A method for dynamic evolution of command and control network based on hypernetwork

By defining evolutionary behavior and constraints in the command and control network and building a hypernetwork dynamic evolution model, the problem of incomplete description of evolutionary behavior in the existing technology is solved, and more powerful anti-attack ability and network performance improvement is achieved.

CN116132305BActive Publication Date: 2024-11-22LINGNAN NORMAL UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211594441.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-13
Publication Date
2024-11-22
Estimated Expiration
2042-12-13

AI Technical Summary

Technical Problem

The existing command and control network evolution model has incomplete description of evolutionary behavior, which is difficult to reflect the evolutionary differences of different types of combat entities, resulting in insufficient anti-attack ability.

Method used

Adopt the dynamic evolution method of command and control network based on hypernetwork, define the evolution behavior and constraints of nodes and edges, give the evolution rules of different types of nodes and edges, and build a dynamic evolution model of command and control hypernetwork, including adding, deleting and recombining behaviors. , and optimize the network structure through the layer interpolation policy and attribute complementary priority principle.

Benefits of technology

It effectively improves the resistance of the command and control network, enhances the network's attack resistance and coordinated combat capabilities, and significantly improves the performance and stability of the network.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116132305B_ABST
    Figure CN116132305B_ABST
Patent Text Reader

Abstract

The invention provides a command and control dynamic evolution method based on a super network, which belongs to the field of information technology. Firstly, the evolution characteristics of the command and control network are analyzed, and the evolution behavior and constraints are defined; secondly, the evolution rules of different types of nodes and edges are given, the network dynamic evolution steps are proposed, and a command and control super network dynamic evolution model is constructed, which effectively improves the anti-destruction capability of the command and control network and provides a new method for constructing a command and control super network dynamic evolution model.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the field of information technology and relates to a command and control hypernetwork modeling method, in particular to a hypernetwork-based command and control dynamic evolution method. Background Art

[0002] The study of command and control network evolution models helps to reveal the internal operation mechanism and external behavior of the network, which is of great significance to improving the network's anti-destruction performance. The existing evolution models have the shortcomings of incomplete description of evolutionary behavior and difficulty in reflecting the evolutionary differences of different types of combat entities. These shortcomings lead to weak anti-attack capabilities of command and control networks. Summary of the invention

[0003] In order to overcome the shortcomings of the prior art, the present invention provides a command and control network dynamic evolution method based on a super network, analyzes the evolution characteristics of the command and control network, defines the evolution behavior and constraints, gives the evolution rules of different types of nodes and edges, proposes the network dynamic evolution steps, constructs a command and control super network dynamic evolution model, improves the anti-destruction capability of the command and control network, and provides a new method for constructing a command and control super network dynamic evolution model.

[0004] In order to achieve the purpose of the above invention, the technical solution adopted by the present invention is as follows:

[0005] A method for dynamic evolution of a command and control network based on a hypernetwork comprises the following steps:

[0006] (1) Define the node evolution behaviors and constraints of the command and control hypernetwork;

[0007] (2) Give the evolution rules of different nodes and edges;

[0008] (3) Based on the dynamic evolution steps, a command and control super network dynamic evolution model is constructed.

[0009] Furthermore, the node evolution behaviors and constraints of the command and control hypernetwork are defined:

[0010] According to the operational requirements of various nodes in the command and control network, the node evolution behaviors of the command and control super network are divided into three categories: addition behavior, deletion behavior and reorganization behavior.

[0011] Definition 1: Adding behavior. According to the latest combat needs, the behavior of adding new combat units and establishing new connection relationships to strengthen combat capabilities or improve network transmission efficiency, specifically manifested in the addition of new nodes and edges to the network.

[0012] Definition 2: Deletion behavior. The behavior of removing nodes or edges from the command and control network after they are attacked or fail. Specifically, it means deleting nodes and edges.

[0013] Definition 3: Reorganization behavior. During the evolution of the command and control network, the network positions and connection relationships of some nodes no longer meet the latest operational requirements. The behavior of adjusting them is specifically manifested in the deletion and addition of existing links.

[0014] As a typical node heterogeneous network, the evolution behavior of the command and control network is subject to various structural and capability constraints. In order to make the evolution rules effectively reflect the node evolution behavior of the command and control network, the following node evolution behavior constraints are defined:

[0015] Constraint 1: Edge type constraint. According to the OODA loop theory, the information transmission between the perception node and the fire strike node must pass through the command node. Therefore, in the evolution process, the establishment of a direct connection relationship between the perception node and the fire strike node is not considered.

[0016] Constraint 2: Command relationship uniqueness constraint. To ensure the uniqueness of the command, the perception node and the fire strike node can only be directly connected to one command node.

[0017] Constraint 3: Combat link integrity constraint. According to the OODA loop theory, combat effectiveness can only be achieved if a complete combat link is formed. Therefore, in the event of an attack, isolated nodes need to be removed and reorganized.

[0018] Furthermore, the evolution rules of different nodes and edges are given:

[0019] Rule 1: Add new command and control nodes. The command and control network has a strict hierarchical characteristic. As the command level increases, the probability of adding command and control nodes decreases.

[0020] First, to ensure the structural stability of the command and control network, the probability of adding new command and control nodes to different levels is:

[0021]

[0022] In formula (1), l represents the command level to which the newly added command node belongs; n c (l) is the number of command nodes with command level l in the current network; N c The total number of accused nodes in the current network.

[0023] Secondly, establish a connection edge for the newly added accusation node. If the level to which the newly added accusation node is added is the highest level, select the accusation node in the same level with the same probability to establish a collaborative edge. Otherwise, select the accusation node with the probability p. c (i) Select the upper level accusation node to establish a connection. The probability of selecting node i is:

[0024]

[0025] In formula (2), d(i) represents the excess value of the command node i; C(l-1) is the command node with command level l-1 in the current network.

[0026] Rule 2: Add firepower nodes. The purpose of adding firepower nodes is to strengthen the combat capability of the command and control network, or to make up for the lack of current combat capability. The attributes of firepower nodes generally include propulsion capability, fire suppression capability, and air defense capability (k = 1, 2, 3). When a certain attribute k does not meet the requirements, when adding new firepower nodes, priority is given to adding nodes that meet this type of attribute. The node attributes are:

[0027] Att f (k) = min(Att f )+(max(Att f )-min(Att f ))*rand() (3)

[0028] In formula (3), Att f is the original firepower node attribute value, Att f (k) is the attribute value of the newly added firepower node.

[0029] At the same time, the newly added firepower node needs to select a charge node for connection. The layer betweenness connection strategy is adopted. According to the attribute value of the newly added firepower node, the node to establish the charge relationship is selected from the charge nodes with a command level of l. The probability that the charge node i is selected to establish the connection is:

[0030]

[0031] In formula (4), n c is the number of command nodes in the current network; M is the number of command nodes with command level l; Deg(i) and C i are the level and betweenness of the command and control node i respectively; H is the total number of levels of the command and control network.

[0032] Rule 3: Add new sensor nodes. The purpose of adding sensor nodes is to increase the types of intelligence reconnaissance or expand the scope of intelligence reconnaissance, thereby improving the network's grasp of battlefield status and making more accurate command decisions. The attributes of sensor nodes generally include communication intelligence acquisition capabilities, image intelligence acquisition capabilities, and artificial intelligence acquisition capabilities (k = 1, 2, 3, 4). When a certain attribute k does not meet the requirements, when adding new sensor nodes, priority is given to adding nodes that meet this type of attribute. The node attributes are:

[0033] Att s (k) = min(Att s )+(max(Att s)-min(Att s ))*rand() (5)

[0034] In formula (5), Att s is the original perception node attribute value, Att s (K) is the attribute value of the newly added perception node;

[0035] At the same time, according to the attribute value of the newly added perception node, the node to establish the charge relationship is selected from the nodes with the command level l. The probability that the charge node i is selected to establish the edge is:

[0036]

[0037] In formula (6), n c is the number of command nodes in the current network; M is the number of command nodes with command level l; Deg(i) and C i are the level and betweenness of the command and control node i respectively; H is the total number of levels of the command and control network;

[0038] The command and control network must have a strict hierarchy and meet the needs of vertical penetration and horizontal interconnection in a tree-like manner. Therefore, the rules for adding links consider adding collaborative links to improve information sharing capabilities and collaborative command capabilities.

[0039] Rule 4: Add a new collaborative edge for the command node. Use the super-selection method to establish a collaborative edge, share the load of the node itself to the collaborative node, and improve the network's collaborative combat capability. The probability of the command node i being selected is:

[0040]

[0041] d(i) represents the excess value of the accused node i.

[0042] Rule 5: Add firepower node collaborative edges. Based on the attribute complementary priority idea, firepower nodes tend to connect with nodes with different attributes. Then the probability of firepower nodes being selected in the local world Ω is p ff (ij) is:

[0043]

[0044]

[0045] In formula (8) and formula (9), Ω is the set of the same command nodes as the firepower node; Att(i,z) represents the zth attribute value of firepower node i; COC(ij) represents the attribute complementarity between nodes; ε(t) is the unit step function, indicating that the nodes can only establish a collaborative relationship when the attribute value of node j is greater than that of node i.

[0046] Rule 6: Add a new sensor node to the collaborative edge. Similar to Rule 5, select a sensor node from the current sensor subnet to add a collaborative edge. The probability of the sensor node being selected is p ss (ij) is:

[0047]

[0048]

[0049] In Equations 10 and 11, Ω is the set of accusation nodes that belong to the same sensing node; Att(i,z) represents the zth attribute value of sensing node i; COC(ij) represents the attribute complementarity between nodes; ε(t) is a unit step function, indicating that a collaborative relationship can be established between nodes only when the attribute value of node j is greater than that of node i;

[0050] During the network evolution process, the deletion of nodes mainly considers two situations: normal elimination of nodes and elimination of nodes due to enemy attacks.

[0051] Rule 7: Normal elimination of nodes. First of all, consider deleting nodes with poor connectivity and weak capabilities. Randomly select a type of node. If the selected node is a perception or firepower node, the probability of node i being eliminated normally is:

[0052]

[0053] In formula (12), d(i) is the excess value of node i; Att is the attribute value corresponding to the node.

[0054] If the selected node is a charge node, the probability of node i being eliminated normally is:

[0055]

[0056] Where d(i) is the excess value of the accusation node i; layer(i) is the layer number corresponding to the accusation node.

[0057] Rule 8: Nodes are eliminated by enemy attacks. Usually there are two ways: random attacks and deliberate attacks. The probability of node i being deleted is:

[0058]

[0059] In formula (14), d(i) is the excess value of node i; 1 is the intentional attack adjustment factor; 2 is the random attack adjustment factor, and there is λ 1 +λ 2 =1; N is the total number of nodes in the current network.

[0060] In the process of network evolution, there are two cases of edge deletion: active deletion and passive deletion. Active deletion refers to the active deletion of edges due to the adjustment of network organizational structure during network evolution; passive deletion means that after a combat unit is attacked, the edges connected to it will become invalid. Here we only consider the case of active deletion.

[0061] Rule 9: Active deletion of edges. The importance of the edge is measured by the excess values ​​of the nodes at both ends of the edge. The weaker the edge importance, the greater the probability of active deletion, and the probability of the edge being selected for deletion is:

[0062]

[0063] During the network evolution process, due to the deletion of some nodes / edges, isolated nodes may be generated. In this case, the isolated nodes need to be reorganized. Common reorganization methods include level by level, leapfrogging, transfer and upgrade. Since the command and control network has a strict hierarchy system, and the need for coordination between nodes at the same level is taken into account, the present invention only considers the node transfer reorganization method.

[0064] Rule 10: Transfer and reorganization of command nodes. When a command node at a certain level is deleted, the subordinate nodes directly connected to it will lose the command relationship and need to be reorganized. The probability of a node being selected is:

[0065]

[0066] In formula (16), d(i) is the excess value of node i; C i is the betweenness of node i.

[0067] Rule 11: Transfer and reorganization of perception / firepower nodes. If a perception / firepower node v i Because other nodes v c If the failure of the node causes it to lose the charge relationship in the network, it is necessary to i attribute value to determine whether reorganization is needed. If the node v i If the attribute value is higher than the average value of nodes of the same type, then the node v i Reorganize, otherwise remove the node v i . Node v i The connection probability formula is as follows:

[0068]

[0069] In formula (17), the local world Θ is the node v c A collection of other connected sibling nodes.

[0070] Furthermore, a dynamic evolution step is proposed to construct a command and control super network dynamic evolution model:

[0071] Step 1: Initialize the network model. Assume that the total number of initial nodes in the network is N, of which the number of charge nodes is N. c , the number of fire attack nodes is Nf, the number of sensing nodes is N s , the command span is K, the command level is H, and the node attributes are assigned to build the initial model of the command and control super network; at the same time, the current step value of the network evolution is set to t 0 =0, the maximum step value is T.

[0072] Step 2: Set the step value t 0 =t 0 +1, the dynamic evolution of the command and control network executes one of the following four different situations in each evolution step:

[0073] 1) With probability p 1 When a node is added to the network, in order to ensure a balanced node ratio, the probability of joining the command, firepower and perception nodes is Among them, n c 、n f 、n s are the number of command nodes, firepower nodes and perception nodes in the current evolving network, N is the total number of nodes, and n c +nf+n s =N.

[0074] If the added node is a command node, set the command level of the node according to Rule 1 and establish the corresponding command relationship; if the added node is a fire node, set the node attribute value according to Rule 2 and select the command node to establish the command edge; if the added node is a perception node, establish the command edge according to Rule 3.

[0075] 2) With probability p 2 Add new collaborative edges in the network according to rules 4, 5, and 6.

[0076] 3) With probability p 3 Delete a node in the network. If it is in the construction phase of the evolution model, the node is deleted according to Rule 7. If it is in the confrontation phase, the node is deleted according to Rule 8. At the same time, when a node is deleted, there is a certain probability that the node will be reorganized according to Rules 10 and 11.

[0077] 4) With probability p 4 Select edges in the network to delete according to rule 9.

[0078] Step 3: Repeat step 2 until the evolution step value t 0 ≥T, the dynamic evolution process ends.

[0079] Since the existing evolutionary models have the disadvantages of incomplete description of evolutionary behavior and difficulty in reflecting the evolutionary differences of different types of combat entities, a dynamic evolution model of command and control network based on hypernetwork is proposed. Firstly, the evolutionary characteristics of command and control network are analyzed, and the evolutionary behavior and constraints are defined; secondly, the evolutionary rules of different types of nodes and edges are given, the steps of network dynamic evolution are proposed, and the dynamic evolution model of command and control hypernetwork is constructed. Based on this, a command and control dynamic evolution method based on hypernetwork is invented.

[0080] The beneficial effects of the present invention relative to the prior art are:

[0081] A comprehensive analysis of the evolutionary characteristics of the command and control network was conducted, and the evolutionary behavior and constraints were defined; the evolutionary rules for different types of nodes and edges were given; the network dynamic evolution steps were proposed, and a command and control super network dynamic evolution model was constructed, which effectively improved the anti-destruction capability of the command and control network. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 Super-degree distribution for command and control hypernetwork model

[0083] Figure 2 For different p 1 and p 2 The average path length of the network under

[0084] Figure 3 For different p 1 and p 2 The average clustering coefficient of the network under

[0085] Figure 4 is the average path change of the three models;

[0086] Figure 5 The changes of average clustering coefficients of the three models;

[0087] Figure 6 This is the change diagram of clustering coefficient under random attack;

[0088] Figure 7 This is the change diagram of combat hyperlink efficiency under random attack;

[0089] Figure 8 This is the change diagram of the clustering coefficient under deliberate attack;

[0090] Fig. 9 This is a graph showing the change in combat hyperlink efficiency under deliberate attack;

[0091] Fig.10 The figure is an overall flow chart of the method of the present invention. DETAILED DESCRIPTION

[0092] The technical solution of the present invention will be described clearly and completely below 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 creative work are within the scope of protection of the present invention.

[0093] The method for dynamic evolution of a command and control network based on a supernetwork provided by the present invention first analyzes the evolution characteristics of the command and control network and defines the evolution behavior and constraints; secondly, the evolution rules of different types of nodes and edges are given; thirdly, the network dynamic evolution steps are proposed to construct a dynamic evolution model of a command and control supernetwork; finally, a model simulation experiment is designed to analyze the network performance and the dynamic characteristics of evolution in the confrontation stage. The simulation shows that the model has obvious anti-attack capabilities, effectively improves the anti-destruction capability of the command and control network, and provides a new method for constructing a dynamic evolution model of a command and control supernetwork.

[0094] Example 1

[0095] In order to verify the feasibility and effectiveness of the present invention, four groups of experiments were conducted, namely, network feature analysis experiment, parameter selection analysis experiment, network performance comparison analysis experiment of different models and network performance comparison analysis experiment under different attack strategies.

[0096] The model parameters are set as follows: the command level of the initial command and control hypernetwork model is set to 4, the command span is set to 3, the number of command nodes, firepower nodes and perception nodes are set to 121, 400 and 150 respectively, and the evolution step value is set to 200. At the same time, in order to overcome the influence of random factors on the simulation results, the experimental results are taken from the average value of 10 independent simulations.

[0097] (1) Network feature analysis experiment

[0098] The scale-free property of the command and control network reflects the uneven distribution of connections between nodes in the network, which makes a few nodes in the network have a lot of connections, while most nodes have only a few connections. Therefore, the organizational structure characteristics of the network can be intuitively reflected by the excess distribution. 1 , the probability of adding a new edge p 2 The simulation experiment of the model super-degree distribution with the evolution step length is shown in Figure 2. 1 =0.3,p 2 =0.6, p 1 =0.5,p 2 = 0.4 and p 1 =0.7,p 2 = 0.2 and other three groups of parameters, and the evolution step lengths are 100, 500 and 1000 respectively. Figure 1It can be seen that after a certain evolution step, the obtained model excess distribution is approximately a power-law distribution, showing an obvious scale-free characteristic, which illustrates the rationality of the evolution method proposed in the present invention.

[0099] (2) Parameter selection analysis experiment

[0100] Figure 2 , 3 The following are the changes in the average path length and clustering coefficient of the network under different parameter values. It can be observed from the figure that the addition of new nodes increases the scale of the network and the sparsity characteristics of the network gradually emerge; at the same time, the average path length of information transmission between nodes increases and the average clustering coefficient of the network decreases. The increase in edges optimizes the internal organizational relationship of the network structure, improves the information interaction ability and closeness between nodes, thereby shortening the average path length of the network and increasing the average clustering coefficient of the network. Considering the two factors of the average path length of the network and the average clustering coefficient of the network, the parameter p is selected. 1 =0.3,p 2 =0.6, and this parameter is used in the following comparative analysis experiments (3) and (4).

[0101] (3) Comparative analysis experiment of network performance of different models

[0102] In order to further analyze the network performance of the evolution model proposed in the present invention, a comparative analysis was conducted with two evolution models, and the model selection is shown in Table 1 below.

[0103] Table 1 Comparison of command and control network evolution models

[0104]

[0105] Figure 4 , 5The figure shows the changes in the average path length and clustering coefficient of the above three models under the same evolution steps. It can be seen from the figure that the change trends of the three models are the same, and as the number of network evolution steps increases, the average path length of the network shows a downward trend. Although the network scale of the three models remains basically consistent during the evolution process, the change range of M1 is better than that of M2 and M3. This is because the proposed model M1 in the present invention emphasizes the attribute complementarity between nodes when constructing collaborative edges, so that nodes with stronger capabilities have shorter network path lengths as much as possible, thereby improving the collaborative efficiency between nodes, and thus showing better network performance. M2 adopts the principle of super-degree selection. Although it can more effectively improve the network clustering coefficient, as the network scale increases, the average path length between some nodes is greater than the initial value of the network. Due to the principle of combining super-degree and betweenness, M3 has a relatively single node selection on the edge of the inter-layer network, and ignores the hierarchical characteristics of the command and control network, showing the characteristics of "the strong are stronger and the weak are weaker", and the sparsity characteristics of the network are gradually revealed.

[0106] (4) Network performance comparison and analysis experiment under different attack strategies

[0107] In the network confrontation stage, the command and control network will be attacked by the outside world, so the network evolution behavior is mainly the removal of nodes and edges. For the two typical attack methods of random attack and deliberate attack, the removal method can be divided into random removal and preferential removal. The following analyzes the network performance of the three models under random attack and deliberate attack strategies.

[0108] Figure 6 , 7 The changes in the clustering coefficient and combat hyperlink efficiency of the three models under random attacks are shown in the figure. It can be seen from the figure that the three network models all show a slow downward trend under random attacks. This is because the probability of removing each node by random attacks is the same, and the command and control network has an approximate tree structure and inter-layer collaborative structure. In the face of random attacks, the damage to the nodes in the network can mostly be compensated by collaborative edges, and only when the key nodes are hit will there be an accelerated decline. At the same time, compared with M2 and M3, model M1 can maintain a higher clustering coefficient and combat hyperlink efficiency in the face of random attacks. This is because M1 takes into account the node reorganization behavior and reorganizes the isolated nodes, which can reduce the damage to the network model and improve the combat capability of the network model.

[0109] Figure 8 , 9The changes in the clustering coefficient and combat hyperlink efficiency of the three network models under deliberate attacks are shown in Figure 1. Compared with the changes in network performance under random attacks, since deliberate attacks prioritize attacking the key nodes with the largest degree in the network, as the number of attacks increases, the decrease in the clustering coefficient and combat hyperlink efficiency of the three network models is greater than that of random attacks. As can be seen from the above figure, the anti-destruction performance of model M1 under deliberate attacks is significantly better than that of M2 and M3. This is because the super-degree preferential connection adopted by M2 and M3 makes the network evolution process too dependent on key nodes, which can easily cause drastic changes in network performance when facing deliberate attacks. In the evolution process of model M1, not only super-degree is considered, but also some isolated nodes with combat capabilities are reorganized, so that the model can still maintain relatively good network performance when facing deliberate attacks.

[0110] Based on the above simulation analysis, the model proposed in the present invention can maintain relatively good network performance when facing random attacks and intentional attacks, and it also verifies that the proposed evolution method improves the network performance of the command and control dynamic evolution model.

[0111] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.

[0112] References:

[0113] 1. The layer-betweenness edge-linking strategy involved in Rule 2: [Reference: B.Chen, H.Tao, et al. A command-and-control hypernetwork modeling approach based on hierarchy-betweenness edge-linking strategy[J]. JT Supercomputing 2022.].

[0114] 2. Rule 9 involves the reorganization methods of step-by-step, skipping levels, transfer and upgrade: [Literature: Wang Xiaokai, Hou Chaozhen. Network reorganization method of tactical Internet [J]. Firepower and Command and Control, 2005.].

Claims

1. A method for dynamic evolution of command and control network based on hypernetwork, characterized in that: The steps include: S1: Define the node evolution behaviors and constraints of the command and control hypernetwork; S2: Give the evolution rules of different nodes and edges; S3: According to the dynamic evolution steps, a command and control super network dynamic evolution model is constructed; Step S1 is specifically as follows: according to the operational requirements of various nodes in the command and control network, the node evolution behavior of the command and control super network is divided into three categories: addition behavior, deletion behavior and reorganization behavior; the node evolution behavior constraints are defined as three categories: edge type constraints, command relationship uniqueness constraints and combat link integrity constraints, so that the evolution rules can effectively reflect the node evolution behavior of the command and control network; The addition behavior refers to the behavior of strengthening combat capabilities or improving network transmission efficiency by adding new combat units and establishing new connection relationships, which is manifested as the addition of new nodes and new edges to the network; The deletion behavior refers to the behavior of removing a node or edge from the network after the node or edge is attacked or fails, which is manifested as the deletion of nodes and edges. The reorganization behavior refers to the behavior of adjusting the network positions and connection relationships of several nodes when they no longer meet the latest operational requirements, which is manifested in the deletion and addition of existing links; The edge type constraint means that during the evolution process, the direct connection relationship established between the perception node and the firepower node is not considered; The command relationship uniqueness constraint means that the sensing node and the firepower node can only be directly connected to one command node to ensure the uniqueness of the command; The operational link integrity constraint refers to the removal and reorganization of isolated nodes in the event of an attack; The evolution rules described in step S2 include adding a new command node, adding a firepower node, adding a perception node, adding a new command node collaborative edge, adding a new firepower node collaborative edge, adding a new perception node collaborative edge, normal elimination of nodes, elimination of nodes attacked by the enemy, active deletion of edges, transfer and reorganization of command nodes, and transfer and reorganization of perception / firepower nodes; The newly added charging nodes are specifically: First, the probability of adding new accusation nodes to different levels is: In formula (1), l represents the command level to which the newly added command node belongs; n c (l) is the number of command nodes with command level l in the current network; N c The total number of accused nodes in the current network; Secondly, establish edges for the newly added accusation nodes. If the level to which the newly added accusation node is added is the highest level, select the accusation node in the same level with the same probability to establish a collaborative edge; otherwise, select the accusation node with probability p. c (i) Select the upper-level accusation node to establish a connection. The probability of selecting node i is: In formula (2), d(i) represents the excess value of the command node i; C(l-1) is the command node with command level l-1 in the current network; The newly added firepower nodes are specifically: the firepower node attributes include propulsion capability, firepower suppression capability and air defense capability. When a certain attribute k (k=1, 2, 3) does not meet the requirements, when adding new firepower nodes, priority is given to adding nodes that meet this type of attribute. The node attributes are: To f (k)=min(At f )+(max(To f )-min(To f ))*rand() (3) In formula (3), Att f is the original firepower node attribute value, Att f (k) is the attribute value of the newly added firepower node; Adopting the layer betweenness edge strategy, according to the attribute value of the newly added firepower node, the nodes to establish the command relationship edge are selected from the command nodes with the command level l. The probability of the command node i being selected to establish the edge is: In formula (4), n c is the number of command nodes in the current network; M is the number of command nodes with command level l; Deg(i) and C i are the level and betweenness of the command and control node i respectively; H is the total number of levels of the command and control network; The newly added sensing nodes are specifically: the sensing node attributes include communication intelligence acquisition capability, image intelligence acquisition capability and artificial intelligence acquisition capability. When a certain attribute k (k=1, 2, 3, 4) does not meet the requirements, when adding a new sensing node, priority is given to adding a node that meets the attribute. The node attribute is: To s (k)=min(At s )+(max(To s )-min(To s ))*rand() (5) In formula (5), Att s is the original perception node attribute value, Att s (K) is the attribute value of the newly added perception node; According to the attribute value of the newly added perception node, the node to establish the charge-and-control relationship edge is selected from the charge-and-control nodes with the command level l. The probability that the charge-and-control node i is selected to establish the edge is: In formula (6), n c is the number of command nodes in the current network; M is the number of command nodes with command level l; Deg(i) and C i are the level and betweenness of the command and control node i respectively; H is the total number of levels of the command and control network; The newly added coordinated edge of the command node is specifically as follows: the coordinated edge is established by using the super-selective method to share the load of the node itself to the coordinated node to improve the network collaborative combat capability. The probability of the command node i being selected is: d(i) represents the excess value of the accused node i; The newly added firepower node collaborative connection is specifically: based on the attribute complementary priority idea, the firepower node tends to be connected to the node with different attributes, then the probability p of the firepower node being selected in the local world Ω ff (ij) is: In formula (8) and formula (9), Ω is the set of the same command nodes as the firepower node; Att(i,z) represents the zth attribute value of firepower node i; COC(ij) represents the attribute complementarity between nodes; ε(t) is the unit step function, indicating that the nodes can only establish a collaborative relationship if the attribute value of node j is greater than that of node i; The newly added sensor node collaborative edge is specifically: select a sensor node from the current sensor subnet to add a collaborative edge, and the sensor node is selected with a probability p ss (ij) is: In formula (10) and formula (11), Ω is the set of the same accusation nodes as the sensing node; Att(i,z) represents the zth attribute value of sensing node i; COC(ij) represents the attribute complementarity between nodes; ε(t) is the unit step function, indicating that the nodes can only establish a collaborative relationship if the attribute value of node j is greater than that of node i; The normal elimination of the node is specifically as follows: giving priority to deleting nodes with poor connectivity and weak capabilities, randomly selecting a class of nodes, and if the selected node is a sensing node or a firepower node, the probability of node i being normally eliminated is: In formula (12), d(i) is the excess value of node i; Att is the attribute value corresponding to the node; If the selected node is a charge node, the probability of node i being eliminated normally is: In formula (13), d(i) is the excess value of the accusation node i; layer(i) is the layer number corresponding to the accusation node i; The elimination of the node by the enemy attack is specifically: including two ways of random attack and deliberate attack. The probability of node i being deleted is: In formula (14), d(i) is the excess value of node i; λ1 is the intentional attack adjustment factor; λ2 is the random attack adjustment factor, and λ1+λ2=1; N is the total number of nodes in the current network; The active deletion of the edge is specifically as follows: the importance of the edge is measured by using the excess values ​​of the nodes at both ends of the edge. The weaker the importance of the edge, the greater the probability of active deletion. The probability of the edge being selected for deletion is: d(i) is the excess value of node i; The transfer and reorganization of the command node is specifically as follows: when a command node at a certain level is deleted, the subordinate nodes directly connected to it lose the command relationship, and the node needs to be reorganized. The probability of the node being selected is: d(i) is the excess value of node i; C i is the betweenness of node i; The transfer and reorganization of the perception / firepower node is specifically as follows: if a perception / firepower node v i Because other nodes v c The failure of the node v causes it to lose the charge relationship in the network. i attribute value to determine whether reorganization is needed. If the node v i If the attribute value is higher than the average value of nodes of the same type, then the node v i Reorganize, otherwise remove the node v i ; Node v i The connection probability formula is as follows: In formula (17), the local world Θ is the node v c The collection of other connected peer nodes; Step S3 is specifically as follows: S31. Initialize the network model: Assume that the total number of initial nodes in the network is N, of which the number of charge nodes is N c , the number of fire attack nodes is N f , the number of sensing nodes is N s , the command span is K, the command level is H, and the node attributes are assigned to build the initial model of the command and control super network; at the same time, the current step value of network evolution is set to t0 = 0, and the maximum step value is T; S32. Assuming that the step value is t0=t0+1, the dynamic evolution of the command and control network executes one of the following four different situations in each evolution step: 1) Add a node to the network with probability p1. To ensure a balanced ratio of nodes, the probability of adding the command, firepower and perception nodes is Among them, n c 、n f 、n s are the number of command nodes, firepower nodes and perception nodes in the current evolving network, N is the total number of nodes, and n c +n f +n s =N; If the added node is a command node, the command level of the node is set according to the newly added command node evolution rules, and the corresponding command relationship is established; if the added node is a firepower node, the attribute value of the node is set according to the newly added firepower node evolution rules, and the command node is selected to establish an edge; if the added node is a perception node, the edge is established according to the newly added perception node evolution rules; 2) Add new collaborative edges in the network with probability p2 according to the collaborative edge evolution rules of new command nodes, new firepower nodes, and new perception nodes; 3) Delete a node from the network with probability p3; if it is in the construction phase of the command and control super network dynamic evolution model, the node is deleted according to the normal elimination evolution rule of the node; if it is in the confrontation phase, the node is deleted according to the elimination evolution rule of the node under enemy attack; at the same time, when a node is deleted, there is a probability that the node is reorganized according to the transfer and reorganization evolution rule of the command node and the transfer and reorganization evolution rule of the perception / firepower node; 4) Select edges to delete in the network according to the evolutionary rule of active edge deletion with probability p4; S33. Repeat step S32 until the evolution step value t0≥T, ending the dynamic evolution process.