vehicles generate repulsive force to move away from each other to prevent collision; when the relative distance is 2r d ≤z<r att , the potential function between the unmanned vehicles is 0, and the unmanned vehicles in the system are mainly completed by the consensus algorithm; when the relative distance is r att ≤z<r a , the unmanned vehicles generate attractive force to approach each other, forming a formation; assuming that the minimum safety radius is 1 m, the safety distance between the unmanned vehicles is 2 m, the obstacle avoidance radius r att = 10 m, and the communication radius r a = 12 m, the obstacle avoidance potential field function of the unmanned vehicle is as shown in Figure 3 (b).

[0113] 3) Formation obstacle avoidance algorithm

[0114] Based on the above potential function and the consensus control algorithm, the obstacle avoidance algorithm for the unmanned vehicle i∈F G is as follows:

[0115]

[0116] wherein, the formation obstacle avoidance input

[0117] The unmanned vehicle i∈F A consensus control algorithm is unchanged:

[0118]

[0119] wherein, is the state of the observer, i.e. x i is the estimated state of the leader q0, K 1i is the feedback coefficient matrix, which makes A i +B i K 1i is a Hurwitz matrix, K 2i = Γ i -K 1i Π i , v i (t) is the input compensation signal of the time-varying formation, h xi represents the state offset vector relative to the follower state x i (t).

[0120] Experimental simulation

[0121] Assume that the UAVs and UGVs are randomly generated in the two-dimensional coordinates [0, 20m], and at t = 0s, the UAVs are all stopped on the ground, the initial position of the virtual leader is [10m, 5m, 0m], and the virtual leader flies at a constant speed of 4m / s; it is assumed that communication is carried out between the virtual leader and the follower, and when the distance between the UAVs and UGVs is greater than the safety distance, the formation is controlled by the consensus algorithm; when the relative safety distance is too small, the primary task of the formation is to avoid obstacles, which is mainly controlled by the artificial potential field method.

[0122] Assume that the UGVs form a square formation, and the UAVs form a "one-word" formation for formation, and a circular obstacle is set at the track (10m, 25m), the obstacle radius r o = 2m, and the dotted line represents r o + r d , wherein r d = 3m is the minimum obstacle radius of the UGV. Figure 4 is the trajectory of the UGV avoiding the obstacle, wherein p0 represents the virtual leader of the formation, p1, p2, p3, and p4 represent the four UGVs in the formation, respectively. It can be seen that the second car (UGV2) avoids obstacles when encountering obstacles, but due to the higher priority of obstacle avoidance, the formation does not converge at this time, and until the obstacle is avoided, the consensus algorithm again plays a major role, forming the expected formation, Figure 5 is the formation tracking trajectory of the UAV and UGV in three-dimensional space. pA1, pA2, and pA3 are the three UAVs in the formation, which finally form a "one-word" formation.

[0123] Figure 6 and Figure 7The position error curve and the speed error curve in the formation are respectively, and p0A2 converges slower because the formation avoids obstacles at t=2.6s. It can be seen that the consistency error finally converges to 0, that is, the unmanned system can finally achieve good obstacle avoidance formation.

[0124] The above merely illustrates the preferred embodiments of the present application, which are used to explain the technical solutions of the present application, and cannot be used to limit the protection scope of the present application. Any modification made according to the technical idea of the present application and on the basis of the technical solutions falls within the protection scope of the present application.

Claims

1. A dynamic topology-based heterogeneous multi-unmanned system formation tracking control method, characterized in that, The method comprises the following steps: Step 1: Analyze the kinematics and dynamics model of the unmanned aerial vehicle and unmanned vehicle, decouple the control of the unmanned aerial vehicle system, control the height channel and the plane position channel respectively, so as to unify the mathematical model of the heterogeneous system; based on the "leader-follower" mode, establish a two-layer architecture, use a distributed observer to estimate the state of the virtual leader, and propose a consensus protocol to realize the tracking of the time-varying trajectory by the heterogeneous system; design a distributed virtual leader observer for each follower: ， wherein is the state of the observer, i.e. in the follower system estimated state of the leader , is a constant to be designed, is a control gain matrix to be designed, is a constant parameter matrix, is an element in the adjacency matrix; To realize the formation control of unmanned systems, the following consistency control algorithm is proposed: The following consistency control algorithm is proposed: ， wherein is a feedback coefficient matrix, such that is a Hurwitz matrix, , is an input compensation signal of the time-varying formation, denotes a state deviation vector from the state of the follower, F is a leader matrix, is a pair of matrices, , is a tracker matrix; Let , , Then the above equation is: ， In the formula: ； Step 2: Abstract the concept of communication neighborhood based on the problem that the communication ability between unmanned vehicles is affected by distance, and establish a state-dependent dynamic joint communication topology structure, the connection strength between each intelligent agent and its neighbors in the communication range changes linearly with the change of relative position; Step 3: define the safety area, obstacle avoidance area, and attractive area of the unmanned vehicle, add virtual artificial potential field, introduce Norm, respectively construct the smooth and continuous attractive field potential energy function and repulsive field potential energy function, and obtain the virtual potential field force received by each unmanned vehicle, realize the obstacle avoidance and collision avoidance control of the unmanned vehicle subsystem during the time-varying tracking formation of the unmanned aerial vehicle and the unmanned vehicle.

2. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method according to claim 1, wherein, The leader state space equation in step 1 is: ， wherein, and are constant matrices; and represent the state and output of the leader, respectively; The height channel dynamics equation of the unmanned aerial vehicle is: ， wherein, represents the first position of the drone in space, represents the velocity of the drone in space, respectively represent the pitch angle of the drone; wherein is the acceleration of gravity; The unmanned vehicle-unmanned aerial vehicle heterogeneous system is located at The dynamics equation of the plane is expressed by the following linear state space equation: ， wherein, represent the position state, control input and system output of the i-th drone or unmanned vehicle, respectively, , , denotes the tracker matrix.​ 3. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method of claim 1, wherein, The communication neighborhood of the unmanned vehicle in step 2 is: ， wherein, is the aggregation radius, and is the communication sensing radius, and is the communication sensing radius, and , is the communication sensing radius, and is the communication sensing radius, and is the communication sensing radius, and is the communication sensing radius, and is the communication sensing radius, and is the communication sensing radius, and 4. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method of claim 3, wherein, The elements in the adjacency matrix are redefined to dynamically adjust in real time depending on the state quantities , as follows: ， wherein, With respect to and Symmetric, i.e. , the time-varying communication topology is defined as , the state-dependent dynamic communication topology is defined as , the system initial state communication structure is The neighbor set of each agent is .

5. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method according to claim 1 or 3, characterized in that, The safety area of the unmanned vehicle in step 3 is: ， wherein, is the minimum safety radius of the unmanned vehicle, if then it means that the unmanned vehicle platoon has collided. The obstacle avoidance area of the unmanned vehicle is: ， wherein, is the maximum obstacle avoidance radius of the unmanned vehicle.

6. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method according to claim 5, wherein, In step 3, to ensure the constructed potential field function is smooth and derivable, the following function is introduced norm, (1) norm) is defined as: ， wherein is a constant.

7. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method according to claim 6, wherein, The potential field force between the unmanned vehicle and the obstacle in step 3 is: 。 8. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method according to claim 6, wherein, The repulsive force between the unmanned vehicles in step 3 is: ， The attractive force between the unmanned vehicles is: ， When the relative distance is greater than a certain value, repulsive force is generated between the unmanned vehicles to keep them away from each other to prevent collision; when the relative distance is equal to zero, the potential energy function between the unmanned vehicles is zero, and the unmanned vehicles in the system are mainly formed into a formation by a consensus algorithm; when the relative distance is less than a certain value, attractive force is generated between the unmanned vehicles to keep them close to each other to form a formation.

9. The dynamic topology based heterogeneous multi-unmanned system formation tracking control method of claim 8, wherein, The obstacle avoidance algorithm of the unmanned vehicle is: ， wherein the platoon obstacle avoidance input .

Images