Hybrid unmanned cluster task reliability evaluation method based on complex network

Abstract

The invention provides a hybrid unmanned cluster task reliability evaluation method based on a complex network, and the method comprises the steps: carrying out the complex network modeling of an interaction relation between unmanned plane nodes in a hybrid unmanned cluster system, and analyzing the structural characteristics and dynamic evolution rules of the network. And setting network robustness evaluation indexes such as a maximum connected subgraph, network connection efficiency, destroy resistance measurement and the like, and researching a connection relationship between nodes and an information transmission rule. A robustness evaluation strategy is constructed based on a random attack algorithm and a deliberate attack algorithm, and robustness analysis of the hybrid unmanned cluster complex network is carried out; and based on the constructed complex network model, providing a hybrid unmanned cluster task reliability evaluation index which comprehensively considers a network topology structure, a node state and communication quality and is used for evaluating the task reliability of a cluster system. According to the method, the task reliability of the hybrid unmanned cluster system can be effectively evaluated, and technical support is provided for design optimization and task planning of the cluster system.

Description

Technical Field

[0001] The present invention belongs to the technical field of unmanned aerial vehicle control, and particularly relates to a method for evaluating the mission reliability of a hybrid unmanned cluster based on complex networks. Background Art

[0002] Clusterization is an important future development trend of unmanned combat equipment such as unmanned aerial vehicles. Through information interaction and cooperative control among individual unmanned aerial vehicles, the flexibility and elasticity of missions can be enhanced, and the mission capabilities and overall combat effectiveness of unmanned aerial vehicles can be improved. However, in the face of a battlefield environment with high antagonism, uncertainty, and dynamics, the cluster system must be stable and reliable and show high resilience to abnormal situations such as failures in order to be deployable in actual combat environments. Compared with traditional combat systems, unmanned aerial vehicle clusters with autonomous cooperative combat capabilities have characteristics such as a large number of nodes and complex information interaction among nodes, and are a typical complex network system. Their system robustness and mission reliability are greatly affected by battlefield conditions. Conducting theoretical research on evaluation methods for the robustness and mission reliability of cluster systems is of great engineering significance for studying the combat patterns of cluster systems to fully exert their overall combat effectiveness.

[0003] Traditional reliability evaluation methods mainly rely on reliability theory and systems engineering methods. By establishing the relationship between mission processes and subsystems, the reliability indicators of the system are calculated. This method performs well in simple systems but has certain limitations when dealing with complex unmanned cluster systems. Traditional mission reliability evaluation methods usually only consider the direct relationship between nodes. However, the interaction among nodes in a hybrid unmanned cluster system is complex and diverse, and there may be non-linear relationships and feedback mechanisms, which cannot be fully captured by traditional methods. In particular, the state and connection relationship of nodes in an unmanned cluster system may change over time. It is difficult for traditional methods to accurately evaluate and analyze the dynamics of the system and fully consider characteristics such as dynamic interaction relationships, complex structures, and large scale in the system.

[0004] Compared with traditional evaluation methods, the reliability evaluation method based on complex networks is a method that combines complex theory and computer science. By modeling and analyzing the network structure and node relationships, it evaluates the reliability of the system. The reliability evaluation method based on complex networks can better capture the complex interaction relationships between nodes, describe the non-linear relationships between nodes and the complex network structure. At the same time, complex network theory can help analyze the changes in topological structure and node states at different time points, so as to better evaluate the dynamic reliability of the system. In addition, complex network theory provides efficient algorithms and methods to handle large-scale network systems, and can effectively process data analysis and calculation of complex network structures and large-scale nodes. Therefore, the task reliability evaluation method based on complex networks has stronger analysis ability and applicability than traditional methods, and can accurately evaluate the reliability of the complex system of hybrid unmanned clusters. Summary of the Invention

[0005] Object of the Invention: Facing the battlefield environment with high antagonism, uncertainty, and dynamics, the cluster system must be stable and reliable, and show high resilience to abnormal situations such as failures in order to be deployable in the actual combat environment. The present invention proposes a method for evaluating the task reliability of a hybrid unmanned cluster based on complex networks. First, a complex network model is established for the interaction relationships between UAV nodes in the hybrid unmanned cluster system, and the structural characteristics and dynamic evolution laws of the network are analyzed. Then, network robustness evaluation indicators such as the largest connected subgraph, network connectivity efficiency, and anti-destruction metric are set, and the connection relationships between nodes and the information transmission laws are studied. Secondly, based on the random attack algorithm and the deliberate attack algorithm, a robustness evaluation strategy is constructed to perform the robustness analysis of the complex network of the hybrid unmanned cluster; finally, based on the constructed complex network model, a task reliability evaluation indicator for the hybrid unmanned cluster that comprehensively considers the network topology structure, node state, and communication quality is proposed to evaluate the task reliability of the cluster system.

[0006] Technical Solution: To achieve the object of the present invention, the technical solution adopted by the present invention is: a method for evaluating the task reliability of a hybrid unmanned cluster based on complex networks, which specifically includes the following steps:

[0007] Step 1, construct a task scenario;

[0008] Step 2, according to the task scenario constructed in Step 1, perform complex network modeling on the hybrid unmanned cluster system to obtain an unmanned cluster complex network model;

[0009] Step 3, based on the unmanned cluster complex network model constructed in Step 2, establish a robustness evaluation index for the complex network model;

[0010] Step 4: Construct a robustness evaluation strategy based on the random attack algorithm and the deliberate attack algorithm, and conduct network fault simulation;

[0011] Step 5: Construct the task reliability evaluation index R.

[0012] In Step 1, the task scenario includes: According to the combat situation in reality, a reconnaissance team includes more than two unmanned reconnaissance aircraft;

[0013] Under the guidance of the command aircraft, the unmanned reconnaissance aircraft enter the combat area to detect and reconnoiter the enemy's defense works and communication bases;

[0014] The unmanned reconnaissance aircraft S conducts reconnaissance and interference in the designated area. Once a target is detected, the unmanned reconnaissance aircraft S will immediately lock and track the target, and then transmit the target information and battlefield situation to the command aircraft D. After the command aircraft D completes the situation assessment, tactical decision-making, and task arrangement, it issues a task instruction to the unmanned attack aircraft I; According to the target information and task instruction feedback by the command aircraft D, the unmanned attack aircraft I completes the coordinated target allocation and route planning tasks, and realizes the fixed-point strike and destruction of the target.

[0015] Step 2 includes:

[0016] Step 2-1: Initialize the nodes;

[0017] Step 2-2: Calculate the comprehensive degree of the nodes;

[0018] Step 2-3: Determine the key nodes;

[0019] Step 2-4: Connect the nodes;

[0020] Step 2-5: Construct the adjacency matrix of the complex network of the unmanned cluster;

[0021] Step 2-6: Output the nodes, edges, and complex network.

[0022] Step 2-1 includes: Generate N nodes, and define the numbers of the command aircraft D, the unmanned reconnaissance aircraft S, and the unmanned attack aircraft I as N D , N S and N I , and satisfy N D +N S +N I =N.

[0023] Step 2-2 includes: Calculate the comprehensive degree c of the i-th node using the following formula i :

[0024] c i =k i +μ i m i (1)

[0025] where k i is the connectivity value of the i-th node, and m i is the number of secondary neighbor nodes of the i-th node, and μ i is the influence coefficient, and its calculation formula is:

[0026]

[0027] where n i is the total number of neighbor nodes of the i-th node.

[0028] Step 2-3 includes: Sort the comprehensive degrees of the nodes according to the principle of arranging from large to small, and select the nodes with the comprehensive degrees ranked in the top 5% to 50% in the unmanned cluster network as the protected nodes to form a key node set.

[0029] Step 2-4 includes: Each command aircraft, unmanned reconnaissance aircraft, and unmanned attack aircraft is regarded as a node. Connect the command node with the key nodes respectively, connect any two points among the reconnaissance nodes, connect any two points among the attack nodes, and connect any two points among the unmanned attack aircraft I; The command aircraft has a maximum working limit, and the workload it responds to does not exceed ω max ; There is no connection edge between the reconnaissance aircraft and the combat aircraft.

[0030] Step 2-5 includes: The adjacency matrix is symmetric and the diagonal elements are all 0. Only N(N - 1) / 2 storage units are required. The adjacency matrix is represented by A, and the expression is:

[0031]

[0032] where the element in the i-th row and j-th column of the adjacency matrix A where (v i , v j ) ∈ E represents that there is a connection edge between the nodes, and a ij is represented as 1, represents that there is no connection edge between the nodes, and a ij is represented as 0, where v i represents the i-th node, and v j represents the j-th node.

[0033] Step 3 includes:

[0034] Step 3-1, for any network, define U as the number of all nodes in the target network, and define U' as the number of nodes in the largest connected subgraph that can be obtained after removing some nodes after being attacked. The formula is:

[0035]

[0036] Among them, M is the degree of network collapse;

[0037] Step 3-2: Define the node efficiency e(i) of the i-th node as the sum of the reciprocals of the shortest paths of all the nodes connected to the i-th node. The formula is:

[0038]

[0039] where d ij is the shortest path experienced by the i-th node to the i-th node;

[0040] According to the node efficiency, define the efficiency E of the whole network as the average value of the sum of the efficiencies of all the nodes in the network. The formula is:

[0041]

[0042] Step 3-3: Define the viscosity CH(i) of the i-th node as the product of the survival probability p(i) of the i-th node and the node degree value k i (the degree value is defined as the number of edges connecting the i-th node and other nodes). The expression is:

[0043] CH(i) = p(i)k i (7)

[0044] Among the nodes under random attack, the survival probability of each node is:

[0045]

[0046] Among the nodes under deliberate attack, the survival probability of each node is:

[0047]

[0048] The calculation formula for the viscosity CH(i) of the i-th node is:

[0049]

[0050] Therefore, the calculation formula for the anti-destruction metric H of the network is:

[0051]

[0052] Step 4 includes:

[0053] Step 4-1: Randomly attack a single node in the network: For each attack on all the nodes in the network, only randomly select one node for attack until all the nodes in the network are removed;

[0054] Step 4-2, deliberately attack a single node with the largest degree value in the network: Each attack selects the node with the largest degree value in the network for attack, and it is set to remove one node each time.

[0055] Step 5 includes: constructing a task reliability evaluation index R using the following formula:

[0056]

[0057] where ω1, ω2, ω3 are weights, and ω1 + ω2 + ω3 = 1.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] 1. The present invention evaluates the task reliability of a hybrid unmanned cluster system based on complex network theory, which can better capture the complex interaction relationships between nodes, can more effectively describe the non-linear relationships and complex network structures between nodes, and can more accurately analyze the impact of the interaction between nodes on the system reliability.

[0060] 2. It can help analyze the changes in the topological structure and node states of the system at different stages, so as to better evaluate the dynamic reliability of the system.

[0061] 3. A task reliability evaluation index for hybrid unmanned clusters that comprehensively considers the network topological structure, node states, and communication quality is proposed, which can comprehensively consider the topological structure of the system, the connection methods between nodes, and system functions, so as to more comprehensively evaluate the system reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The following further specific description of the present invention is made in conjunction with the accompanying drawings and specific embodiments, and the above and / or other advantages of the present invention will become clearer.

[0063] Figure 1 is the flowchart of the method of the present invention.

[0064] Figure 2 is a schematic diagram of a hybrid unmanned cluster system.

[0065] Figure 3 is a complex network diagram of a hybrid unmanned cluster.

[0066] Figure 4 is the flowchart of network fault simulation.

[0067] Figure 5 is a schematic diagram of the change in the task reliability of the cluster system.

[0068] Figure 6 is a schematic diagram of the simulation results. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0069] As Figure 1 shown, this embodiment provides a method for evaluating the reliability of a hybrid unmanned cluster mission based on complex networks, including the following steps:

[0070] Step 1, construct a mission scenario: According to the combat situation in reality, a reconnaissance team consists of multiple unmanned reconnaissance aircraft. Under the guidance of a command aircraft, they enter the combat area to detect and reconnoiter the enemy's defense works and communication bases. To protect the command aircraft and reduce the losses of the entire system, the unmanned reconnaissance aircraft S conducts reconnaissance and interference in the designated area. Once a target is discovered, they will immediately lock and track the target, and then transmit the target information and battlefield situation to the command aircraft D. After the command aircraft completes the situation assessment, tactical decision-making, and task arrangement, it issues a mission instruction to the unmanned attack aircraft I. According to the target information and mission instruction feedback by the command aircraft D, the unmanned attack aircraft I completes the coordinated target allocation and route planning tasks, and realizes the fixed-point strike and destruction of the target.

[0071] Step 2, based on the mission scenario constructed in Step 1, conduct complex network modeling on the hybrid unmanned cluster system.

[0072] Step 2-1, initialize nodes, generate N nodes, and define the numbers of the command aircraft D, unmanned reconnaissance aircraft S, and unmanned attack aircraft I as N D , N S and N I , and satisfy N D +N S +N I =N.

[0073] Step 2-2, calculate the node comprehensive degree. The node comprehensive degree c i is used to measure the information transmission ability of the node in the network. The larger the value, the higher the information transmission efficiency of the node. Its definition is:

[0074] c i =k i +μ i m i (1)

[0075] where k i is the node degree, m i is the number of the secondary neighbor nodes of the node, and μ i is the influence coefficient.

[0076]

[0077] where n i is the total number of the neighbor nodes of the node.

[0078] Step 2-3: Determine the key nodes, sort them based on the node comprehensiveness, select the nodes with higher comprehensiveness in the unmanned cluster network as the protected nodes, and form a set of key nodes. The selection of the proportion of key nodes selected as protected nodes in a complex network depends on the specific network structure, application scenario, and protection requirements. Generally, select the nodes with comprehensiveness ranking in the top 5% to 50% in the unmanned cluster network as key nodes. The more the number of key nodes, the better the robustness and stability of the network, but at the same time, it will also increase the protection cost and resource consumption. Therefore, when determining the key nodes, it is necessary to appropriately adjust the proportion of protected nodes according to the characteristics of the network and the protection requirements.

[0079] Step 2-4: Connections between nodes. The command aircraft D is connected to the key nodes, and any two points of the unmanned reconnaissance aircraft S are connected, and any two points of the unmanned attack aircraft I are connected; in order to avoid excessive load on the leader, the command aircraft has a maximum working limit, and the corresponding workload does not exceed ω max ; in order to avoid excessive load on the followers, there are no edges between the reconnaissance aircraft and the combat aircraft;

[0080] Step 2-5: Construct the adjacency matrix of the unmanned cluster complex network. The adjacency matrix is symmetric and the diagonal elements are all 0, so only the upper triangular or lower triangular data needs to be stored, that is, only N(N - 1) / 2 storage units are required. Then the general expression is:

[0081]

[0082] where where (v i , v j ) ∈ E represents that there is an edge between nodes and is represented as 1, represents that there is no edge between nodes and is represented as 0.

[0083] Step 2-6: Output the nodes, edges, and complex network, as shown in Figure 2 、 Figure 3 shown, Figure 2 where S represents the unmanned reconnaissance aircraft, I represents the unmanned attack aircraft, and D represents the command aircraft.

[0084] Step 3: Based on the unmanned cluster complex network model constructed in Step 2, establish a robustness evaluation index for the complex network model.

[0085] Step 3-1: The connected subgraph of a network refers to the set of all nodes that are directly or indirectly connected along a random edge. For any network, define N as the number of all nodes in the target network, and define N' as the number of nodes in the largest connected subgraph that can be obtained after removing some nodes under attack. The formula is:

[0086]

[0087] Among them, M is the degree of network collapse, which is the ratio of the number of nodes in the original network to the number of nodes in the largest connected subgraph.

[0088] Step 3-2, the index of the shortest path in a complex network can describe the reliability of the network. The larger the value, the worse the reliability. To measure the connectivity efficiency of the network, the node efficiency is defined as the sum of the reciprocals of the shortest paths of all the nodes connected to node i. The formula is:

[0089]

[0090] where d ij is the shortest path from node i to node j. According to the definition of node efficiency, the efficiency of the whole network is defined as the average value of the sum of the efficiencies of all nodes in the network. The formula is:

[0091]

[0092] Step 3-3, in a complex network, after being attacked, the network still needs to maintain a relatively stable state. This is one of the measures of robustness, namely anti-destruction, which reflects the robustness of the network system.

[0093] After the nodes in the network are attacked, successive failures will occur. In addition to considering the efficiency of the nodes, it is also necessary to consider the impact of the survival probability of the nodes on the anti-destruction of the network. Because the degree value k of the nodes in the network i is larger, the survival probability p(i) of the corresponding node is higher, and the difficulty of completely deleting this node is greater. Based on this property, viscosity is proposed. The viscosity is defined as the product of the survival probability of this node and the node degree value. Its expression is:

[0094] CH(i) = p(i)k i (7)

[0095] The ratio of the viscosity CH(i) of node i to the total number of edges completely connected to all nodes in the network is the anti-destruction of node i in the network. Its expression is:

[0096]

[0097] Therefore, the measure of the anti-destruction of the network is the sum of the anti-destructions of all nodes in the network. Its expression is:

[0098]

[0099] The value range of the measure function of anti-destruction is 0 ≤ D ≤ 1. In a real network, even if the network is fully connected, there is still a probability that each node will be destroyed. Therefore, the best effect of anti-destruction cannot be achieved, that is, D = 1.

[0100] Step 4: Construct a robustness evaluation strategy based on the random attack algorithm and the deliberate attack algorithm to simulate network failures.

[0101] Step 4-1: Randomly attack a single node in the network: For each attack on all nodes in the network, only randomly select one node for attack until all nodes in the network are removed.

[0102] Step 4-2: Deliberately attack a single node with a higher node degree value in the network: For each attack, select the node with the largest degree of connection value in the network for attack, and set to remove one node each time for easy statistical analysis.

[0103] The network failure simulation process is as Figure 4 shown.

[0104] Step 5: Based on the robustness evaluation indicators in Step 3, namely the network collapse degree M, the overall network efficiency E, and the anti-destruction metric D, use the sensitivity analysis method to construct a task reliability evaluation indicator R, and the expression is:

[0105]

[0106] where ω1, ω2, and ω3 are the weights corresponding to each indicator, and ω1 + ω2 + ω3 = 1.

[0107] Step 6: Apply the model. Apply the hybrid unmanned cluster task reliability evaluation method based on complex networks to obtain the change in the task reliability of the cluster system. The simulation process is as Figure 5 .

[0108] Step 6-1: Taking the reconnaissance task as an example, assume that the task area is a two-dimensional plane and all unmanned aerial vehicles move on the two-dimensional plane.

[0109] Step 6-2: Before the task starts, perform parameter settings, and the parameter settings are as follows: the area of the task area, the number of command aircraft, the number of unmanned reconnaissance aircraft, and the number of unmanned attack aircraft.

[0110] Step 6-3: Based on the unmanned cluster complex network model constructed in Step 2.

[0111] Step 6-4: From the initial moment to the end of the simulation, each unmanned aerial vehicle acts according to the following sequential process: 1) Determine the candidate mobile unit area; 2) Select the next mobile area according to the set motion rules.

[0112] Step 6-5: When the simulation progresses to the time period when an attack is triggered, based on the robustness evaluation strategy constructed in Step 4, select the random attack algorithm, and the unmanned aerial vehicle cluster will be attacked and damaged. The attacked unmanned aerial vehicles will be removed from the unmanned cluster complex network, and the corresponding network topology relationship will be damaged.

[0113] Step 6-6: Based on the task reliability evaluation index R constructed in Step 5, continuously calculate the task reliability of the unmanned cluster system at each moment.

[0114] Step 6-7: The drone cluster continuously repeats Step 6-4, Step 6-5, Step 6-6, and Step 6-7 during the entire mission until the simulation proceeds to the preset termination time, at which point the simulation is aborted, and the simulation results are as Figure 6 .

[0115] The present invention provides a method for evaluating the task reliability of a hybrid unmanned cluster based on complex networks. There are many methods and approaches to specifically implement this technical solution. The above description is only the preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art of this technology, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention. Each component not clearly defined in this embodiment can be implemented using existing technologies.

Claims

1. A reliability assessment method for mixed unmanned cluster tasks based on complex networks, characterized in that: The following steps are involved: Step 1: Build a task scenario; Step 2: Based on the mission scenario constructed in step 1, complex network modeling is performed on the mixed unmanned cluster system to obtain a complex network model of the unmanned cluster; Step 3, based on the unmanned cluster complex network model constructed in step 2, establish a robustness evaluation index of the complex network model; Step 4: Construct a robustness evaluation strategy based on the random attack algorithm and the intentional attack algorithm to simulate network failures; Step 5: Construct the task reliability evaluation index R.

2. The method according to claim 1, characterized in that In step 1, the mission scenario includes: according to the actual combat situation, a reconnaissance team includes more than two unmanned reconnaissance aircraft; Under the guidance of the command aircraft, the unmanned reconnaissance aircraft entered the combat area to detect and scout the enemy's fortifications and communication bases; The unmanned reconnaissance aircraft S conducts reconnaissance and interference in the designated area. Once a target is found, the unmanned reconnaissance aircraft S will immediately lock and track the target, and then transmit the target information and battlefield situation to the command aircraft D. After the command aircraft D completes the situation assessment, tactical decision-making and task arrangement, it issues a mission instruction to the unmanned attack aircraft I; based on the target information and mission instructions fed back by the command aircraft D, the unmanned attack aircraft I completes the task of coordinating target allocation and route planning to achieve fixed-point strikes and destruction of the target.

3. The method according to claim 2, characterized in that Step 2 includes: Step 2-1, initialize the node; Step 2-2, calculate the node comprehensiveness; Step 2-3, determine the key nodes; Step 2-4, connect the nodes; Step 2-5, construct the adjacency matrix of the unmanned cluster complex network; Step 2-6, output nodes, edges and complex networks.

4. The method according to claim 3, characterized in that Step 2-1 includes: generating N nodes and defining the command machine D , unmanned reconnaissance aircraft S and unmanned attack aircraft I The number of D 、N S and N I , and satisfies N D +N S +N I =N; Step 2-2 includes: calculating the comprehensive degree c of the i-th node using the following formula: i : c i =k i +m i m i (1) where k i is the connectivity value of the ith node, m i is the number of secondary neighbor nodes of the ith node, μ i is the influence coefficient, and the calculation formula is: where n i is the total number of neighbor nodes of the ith node.

5. The method according to claim 4, characterized in that Step 2-3 includes: sorting the comprehensiveness of the nodes according to the principle of arrangement from large to small, selecting the nodes with comprehensiveness ranking in the top 5% to 50% in the unmanned cluster network as protected nodes, and forming a key node set.

6. The method according to claim 5, characterized in that Steps 2-4 include: each command aircraft, unmanned reconnaissance aircraft, and unmanned attack aircraft are regarded as a node, and the command node is connected to the key node respectively, any two points in the reconnaissance node are connected, any two points in the attack node are connected, and any two points in the unmanned attack aircraft are connected; the command aircraft has a maximum working limit, and the response workload does not exceed ω max ; There is no connection between reconnaissance aircraft and combat aircraft.

7. The method according to claim 6, characterized in that Step 2-5 includes: the adjacency matrix is ​​symmetrical and the diagonal elements are all 0, only N(N-1) / 2 storage units are required, the adjacency matrix is ​​represented by A, and the expression is: The element in the i-th row and j-th column of the adjacency matrix A is Where (v i ,v j )∈E represents the existence of edges between nodes, a ij It is represented as 1, Indicates that there is no edge between nodes, a ij Represented as 0, where v i represents the i-th node, v j represents the jth node.

8. The method according to claim 7, characterized in that Step 3 includes: Step 3-1: For any network, define U as the number of all nodes in the target network, and define U' as the number of nodes in the maximum connected subgraph that can be obtained after removing some nodes after being attacked. The formula is: Among them, M is the degree of network collapse; Step 3-2, define the node efficiency e(i) of the i-th node as the sum of the reciprocals of the shortest paths of all nodes connected to the i-th node, and the formula is: Among them, d ij is the shortest path from the i-th node to the i-th node; According to the node efficiency, the efficiency E of the entire network is defined as the average value of the sum of the efficiencies of all nodes in the network. The formula is: Step 3-3, define the viscosity CH(i) of the i-th node as the survival probability p(i) of the i-th node and the node degree k i The product of is: CH(i)=p(i)k i (7) Among the nodes that are randomly attacked, the survival probability of each node is: Among the nodes that are intentionally attacked, the survival probability of each node is: The calculation formula of the viscosity CH(i) of the i-th node is: Therefore, the network's anti-destruction metric H is calculated as:

9. The method according to claim 8, characterized in that Step 4 includes: Step 4-1, randomly attack a single node in the network: randomly select one node to attack each time for all nodes in the network until all nodes in the network are removed; Step 4-2, deliberately attack the single node with the largest node degree in the network: each attack selects the node with the largest node degree in the network to attack, and sets to remove one node each time.

10. The method according to claim 9, characterized in that Step 5 includes: constructing the task reliability evaluation index R using the following formula: Among them, ω1, ω2, ω3 are weights, and ω1+ω2+ω3=1.

Images