Unmanned aerial vehicle cluster configuration dynamic self-repairing optimal control method

Abstract

The invention discloses an unmanned aerial vehicle cluster configuration dynamic self-repairing optimal control method, which comprises the steps of selecting a local optimal unmanned aerial vehicle based on a distributed decision, generating a grade base of the local optimal unmanned aerial vehicle and fixing a set value, and determining unmanned aerial vehicle cluster grade distribution based on the grade base of the local optimal unmanned aerial vehicle; and constructing a repair unmanned aerial vehicle election mechanism, when an invalid unmanned aerial vehicle node appears, selecting all repair unmanned aerial vehicles based on the unmanned aerial vehicle cluster level distribution and the repair unmanned aerial vehicle election mechanism, generating an unmanned aerial vehicle repair path, and reconstructing a network topology structure of the unmanned aerial vehicle cluster based on the unmanned aerial vehicle repair path. And after reconstruction is completed, unmanned aerial vehicle cluster grade distribution is determined again. According to the method, network structure damage caused by node failure is effectively recovered, so that the inherent network topology structure and cluster configuration of the unmanned aerial vehicle cluster are rapidly recovered, and an unmanned aerial vehicle cluster system has relatively good survivability and capability of continuously executing tasks in a confrontation environment.

Description

Technical Field

[0001] The present invention belongs to the technical field of fault-tolerant control of unmanned aerial vehicle clusters, and in particular relates to a dynamic self-repair optimal control method for a unmanned aerial vehicle cluster configuration. Background Art

[0002] At present, the field of UAV self-repair algorithms is in the initial stage of research, and there are few systematic research results. It can be mainly divided into three typical categories: direct self-repair method, density-based self-repair method, and recursive self-repair method.

[0003] Direct self-repair methods completely rely on pre-defined self-repair rules, which determine the actions that should be taken after a failed individual is found. A common self-repair method is to select a backup individual to replace the failed individual and re-establish the network topology to ensure the connectivity of the network. However, such methods usually require a global or high-level planner to guide the self-repair process, or require each individual to expand its communication range during the self-repair process. Therefore, there are still some challenges and difficulties in implementing such methods in fully distributed applications. For example, the global planner may require a lot of computing and communication resources to effectively guide the self-repair process, while each individual expanding the communication range may increase energy consumption and communication burden.

[0004] In density-based self-repair methods, the density distribution of clusters is used as an important evaluation indicator, and the self-repair task is mainly achieved through density diffusion and rebalancing processes. Inspired by the contained gas model, Cheng et al. proposed a distributed self-aggregation method to coordinate the dispersion of clusters within a predetermined specific shape, and further developed a self-repair algorithm to solve the problem of individual death and displacement by maintaining a constant pressure balance of the entire shape model throughout the task. Shen et al. proposed a hormone-inspired algorithm that uses the different distribution characteristics of hormone concentrations to complete self-organization and self-repair tasks. Derbakova et al. proposed a decentralized self-repair algorithm. By moving to a special gateway node and redispersing within a predefined range, individuals can restore the connection of the wireless mesh network. Density-based methods rely heavily on predefined reaction behaviors, making it difficult to mathematically predict control performance and ensure the convergence of the self-repair process. In addition, this type of individuals will be affected by a large number of topological changes and long-term node movements, and the number of individuals involved is large, so they are mainly deployed and implemented in certain static tasks.

[0005] In the recursive self-repair method, a series of robots are selected to recursively repair the network topology through cascade motion. Younis et al. achieved the self-repair task by commanding the neighbors of the damaged nodes to move to each other, thereby reconstructing the necessary connections. However, the proposed method suffers from large topological changes and a large number of moving roles during the connection repair process. Abbasi et al. proposed a recursive self-repair algorithm to restore the connection of a wireless sensor-individual network after individual failures by relocating some cascaded individual nodes. Subsequently, they further optimized the node movement and proposed an advanced self-repair algorithm with minimal topological changes. Liu et al. used gradients to estimate distances in distributed multi-agent networks and proposed a fully distributed gradient-based self-repair algorithm that uses local minimum robots to generate gradients and solves the self-repair problem through optimal repair paths. However, the definition of local minimum robots imposes strict requirements on the network topology and lacks universality. Summary of the invention

[0006] In order to solve the above technical problems, the present invention proposes a dynamic self-repair optimal control method for a UAV cluster configuration to solve the problems existing in the above-mentioned prior art.

[0007] To achieve the above object, the present invention provides a method for optimal control of a dynamic self-repairing configuration of a drone cluster, comprising:

[0008] Selecting a local optimal drone based on distributed decision making, generating a rank base of the local optimal drone, and determining a rank distribution of the drone cluster based on the rank base of the local optimal drone;

[0009] A repair drone election mechanism is constructed. When an invalid drone node appears, all repair drones are selected and drone repair paths are generated based on the drone cluster level distribution and the repair drone election mechanism. The network topology of the drone cluster is rebuilt based on the drone repair path. After the reconstruction is completed, the drone cluster level distribution is redetermined.

[0010] Optionally, the process of selecting the local optimal drone includes: determining the relationship between the number of communication links of each drone node and the number of communication links of the corresponding neighboring node, and if the drone node meets a preset condition, determining it as the local optimal drone; wherein the preset condition is as follows:

[0011]

[0012] Among them, U LoU is the local optimal UAV, and d represents the number of communication links of the UAV.

[0013] Optionally, the process of determining the distribution of drone cluster levels includes:

[0014] The level of the local optimal drone is generated and fixed, a drone level diffusion mechanism is constructed, and the level distribution of the drone cluster is determined based on the drone level diffusion mechanism and the level of the local optimal drone; wherein the drone level diffusion mechanism includes: each drone node receives the level base of the corresponding neighbor node, determines the level base of the drone node based on the level base and the level base increment of the neighbor node, and determines the level based on the level base and the number of communication links of the drone node, until all drones in the drone cluster are determined.

[0015] Optionally, the repair drone election mechanism includes a direct repair drone election mechanism for failed nodes and a subsequent repair drone election mechanism.

[0016] Optionally, the direct repair drone election mechanism includes:

[0017] All neighboring nodes of the failed node are taken as candidate drone nodes, and each candidate drone node initializes a negotiation package and formulates election rules; the candidate drone nodes send and receive each other's negotiation packages, and update the negotiation packages according to the information included in the negotiation packages until the contents of all negotiation packages are the same, and a direct repair drone of the failed node is obtained; wherein, the negotiation package includes a node number and a node level.

[0018] Optionally, the election rules include:

[0019] Each drone node compares its own negotiation package with the received negotiation package. If the node level of the drone node is greater than the received node level, the negotiation package of the drone node is updated based on the received negotiation package; if the node level of the drone node is equal to the received node level and the node number is greater than the received node number, the negotiation package of the drone node is updated based on the received negotiation package; otherwise, the negotiation package of the drone node is not updated.

[0020] Optionally, the subsequent repair of the drone election mechanism includes:

[0021] All neighboring nodes of the direct repair drone initialize a negotiation packet, and send and receive each other's negotiation packets; and update the negotiation packet according to the information included in the negotiation packet until the contents of all negotiation packets are the same, and the subsequent repair drone of the failed node is obtained.

[0022] Optionally, the method further includes: if the number of failed nodes is greater than one, determining whether the drone node has been selected as a repair drone or a subsequent repair drone; if it has been selected, not participating in the election when other failed nodes select a repair drone.

[0023] Compared with the prior art, the present invention has the following advantages and technical effects:

[0024] The present invention provides a method for dynamic self-repair optimal control of a drone cluster configuration. First, a local optimal drone is selected based on distributed decision-making, a level base of the local optimal drone is generated and a set value is fixed, and then, the level distribution of the drone cluster is determined based on the level base of the local optimal drone; thirdly, a repair drone election mechanism is constructed. When an invalid drone node appears, all repair drones are selected based on the drone cluster level distribution and the repair drone election mechanism and a drone repair path is generated. Finally, the network topology of the drone cluster is rebuilt based on the drone repair path, and the drone cluster level distribution is re-determined after the reconstruction is completed. The present invention effectively restores the network structure damage caused by node failure, and then quickly restores the inherent network topology and cluster configuration of the drone cluster, so that the drone cluster system has better anti-destruction and continuous mission execution capabilities in a confrontation environment, which has great research significance and application value for the practical application of drone clusters in the military field. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0026] Figure 1 The figure is a flow chart of a method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0027] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0028] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0029] Embodiment 1

[0030] like Figure 1 As shown, this embodiment provides a method for optimal control of a dynamic self-repair of a drone cluster configuration, including:

[0031] Selecting a local optimal drone based on distributed decision making, generating a rank base of the local optimal drone, and determining a rank distribution of the drone cluster based on the rank base of the local optimal drone;

[0032] Furthermore, the process of selecting the local optimal UAV includes: determining the relationship between the number of communication links of each UAV node and the number of communication links of the corresponding neighboring nodes, and if the UAV node meets the preset conditions, it is determined to be the local optimal UAV; wherein the preset conditions are as follows:

[0033]

[0034] Among them, U LoU is the local optimal UAV, and d represents the number of communication links of the UAV.

[0035] Specifically, each drone node U i By judging one's own i and its neighbors j ,j∈N(U i ), select the local optimal UAV in a distributed decision-making manner, generate a hierarchy from it, and use it as the final repair UAV for the failed UAV. The above method is used to determine the local optimal UAV and cleverly use the relationship between the UAV node and its neighbors to generate a series of local optimal UAVs in the outermost layer of the cluster. Compared with previous technologies, it has better universality and is suitable for UAV network topologies of various structures.

[0036] Furthermore, the process of determining the distribution of the drone cluster hierarchy includes:

[0037] The level of the local optimal drone is generated and fixed, a drone level diffusion mechanism is constructed, and the drone cluster level distribution is determined based on the drone level diffusion mechanism and the level of the local optimal drone; wherein the drone level diffusion mechanism includes: each drone node receives the level base of the corresponding neighbor node, determines the level base of the drone node based on the level base and the level base increment of the neighbor node, and determines the level based on the level base and the number of communication links of the drone node, until all drones in the drone cluster are determined.

[0038] Specifically, the level is used to describe the importance of the UAV node, which is determined by the number of communication links of the node and the level of the node in the cluster. i , whose level l i The calculation formula is as follows:

[0039] l i =d i Lr i

[0040] where d i It's U i Degree, Lr i It's U iThe level-root of the .

[0041] The rank basis is a quantity used to estimate the distance between two points in a distributed system, that is, the rank of the drone node. The rank is generated from the previously generated local optimal drone, U LoU Initialize Lr LoU = 1 and remain unchanged, then based on d LoU Calculate the rank. The remaining drone nodes of non-local optimal drones initialize their rank base to infinity. Then, the rank will be recursively transferred and diffused along the communication links between nodes. For each time t, except for the local optimal drone, all nodes will update their rank base values ​​according to the following mechanism:

[0042]

[0043] Where Δ is the level basis increment at each time t.

[0044] When a drone node other than the local optimal drone receives a rank base from its neighbors, it will update its own rank base based on the above update mechanism and recalculate its rank. Subsequently, it will send the updated rank base to its neighbors again. The whole process will continue recursively until the entire drone cluster has a stable rank distribution.

[0045] A repair drone election mechanism is constructed. When an invalid drone node appears, all repair drones are selected and drone repair paths are generated based on the drone cluster level distribution and the repair drone election mechanism. The network topology of the drone cluster is rebuilt based on the drone repair path. After the reconstruction is completed, the drone cluster level distribution is redetermined.

[0046] Furthermore, the repair drone election mechanism includes a direct repair drone election mechanism for failed nodes and a subsequent repair drone election mechanism.

[0047] Furthermore, the direct repair drone election mechanism includes:

[0048] All neighboring nodes of the failed node are taken as candidate drone nodes, and each candidate drone node initializes a negotiation package and formulates election rules; the candidate drone nodes send and receive each other's negotiation packages, and update the negotiation packages according to the information included in the negotiation packages until the contents of all negotiation packages are the same, and a direct repair drone of the failed node is obtained; wherein, the negotiation package includes a node number and a node level.

[0049] Furthermore, the election rules include:

[0050] Each drone node compares its own negotiation package with the received negotiation package. If the node level of the drone node is greater than the received node level, the negotiation package of the drone node is updated based on the received negotiation package; if the node level of the drone node is equal to the received node level and the node number is greater than the received node number, the negotiation package of the drone node is updated based on the received negotiation package; otherwise, the negotiation package of the drone node is not updated.

[0051] Specifically, first, for the failed node U f (U f ≠U LoU ). All its neighbors will be U i ∈N(U f )Initialize a negotiation packet in Secondly, U i To U j ∈N(U i )Send p i And receive from U j p j Then, U i Compare j According to the following rules, p i To update:

[0052] Rule 1: If but

[0053] Rule 2: If but

[0054] Rule 3: If p j Neither rule 1 nor rule 2 is satisfied, then p i Not updating

[0055] Finally, U i Send p again to all its neighbors i Repeat this process until all negotiation packets have the same value. This means that the best drone node U to repair the drone as the first step has the minimum level. b ∈N(U f ) was elected, in, and It represents the identity serial number of drone nodes i and j. The purpose of comparing and distinguishing them by the size of the number is to quickly select one as the repair drone when the levels of two drone nodes are the same, and this approach is more in line with actual engineering.

[0056] Furthermore, the subsequent repair of the drone election mechanism includes:

[0057] All neighboring nodes of the direct repair drone initialize a negotiation packet, and send and receive each other's negotiation packets; and update the negotiation packet according to the information included in the negotiation packet and the election rules until the contents of all negotiation packets are the same, and the subsequent repair drone of the failed node is obtained.

[0058] Specifically, select the first step to repair the drone After that, since the subsequent node communication link still exists, the drone can be repaired by step (n-1) Directly from Select the nth step to repair the drone First, U i Initialize and send only to. Secondly, Compare i and negotiate its own package according to the following rules To update:

[0059] Rule 4: If but

[0060] Rule 5: If but

[0061] Rule 6: If p i Neither rule 4 nor rule 5 is satisfied, then Not updating

[0062] Repeat this process until all p i are received and processed. This means that the best drone node with the minimum level is the best drone node for repairing the drone in step n. was chosen, The above process will continue recursively until the local optimal UAV is selected as the final repair UAV.

[0063] A series of selected repair drones are moved to the position of the previous repair drone in turn, re-establishing the topological connection, and then re-generating and diffusing the levels according to the new network topology to realize the closed loop of the algorithm.

[0064] Furthermore, the method also includes: if the number of failed nodes is greater than one, determining whether the drone node is selected as a repair drone or a subsequent repair drone, and if it has been selected, not participating in the election when other failed nodes select a repair drone.

[0065] After all repair drones are selected, each repair drone recursively moves to the position of the previous drone node according to the repair path, reestablishes the network topology connection with its respective neighbors, forms a communication link, and thus restores the original network topology.

[0066] After the network topology is repaired, the levels are generated and diffused again according to the previous steps, so that the entire drone cluster obtains a new level distribution after reconstruction, completing the closed loop of the repair process and being ready to deal with new drone node failures at any time.

[0067] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A method for optimal control of dynamic self-repair of UAV cluster configuration, characterized in that: The following steps are involved: Selecting a local optimal drone based on distributed decision making, generating a rank base of the local optimal drone, and determining a rank distribution of the drone cluster based on the rank base of the local optimal drone; A repair drone election mechanism is constructed. When an invalid drone node appears, all repair drones are selected and drone repair paths are generated based on the drone cluster level distribution and the repair drone election mechanism. The network topology of the drone cluster is rebuilt based on the drone repair path. After the reconstruction is completed, the drone cluster level distribution is redetermined.

2. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 1 is characterized in that: The process of selecting the local optimal UAV includes: determining the relationship between the number of communication links of each UAV node and the number of communication links of the corresponding neighboring nodes. If the UAV node meets the preset conditions, it is determined to be the local optimal UAV; wherein the preset conditions are as follows: Among them, U LoU is the local optimal UAV, and d represents the number of communication links of the UAV.

3. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 1 is characterized in that: The process of determining the distribution of drone cluster levels includes: The level of the local optimal drone is generated and fixed, a drone level diffusion mechanism is constructed, and the level distribution of the drone cluster is determined based on the drone level diffusion mechanism and the level of the local optimal drone; wherein the drone level diffusion mechanism includes: each drone node receives the level base of the corresponding neighbor node, determines the level base of the drone node based on the level base and the level base increment of the neighbor node, and determines the level based on the level base and the number of communication links of the drone node, until all drones in the drone cluster are determined.

4. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 1 is characterized in that: The repair drone election mechanism includes a direct repair drone election mechanism for failed nodes and a subsequent repair drone election mechanism.

5. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 4 is characterized in that: The direct repair drone election mechanism includes: All neighboring nodes of the failed node are taken as candidate drone nodes, and each candidate drone node initializes a negotiation package and formulates election rules; the candidate drone nodes send and receive each other's negotiation packages, and update the negotiation packages according to the information included in the negotiation packages until the contents of all negotiation packages are the same, and a direct repair drone of the failed node is obtained; wherein, the negotiation package includes a node number and a node level.

6. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 5 is characterized in that: The election rules include: Each drone node compares its own negotiation package with the received negotiation package. If the node level of the drone node is greater than the received node level, the negotiation package of the drone node is updated based on the received negotiation package; if the node level of the drone node is equal to the received node level and the node number is greater than the received node number, the negotiation package of the drone node is updated based on the received negotiation package; otherwise, the negotiation package of the drone node is not updated.

7. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 6 is characterized in that: The subsequent repair of the drone election mechanism includes: All neighboring nodes of the direct repair drone initialize a negotiation packet, and send and receive each other's negotiation packets; and update the negotiation packet according to the information included in the negotiation packet until the contents of all negotiation packets are the same, and the subsequent repair drone of the failed node is obtained.

8. The method for optimal control of the dynamic self-repair of the UAV cluster configuration according to claim 1 is characterized in that: The method further includes: if the number of failed nodes is greater than one, determining whether the drone node is selected as a repair drone or a subsequent repair drone, and if it has been selected, not participating in the election when other failed nodes select a repair drone.

Images