Task-oriented unmanned aerial vehicle cluster super network modeling and reliability evaluation method

Abstract

The invention discloses a task-oriented unmanned aerial vehicle cluster super-network modeling and reliability evaluation method, and belongs to the field of unmanned aerial vehicle task execution evaluation, and the method comprises the following steps: constructing a multi-layer super-network model of an unmanned aerial vehicle cluster based on the task execution characteristics and function requirements of the unmanned aerial vehicle cluster; based on a dynamic failure strategy and a reliability evaluation index, establishing an unmanned aerial vehicle cluster task reliability evaluation framework oriented to the multi-layer hypernetwork model; and performing reliability evaluation on the unmanned aerial vehicle cluster task based on the unmanned aerial vehicle cluster task reliability evaluation framework to obtain an evaluation result.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle (UAV) mission execution evaluation, and in particular relates to a task-oriented UAV cluster hypernetwork modeling and reliability evaluation method. Background Art

[0002] With the rapid development of technologies such as artificial intelligence and autonomous control, drones have been widely used in a variety of fields, including disaster relief, intelligent logistics, and agriculture. Compared to single drone operations, drone swarms, relying on multi-drone collaboration, intelligent scheduling, and task division, can effectively improve operational efficiency, cover wider areas, and enhance adaptability to complex environments, demonstrating great potential for the future. However, as the scale of drone swarms continues to expand, challenges such as task scheduling, information exchange, resource coordination, and reliability assurance are becoming increasingly prominent. During mission execution, drone swarms may be affected by factors such as individual drone failures, communication interference, and environmental changes, which can reduce mission reliability. Drone swarm modeling is the prerequisite and foundation for drone swarm research and is crucial for achieving efficient swarm coordination. Therefore, establishing scientific and effective drone swarm models that accurately describe the dynamic behavior of drone swarms and quantitatively evaluate their stability has become a key research direction in the current development of drone swarm technology. Summary of the Invention

[0003] In order to solve the above technical problems, the present invention proposes a task-oriented UAV cluster hypernetwork modeling and reliability assessment method to solve the problems existing in the above-mentioned prior art.

[0004] To achieve the above objectives, the present invention provides a task-oriented UAV cluster hypernetwork modeling and reliability assessment method, comprising:

[0005] A multi-layer hyper-network model of UAV swarm is constructed based on the mission execution characteristics and functional requirements of UAV swarm;

[0006] Based on dynamic failure strategies and reliability evaluation indicators, a UAV swarm mission reliability evaluation framework for the multi-layer hypernetwork model is established;

[0007] Based on the UAV swarm mission reliability assessment framework, the reliability of the UAV swarm mission is evaluated to obtain the evaluation results.

[0008] Optionally, the process of building a multi-layer hypernetwork model of a drone swarm includes:

[0009] Build a business network based on the mission execution relationship of drone clusters;

[0010] Construct a mission network based on the actual collaborative relationship between the payloads of the UAV cluster during the mission;

[0011] Build a communication network based on the communication connection relationship between aircraft within the UAV cluster;

[0012] Construct a resource network based on the equipment resource collection of aircraft in the UAV cluster;

[0013] A multi-layer super network model of the drone cluster is constructed based on the interactive relationship and mapping mechanism of the business network, task network, communication network and resource network.

[0014] Optionally, the process of constructing a multi-layer super-network model of the drone cluster based on the interactive relationship and mapping mechanism among the business network, the task network, the communication network, and the resource network includes:

[0015] Decompose the tasks performed by the drone cluster into several sub-services, each sub-service corresponds to a service node, and map the service nodes to several groups of task links to build a service-task relationship;

[0016] Connect the communication nodes of the aircraft in the UAV cluster to several mission payload nodes in the mission network to build a mission-communication relationship;

[0017] Mapping the communication node to resource network nodes in a plurality of resource networks to establish a communication-resource relationship;

[0018] A multi-layer super network model is constructed based on the business-task relationship, task-communication relationship and communication-resource relationship.

[0019] Optionally, the business network is a directed graph, wherein the nodes of the directed graph represent sub-businesses and the edges represent the execution order between the sub-businesses;

[0020] The task network is an unweighted directed network, wherein the nodes of the unweighted directed network corresponding to the task network represent task loads, and the edges represent the dependency relationships between task loads;

[0021] The communication network is a weighted undirected network, wherein the nodes of the weighted undirected network represent communication nodes, the edges represent communication links, and the weights represent the transmission efficiency of the communication links;

[0022] The resource network is an unweighted directed network, wherein the nodes of the unweighted directed network corresponding to the resource network represent resource nodes, and the edges represent logical dependency relationships between resource nodes.

[0023] Optionally, the dynamic failure strategy includes: a random attack strategy and a cascading failure strategy.

[0024] Optionally, the reliability evaluation indicators include average degree, average betweenness and vulnerability.

[0025] Optionally, the calculation expression of the average degree is:

[0026]

[0027] Where K represents the average degree in the hypernetwork, k i represents the node degree of any node i in the hypernetwork, and N represents the total number of nodes in the network.

[0028] Optionally, the calculation expression of the average betweenness is:

[0029]

[0030] Where δ represents the average betweenness in the hypernetwork, N represents the total number of nodes in the network, and δ i Indicates betweenness.

[0031] Optionally, the vulnerability calculation expression is:

[0032]

[0033] Where P represents vulnerability, k i represents the degree of node i, and α represents the weight parameter.

[0034] Compared with the prior art, the present invention has the following advantages and technical effects:

[0035] The present invention's task-oriented UAV cluster hypernetwork modeling and reliability assessment method comprehensively depicts the complex interactive relationships of UAV clusters during task execution by constructing a multi-layer hypernetwork model, accurately reflecting the cluster's dynamic characteristics and task requirements. This method combines dynamic failure strategies and reliability assessment indicators to establish a systematic evaluation framework that can quantitatively analyze the impact of key node failures on network topology and task reliability. Compared with existing technologies, the present invention not only improves the scientific nature of UAV cluster task planning and scheduling, but also enhances the cluster's anti-interference ability and stability in complex environments, providing strong technical support for the efficient collaborative operation of UAV clusters and significantly improving the reliability and success rate of task execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which constitute part of this application, are intended to provide a further understanding of this application. The exemplary embodiments and descriptions of this application are intended to explain this application and do not constitute an improper limitation on this application. In the accompanying drawings:

[0037] Figure 1 A schematic diagram of a task-oriented service network according to an embodiment of the present invention;

[0038] Figure 2 A schematic diagram of a task link according to an embodiment of the present invention;

[0039] Figure 3A schematic diagram of a resource network according to an embodiment of the present invention;

[0040] Figure 4 This is a schematic diagram of the failure and reconstruction of a drone cluster according to an embodiment of the present invention;

[0041] Figure 5 This is the UAV swarm forest rescue business flow of an embodiment of the present invention;

[0042] Figure 6 This is a hypernetwork model of a UAV cluster in the T4 phase of an embodiment of the present invention;

[0043] Figure 7 Parameter change trends at each stage of the embodiment of the present invention; where (a) is the number of nodes and edges, (b) is the average degree, (c) is the average betweenness, and (d) is the network vulnerability;

[0044] Figure 8 Figure 2 shows the parameter change trends at each moment in the T4 phase of the embodiment of the present invention; (a) represents the number of nodes and edges, (b) represents the average degree, (c) represents the average betweenness, and (d) represents the network vulnerability.

[0045] Figure 9 The figure is a task-oriented flowchart of an embodiment of the present invention. DETAILED DESCRIPTION

[0046] It should be noted that, in the absence of conflict, the embodiments and features of the embodiments in this 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.

[0047] 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.

[0048] Example 1

[0049] With the development of artificial intelligence and drone technology, drone swarms are increasingly being used in diverse missions. However, existing research has inadequately characterized the dynamic mission requirements and complex multi-layer interactions of drone swarms, as well as their reliability measurement. To address this issue, this study proposes a task-oriented, multi-level hypernetwork model and reliability assessment framework. First, heterogeneous subnetworks are constructed from four dimensions: business, mission, communication, and resources, characterizing the multi-layer interactions of the cluster. Second, considering the dependencies and cascading failures between multi-layer networks, a random failure and network reconstruction strategy is proposed to quantitatively analyze the impact of key node failures on network topology, vulnerability, and other indicators. Finally, a fire rescue case study is used to verify the effectiveness and accuracy of the proposed method.

[0050] This paper introduces the hypernetwork theory and, based on the idea of multi-layer modeling, proposes a task-oriented hypernetwork modeling and reliability assessment method for UAV clusters to address the difficulties in modeling multi-layer complex interactive relationships and measuring task reliability of UAV clusters.

[0051] like Figure 9 As shown, this embodiment provides a task-oriented UAV cluster super network modeling and reliability assessment method, including the following steps: constructing a multi-layer super network model of the UAV cluster based on the task execution characteristics and functional requirements of the UAV cluster; establishing a UAV cluster task reliability assessment framework oriented to the multi-layer super network model based on dynamic failure strategy and reliability assessment indicators; and performing reliability assessment on the UAV cluster task based on the UAV cluster task reliability assessment framework to obtain an assessment result.

[0052] S1. Multi-level modeling of drone swarms based on hypernetwork.

[0053] To effectively describe the complex, multi-layered, multi-element, high-dimensional relationships within drone swarms, this step constructs a multi-layered hypernetwork model of drone swarms based on their mission execution characteristics and functional requirements. This model, designed to meet mission requirements, systematically connects equipment nodes of varying types and performance through various information relationships, such as communication and missions, to establish a multi-layered functional subnetwork. These networks intersect and integrate with each other, forming a multi-layered, heterogeneous hypernetwork model. The model comprises four heterogeneous subnetworks: the business network, the mission network, the communication network, and the resource network, each fulfilling a distinct mission and role at a specific level.

[0054] S1.1 Business Network

[0055] The business network describes the execution relationship of cluster tasks, decomposing the cluster's mission into executable task activities, which can be abstracted as a directed graph. The business network is denoted as:

[0056] G a = <V a , E a >.

[0057] Among them, V a Represents a collection of sub-business nodes. I is the number of sub-businesses. a Represents a collection of sub-business execution relationships, such as Figure 1 , an example of a business network is described.

[0058] S1.2 Task Network

[0059] The mission network represents the actual collaborative relationship between the payloads of the UAV cluster during the mission. Each mission payload is considered a node, and the mission payload nodes are connected by dependency relationships, which represent the logical order between tasks. The mission network is an unweighted directed network:

[0060] G b = <V b , E b >.

[0061] Among them, V b Represents the set of task nodes, E b Represents the actual dependency set of the task node. There are:

[0062] V b =<Type,Cap> .

[0063] Among them, Type represents the type of equipment, and Type = S|D|A, where S, D, and A represent reconnaissance, decision-making, and attack nodes respectively. Cap represents the current capability attributes of each task node. Figure 2 As shown in the figure, a schematic diagram of several task links is shown.

[0064] S1.3 Communication Network

[0065] The communication network represents the communication connections between aircraft within a cluster. In an ad hoc network, each aircraft is a communication node, and nodes establish communication links through undirected edges. During the mission, the cluster structure is constantly changing, which has a certain impact on the information exchange within it. Therefore, the communication network is abstracted as a weighted undirected network, denoted as:

[0066] G c = <V c , E c , W c >.

[0067] Among them, V c represents the set of communication nodes, E c represents the set of communication links between nodes, W c Indicates the transmission efficiency of the communication link.

[0068] V c =<Location,Scope> .

[0069] Among them, Location is used to describe the deployment location of the node during the mission process, and Scope represents the maximum transmission distance of the node. If the equipment deployment location is outside the communication range, an effective communication link cannot be formed.

[0070] S1.4 Resource Network

[0071] The resource network reflects the collection of equipment resources within the swarm. Different types of equipment have relationships such as reconnaissance, command, and execution, and the same type of equipment may also have collaborative relationships. Compared to the mission network, the resource network is not dynamic and is not affected by equipment performance, location, etc. The resource network is abstracted as an unweighted directed network and defined as:

[0072] G d = <V d , E d >.

[0073] Among them, V d Represents a set of resource nodes, E d Represents a collection of logical dependencies between resources. They are:

[0074] V d =<Type,Status,Hp> .

[0075] Type represents the equipment type, with Type = S|D|A, where S, D, and A represent reconnaissance, decision-making, and strike nodes, respectively. Status indicates whether the corresponding equipment is being used in the mission. Status:: = work|free, where work indicates the equipment is being used and free indicates it is not currently participating in the mission. Hp indicates the equipment's mission and communication capabilities. Figure 3 A schematic diagram depicting the resource network.

[0076] S1.5 UAV Cluster Multi-Heterogeneous Network Model

[0077] On the basis of the four-layer heterogeneous sub-network, the interaction relationship and mapping mechanism between sub-networks are further established.

[0078] Business-Task Relationship L a-b :The business is decomposed into several sub-businesses at the business layer. Each sub-business needs to match one or more groups of specific task links (minitor tasks, communication guarantees, execution capabilities, etc.) to complete. The mapping relationship is reflected in that one business node corresponds to one or more groups of task links, which is represented by L a-b ;

[0079] Task-communication relationship L b-c :Multiple mission payload nodes carried by the aircraft exchange information through their embedded communication nodes. Each communication node in the communication network is connected to one or more mission payload nodes in the mission network, forming a one-to-one or one-to-many mapping relationship. b-c For example, for a reconnaissance and execution integrated UAV, a node in the communication network corresponds to a reconnaissance node and an execution node in the mission network.

[0080] Each mission node relies on the communication link of the communication network to complete information interaction, data transmission, command synchronization and other operations. Multiple mission payload nodes carried by the aircraft realize information exchange through its embedded communication nodes. The mapping relationship is reflected in the connection between a communication node in the communication network and one or more mission payload nodes in the mission network, which is represented by L b-c : For example, for a reconnaissance and execution integrated drone, a node in the communication network corresponds to a reconnaissance node and an execution node in the mission network.

[0081] Communication-Resource Relationship c-d :The aircraft entity usually contains one or more communication modules. The communication nodes in the communication network are mapped to the aircraft resource nodes in the resource network, forming a one-to-one or one-to-many relationship. c-d .

[0082] The resource network provides the hardware foundation to support communication. The communication network node is the functional unit embedded in the resource network node (i.e., the aircraft). The aircraft entity usually contains one or more communication modules. The mapping relationship is reflected in the connection between an aircraft resource node in the resource network and one or more communication nodes in the communication network, which is represented by L c-d :

[0083] The four layers of networks intersect and integrate with each other. The business network, at the top, describes the logical structure and execution sequence of mission objectives and expresses the strategic intent of the cluster in carrying out missions. The mission network, as the middle layer, is responsible for the functional implementation of business objectives, decomposing and mapping high-level tasks onto specific drones with different functional capabilities, thus building a collaborative execution structure. The communication network, as the support layer, provides the necessary information exchange capabilities during mission execution and ensures the connectivity and real-time communication between mission nodes. The resource network, at the bottom layer, reflects the configuration status of actual equipment resources and determines the system's available capacity boundaries and mission-carrying capacity. The four layers of networks achieve logical and physical connectivity through a cross-layer mapping mechanism, forming a complete mapping link from "goal → function → information → entity".

[0084] S2. UAV swarm mission reliability assessment framework for hyper-network model

[0085] This step constructs a mission reliability assessment framework for the drone swarm hypernetwork model. First, we describe in detail the failure and reorganization process of the drone swarm hypernetwork model after an attack. Next, we provide node degree, average betweenness, and vulnerability to analyze the mission reliability of drone swarms.

[0086] S2.1 Dynamic Failure Strategy of the UAV Swarm Hypernetwork Model

[0087] During mission execution, a drone swarm can experience environmental or self-failures, leading to the disappearance or weakening of nodes or edges, and consequently, reduced cluster reliability. To describe the dynamic failure process of a drone swarm during mission execution, this example uses a typical attack strategy, random attack, to randomly select a node for attack.

[0088] Due to the hierarchical interconnectedness of the drone swarm hypernetwork, failed nodes can cause cascading failures. This embodiment primarily considers two types of cascading failures: (a) The communication network relies on the resource network. When a node in the resource network disappears, the corresponding node in the communication network also disappears. (b) The communication network supports the mission network. When a node in the communication network disappears, the corresponding node in the mission network also disappears. Furthermore, if an isolated node exists in the communication network, it is considered failed.

[0089] To describe the failure process of a drone cluster, this embodiment takes a cluster of 12 drones consisting of 4 monitor drones, 3 decision drones, and 5 execution drones as an example to attack the nodes of the resource network. It is assumed that each drone only carries one communication device and one mission payload. Based on the drone cluster super network modeling method provided above, a functional network under a certain sub-service is established, and its dynamic failure and reconstruction process is given as follows: Figure 4 shown.

[0090] (1) In the resource network, decision node F is attacked and becomes ineffective;

[0091] (2) The failure of node F1 causes cascading failure in the hypernetwork model. Nodes F1 and F2 corresponding to node F in the communication network and task network, as well as their edges, fail. The failed nodes and edges are removed.

[0092] (3) Identify and detect isolated node B1 in the communication network, and node B1 fails;

[0093] (4) Node B1 further causes the associated nodes B2 and B in the task network and resource network, as well as their edges, to fail. The failed nodes and edges are then removed.

[0094] (5) Perform integrity checks on the communication network to ensure that there are no isolated nodes and edges in the communication network, and obtain a supernetwork model under dynamic failure;

[0095] (6) Identify and detect isolated nodes in the task network, and reconnect and construct a new task chain based on the communication links in the communication network for the nodes that cannot form a complete task chain, until there are no isolated nodes in the task network.

[0096] S2.2 Mission reliability evaluation indicators for the UAV swarm dynamic hypernetwork model

[0097] To further quantify the mission reliability of the drone swarm hypernetwork in complex environments, this embodiment, based on existing complex system analysis indicators, focuses on three core indicators: node degree, betweenness, and vulnerability, from the perspective of network topology characteristics.

[0098] (1) Average

[0099] Degree is a fundamental concept in complex networks. The degree of a node refers to the number of edges connected to it, indicating the node's importance in the network. A larger degree indicates more connections between the node and other nodes. The node degree in the drone swarm hypernetwork can be divided into two parts:

[0100] 1) The node degree k calculated by the node connection of node i in its network layer 1 i , including the degree k of the task network layer bi , the degree k of the communication network layer ci and the degree k of the resource network layer di , can be calculated by the single-layer network indicators:

[0101] k 1 i＝ k bi +k ci +k di

[0102] 2) The node degree k calculated by connecting node i to other layer nodes through the hierarchical edges 2 i , including the degree k of the task-communication network layer b-ci and the degree k of the communication-resource network layer c-di If each UAV carries only one communication model and one mission payload, then:

[0103]

[0104] Where N is the total number of nodes in the network.

[0105] Therefore, the node degree of any node i in the hypernetwork can be expressed as k i :

[0106] k i =k 1 i +k 2 i

[0107] Let K represent the average degree in the hypernetwork, then:

[0108]

[0109] (2) Betweenness

[0110] The betweenness reflects the global importance of a node in the network and can be expressed as δ, which is the degree to which a node acts as a "bridge" in the shortest path of the network.

[0111]

[0112] Among them, γ qp represents the number of all shortest paths from node q to node p, is the number of paths passing through node i among these shortest paths. Using δ to represent the average betweenness in the hypernetwork, we have:

[0113]

[0114] (3) Vulnerability

[0115] Vulnerability is defined as the ability of a system to recover and stabilize due to system failure or malfunction when encountering external interference or threats. The vulnerability of the drone swarm hypernetwork reflects the overall robustness of the hypernetwork and is characterized by the normalized values of degree and betweenness, denoted by P:

[0116]

[0117] Among them, k i is the degree of node i, δ i is the betweenness centrality of node i, and α is a weight parameter used to control the impact of node degree and betweenness on vulnerability.

[0118] Simulation Verification: Using a forest fire rescue mission as an example, we applied and analyzed the hypernetwork-based multi-level modeling of drone swarms and their mission reliability assessment methods. Assume a sudden forest fire breaks out in a certain area, requiring fire monitoring and on-site firefighting. To this end, a swarm of 24 drones, consisting of 8 reconnaissance drones, 6 decision-making drones, and 10 execution drones, is deployed. These drones are divided into two groups, each carrying a set of communications equipment and a type of mission payload. During the mission, drones may malfunction or fail due to environmental or internal factors, impacting the overall reliability of the swarm's operations.

[0119] Furthermore, the hypernetwork model is constructed:

[0120] (1) Business decomposition.

[0121] Decompose fire rescue into atomic business and get the business set: Va = {Va1, Va2, Va3, Va4, Va5} = {mission planning, monitoring and positioning, command decision-making, fire fighting and rescue, evaluation and settlement}; according to the time sequence and logical relationship of each sub-business, generate the UAV cluster forest rescue business flow as follows: Figure 5shown.

[0122] (2) Network element abstraction.

[0123] In this forest firefighting mission, the drone swarm's resource network layer included eight monitoring nodes, six decision-making nodes, and ten execution nodes. At the mission and communication network layers, the nodes actually participating in different business phases and their interactions dynamically evolve with business needs. Table 1 shows the statistics of the nodes actually participating in each phase.

[0124] Table 1

[0125]

[0126]

[0127] (3) Hypernetwork modeling.

[0128] The abstracted nodes and interaction relationships are connected and combined according to the association mapping rules to build a super network model of the drone cluster. Taking the firefighting and rescue business stage (T4 stage) as an example, the super network model of this stage is as follows: Figure 6 As shown in the figure, purple A~H represent reconnaissance nodes, red I~N represent decision nodes, and blue O~X represent execution nodes.

[0129] Furthermore, reliability assessment:

[0130] (1) Reliability assessment at different business stages.

[0131] The number of nodes, number of edges, average degree, average betweenness, and vulnerability of the UAV cluster dynamic hypernetwork at different business stages are evaluated and statistically analyzed, and the relevant parameters are shown in Table 2.

[0132] Table 2

[0133]

[0134] The changing trends of the number of nodes N and the number of edges E in different business stages are as follows: Figure 7 As shown in a, the changing trend of the average degree K is as follows Figure 7 As shown in b, the changing trend of the average betweenness δ is as follows Figure 7 As shown in c, the changing trend of network vulnerability P is as follows Figure 7 d. The results show that:

[0135] 1) The number of nodes, edges, and average degree all showed an increasing trend from mission planning (T1) to firefighting and rescue operations (T4), indicating that the network structure was gradually expanding and the connections between nodes were continuously strengthening. During the firefighting and rescue phase (T4), the number of nodes, edges, and average degree reached their peak (72 nodes, 114 edges, and a degree of 3.1667). This phase showed the highest network complexity and the most intensive information exchange. During T5 and T6, the number of nodes and edges decreased, and the average degree also dropped to 2.25, indicating that the network tended to simplify and converge after mission completion, with some nodes exiting the service and the number of connections decreasing.

[0136] 2) The average betweenness gradually decreases as the business phase progresses, from 0.2 in T1 to 0.0308 in T6, indicating a decrease in the importance of key nodes. Information flow becomes more balanced, and the network tends to be decentralized. In the early stages (such as T1), a few key nodes bear the majority of communication tasks. However, as the network structure becomes more complex, the load becomes more distributed, and the overall betweenness value decreases.

[0137] 3) It dropped from 0.5333 in T1 to 0.1632 in T4, indicating that the network has become more robust and has improved its anti-interference capabilities during its evolution. In T5 and T6, the vulnerability slightly increased to 0.1901, possibly due to the reduction in network size and the continued burden of critical tasks on some nodes.

[0138] (2) Reliability assessment under dynamic failure.

[0139] Taking the firefighting and rescue operation (time T4) as an example, we simulated the dynamic changes in network performance after an attack by randomly deleting resource network nodes and considering cascading failures. The specific failure process is as follows: A. Decision node J fails at time T42. B. Monitoring node B fails at time T43. C. Execution nodes O and U fail at time T44. D. Decision node M fails at time T45. E. Monitoring node H and execution node X fail at time T46. F. Execution nodes Q and T fail at time T47.

[0140] The number of nodes, number of edges, average degree, average betweenness, and vulnerability of the UAV cluster dynamic supernetwork at each moment are evaluated and counted. The results are shown in Table 3.

[0141] Table 3

[0142]

[0143]

[0144] At each moment in stage T4, the changing trends of the number of nodes N and the number of edges E are as follows Figure 8 As shown in a, the changing trend of the average degree K is as follows Figure 8 As shown in b, the changing trend of the average betweenness δ is as follows Figure 8 As shown in c, the changing trend of network vulnerability P is as follows Figure 8 d. The results show that:

[0145] 1) During the firefighting and rescue mission, as key nodes were removed, the cluster size and information exchange gradually decreased. However, the number of edges remained unchanged between T42 and T43, indicating that the network was restructured during this phase, avoiding a serious chain reaction.

[0146] 2) The average degree showed an overall downward trend. Between T41 and T42, due to the removal of decision node J, some drones lost key connections, causing the average degree to drop from 3.1667 to 2.8116. Subsequently, at T43, despite the disappearance of reconnaissance node B, the network was restructured, causing the average degree to rebound to 2.9394. However, with the disappearance of execution nodes O and U at T44, the average degree dropped again.

[0147] 3) The average betweenness initially remained relatively flat, but gradually increased as the business progressed, rising from 0.0348 in T41 to 0.0566 in T47. After some nodes were removed, the remaining intermediary nodes carried more path transmission tasks. This indicates that the network's degree of decentralization has decreased, and some nodes have become more important.

[0148] 4) The overall vulnerability increases, indicating that the network's stability gradually decreases after the attack. During the T41-T43 period, the vulnerability increases from 0.1632 to 0.1755 due to the failure of decision node J. Subsequently, the vulnerability decreases slightly to 0.1691 after the execution nodes O and U disappear at T45, possibly due to the network's redundancy at this stage. However, during the T46-T47 period, as more task nodes (such as reconnaissance node H and execution node X) are removed, the vulnerability increases sharply, indicating that the overall anti-interference capability of the drone swarm decreases after multiple rounds of attacks.

[0149] This embodiment addresses the problems of high complexity in drone cluster modeling and difficulty in reliability measurement. It conducts multi-layer linkage modeling of drone clusters, focuses on analyzing the dynamic evolution process and overall reliability of the cluster when performing collaborative tasks, and proposes a set of task-oriented multi-level hypernetwork modeling and reliability assessment methods.

[0150] The main research results of this invention are: (1) Task-oriented UAV cluster super-network model: driven by tasks and based on super-network theory, a super-network model of UAV cluster is constructed from four levels: business network, task network, communication network and resource network. (2) Task reliability evaluation framework for super-network model: considering the dependency relationship between tasks, communications and resources, a set of dynamic failure strategies for UAV cluster super-network model is proposed, and on this basis, a task reliability analysis framework for UAV cluster is further proposed. (3) Case verification: taking the fire rescue mission as an example, the topological evolution process of UAV cluster in different business stages is simulated, and by introducing the dynamic failure mechanism, the impact of key node failure on network structure and task reliability is quantitatively analyzed. The experimental results verify the practicality and effectiveness of the proposed modeling and analysis methods.

[0151] This invention addresses key challenges in modeling complex, multi-layered, and multi-element relationships within drone swarms, as well as the difficulty in measuring reliability under dynamic missions. The proposed method covers the entire design and deployment phase of drone swarms, providing a theoretical basis and technical support for mission planning, dynamic scheduling, and optimal resource allocation. Future research will explore the integration of dynamic and complex environments and the introduction of intelligent algorithms to further research cluster modeling and reliability analysis.

[0152] The above are merely preferred embodiments of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. A task-oriented UAV swarm hypernetwork modeling and reliability assessment method, characterized by: The following steps are involved: A multi-layer hyper-network model of UAV swarm is constructed based on the mission execution characteristics and functional requirements of UAV swarm; Based on dynamic failure strategies and reliability evaluation indicators, a UAV swarm mission reliability evaluation framework for the multi-layer hypernetwork model is established; Based on the UAV swarm mission reliability assessment framework, the reliability of the UAV swarm mission is evaluated to obtain the evaluation results.

2. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 1 is characterized in that: The process of building a multi-layer hypernetwork model for drone swarms includes: Build a business network based on the mission execution relationship of drone clusters; Construct a mission network based on the actual collaborative relationship between the payloads of the UAV cluster during the mission; Build a communication network based on the communication connection relationship between aircraft within the UAV cluster; Construct a resource network based on the equipment resource collection of aircraft in the UAV cluster; A multi-layer super network model of the drone cluster is constructed based on the interactive relationship and mapping mechanism of the business network, task network, communication network and resource network.

3. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 2 is characterized in that: The process of constructing a multi-layer super network model of a drone cluster based on the interactive relationship and mapping mechanism among the business network, task network, communication network and resource network includes: Decompose the tasks performed by the drone cluster into several sub-services, each sub-service corresponds to a service node, and map the service nodes to several groups of task links to build a service-task relationship; Connect the communication nodes of the aircraft in the UAV cluster to several mission payload nodes in the mission network to build a mission-communication relationship; Mapping the communication node to resource network nodes in a plurality of resource networks to establish a communication-resource relationship; A multi-layer super network model is constructed based on the business-task relationship, task-communication relationship and communication-resource relationship.

4. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 3 is characterized in that: The business network is a directed graph, wherein the nodes of the directed graph represent sub-businesses and the edges represent the execution order between sub-businesses; The task network is an unweighted directed network, wherein the nodes of the unweighted directed network corresponding to the task network represent task loads, and the edges represent the dependency relationships between task loads; The communication network is a weighted undirected network, wherein the nodes of the weighted undirected network represent communication nodes, the edges represent communication links, and the weights represent the transmission efficiency of the communication links; The resource network is an unweighted directed network, wherein the nodes of the unweighted directed network corresponding to the resource network represent resource nodes, and the edges represent logical dependency relationships between resource nodes.

5. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 1 is characterized in that: The dynamic failure strategy includes: random attack strategy and cascading failure strategy.

6. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 1 is characterized in that: The reliability evaluation indicators include average degree, average betweenness and fragility.

7. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 6 is characterized in that: The calculation expression of the average degree is: Where K represents the average degree in the hypernetwork, k i represents the node degree of any node i in the hypernetwork, and N represents the total number of nodes in the network.

8. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 6 is characterized in that: The calculation expression of the average betweenness is: Where δ represents the average betweenness in the hypernetwork, N represents the total number of nodes in the network, and δ i Indicates betweenness.

9. The task-oriented UAV cluster super-network modeling and reliability assessment method according to claim 8 is characterized in that: The calculation expression of the vulnerability is: Where P represents vulnerability, k i represents the degree of node i, and α represents the weight parameter.

Images