Multi-domain unmanned cluster toughness improvement method based on block segmentation under multi-source disturbance

Abstract

The invention belongs to the field of unmanned clusters, system engineering and toughness regulation and control, and provides a multi-domain unmanned cluster toughness improvement method based on block segmentation under multi-source disturbance. Comprising the steps of multi-domain unmanned cluster multi-source disturbance failure model construction, unmanned cluster multi-layer coloring network model construction, block segmentation and node state judgment, block deconstruction strategy execution, block reconstruction strategy execution and block switching strategy execution. Through unmanned cluster topological transformation under the block deconstruction strategy, the block reconstruction strategy and the block switching strategy, the network topological structure is improved, and the cluster toughness is improved.

Description

Technical Field

[0001] The present invention relates to the fields of unmanned clusters, system engineering and resilience regulation, and in particular to a method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances. Background Art

[0002] With the deep integration of information technology, modern mission space is rapidly evolving toward full-domain, multi-dimensional, and multi-functional capabilities. In this context, resilience is no longer a single performance metric for unmanned swarms, but rather a systematic regulatory capability for the swarm to cope with complex internal and external disturbances. Multi-domain unmanned swarm systems exhibit distinct characteristics: node heterogeneity, cross-domain collaboration, dynamic missions, and network coupling. Whether due to internal issues such as node failures and communication interruptions, or external challenges such as mission changes, environmental changes, and enemy attacks, localized disturbances can propagate through both intra-domain and inter-domain pathways, leading to task chain disruptions, network collapse, and system structure disintegration, ultimately resulting in swarm performance degradation or even mission failure. Therefore, strengthening the research and development of technologies to enhance the resilience of multi-domain unmanned swarms and improve their anti-interference and rapid recovery capabilities is of great strategic significance for ensuring the stable and efficient operation of swarms in complex and changing environments.

[0003] Existing research has achieved real-time dynamic reconstruction of clusters after failures, but different reconstruction strategies can only be applied in specific scenarios. Research on scalable, general, and efficient dynamic reconstruction strategies is still lacking. Reconstruction strategies designed based on deep reinforcement learning methods, which do not require prior data, are an important approach for future research on strategies to improve the resilience of unmanned clusters. However, this approach is complex and difficult to apply. Summary of the Invention

[0004] In order to overcome the shortcomings of the existing technology, the purpose of the present invention is to provide a multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbances, so as to solve the problem that multi-domain unmanned clusters are prone to collapse and difficult to recover under strong disturbances.

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

[0006] A method for improving the resilience of multi-domain unmanned swarms under multi-source disturbances based on block segmentation, comprising:

[0007] Construct a multi-domain unmanned swarm multi-source disturbance failure model;

[0008] Construct an unmanned cluster multi-layer coloring network model;

[0009] Dividing the unmanned cluster multi-layer colored network model into a plurality of cluster sub-blocks, and using the multi-domain unmanned cluster multi-source disturbance failure model to determine the node status of each cluster sub-block;

[0010] When the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, performing an unmanned cluster topology transformation using a block deconstruction strategy;

[0011] When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, performing an unmanned cluster topology transformation using a block reconstruction strategy;

[0012] When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, an unmanned cluster topology transformation is performed using a block switching strategy.

[0013] Preferably, a multi-domain unmanned swarm multi-source disturbance failure model is constructed, including:

[0014] Set the fault parameters of all nodes to obtain the fault parameter set;

[0015] Set the repair parameters of all nodes to obtain the repair parameter set;

[0016] Set the degree values ​​of all nodes to obtain the node degree set;

[0017] Set the task set, task sequence set and task type set to obtain the task layer;

[0018] Set the coordinate parameter set, radius parameter set and occurrence probability parameter set to obtain the collapse circle;

[0019] The fault parameter set, the repair parameter set, the node degree set, the task layer, and the collapse circle are integrated to obtain the multi-domain unmanned cluster multi-source disturbance failure model; the expression of the multi-domain unmanned cluster multi-source disturbance failure model includes:

[0020]

[0021]

[0022] G m =(τ,ρ,θ),

[0023] O = {x, y, r, p};

[0024] Where τ={τ h |h=1,2,...,δ}; θ={θ s .θ w}; λ represents the fault parameter set; Represents nodes respectively Fault parameter; μ represents the repair parameter set; Representation node The repair parameter; k represents the node degree set; Representation node The degree value of G m represents the task layer; τ represents the task set; ρ represents the task sequence set; ρ i represents the task sequence set of node i; θ represents the task type set; O represents the collapse circle; {x, y, r, p} represents the set of coordinates, radius, and occurrence probability parameters of the collapse circle; τ h represents the hth task; δ represents the total number of tasks; represents the set of task sequences for node i; represents the task sequence set of node j; ε and γ represent the total number of sensing nodes and attack nodes respectively; s represents the node type is sensing class; w represents the node type is attack class; represents the Zth task of node i; represents the i-th sensor node; represents the jth decision node; Indicates the mth attack node.

[0025] Preferably, constructing an unmanned cluster multi-layer coloring network model includes:

[0026] Set the perception node set, the decision node set and the strike decision set to obtain the network node set;

[0027] Set the hyperedge set, edge weight set, node color vector set, and edge color vector level set;

[0028] The network node set, the hyperedge set, the edge weights, the node color vector set, and the edge color vector level set are integrated to obtain the unmanned cluster multi-layer colored network model; the expression of the unmanned cluster multi-layer colored network model is:

[0029]

[0030] Where V = {V S ,V D ,V W}; Γ represents the unmanned cluster multi-layer colored network model; V represents the network node set; U represents the hyperedge set; represents the node weight set; represents the edge weight set; κ V represents the node color vector set; κ URepresents the edge color vector level set; V S represents the set of sensing nodes; V D represents the decision node set; V W represents the strike decision set; represents the xth hyperedge of node i; u represents the number of hyperedges connected to node i; represents the color of the jth hyperedge of node i; k represents the total number of colors; represents the color weight of the e-th hyperedge of node i; l represents the total number of hyperedges with color weights connected to node i.

[0031] Preferably, when the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, performing an unmanned cluster topology transformation using a block deconstruction strategy includes:

[0032] Determine the faulty node and remove the edges connected to the faulty node;

[0033] When the faulty node is not the only node of the same type in the cluster sub-block, remove the faulty node and connect the nodes connected to the faulty node to nodes of the same type as the faulty node;

[0034] After the connection is completed, the cluster sub-block is updated.

[0035] Preferably, when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, performing an unmanned cluster topology transformation using a block reconstruction strategy includes:

[0036] Determine the faulty node and remove the edges connected to the faulty node;

[0037] When the faulty node is the only node of the same type in the cluster sub-block, determining whether there is a node of the same type as the faulty node in an adjacent block of the cluster sub-block, and obtaining a type determination status;

[0038] When the priority of the task of the cluster sub-block is greater than that of the adjacent block, executing the block switching strategy;

[0039] When the type judgment state indicates that there are nodes of the same type, a node of the same type as the faulty node in the adjacent block is determined as a node to be replaced; if the node to be replaced does not belong to the faulty node, the node to be replaced is removed and the faulty node is replaced by the node to be replaced;

[0040] The cluster sub-block is merged into the adjacent block, and the cluster sub-block is updated.

[0041] Preferably, when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, performing an unmanned cluster topology transformation using a block switching strategy includes:

[0042] Determine the faulty node and remove the edges connected to the faulty node;

[0043] When the block deconstruction strategy and the block reconstruction strategy cannot be executed, if the cluster sub-block is the sub-block with the highest mission value, then the cluster sub-block is switched to a sub-block with complete node elements in the unmanned cluster multi-layer colored network model; the sub-block with the highest mission value is the cluster sub-block that ranks first in descending order of mission value; the completeness of the node elements indicates that the kill chain from reconnaissance to decision-making to strike is closed;

[0044] The cluster sub-block is updated.

[0045] Preferably, a multi-domain unmanned cluster resilience enhancement system based on block segmentation under multi-source disturbances includes:

[0046] Failure model construction module, used to build multi-domain unmanned swarm multi-source disturbance failure models;

[0047] Network model building module, used to build unmanned cluster multi-layer coloring network model;

[0048] a sub-block division module, configured to divide the unmanned cluster multi-layer colored network model into a plurality of cluster sub-blocks, and determine the node status of each cluster sub-block using the multi-domain unmanned cluster multi-source disturbance failure model;

[0049] a block deconstruction module, configured to perform an unmanned cluster topology transformation using a block deconstruction strategy when the node status indicates that a faulty node exists in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block;

[0050] a block reconstruction module, configured to perform an unmanned cluster topology transformation using a block reconstruction strategy when the node status indicates that the faulty node exists in the cluster sub-block, the faulty node is the only node of its type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero;

[0051] The block switching module is configured to perform an unmanned cluster topology transformation using a block switching strategy when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero.

[0052] Preferably, an electronic device comprises: at least one processor, and a memory communicatively connected to the processor; wherein the memory stores instructions that can be executed by the processor, and the instructions are executed by the processor so that the processor can execute the aforementioned multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbance.

[0053] Preferably, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable a computer to execute the aforementioned multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbances.

[0054] The present invention discloses the following technical effects:

[0055] The present invention provides a method for improving the resilience of multi-domain unmanned clusters under multi-source disturbances based on block segmentation. Through the unmanned cluster topology transformation under block deconstruction strategy, block reconstruction strategy and block switching strategy, the problem that multi-domain unmanned clusters are prone to collapse and difficult to recover under strong disturbances is solved, and the network topology structure is improved and the cluster resilience is enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. 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 any creative work.

[0057] Figure 1 A schematic diagram of a multi-domain unmanned cluster resilience improvement process based on block segmentation under multi-source disturbances provided by an embodiment of the present invention;

[0058] Figure 2 A multi-domain unmanned cluster colored directed graph provided by an embodiment of the present invention;

[0059] Figure 3 The unmanned cluster network topology transformation diagram based on block segmentation provided by an embodiment of the present invention;

[0060] Figure 4 A graph showing the resilience change trend of an unmanned cluster under multi-source disturbances provided by an embodiment of the present invention;

[0061] Figure 5 A schematic diagram of the coordinates of an unmanned cluster provided by an embodiment of the present invention;

[0062] Figure 6 A schematic diagram of an unmanned cluster initial coupling network provided by an embodiment of the present invention;

[0063] Figure 7A schematic diagram comparing the resilience changes of a multi-domain unmanned cluster provided by an embodiment of the present invention;

[0064] Figure 8 A schematic diagram comparing resilience improvement indicators for multi-domain unmanned clusters provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0066] The purpose of the present invention is to provide a method for improving the resilience of a multi-domain unmanned cluster based on block segmentation under multi-source disturbances, so as to solve the problem that the multi-domain unmanned cluster is prone to collapse and difficult to recover under strong disturbances.

[0067] 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.

[0068] Figure 1 A schematic diagram of a multi-domain unmanned cluster resilience improvement process based on block segmentation under multi-source disturbances provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbances, comprising:

[0069] Construct a multi-domain unmanned swarm multi-source disturbance failure model;

[0070] Construct an unmanned cluster multi-layer coloring network model;

[0071] Dividing the unmanned cluster multi-layer colored network model into a plurality of cluster sub-blocks, and using the multi-domain unmanned cluster multi-source disturbance failure model to determine the node status of each cluster sub-block;

[0072] When the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, performing an unmanned cluster topology transformation using a block deconstruction strategy;

[0073] When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, performing an unmanned cluster topology transformation using a block reconstruction strategy;

[0074] When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, an unmanned cluster topology transformation is performed using a block switching strategy.

[0075] Specifically, a multi-domain unmanned swarm multi-source disturbance failure model is constructed, including:

[0076] Set the fault parameters of all nodes to obtain the fault parameter set;

[0077] Set the repair parameters of all nodes to obtain the repair parameter set;

[0078] Set the degree values ​​of all nodes to obtain the node degree set;

[0079] Set the task set, task sequence set and task type set to obtain the task layer;

[0080] Set the coordinate parameter set, radius parameter set and occurrence probability parameter set to obtain the collapse circle;

[0081] The fault parameter set, the repair parameter set, the node degree set, the task layer, and the collapse circle are integrated to obtain the multi-domain unmanned cluster multi-source disturbance failure model; the expression of the multi-domain unmanned cluster multi-source disturbance failure model includes:

[0082]

[0083] G m =(τ,ρ,θ),

[0084] O = {x, y, r, p};

[0085] Where τ={τ h |h=1,2,...,δ}; θ={θ s .θ w}; λ represents the fault parameter set; Represents nodes respectively Fault parameter; μ represents the repair parameter set; Representation node The repair parameter; k represents the node degree set; Representation node The degree value of G m represents the task layer; τ represents the task set; ρ represents the task sequence set; ρ irepresents the task sequence set of node i; θ represents the task type set; O represents the collapse circle; {x, y, r, p} represents the set of coordinates, radius, and occurrence probability parameters of the collapse circle; τ h represents the hth task; δ represents the total number of tasks; represents the set of task sequences for node i; represents the task sequence set of node j; ε and γ represent the total number of sensing nodes and attack nodes respectively; s represents the node type is sensing class; w represents the node type is attack class; represents the Zth task of node i; represents the i-th sensor node; represents the jth decision node; Indicates the mth attack node.

[0086] Preferably, constructing an unmanned cluster multi-layer coloring network model includes:

[0087] Set the perception node set, the decision node set and the strike decision set to obtain the network node set;

[0088] Set the hyperedge set, edge weight set, node color vector set, and edge color vector level set;

[0089] The network node set, the hyperedge set, the edge weights, the node color vector set, and the edge color vector level set are integrated to obtain the unmanned cluster multi-layer colored network model; the expression of the unmanned cluster multi-layer colored network model is:

[0090]

[0091] Where V = {V S ,V D ,V W}; Γ represents the unmanned cluster multi-layer colored network model; V represents the network node set; U represents the hyperedge set; represents the node weight set; represents the edge weight set; κ V represents the node color vector set; κ U Represents the edge color vector level set; V S represents the set of sensing nodes; V D represents the decision node set; V W represents the strike decision set; represents the xth hyperedge of node i; u represents the number of hyperedges connected to node i; represents the color of the jth hyperedge of node i; k represents the total number of colors; represents the color weight of the e-th hyperedge of node i; l represents the total number of hyperedges with color weights connected to node i.

[0092] Specifically, when the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, an unmanned cluster topology transformation is performed using a block deconstruction strategy, including:

[0093] Determine the faulty node and remove the edges connected to the faulty node;

[0094] When the faulty node is not the only node of the same type in the cluster sub-block, remove the faulty node and connect the nodes connected to the faulty node to nodes of the same type as the faulty node;

[0095] After the connection is completed, the cluster sub-block is updated.

[0096] Furthermore, when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, an unmanned cluster topology transformation is performed using a block reconstruction strategy, including:

[0097] Determine the faulty node and remove the edges connected to the faulty node;

[0098] When the faulty node is the only node of the same type in the cluster sub-block, determining whether there is a node of the same type as the faulty node in an adjacent block of the cluster sub-block, and obtaining a type determination status;

[0099] When the priority of the task of the cluster sub-block is greater than that of the adjacent block, executing the block switching strategy;

[0100] When the type judgment state indicates that there are nodes of the same type, a node of the same type as the faulty node in the adjacent block is determined as a node to be replaced; if the node to be replaced does not belong to the faulty node, the node to be replaced is removed and the faulty node is replaced by the node to be replaced;

[0101] The cluster sub-block is merged into the adjacent block, and the cluster sub-block is updated.

[0102] Furthermore, when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, an unmanned cluster topology transformation is performed using a block switching strategy, including:

[0103] Determine the faulty node and remove the edges connected to the faulty node;

[0104] When the block deconstruction strategy and the block reconstruction strategy cannot be executed, if the cluster sub-block is the sub-block with the highest mission value, then the cluster sub-block is switched to a sub-block with complete node elements in the unmanned cluster multi-layer colored network model; the sub-block with the highest mission value is the cluster sub-block that ranks first in descending order of mission value; the completeness of the node elements indicates that the kill chain from reconnaissance to decision-making to strike is closed;

[0105] The cluster sub-block is updated.

[0106] Specifically, the method for improving the resilience of multi-domain unmanned clusters under multi-source disturbances based on block segmentation includes the following steps:

[0107] Considering that local disturbances may cause risks to spread within and between layers, triggering chain reactions and resulting in the diversity, complexity, and uncertainty of failure modes, a multi-domain unmanned swarm multi-source disturbance failure model is constructed by analyzing the failure modes of unmanned swarms to provide parameter input for equipment status.

[0108] Considering that the kill network is a multi-layered complex network with networks within networks and heterogeneous networks, a single-layer task network cannot clearly depict the correlation between cross-domain elements. Therefore, a multi-layer colored network model of unmanned swarms is constructed, which includes multiple functional networks such as perception, decision-making, attack, and target layers. This provides a basic network for improving the resilience of unmanned swarms based on block segmentation.

[0109] Divide the multi-layer network into multiple cluster sub-blocks. When a faulty node exists within a block, it is replaced by a similar node within the block to maintain topological stability. The faulty node is directly removed from the block, achieving unmanned cluster topology transformation under the block deconstruction strategy.

[0110] When there is a faulty node in a block and the faulty node is the only node of the same type in the block, and block deconstruction cannot be performed, the unmanned cluster topology transformation under the block reconstruction strategy is achieved by replacing adjacent block nodes;

[0111] When there are no healthy nodes in a block or in adjacent blocks that can replace a failed node, the block task priority is sorted to match the blocks with high integrity to important tasks, thus achieving unmanned cluster topology transformation under the block switching strategy.

[0112] Considering the damage caused by multi-source disturbances to unmanned clusters, a method to improve the resilience of unmanned clusters based on block segmentation is implemented by realizing block deconstruction, block reconstruction and block switching.

[0113] Furthermore, the multi-domain unmanned cluster multi-source disturbance failure model specifically adopts the following formula:

[0114]

[0115] G m=(τ,ρ,θ)

[0116] O={x,y,r,p}

[0117] Preferably, the unmanned cluster multi-layer network model based on the colored graph specifically adopts the following formula and obtains Figure 2 The multi-layer shading network model shown:

[0118]

[0119] V={V S ,V D ,V W}

[0120]

[0121] Specifically, the unmanned cluster topology transformation under the block deconstruction strategy adopts the following steps and obtains the following: Figure 3 The topology transformation diagram shown:

[0122] Faulty node v i ∈Bτ i , delete v i Connected edges where Bτ i where represents the block cluster i, Indicates deleting v from the set of perception decision link and decision attack link super edge i The edge,

[0123] There is a node v′ of the same type in the block i ∈Bτ i , which is the premise of deconstruction; if it does not exist, the block reconstruction strategy is used.

[0124] Remove the failed node Its edges connect to nodes of the same type Represents nodes of the same type v′ i Connecting to the faulty node v i The nodes before and after the kill chain,

[0125] Update block cluster Bτ={Bτ i |i=1,2,3...,δ}.

[0126] Furthermore, the unmanned cluster topology transformation under the block reconstruction strategy specifically adopts the following steps:

[0127] Faulty node v i ∈Bτ i , delete v i Connected edges

[0128] The faulty node is the only type of node in the block, and the block deconstruction cannot be executed.

[0129] There are nodes v′ of the same type in adjacent blocks i ∈Bτ j If it does not exist, go to the priority comparison step.

[0130] Node v′ i Satisfy the attribute restrictions. If not, proceed to the priority comparison step.

[0131] Remove in adjacent blocks And add it to this block,

[0132] Block task τ i The priority of the task is less than that of the adjacent block τ j If it is greater than, the block switching strategy is used.

[0133] Block cluster Bτ i Integrate into adjacent blocks,

[0134] Update block cluster Bτ={Bτ i |i=1,2,3...,δ}.

[0135] Furthermore, the unmanned cluster topology transformation under the block switching strategy specifically adopts the following steps:

[0136] Faulty node v i ∈Bτ i , delete v i Connected edges

[0137] Block deconstruction and block reconstruction cannot be performed,

[0138] If Bτ i The corresponding task value is the highest, switch Bτ i Bτ is a node element complete j The cluster performs the task.

[0139] Update block cluster Bτ={Bτ i |i=1,2,3...,δ}.

[0140] Preferably, taking the maritime unmanned swarm countering low, small and slow targets in the air as an example, the correctness and superiority of the unmanned swarm resilience improvement method based on block segmentation is verified, including:

[0141] Considering that local disturbances may cause risks to spread within and between layers, triggering chain reactions and resulting in diverse, complex, and uncertain failure modes, a multi-domain, multi-source disturbance failure model for unmanned swarms is constructed by analyzing the failure modes of unmanned swarms to provide parameter input for equipment status. The specific implementation of this step is as follows:

[0142] Based on the failure mode of unmanned swarm, input failure rate and other parameters, the natural failure model, intentional attack model and collapse circle attack model of unmanned swarm are constructed, as shown in Table 1 and Figure 4 Analysis of the impact of multi-source disturbances and the temporal evolution of resilience are shown.

[0143] Table 1

[0144]

[0145] By controlling variables, single and multiple perturbation experiments were conducted, yielding resilience values ​​under multi-source perturbations, as shown in Table 1. The number of simulations was set to 5000. Perturbations 1 to 3 in the table correspond to self-perturbation failure, deliberate attack, and disintegration circle perturbation, respectively. Table 1 shows that among single perturbations, a single disintegration circle perturbation caused an average of four nodes to fail, and the resilience value dropped to 0.6144, significantly higher than other single perturbations. This indicates that the disintegration circle perturbation, by simultaneously affecting multiple node elements, increases its destructiveness to the unmanned cluster, significantly reducing resilience. Among multiple perturbations, the combined deliberate attack and disintegration circle perturbation caused an average of five nodes to fail, and the resilience value dropped to 0.5311. This indicates that deliberate attacks weakened the cluster's defense capabilities, while disintegration circle perturbations further amplified the destructive effects. Meanwhile, the combined self-perturbation failure, deliberate attack, and disintegration circle perturbation caused an average of six nodes to fail, and the resilience value dropped to 0.5378. This indicates that self-perturbation failure has a relatively small impact on system resilience, while deliberate attacks and disintegration circle perturbations remain the primary destructive factors. Combine Figure 4 Analyzing multidimensional resilience trends under multi-source perturbations reveals that multiple perturbations (combined deliberate attacks and the disruption circle) lead to greater node failures and a greater decrease in resilience than single perturbations. This suggests a synergistic effect between multiple perturbations, amplifying the destructive effects on the system. By establishing a multi-source perturbation model, we can more comprehensively assess system resilience and provide a scientific basis for system optimization.

[0146] Furthermore, considering that the kill network is a multi-layered complex network with networks within networks and heterogeneous networks, a single-layer task network cannot clearly depict the correlation between cross-domain elements. Therefore, a multi-layer colored network model of unmanned swarms is constructed, which includes multiple functional networks such as perception, decision-making, attack, and target layers. This provides a basic network for improving the resilience of unmanned swarms based on block segmentation. The specific implementation of this step is as follows:

[0147] The multi-domain unmanned swarm is abstracted as a network node consisting of 14 reconnaissance nodes, 8 command nodes, 30 attack nodes and 10 target nodes, such as Figure 5-6The properties are set as follows: detection intensity is {1, 2, 3, 4}, detection distance is 300, number of detections is {6, 7, 8}, communication distance is {9, 10, 11, 12}, maximum batch size is 20, attack intensity is {1, 2, 3}, attack distance is 100, detection difficulty is {1, 2, 3}, and attack difficulty is {1, 2, 3}. This results in a multi-domain unmanned cluster multi-layer colored evolution network.

[0148] The multi-source disturbance failure model provides input for the equipment status, which determines whether the equipment is faulty. If a fault occurs, the unmanned cluster topology transformation based on the block strategy needs to be enabled.

[0149] The unmanned cluster multi-layer colored network model is the basic network for topology transformation.

[0150] Specifically, the multi-layer network is divided into multiple cluster sub-blocks. If a faulty node exists within a block, it is replaced with a similar node within the block to maintain topological stability. The faulty node is then directly removed from the block, achieving unmanned cluster topology transformation under the block deconstruction strategy. This step is performed simultaneously with the unmanned cluster topology transformation method described below.

[0151] Furthermore, when there is a faulty node in the block and the faulty node is the only node of the same type in the block, and block deconstruction cannot be performed, the unmanned cluster topology transformation under the block reconstruction strategy is achieved by replacing the adjacent block nodes.

[0152] Furthermore, when there are no normal nodes in the block or in the adjacent blocks that can replace the faulty node, the blocks with high integrity are matched with important tasks through block task priority sorting, realizing the unmanned cluster topology transformation under the block switching strategy.

[0153] Preferably, considering the damage caused by multi-source disturbances to unmanned clusters, an unmanned cluster resilience improvement method based on block segmentation is implemented by implementing block deconstruction, block reconstruction and block switching.

[0154] Furthermore, by analyzing the average recovery time, code running time, recovery effect and other indicators, the effectiveness of the unmanned cluster resilience improvement method based on the block optimization strategy is verified. Considering multi-source disturbances and block strategy optimization, a comparative analysis of the changes in the resilience of multi-domain unmanned clusters is conducted, and the following results are obtained: Figure 7 The resilience improvement comparison chart is shown. In the block strategy optimization, the genetic algorithm parameters are set as follows: population size is 50, number of iterations is 100, and mutation probability is 0.05.

[0155] Specifically, Figure 7The rule-based reconstruction method for optimizing resilience is derived from relevant existing literature. The method is briefly described as taking into account the limited resources and adopting intra-cluster, inter-cluster and self-repair reconstruction strategies for faulty nodes to maintain task reliability. This embodiment reproduces the rule-based reconstruction method in cluster resilience evaluation and makes an objective comparative analysis. Figure 7 It can be seen that the average resilience improvement of the block strategy optimization method is 0.1210, and the average resilience improvement of the rule reconstruction method is 0.0928. The improvement of the block strategy optimization method is significantly higher than that of the rule reconstruction method. Figure 8 It can be seen that compared with the rule reconstruction method, the block strategy recovery effect is relatively improved by 30.39%. This shows that the block strategy optimization method has relative flexibility and adaptability, while the rule reconstruction method is limited by the shortcomings of fixed rules and global optimization, and the recovery effect is relatively poor. Figure 8 It can be seen that the mean time to recovery (MTTR) and running time of the block strategy optimization are both higher than those of the rule reconstruction method, but the maximum difference is only 0.0075 and 0.0076 seconds. Considering the significant advantages of the block strategy optimization method in improving resilience and recovery effects, this small difference can be ignored in practical applications.

[0156] The beneficial effects of the present invention are as follows:

[0157] The present invention improves the network topology structure and enhances cluster resilience through unmanned cluster topology transformation under block deconstruction strategy, block reconstruction strategy and block switching strategy.

[0158] 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.

[0159] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The above examples are only intended to help understand the method and core concept of the present invention. At the same time, those skilled in the art will find that the specific implementation methods and application scopes may vary based on the concept of the present invention. In summary, the contents of this specification should not be construed as limiting the present invention.

Claims

1. A method for improving the resilience of multi-domain unmanned clusters under multi-source disturbances based on block segmentation, characterized in that: include: Construct a multi-domain unmanned swarm multi-source disturbance failure model; Construct an unmanned cluster multi-layer coloring network model; Dividing the unmanned cluster multi-layer colored network model into a plurality of cluster sub-blocks, and using the multi-domain unmanned cluster multi-source disturbance failure model to determine the node status of each cluster sub-block; When the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, performing an unmanned cluster topology transformation using a block deconstruction strategy; When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, performing an unmanned cluster topology transformation using a block reconstruction strategy; When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, an unmanned cluster topology transformation is performed using a block switching strategy.

2. The method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances according to claim 1 is characterized in that: Construct a multi-domain unmanned swarm multi-source disturbance failure model, including: Set the fault parameters of all nodes to obtain the fault parameter set; Set the repair parameters of all nodes to obtain the repair parameter set; Set the degree values ​​of all nodes to obtain the node degree set; Set the task set, task sequence set and task type set to obtain the task layer; Set the coordinate parameter set, radius parameter set and occurrence probability parameter set to obtain the collapse circle; The fault parameter set, the repair parameter set, the node degree set, the task layer, and the collapse circle are integrated to obtain the multi-domain unmanned cluster multi-source disturbance failure model; the expression of the multi-domain unmanned cluster multi-source disturbance failure model includes: G m =(t,r,θ)、 O = {x, y, r, p}; Where τ={τ h |h=1,2,...,δ}; θ={θ s .θ w }; λ represents the fault parameter set; Represents nodes respectively Fault parameter; μ represents the repair parameter set; Representation node The repair parameter; k represents the node degree set; Representation node The degree value of G m represents the task layer; τ represents the task set; ρ represents the task sequence set; ρ i represents the task sequence set of node i; θ represents the task type set; O represents the collapse circle; {x, y, r, p} represents the set of coordinates, radius, and occurrence probability parameters of the collapse circle; τ h represents the hth task; δ represents the total number of tasks; represents the set of task sequences for node i; represents the task sequence set of node j; ε and γ represent the total number of sensing nodes and attack nodes respectively; s represents the node type is sensing class; w represents the node type is attack class; represents the Zth task of node i; represents the i-th sensor node; represents the jth decision node; Indicates the mth attack node.

3. The method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances according to claim 1 is characterized in that: Construct an unmanned cluster multi-layer coloring network model, including: Set the perception node set, the decision node set and the strike decision set to obtain the network node set; Set the hyperedge set, edge weight set, node color vector set, and edge color vector level set; The network node set, the hyperedge set, the edge weights, the node color vector set, and the edge color vector level set are integrated to obtain the unmanned cluster multi-layer colored network model; the expression of the unmanned cluster multi-layer colored network model is: Where V = {V S ,V D ,V W }; Γ represents the unmanned cluster multi-layer colored network model; V represents the network node set; U represents the hyperedge set; represents the node weight set; represents the edge weight set; κ V represents the node color vector set; κ U Represents the edge color vector level set; V S represents the set of sensing nodes; V D represents the decision node set; V W represents the strike decision set; represents the xth hyperedge of node i; u represents the number of hyperedges connected to node i; represents the color of the jth hyperedge of node i; k represents the total number of colors; represents the color weight of the e-th hyperedge of node i; l represents the total number of hyperedges with color weights connected to node i.

4. The method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances according to claim 1 is characterized in that: When the node status indicates that there is a faulty node in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block, performing an unmanned cluster topology transformation using a block deconstruction strategy includes: Determine the faulty node and remove the edges connected to the faulty node; When the faulty node is not the only node of the same type in the cluster sub-block, remove the faulty node and connect the nodes connected to the faulty node to nodes of the same type as the faulty node; After the connection is completed, the cluster sub-block is updated.

5. The method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances according to claim 1 is characterized in that: When the node status indicates that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, performing an unmanned cluster topology transformation using a block reconstruction strategy includes: Determine the faulty node and remove the edges connected to the faulty node; When the faulty node is the only node of the same type in the cluster sub-block, determining whether there is a node of the same type as the faulty node in an adjacent block of the cluster sub-block, and obtaining a type determination status; When the priority of the task of the cluster sub-block is greater than that of the adjacent block, executing the block switching strategy; When the type judgment state indicates that a node of the same type exists, a node of the same type as the faulty node in the adjacent block is determined as a node to be replaced; if the node to be replaced does not belong to the faulty node, the node to be replaced is removed and the faulty node is replaced by the node to be replaced; The cluster sub-block is merged into the adjacent block, and the cluster sub-block is updated.

6. The method for improving the resilience of multi-domain unmanned clusters based on block segmentation under multi-source disturbances according to claim 1 is characterized in that: When the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero, performing an unmanned cluster topology transformation using a block switching strategy includes: Determine the faulty node and remove the edges connected to the faulty node; When the block deconstruction strategy and the block reconstruction strategy cannot be executed, if the cluster sub-block is the sub-block with the highest mission value, then the cluster sub-block is switched to a sub-block with complete node elements in the unmanned cluster multi-layer colored network model; the sub-block with the highest mission value is the cluster sub-block that ranks first in descending order of mission value; the completeness of the node elements indicates that the kill chain from reconnaissance to decision-making to strike is closed; The cluster sub-block is updated.

7. A multi-domain unmanned cluster resilience enhancement system based on block segmentation under multi-source disturbances, characterized by: The method for improving the resilience of a multi-domain unmanned cluster based on block segmentation under multi-source disturbances according to claim 1 comprises: Failure model construction module, used to build multi-domain unmanned swarm multi-source disturbance failure models; Network model building module, used to build unmanned cluster multi-layer coloring network model; a sub-block division module, configured to divide the unmanned cluster multi-layer colored network model into a plurality of cluster sub-blocks, and determine the node status of each cluster sub-block using the multi-domain unmanned cluster multi-source disturbance failure model; a block deconstruction module, configured to perform an unmanned cluster topology transformation using a block deconstruction strategy when the node status indicates that a faulty node exists in the cluster sub-block and the faulty node is not the only node of the same type in the cluster sub-block; a block reconstruction module, configured to, when the node status is that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is not zero, perform an unmanned cluster topology transformation using a block reconstruction strategy; The block switching module is configured to perform an unmanned cluster topology transformation using a block switching strategy when the node status indicates that the faulty node exists in the cluster sub-block, the faulty node is the only node of the same type in the cluster sub-block, and the number of adjacent nodes of the faulty node is zero.

8. An electronic device, characterized in that: include: At least one processor, and a memory communicatively connected to the processor; wherein the memory stores instructions that can be executed by the processor, and the instructions are executed by the processor so that the processor can execute a multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbances as described in any one of claims 1 to 7.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to enable a computer to execute a multi-domain unmanned cluster resilience improvement method based on block segmentation under multi-source disturbances according to any one of claims 1 to 7.

Images