An unmanned cluster topology repair method in highly dynamic environments

Abstract

The present invention relates to a method for repairing unmanned swarm topology in a highly dynamic environment. The method belongs to the field of control technology and is intended to address the problem of reduced communication efficiency and mission execution capability caused by network topology damage after a large-scale node failure in a drone swarm. The method constructs a network model and utilizes virtual forces such as guiding force, aggregation force, topological force, and isolated node boundary force to guide the movement of surviving nodes, restore network connectivity, and optimize the network topology. Furthermore, through triangulation based on the particle swarm-simulated annealing algorithm, network coverage is further optimized and network performance is enhanced. The present invention effectively improves the stability and reliability of drone swarm networks, reduces topology repair costs, and improves network coverage, providing strong support for the application of drone swarm networks in complex environments.

Description

Technical Field

[0001] The present invention belongs to the field of control technology and relates to an unmanned cluster topology repair method in a highly dynamic environment. Background Art

[0002] Drones, due to their flexible mobility, easy deployment, low networking costs, and multi-sensor integration, possess a high degree of autonomy and adaptability. They have found widespread application in relay communications, agriculture, logistics, disaster monitoring, environmental protection and monitoring, infrastructure inspection, and other fields. Furthermore, with the expansion of drone applications and the complex and ever-changing application environments, drone swarm networks, which combine multiple drones into a network, are enabling collaborative work and information sharing. Drone swarm networks feature distributed control, self-organizing collaboration, task division and cooperation, strong robustness, and wide coverage. These networks enable more complex and efficient mission execution and are adaptable to a wide range of application scenarios and environments.

[0003] During the execution of a UAV swarm mission, drone nodes may become damaged due to their own energy limitations or external environmental factors, impacting the swarm's topology and mission-completion capabilities. Therefore, to ensure that the UAV swarm can quickly recover network performance and maintain data transmission and information sharing after topology changes, a topology repair method for UAV swarm communication networks is needed. First, it can improve network reliability and stability. UAV swarm networks are often used in critical tasks such as search and rescue, disaster monitoring, and other tasks that require high connectivity. Topology repair can quickly recover from network connectivity issues caused by failed nodes, ensuring reliable network transmission, thereby better responding to emergencies and protecting people's lives. Second, topology repair can improve the efficiency and performance of UAV swarm networks. Given the large scale and frequent topology changes of UAV swarm networks, topology repair can optimize network coverage, extend network life, and optimize data transmission paths, thereby improving overall network performance. Third, topology repair can enhance the adaptability of swarm networks. In the face of complex and changing environments and mission requirements, topology repair methods can enable swarm networks to acquire self-awareness, information sharing, and autonomous decision-making capabilities, rapidly adjust network structure, and improve fault tolerance and adaptability.

[0004] On the one hand, the inventors' existing research on topology repair has focused on wireless sensor networks and mobile ad hoc networks. Relatively little research has been conducted on flying ad hoc networks, which feature high mobility, dramatic topology dynamics, high connectivity requirements, and long link communication distances. Most research has focused on formation control, mission planning, and topology construction in specific scenarios. On the other hand, most research focuses on repairing static scenarios and specific network node failures, while research on dynamic environments and large-scale drone node failures is limited. Therefore, a method for topology repair in a dynamic environment for drone swarm networks after large-scale node failures is needed to improve the communication efficiency and fault tolerance of drone swarm networks. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide an unmanned cluster topology repair method in a highly dynamic environment, which can repair the topological connectivity of the drone cluster network in the event of large-scale node failure and optimize the network topology structure.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A method for repairing the topology of a drone cluster network includes the following steps:

[0008] S1: Load network parameters;

[0009] S2: Establish communication links between drone nodes and build normal flight and damaged topology models of the cluster network;

[0010] S3: Normal flight phase: Calculate the virtual force of the drone cluster during normal flight based on the predefined path and update the position information;

[0011] S4: Restoring network 1-connectivity: Based on the neighborhood information of the surviving nodes, multiple virtual forces are calculated to repair the network topology and restore network connectivity; where network 1-connectivity means that the network connectivity is 1;

[0012] S5: Optimize network coverage: Optimize network topology coverage performance based on global network node location information;

[0013] According to the constructed network topology structure, with the goal of repairing the damaged network connectivity and optimizing the network coverage, a UAV network topology repair method based on improved multiple virtual forces is adopted to update the moving direction and speed of the UAV nodes.

[0014] Furthermore, in said S1, the network parameters loaded include communication range, sensing range and UAV generation position information, the UAV predefined trajectory, and the initial generation of the UAV obeys the Poisson disk distribution.

[0015] Furthermore, in S2, a topology model of a normal flight and a damaged network is constructed, and the specific steps are as follows:

[0016] For a UAV cluster V consisting of N UAVs in free space, N}, N is the number of clustered drones, and two drone nodes u i and u j The distance between them is d ij , when d ij Less than the communication distance R c When there is a bidirectional link between two nodes, the entire network is abstracted as a set of communication relationships between the node set V and the drones E = {(u i ,u j )|u i ,u j ∈V,d ij ≤R c} is composed of a network G=(V,E); when the network nodes of the drone cluster are damaged by a large-scale attack during the movement, the network topology structure is divided into G'={G i =(V i ,E i )|i=1,2,...,m}, including m connected sub-regions.

[0017] Furthermore, in said S3, the position information of the drone node in the normal flight phase is updated, and the drone node u i Along the predefined path P i ={p i,n =(x i,n ,y i,n )|n=1,2,...,k} move forward to the target area, k is the number of waypoints in the predefined path; node u i The current position is p i,n =(x i,n ,y i,n ), the next moment position is p i,n+1 =(x i,n+1 ,y i,n+1 ), and its guiding force expression is:

[0018]

[0019] k g is the guiding force adjustment coefficient, The unit vector from the current waypoint to the next waypoint, and the net force acting on the node Calculate the node speed:

[0020]

[0021] Furthermore, in S4, the type and size of the virtual force of the surviving nodes in the network during the network damage phase are designed, specifically including:

[0022] S41: Setting the guiding force: The guiding force setting is consistent with the setting in S3, and the surviving nodes move forward along the predefined path;

[0023] S42: Set aggregation force: The node whose neighboring node is attacked is used as the repair node to complete the repair process. The node determines the direction of the initial aggregation force by the failure of other nodes in its neighborhood. The node receives the location information from the neighbor and the one-hop neighbor list maintained by the neighbor to update its own node neighborhood list, including the node's one-hop neighbor list Γ i ={(id j ,pos j )|d ij <R c} and the corresponding two-hop neighbor list I i ={(id j ,pos j ,Γ j )|d ij <R c}; When the network fails, the node will calculate the aggregate position set U based on the particle swarm-simulated annealing algorithm uc ={o1,o2,...,o l}, l is the number of node aggregation positions, and the node aggregation force is defined as:

[0024]

[0025] k c is the cohesive force adjustment coefficient, The node disconnected location set is U uc ={o1,o2,...,o l}'s position vector;

[0026] S43: Set topological forces: To ensure the overall topology is roughly stable, set topological forces between neighbors to maintain the connection between neighbors, including attractive forces to prevent disconnection and repulsive forces to prevent nodes from being too close. The specific definitions are:

[0027]

[0028] is attractiveness, defined as:

[0029]

[0030] is the repulsive force, defined as:

[0031]

[0032] k aand k r is the topological force adjustment coefficient, α∈(0,1], αR c is the ideal distance to exert attraction on the node, β∈(0,1], βR d is the ideal distance at which a repulsive force is expected to be applied to the node;

[0033] S44: Set isolated node boundary force: An isolated node is a severely damaged node in the neighborhood. It constitutes a connected branch alone and is not affected by the neighborhood topological force. A boundary force is applied to it through the region boundary to control its deviation from the cluster. The region boundary is defined as a rectangular area centered on the predefined path of the node at the next moment. The specific definition is:

[0034]

[0035] k b is the boundary force adjustment coefficient;

[0036] S45: Set the resultant force: Restore the resultant force on the node in the 1-connection stage to:

[0037]

[0038] Set k according to the neighbor survival rate of the node c , defines the neighbor survival rate of a node is the number of real-time neighbors of the node, is the number of neighbors of the node at the initial deployment, and the aggregation coefficient is defined as:

[0039]

[0040] Furthermore, in S5, the network coverage is optimized, and the coverage redundancy or void of the network is calculated by triangulation based on the particle swarm-simulated annealing algorithm, and the coverage repair force is applied to optimize the network. The coverage repair force applied to the node is specifically defined as:

[0041]

[0042] R oj is the radius of the empty circle of the triangle where the node is located after the network triangulation, is the position vector of the center of the empty circle;

[0043] The resultant force on the node in stage S5 is:

[0044] The beneficial effects of the present invention are:

[0045] (1) The present invention proposes a topology repair solution for the situation where large-scale node failures in drone cluster networks cause serious network damage in dynamic scenarios, thereby improving the stability and reliability of drone cluster networks in complex environments.

[0046] (2) The present invention is based on an improved multiple virtual force algorithm to enable UAV nodes to distribute the calculation of movement direction and speed, reduce the topology repair time and node movement cost, and impose restriction force on isolated nodes to prevent them from leaving the cluster, thereby maintaining the communication performance of the entire UAV cluster network.

[0047] (3) While repairing the network connectivity, the present invention improves the coverage performance of the damaged network by introducing coverage repair capabilities, which helps to optimize the network topology in many aspects.

[0048] Other advantages, objects, and features of the present invention will be described in part in the following description and, in part, will be apparent to those skilled in the art upon examination of the following description or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below with reference to the accompanying drawings, in which:

[0050] Figure 1 This is a diagram of the implementation steps of the method for repairing the topology of a drone cluster in a highly dynamic environment according to the present invention;

[0051] Figure 2 Schematic diagram of the topological structure of the network in normal and damaged stages of the present invention;

[0052] Figure 3 This is a flow chart of a triangulation method based on an improved particle swarm-simulated annealing algorithm. DETAILED DESCRIPTION

[0053] The following describes the embodiments of the present invention by means of specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0054] Among them, the accompanying drawings are only for illustrative purposes and represent only schematic diagrams rather than actual pictures, and should not be understood as limiting the present invention. In order to better illustrate the embodiments of the present invention, some parts of the accompanying drawings may be omitted, enlarged or reduced, and do not represent the dimensions of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions may be omitted in the accompanying drawings.

[0055] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if there are terms such as "upper", "lower", "left", "right", "front", "back", etc. indicating directions or positional relationships, they are based on the directions or positional relationships shown in the drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operate in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0056] Figure 1 This is a diagram of the implementation steps of the method for repairing the topology of a drone cluster in a highly dynamic environment of the present invention. The specific steps include:

[0057] S1: Load network parameters;

[0058] S2: Establish communication links between drone nodes and build normal flight and damaged topology models of the cluster network;

[0059] S3: Normal flight phase: Calculate the virtual force of the drone cluster during normal flight based on the predefined path and update the position information;

[0060] S4: Restoring network 1-connectivity phase: Based on the neighborhood information of surviving nodes, multiple virtual forces are calculated to repair the network topology and restore network connectivity;

[0061] S5: Optimize network coverage: Optimize network topology coverage performance based on global network node location information;

[0062] Furthermore, S1 loads the network parameters, including the communication range, sensing range, UAV generation location information, UAV predefined trajectory, and the initial generation of UAVs follows the Poisson disk distribution;

[0063] further, Figure 2 To instantiate the normal flight and damaged topology of the S2 network, the specific construction steps are as follows:

[0064] For a UAV cluster V consisting of 60 UAVs in free space, 60}, assuming that two drone nodes ui and u j The distance between them is d ij , when d ij Less than the communication distance R c When there is a bidirectional link between two nodes, the entire network is abstracted as a set of communication relationships between the node set V and the drones E = {(u i ,u j )|u i ,u j ∈V,d ij ≤R c} is composed of a network G=(V,E). When the network nodes of the drone cluster are damaged by a large-scale attack during the movement, the network topology is divided into G'={G i =(V i ,E i )|i=1,2,3,4}, including 4 connected sub-regions.

[0065] Furthermore, S3 updates the position information of the drone node during the normal flight phase. i Along the predefined path P i ={p i,n =(x i,n ,y i,n )|n=1,2,...,k} to the target area, where k is the number of waypoints in the predefined path. i The current position is p i,n =(x i,n ,y i,n ), the next moment position is p i,n+1 =(x i,n+1 ,y i,n+1 ), and its guiding force expression is:

[0066]

[0067] k g is the guiding force adjustment coefficient, The unit vector from the current waypoint to the next waypoint, and the net force acting on the node Calculate the node speed:

[0068]

[0069] Furthermore, S4 designs the types and sizes of virtual forces of surviving nodes in the network during the network damage phase. The specific contents include:

[0070] S41: Setting the guiding force: The setting of the guiding force is consistent with that in S3, and the surviving nodes still move forward along the predefined path;

[0071] S42: Set cohesion: For a node whose neighboring nodes are attacked, it will serve as the main repair node to complete the repair process. Since the damage level of each node is different, the node needs to determine the direction of the initial cohesion based on the failure status of other nodes in its neighborhood. The node will periodically receive location information from its neighbors and the one-hop neighbor list maintained by its neighbors to update its own node neighborhood list, including the node's one-hop neighbor list Γ i ={(id j ,pos j )|d ij <R c} and the corresponding two-hop neighbor list I i ={(id j ,pos j ,Γ j )|d ij <R c When the network fails, the node will calculate the aggregate position set U based on the particle swarm-simulated annealing algorithm. uc ={o1,o2,...,o l}, l is the number of node aggregation positions, and the node aggregation force is defined as:

[0072]

[0073] k c is the cohesive force adjustment coefficient, The node disconnected location set is U uc ={o1,o2,...,o l} position vector.

[0074] Figure 3 This is a flowchart of the present invention for calculating the aggregated position set of surviving nodes based on triangulation of the particle swarm-simulated annealing algorithm. The specific steps are:

[0075] 1) Initialize particle swarm parameters: set the particle swarm size n and the maximum number of iterations T according to the number of drones in the cluster. max , set the particle initial value x i and speed v i , limiting the initial particle value to one side of the baseline is conducive to the particles quickly approaching the optimal solution, the initial value x i Initialization follows:

[0076]

[0077] e1 and e2 are the endpoints of a baseline that has been added to a triangle in the triangulation network, and v is the point that has formed a triangle with the baseline;

[0078] 2) Set the fitness function f(x i ), the specific expression is:

[0079] f(x i )=g(x i )+mp(x i )

[0080]

[0081] Point o is the baseline and particle x i The center of the outer circle of the triangle, points e1, e2, x i The coordinates of the circle are (x1, y1), (x2, y2), (x3, y3), and the coordinates of the center of the circle (x o ,y o )for:

[0082]

[0083] 3) Initialize the population individual extreme value pbest and the global extreme value gbest;

[0084] 4) Initialize the simulated annealing algorithm parameters: set the initial temperature T, temperature attenuation coefficient α, and initial energy state E = gbest;

[0085] 5) Generate new solutions w', c1', c2' for particle update parameters and calculate E'=max[f(x i ),i=1,2,...,n], let ΔE=E'-E;

[0086] 6) Determine the energy difference ΔE between the new state and the previous state, and accept the new solution with a certain probability p, which is defined as:

[0087]

[0088] 7) Update particle velocity v i and particle position x i , and add a random compensation factor during the update to further prevent the particle swarm from reaching local optimality;

[0089] 8) Update the individual extreme value pbest and the global extreme value gbest;

[0090] 9) If the maximum number of iterations is reached, the algorithm terminates, otherwise continue to execute 5)-9);

[0091] After triangulating the neighborhood of the surviving nodes, the nodes can calculate the aggregation position set of the surviving nodes by comparing the side length of the triangle with the communication range of the nodes, and then set the direction and size of the virtual aggregation force. The specific instantiation is as follows: let a triangle be ΔABC, A is the main node, B and C are nodes in its neighborhood, when AB and AC are smaller than the communication range R c When AC<Rc <AB,BC>R c When AC>R c >AB,BC>R c When AC>R c ,AB>R c When the cohesive force is directed toward the center of the circumcircle of the triangle (x o ,y o ). Let the set of disconnected locations generated by the node after splitting be U uc ={o1,o2,...,o k}.

[0092] S43: Set topological forces: To ensure the overall topology is roughly stable, set topological forces between neighbors to maintain the connection between neighbors, including attractive forces to prevent disconnection and repulsive forces to prevent nodes from being too close. The specific definitions are:

[0093]

[0094] is attractiveness, defined as:

[0095]

[0096] is the repulsive force, defined as:

[0097]

[0098] k a and k r is the topological force adjustment coefficient, α∈(0,1], αR c is the ideal distance to exert attraction on the node, β∈(0,1], βR d The ideal distance at which you want to exert a repulsive force on the node.

[0099] S44: Set isolated node boundary force: An isolated node is a severely damaged node in the neighborhood. It constitutes a connected branch alone and is not affected by the neighborhood topological force. A boundary force is applied to it through the region boundary to control its deviation from the cluster. The region boundary is defined as a rectangular area centered on the predefined path of the node at the next moment. The specific definition is:

[0100]

[0101] k b is the boundary force adjustment coefficient.

[0102] S45: Set the resultant force: Restore the resultant force on the node in the 1-connection stage to:

[0103]

[0104] Set k according to the neighbor survival rate of the node c , defines the neighbor survival rate of a node is the number of real-time neighbors of the node, is the number of neighbors of the node at the initial deployment, and the aggregation coefficient is defined as:

[0105] k c =w c (1-η i )

[0106]

[0107] Furthermore, S5 optimizes network coverage by Figure 3 The triangulation based on the particle swarm-simulated annealing algorithm shown in the figure calculates the coverage redundancy or void of the network and applies the coverage repair force to optimize the network. The coverage repair force on the node is specifically defined as:

[0108]

[0109] R oj is the radius of the empty circle of the triangle where the node is located after the network triangulation, is the position vector of the center of the empty circle.

[0110] The resultant force on node S5 is:

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solutions, which should all be included in the scope of the claims of the present invention.

Claims

1. An unmanned cluster topology repair method in a highly dynamic environment, characterized by: The method comprises the following steps: S1: Load network parameters; S2: Establish communication links between drone nodes and build normal flight and damaged topology models of the cluster network; S3: Normal flight phase: Calculate the virtual force of the drone cluster during normal flight based on the predefined path and update the position information; S4: Restoring network 1-connectivity: Based on the neighborhood information of the surviving nodes, multiple virtual forces are calculated to repair the network topology and restore network connectivity; where network 1-connectivity means that the network connectivity is 1; S5: Optimize network coverage: Optimize network topology coverage performance based on global network node location information; According to the constructed network topology structure, with the goal of repairing the damaged network connectivity and optimizing the network coverage, a UAV network topology repair method based on improved multiple virtual forces is adopted to update the moving direction and speed of the UAV nodes.

2. The unmanned cluster topology repair method in a highly dynamic environment according to claim 1, characterized in that: In S1, network parameters are loaded including communication range, sensing range and UAV generation location information, UAV predefined trajectory, and UAV initial generation obeys Poisson disk distribution.

3. The unmanned cluster topology repair method in a highly dynamic environment according to claim 2, characterized in that: In S2, a topology model of a normal and damaged network is constructed. The specific steps are: For a UAV cluster V consisting of N UAVs in free space, N }, N is the number of clustered drones, and two drone nodes u i and u j The distance between them is d ij , when d ij Less than the communication distance R c When there is a bidirectional link between two nodes, the entire network is abstracted as a set of communication relationships between the node set V and the drones E = {(u i ,u j )|u i ,u j ∈V,d ij ≤R c } is composed of a network G=(V,E); when the network nodes of the drone cluster are damaged by a large-scale attack during the movement, the network topology structure is divided into G'={G i =(V i ,E i )|i=1,2,...,m}, including m connected sub-regions.

4. The unmanned cluster topology repair method in a highly dynamic environment according to claim 3, characterized in that: In said S3, the position information of the drone node in the normal flight phase is updated, and the drone node u i Along the predefined path P i ={p i,n =(x i,n ,y i,n )|n=1,2,...,k} move forward to the target area, k is the number of waypoints in the predefined path; node u i The current position is p i,n =(x i,n ,y i,n ), the next moment position is p i,n+1 =(x i,n+1 ,y i,n+1 ), and its guiding force expression is: k g is the guiding force adjustment coefficient, The unit vector from the current waypoint to the next waypoint, and the net force acting on the node Calculate the node speed:

5. The unmanned cluster topology repair method in a highly dynamic environment according to claim 4, characterized in that: In S4, the type and size of the virtual force of the surviving nodes in the network during the network damage phase are designed, specifically including: S41: Setting the guiding force: The guiding force setting is consistent with the setting in S3, and the surviving nodes move forward along the predefined path; S42: Set aggregation force: The node whose neighboring node is attacked is used as the repair node to complete the repair process. The node determines the direction of the initial aggregation force by the failure of other nodes in its neighborhood. The node receives the location information from the neighbor and the one-hop neighbor list maintained by the neighbor to update its own node neighborhood list, including the node's one-hop neighbor list Γ i ={(id j ,pos j )|d ij <R c } and the corresponding two-hop neighbor list I i ={(id j ,pos j ,Γ j )|d ij <R c }; When the network fails, the node will calculate the aggregate position set U based on the particle swarm-simulated annealing algorithm uc ={o1,o2,...,o l }, l is the number of node aggregation positions, and the node aggregation force is defined as: k c is the cohesive force adjustment coefficient, The node disconnected location set is U uc ={o1,o2,...,o l }'s position vector; S43: Set topological forces: To ensure the overall topology is roughly stable, set topological forces between neighbors to maintain the connection between neighbors, including attractive forces to prevent disconnection and repulsive forces to prevent nodes from being too close. The specific definitions are: is attractiveness, defined as: is the repulsive force, defined as: k a and k r is the topological force adjustment coefficient, α∈(0,1], αR c is the ideal distance to exert attraction on the node, β∈(0,1], βR d is the ideal distance at which a repulsive force is expected to be applied to the node; S44: Set isolated node boundary force: An isolated node is a severely damaged node in the neighborhood. It constitutes a connected branch alone and is not affected by the neighborhood topological force. A boundary force is applied to it through the region boundary to control its deviation from the cluster. The region boundary is defined as a rectangular area centered on the predefined path of the node at the next moment. The specific definition is: k b is the boundary force adjustment coefficient; S45: Set the resultant force: Restore the resultant force on the node in the 1-connection stage to: Set k according to the neighbor survival rate of the node c , defines the neighbor survival rate of a node is the number of real-time neighbors of the node, is the number of neighbors of the node at the initial deployment, and the aggregation coefficient is defined as: k c =w c (1st) i ) 6. The unmanned cluster topology repair method in a highly dynamic environment according to claim 1, characterized in that: In S5, the network coverage is optimized by calculating the coverage redundancy or voids of the network through triangulation based on the particle swarm-simulated annealing algorithm, and applying coverage repair force to optimize the network. The coverage repair force applied to the node is specifically defined as: R oj is the radius of the empty circle of the triangle where the node is located after the network triangulation, is the position vector of the center of the empty circle; The resultant force on the node in stage S5 is:

Images