Event-driven privacy protection method for cooperative formation of unmanned aerial vehicle cluster system

Abstract

The invention provides an event-driven privacy protection method for cooperative formation of unmanned aerial vehicle cluster systems, and relates to the technical field of unmanned aerial vehicle cluster control. Comprising the following steps: modeling a communication topology of an unmanned aerial vehicle cluster by using a graph theory, and determining a formation center of a directed strong connectivity balance topology; establishing a motion model of the unmanned aerial vehicle cluster according to the topology of the unmanned aerial vehicle cluster and the formation center; on the basis of a traditional privacy protection collaborative algorithm, a control item based on an event triggering strategy is introduced, real communication data are protected by utilizing a finite time mask function, and a novel privacy protection composite control law is provided; by using the design, the cooperative formation of the unmanned aerial vehicle cluster system is realized, and the privacy of the initial position and speed of each unmanned aerial vehicle is ensured. According to the invention, the problem of network topology limitation in the unmanned aerial vehicle privacy protection formation realized by using deterministic interference in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of drone cluster control, and in particular to an event-driven privacy protection method for a collaborative formation of a drone cluster system. Background Art

[0002] The formation control of multi-agent systems has attracted more and more attention in multiple scientific fields, such as mobile robot formation systems, drones, and satellite formation systems. However, due to the open nature of wireless communication channels in formation cooperative control, when drone cluster systems use real information for communication, their information can be easily eavesdropped, resulting in individual privacy leakage. For example, when multiple drones from different stakeholders perform formation tasks, each drone may carry its own initial sensitive information, which is not allowed to be accessed by other drones or attackers. Therefore, it is of great significance to study the issue of privacy protection.

[0003] At present, privacy protection methods can be divided into three categories: cryptographic methods, state decomposition methods, and methods of adding interference signals. Cryptography-based methods mainly use homomorphic cryptographic systems for encryption, such as the Paillier cryptographic system; however, since the encryption process will cause huge computational costs, it is not suitable for small multi-agent clusters such as drone clusters. The state decomposition-based method is to decompose individual information into two parts, and only one part of the information is used for system formation control. This method is currently often applied to discrete systems. The method of adding interference signals refers to introducing some interference into the real signal to obtain a false state value. Each agent uses the communication network to exchange the false information and use it for the design of the formation controller. In current research, there are usually two types of interference signals, namely random interference and deterministic interference. The method using random interference signals is represented by differential privacy methods, which often cannot achieve precise formation control, while the method based on deterministic interference requires adding certain topological restrictions to achieve effective privacy protection, which greatly limits its scope of application. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide an event-driven privacy protection method for the collaborative formation of a drone cluster system.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] An event-driven privacy protection method for cooperative formation of drone swarm system, comprising:

[0007] Use graph theory to model the communication topology of drone clusters and determine the formation center of their directed strongly connected equilibrium topology;

[0008] Establishing a motion model of the drone cluster formation according to the directed strongly connected balanced topology of the drone cluster and the formation center;

[0009] According to the motion model, a finite-time mask function is designed to establish a privacy protection control law based on false information for the coordinated formation of the UAV swarm system;

[0010] According to the motion model and the above control law, a class of control laws based on event triggering strategy is introduced to obtain the privacy protection composite control law of the system formation;

[0011] The target UAV cluster is formed and controlled according to the designed privacy-preserving composite control law.

[0012] Preferably, the communication topology of the drone cluster is modeled using graph theory, including:

[0013] Determine the set of drone nodes;

[0014] Determine the out-degree and in-degree of the drone node according to the drone node set;

[0015] Determine the adjacency matrix of the network topology;

[0016] Determine the Laplacian matrix of the network topology based on the out-degree and adjacency matrix of each drone node;

[0017] A directed strongly connected balanced communication topology of the drone cluster is constructed according to the Laplace matrix and the out-degree and in-degree of each drone node.

[0018] Preferably, the method for determining the formation center is:

[0019] Determine the location of the initial formation center;

[0020] Based on the directed strongly connected balanced topology, determine the coordinate value of the drone node in the inertial coordinate system and the coordinate value of the neighboring drone node of the current drone node in the inertial coordinate system;

[0021] Determine the position vector coordinates of the drone node and the position vector coordinates of the neighboring drone nodes of the current drone node according to the coordinate values ​​of the drone node in the inertial coordinate system and the coordinate values ​​of the neighboring drone nodes of the current drone node in the inertial coordinate system;

[0022] Determine a first position deviation and a second position deviation according to the position vector coordinates of the drone node, the position vector coordinates of the neighboring drone node of the current drone node and the position of the initial formation center, wherein the first position deviation is the vector coordinates of the deviation between the drone node position and the initial formation center position; the second position deviation is the vector coordinates of the deviation between the neighboring drone node position of the current drone node and the initial formation center position;

[0023] The first position deviation and the second position deviation are used to adjust the position of the initial formation center to determine the final position of the formation center.

[0024] Preferably, the expression of the motion model of the drone cluster formation is:

[0025]

[0026] Among them, m i For drone node d i The quality of i (t) represents the drone node d i The 3*1 position vector that satisfies the inertial coordinate system at time t during the flight; v i (t) represents the drone node d i During the flight, at time t, the 3*1 velocity vector of the inertial coordinate system is satisfied. Represents the drone node d i The 3*1 acceleration vector at time t during flight; u i1 (t) and u i2 (t) respectively represent the drone node d i The control law based on local false information communication and the control law based on event trigger strategy; d i is the current drone node.

[0027] Preferably, the drone node d i The expression of the local false information exchange control law is:

[0028]

[0029] in, and Represents the drone node d i The false position information and velocity information used for information exchange can be obtained by transforming the real position and velocity information in the inertial coordinate system through a finite time-varying mask function; i Represents the drone node d i The vector coordinates of the deviation between the position of the formation and the center position in the ground coordinate system. Let d jis the neighboring drone node of the current drone node. Similarly, and Respectively represent the neighboring drone nodes d j The false position and velocity information after transformation by the finite time-varying mask function; o j Represents the drone node d j The vector coordinates of the deviation between the position in the ground coordinate system and the center position of the formation. ij For drone node d i Its neighboring drone node d j The adjacency matrix coefficient of ; α is the position control coefficient, and β is the speed control coefficient.

[0030] Preferably, all drones are set to be triggered at the initial time, that is, t1=0, and the expression of the control law based on the event trigger strategy is:

[0031]

[0032] Among them, v i (t) and v i (t k ) are drone nodes d i At time t and t k The true velocity vector at time t k It is a series of trigger time points that meet the event trigger control conditions, which are determined by the data center according to the event trigger control conditions; t k+1 Indicates that at the trigger time t k T is the latest triggering time point after the event; T is the specific time point selected by the designer, reflecting the action range of the event-triggered protection control rate;

[0033] Preferably, the expression of the composite control law is:

[0034]

[0035] The present invention discloses the following technical effects:

[0036] The present invention provides an event-driven privacy protection method for collaborative formation of a drone cluster system, comprising: using graph theory to establish a directed strongly connected balanced communication topology of a drone cluster and determine a formation center; establishing a motion model of a drone cluster formation according to the directed strongly connected balanced communication topology of the drone cluster and the formation center; establishing a control law based on local false information exchange of drone nodes and a control law based on event triggering strategy of drone nodes according to the motion model; determining a composite control law according to the control law based on local false information exchange of drone nodes and the control law based on event triggering strategy of drone nodes; and controlling a target drone cluster according to the composite control law. The method of the present invention uses graph theory to establish a UAV formation motion model and a directed strongly connected balanced communication topology of the formation system, constructs a privacy-preserving collaborative formation control law, solves the privacy leakage problem caused by direct communication in traditional control methods, and achieves the desired formation flight goal at the same time; the present invention uses false information transformed by a finite time-varying mask function for local communication exchange control within a period of time, prevents the leakage of the real initial position and speed information of the UAV during the communication process and achieves the desired collaborative formation flight goal; the present invention adopts an event triggering strategy, each UAV node transmits the real speed to the data center and determines the triggering time, thereby constructing a private non-trivial dynamic behavior inside the UAV node, which is used to solve the network topology condition restriction problem generated in existing research, prevents other UAVs or attackers from inferring the real initial position information of the UAV based on the observer method, and has good feasibility and applicability; the method of the present invention has low complexity and is easy to implement. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0038] Figure 1 A flow chart of an event-driven privacy protection method for cooperative formation of a drone cluster system provided by an embodiment of the present invention;

[0039] Figure 2 A schematic diagram of an event-driven privacy protection method strategy for a drone cluster system collaborative formation provided by an embodiment of the present invention;

[0040] Figure 3 A cluster communication topology diagram of five drones provided in an embodiment of the present invention;

[0041] Figure 4A 3D schematic diagram of the real initial positions of takeoff of five unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention and the false initial positions after transformation by a finite time-varying mask function;

[0042] Figure 5 The trajectory diagram of 5 unmanned aerial vehicles (UAVs) provided in the embodiment of the present invention, wherein: Figure 5 (a) The x-axis, y-axis, and z-axis position trajectory diagrams of the true position and the false position after transformation by the finite time-varying mask function minus the deviation, respectively. Figure 5 (b) The true speeds of the x-axis, y-axis, and z-axis of the five unmanned aerial vehicles (UAVs) and the false speed trajectories after being transformed by the finite time-varying mask function;

[0043] Figure 6 A triggering time interval diagram of the event-triggered controller based on five unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention,

[0044] Figure 7 The actual speed error of the five drones provided in the embodiment of the present invention picture;

[0045] Figure 8 The x-axis privacy trigger error diagram of 5 unmanned aerial vehicles (UAVs) provided in the embodiment of the present invention;

[0046] Fig. 9 A 3D trajectory simulation diagram of five unmanned aerial vehicles (UAVs) in flight provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0049] like Figure 1 As shown, the present invention provides an event-driven privacy protection method for cooperative formation of a drone cluster system, comprising:

[0050] Step 100: Use graph theory to model the communication topology of the drone cluster and determine the formation center of its directed strongly connected equilibrium topology;

[0051] Step 200: establishing a motion model of the drone cluster formation according to the directed strongly connected balanced topology of the drone cluster and the formation center;

[0052] Step 300: According to the motion model, a finite-time mask function is designed to establish a privacy protection control law based on false information for the coordinated formation of the drone cluster system;

[0053] Step 400: According to the motion model and the above control laws, a class of control laws based on event triggering strategies is introduced to obtain a privacy protection composite control law for the system formation;

[0054] Step 500: Perform formation control on the target UAV cluster according to the designed privacy-preserving composite control law.

[0055] Furthermore, graph theory is used to establish a directed strongly connected balanced communication topology for the drone cluster, including:

[0056] Determine the set of drone nodes;

[0057] Determine the out-degree and in-degree of the drone node according to the drone node set;

[0058] Determine the adjacency matrix of the network topology;

[0059] Determine the Laplacian matrix of the network topology structure based on the out-degree and adjacency matrix of each drone node;

[0060] A directed strongly connected balanced communication topology structure of the drone cluster is constructed according to the Laplace matrix and the out-degree and in-degree of each drone node.

[0061] The network communication topology structure assumed in the present invention is a directed strongly connected balanced graph, which is a widely used graph in graph theory.

[0062] Specifically, Figure 2 As shown, in order to realize the state representation of the UAV node, the inertial coordinate system E is used 地 -OXYZ;

[0063] Among them, the inertial coordinate system E 地 -OXYZ is a coordinate system fixed to the earth's surface. The origin O of the coordinate system is selected at a point on the ground plane. The OX axis is the direction pointing to the target. The OY axis is perpendicular to the OX axis. The OZ axis is perpendicular to the other two axes and forms a right-handed rectangular coordinate system. The directed strongly connected balanced communication topology of the drone cluster is:

[0064] G = (V, E);

[0065] Where V is the set of drone nodes, V = {d1, d2, ..., d n}, n is the total number of drone cluster nodes; E = V × V represents the set of edges between drone nodes; if drone node d j To drone node d i To transmit information, the drone node d j is the drone node d i The neighbor node of the drone node d i is the drone node d j The neighboring nodes of drone node d are i∈n, j∈n, i≠j. i The neighboring drone node d j The set is denoted by N i out , N i out ={(d j |j=1,2,…,s,(i,j)∈E}, s is the drone node d i The neighboring drone node d j The total number is called drone node d i out-degree; drone node d i The neighboring drone node d j The set is denoted by N i in , N i in ={(d j |j=1,2,…,k,(j,i)∈E}, k is the drone node d i The neighboring drone node d j The total number is called drone node d i If (i,j)∈E, then the corresponding adjacency matrix coefficient a is called ij =1∈A, A is the adjacency matrix of the network topology. Define L=DA as the Laplace matrix of the network topology, where D=diag{s1,…,s n};L T is the transposed matrix of L; And λ2(L s ) is L s The second smallest eigenvalue other than the zero eigenvalue. If the out-degree of any drone node is equal to the in-degree and there is a communication channel between any two drone nodes, the corresponding communication topology is called a directed strongly connected balanced communication topology. This study assumes that the communication topology is a directed strongly connected balanced communication topology;

[0066] Furthermore, the method for determining the formation center is:

[0067] Determine the location of the initial formation center;

[0068] Based on the directed strongly connected balanced topology, determine the coordinate value of the drone node in the inertial coordinate system and the coordinate value of the neighboring drone node of the current drone node in the inertial coordinate system;

[0069] Determine the position vector coordinates of the drone node and the position vector coordinates of the neighboring drone nodes of the current drone node according to the coordinate values ​​of the drone node in the inertial coordinate system and the coordinate values ​​of the neighboring drone nodes of the current drone node in the inertial coordinate system;

[0070] Determine a first position deviation and a second position deviation according to the position vector coordinates of the drone node, the position vector coordinates of the neighboring drone node of the current drone node and the position of the initial formation center, wherein the first position deviation is the vector coordinates of the deviation between the drone node position and the initial formation center position; the second position deviation is the vector coordinates of the deviation between the neighboring drone node position of the current drone node and the initial formation center position;

[0071] The first position deviation and the second position deviation are used to adjust the position of the initial formation center to determine the final position of the formation center.

[0072] Specifically, Figure 3 As shown, the drone node d i In the inertial coordinate system E 地 -OXYZ's position and deviation from the formation center are defined as follows:

[0073] p i Represents the drone node d i In the ground coordinate system E 地 -Position vector coordinates in OXYZ, p i =[x i ,y i ,z i ].

[0074] x i Represents the drone node d i In the ground coordinate system E 地 - The X-coordinate in OXYZ.

[0075] y i Represents the drone node d i In the ground coordinate system E 地 -Y coordinate in OXYZ.

[0076] z i Represents the drone node d i In the ground coordinate system E 地 - The coordinate in the Z direction in OXYZ.

[0077] o i Represents the drone node d i In the ground coordinate system E 地 -The vector coordinates of the deviation between the position in OXYZ and the center position of the formation, o i =[o xi ,o yi ,,o zi ].

[0078] o xi Represents the drone node d i In the ground coordinate system E 地 - Deviation in the X direction in OXYZ.

[0079] o yi Represents the drone node d i In the ground coordinate system E 地 - Deviation in the Y direction in OXYZ.

[0080] o zi Represents the drone node d i In the ground coordinate system E 地 - Deviation in Z direction in OXYZ.

[0081] Therefore, the drone node d i The neighboring drone node d j In the inertial coordinate system E 地 -OXYZ's position and deviation from the formation center are defined as follows:

[0082] p j Represents the neighboring drone node d j In the ground coordinate system E 地 -Position vector coordinates in OXYZ, p j =[x j ,y j ,z j ].

[0083] x j Represents the neighboring drone node d j In the ground coordinate system E 地 - The X-coordinate in OXYZ.

[0084] y j Represents the neighboring drone node d j In the ground coordinate system E 地 -Y coordinate in OXYZ.

[0085] z j Represents the neighboring drone node d j In the ground coordinate system E 地 - The coordinate in the Z direction in OXYZ.

[0086] o j Represents the drone node d j In the ground coordinate system E 地 -The vector coordinates of the deviation between the position in OXYZ and the center position of the formation, o j =[o xj ,o yj ,o zj ].

[0087] o xj Represents the drone node d j In the ground coordinate system E 地 - Deviation in the X direction in OXYZ.

[0088] o yj Represents the drone node d j In the ground coordinate system E 地 - Deviation in the Y direction in OXYZ.

[0089] o zj Represents the drone node d j In the ground coordinate system E 地 - Deviation in Z direction in OXYZ.

[0090] In this embodiment, the formation center of the drone cluster is expressed as follows: the formation center is in the ground coordinate system E 地 The position vector coordinates in -OXYZ are The center of the formation is in the ground coordinate system E 地 The velocity vector coordinates in -OXYZ are where x i (0), y i (0), z i (0) are drone nodes d i In the ground coordinate system E 地 - Initial position vector coordinates in OXYZ; They are respectively UAV nodes d i In the ground coordinate system E 地 - Initial velocity vector coordinates in OXYZ.

[0091] Furthermore, the expression of the motion model of the UAV cluster formation is:

[0092]

[0093] Among them, d i is the current drone node, m i For drone node d i mass; in this case, each drone has the same mass; p iRepresents the drone node d i During the flight, the 3*1 position vector (x i ,y i ,z i );v i Represents the drone node d i During the flight, the 3*1 velocity vector that satisfies the inertial coordinate system where x i and Correspondingly, it represents the drone node d i Position and velocity information on the x-axis, similarly, y i and Correspondingly, z i and Correspondingly, they represent the drone nodes d i Position information and speed information on the y-axis and z-axis; Represents the drone node d i The acceleration during flight, u i1 (t) and u i2 (t) are drone nodes d i The control law based on local false information exchange and the control law based on event-triggered strategy.

[0094] Furthermore, the drone node d i The expression of the local false information exchange control law is:

[0095]

[0096] in, and Represents the drone node d i The real position and velocity information in the inertial coordinate system is transformed by a finite time-varying mask function to become false position and velocity information used for information exchange; i Represents the drone node d i The vector coordinates of the deviation between the position of the formation and the center position in the ground coordinate system. Let d j is the neighboring drone node of the current drone node. Similarly, and Respectively represent the neighboring drone nodes d j The false position and velocity information after transformation by the finite time-varying mask function; o j Represents the drone node d j The vector coordinates of the deviation between the position in the ground coordinate system and the center position of the formation;

[0097] Specifically, The expression is:

[0098]

[0099] The expression is:

[0100]

[0101] Among them, μ(t) represents the time transformation scale function, which is expressed as is the specific time point chosen by the designer, reflecting the effective interval of the mask function; h is the parameter of μ(t); and further, γ i μ(t) -r Indicates that the drone node d i The selected finite time-varying mask function, γ i For drone node d i The privacy coefficient is selected by the current drone node and is not disclosed to any other drone nodes. r is the parameter of the finite time-varying mask function. It should be noted that the finite time-varying mask function is Memory exists, in disappears; α>0 is the position control coefficient, β>0 is the speed control coefficient, α and β satisfy the constraint conditions a ij Represents the drone node d i With drone node d j The adjacency matrix coefficient of the drone is i The initial position information of the drone is sensitive, so we only need to protect the position information within a period of time so that other drones or attackers cannot infer the initial position, and there is no need to protect the position information during the entire control process; Therefore, in the control process, it is necessary to protect not only the position information of the UAV, but also the speed information of the UAV; since the final goal of the formation task is to i The position and formation center position remain o i Position deviation of drone node d i The speed is kept the same as the center speed of the formation, so in the controller design process, it is only necessary to subtract the corresponding deviation from the position variable to achieve the formation control effect.

[0102] Furthermore, a control law based on event-triggered strategy for drone nodes is established to ensure privacy protection while removing the restrictions of existing network topology conditions:

[0103]

[0104] Among them, t kIt is a series of trigger time points that meet the event trigger control conditions, which are determined by the data center according to the event trigger control conditions. Assume that the control law based on the event trigger strategy is triggered for the first time at t0 = 0, let ∈ i (t) = v i (t)-v i (t k ) is the drone node d i The privacy trigger error is , then the event trigger control condition is:

[0105] Where μ(t) represents the time transformation scale function, h is the parameter of μ(t), r represents the parameter of the finite time-varying mask function, and h and r satisfy the relationship It is understandable that u i2 The update of (t) is determined by the data center. Each drone node sends the real speed information to the data center, and the data center determines whether to trigger according to the event trigger control conditions. If triggered, it sends a message to each drone, such as drone node d i , transmission t k , UAV node d i Update i2 (t). In addition, due to u i2 (t) Only drone node d i Therefore, other drones or attackers cannot infer the drone node d based on the observer method. i Initial position, thus solving the network topology condition limitation that needs to be set additionally in the existing methods At the same time, i2 (t) can be regarded as a drone node d i Its own non-trivial dynamics, its introduction will not change the original position and speed of the formation center, and can realize the UAV cluster formation flight mission; in addition, since only a period of position information needs to be protected, it is the same as adding a time period to the finite time-varying mask function, and only needs to be added in Memory exists when When i2 (t)=0.

[0106] Furthermore, the expression of the composite control law is:

[0107]

[0108] Furthermore, the present invention also provides another embodiment for performing formation control on a cluster of 5 drones. When the 5 drones are performing tasks, the formation control is performed according to the drone formation control method based on event-driven control of privacy-preserving collaborative formation. According to the communication topology established in step 1, Figure 3 It is worth noting that Figure 3 The corresponding network topology does not meet the network topology condition constraints required by existing methods Right now However, under the design of the present invention, the drone cluster system under this network topology can achieve the privacy protection goal, and the initial positions of the five drones are set to: p1(0) = [2, 3, 5] T m, p2(0) = [3, 4, 6] T m, p3(0) = [4, 5, 7] T m, p4(0) = [3, 3.5, 4.5] T m,p5(0)=[5.5,5,6] T m; the initial speed of the 5 drones is: v1(0) = [1, 2, 1] T m / s,v2(0)=[2,3,1.6] T m / s, v3(0)=[1.2, 2.3, 1] T m / s, v4(0)=[2.5, 3.5, 1.6] T m / s, v5(0)=[1.8, 1.7, 0.8] T m / s; the position vector trajectory of the formation center is p0=[3.5+1.7*t,4.1+2.5*t,5.7+1.2*t]; the velocity vector coordinates of the formation center are v p0 =[1.7,2.5,1.2];

[0109] Establish the motion model of the UAV cluster formation and set the mass of the UAV model to m i =50kg;

[0110] Design a control law based on local false information exchange. In order to calculate the control law, set a finite time r = 2.1; h = 5; position control coefficient α = 2.9; speed control coefficient β = 3.1; privacy protection coefficient γ i γ1=3.04; γ2=3.23; γ3=3.5; γ4=3.37; γ5=3.57; The position deviations of the five drones from the formation center are given as follows; o1=[1,0,0] T m;o2=[3,2,0] T m;o3=[5,0,0] T m;o4=[4,-1,0] T m;o5=[2,-1,0] T m;

[0111] The control law based on local false information communication and the control law based on event trigger strategy are combined to obtain the UAV node d i The composite control law.

[0112] Simulation results analysis, from Figure 4 It can be seen that the finite time-varying mask function designed by the present invention makes the real initial positions and false initial positions of the five drones different, thus achieving the goal of protecting the initial positions. Figure 5 As shown, Figure 5 Each small figure in a represents the position trajectory of the five drones along the x-axis, y-axis, and z-axis directions, respectively, and the solid line trajectories in the small figures represent the trajectory of the real position minus the deviation, and the dotted line trajectories represent the trajectory of the false position minus the deviation. Figure 5 a It can be seen that minus the deviation o i The position trajectory of the last five drones is the same as the position trajectory of the formation center, proving that the present invention can enable the drone cluster formation to achieve collaborative formation control. In addition, the false position trajectory transformed by the finite time mask function within 0-5 seconds is different from the real position trajectory, proving that the present invention can protect the position trajectory for a period of time to prevent eavesdroppers from inferring the real initial position of the drone based on the position information at other times. Similarly, Figure 5 Each small graph in b represents the velocity trajectory diagrams of the real speed and false speed of the five drones along the x-axis, y-axis, and z-axis directions, where the solid line trajectory in the small graph represents the real speed trajectory diagram, and the dotted line represents the false speed trajectory diagram. The speed trajectory of the five drones is the same as the speed trajectory of the formation center, proving that the present invention can enable the drone cluster formation to achieve collaborative formation control. In addition, the false speed trajectory and the real speed trajectory transformed by a finite time-varying mask function within 0-5 seconds are different, proving that the present invention can protect the speed trajectory for a period of time, thereby preventing eavesdroppers from inferring the real position information of the drone based on the real speed information. Figure 6-7 It can be seen that the controller of the present invention triggers a limited number of times within the control time. In addition, the speed error is defined as It shows that the speed error of the design of the present invention is small and can meet the control accuracy requirement. Figure 6 represents the controller trigger frequency t based on the event trigger strategy k+1 -t k Since the drone cluster is triggered uniformly according to the trigger command of the data center, it represents the uniform trigger interval of the five drones. Figure 7 It shows the true speed error trajectory of 5 UAVs. Figure 8 It can be seen that the privacy trigger error of each drone in the present invention is different, which proves that under the design of the present invention, the eavesdropper cannot infer the initial position of the drone node based on the observer-based method, thereby solving the topological condition limitations in the existing design. Fig. 9 It can be seen that the UAV cluster formation control effect is good and the expected formation effect can be achieved.

[0113] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0114] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. An event-driven privacy protection method for cooperative formation of drone swarm system, characterized in that: include: Use graph theory to model the communication topology of drone clusters and determine the formation center of their directed strongly connected equilibrium topology; Establishing a motion model of the drone cluster formation according to the directed strongly connected balanced topology of the drone cluster and the formation center; According to the motion model, a finite-time mask function is designed to establish a privacy protection control law based on false information for the coordinated formation of the UAV swarm system; According to the motion model and the above control law, a class of control laws based on event triggering strategy is introduced to obtain the privacy protection composite control law of the system formation; The target UAV cluster is formed and controlled according to the designed privacy-preserving composite control law.

2. The event-driven privacy protection method for cooperative formation of drone cluster system according to claim 1 is characterized in that: Use graph theory to model the communication topology of drone clusters, including: Determine the set of drone nodes; Determine the out-degree and in-degree of the drone node according to the drone node set; Determine the adjacency matrix of the network topology; Determine the Laplacian matrix of the network topology based on the out-degree and adjacency matrix of each drone node; A directed strongly connected balanced communication topology of the drone cluster is constructed according to the Laplace matrix and the out-degree and in-degree of each drone node.

3. The event-driven privacy protection method for cooperative formation of drone cluster system according to claim 1 is characterized in that: The method for determining the formation center is: Determine the location of the initial formation center; Based on the directed strongly connected balanced topology, determine the coordinate value of the drone node in the inertial coordinate system and the coordinate value of the neighboring drone node of the current drone node in the inertial coordinate system; Determine the position vector coordinates of the drone node and the position vector coordinates of the neighboring drone nodes of the current drone node according to the coordinate values ​​of the drone node in the inertial coordinate system and the coordinate values ​​of the neighboring drone nodes of the current drone node in the inertial coordinate system; Determine a first position deviation and a second position deviation according to the position vector coordinates of the drone node, the position vector coordinates of the neighboring drone node of the current drone node and the position of the initial formation center, wherein the first position deviation is the vector coordinates of the deviation between the drone node position and the initial formation center position; the second position deviation is the vector coordinates of the deviation between the neighboring drone node position of the current drone node and the initial formation center position; The first position deviation and the second position deviation are used to adjust the position of the initial formation center to determine the final position of the formation center.

4. The event-driven privacy protection method for cooperative formation of drone swarm system according to claim 1 is characterized in that: The expression of the motion model of the UAV cluster formation is: Among them, m i For drone node d i The quality of i (t) represents the drone node d i The 3*1 position vector that satisfies the inertial coordinate system at time t during the flight; v i (t) represents the drone node d i During the flight, at time t, the 3*1 velocity vector of the inertial coordinate system is satisfied. Represents the drone node d i The 3*1 acceleration vector at time t during flight; u i1 (t) and u i2 (t) respectively represent the drone node d i The control law based on local false information communication and the control law based on event trigger strategy; d i is the current drone node.

5. The event-driven privacy protection method for cooperative formation of drone cluster system according to claim 4 is characterized in that: The drone node d i The expression of the local false information exchange control law is: in, and Represents the drone node d i The false position information and velocity information used for information exchange can be obtained by transforming the real position and velocity information in the inertial coordinate system through a finite time-varying mask function; i Represents the drone node d i The vector coordinates of the deviation between the position of the formation and the center position in the ground coordinate system. Let d j is the neighboring drone node of the current drone node. Similarly, and Respectively represent the neighboring drone nodes d j The false position and velocity information after transformation by the finite time-varying mask function; o j Represents the drone node d j The vector coordinates of the deviation between the position in the ground coordinate system and the center position of the formation. ij For drone node d i Its neighboring drone node d j The adjacency matrix coefficient of ; α is the position control coefficient, and β is the speed control coefficient.

6. The event-driven privacy protection method for cooperative formation of drone cluster system according to claim 5 is characterized in that: Assuming that all drones are triggered at the initial time, that is, t1 = 0, the expression of the control law based on the event trigger strategy is: Among them, v i (t) and v i (t k ) are drone nodes d i At time t and t k The true velocity vector at time t k It is a series of trigger time points that meet the event trigger control conditions, which are determined by the data center according to the event trigger control conditions; t k+1 Indicates that at the trigger time t k T is the latest triggering time point after the event; T is the specific time point selected by the designer, reflecting the action range of the event-triggered protection control rate.

7. The event-driven privacy protection method for cooperative formation of drone cluster system according to claim 6 is characterized in that: The expression of the composite control law is:

Images