Target Importance Evaluation Method Based on Complex Network

Abstract

A method for evaluating the importance of targets based on complex networks, comprising: Step 1: Operational capability modeling, that is, quantifying the operational capabilities of targets according to target characteristics; Step 2: Modeling the network topology structure of the operational system, that is, quantifying the role and status of targets in the network according to the operational network topology structure; Step 3: Expansion of the target operational capability model, that is, evaluating targets according to operational intentions, and converting the strike intention into the capabilities that the targets to be struck possess and the role and status of the targets in the operational network; Step 4: Giving specific evaluation scores, that is, setting target constraints and solving them; Combining with other structures effectively avoids the defects of insufficient effectiveness and quantification degree of the existing methods for evaluating operational targets.

Description

Technical Field

[0001] The present invention relates to the technical field of target importance evaluation, and particularly relates to a method for evaluating target importance based on complex networks. Background Art

[0002] The networked combat system is the various weapons, equipment, facilities, forces in joint operations and the connections between them. According to different combat tasks, combat elements are usually dynamically reorganized. Classifying various targets in the combat system into levels according to specific rules, quantifying the target capabilities based on classification and grading, determining the key nodes and vital parts in the combat network, and conducting combat target selection, ranking and evaluation are important bases for performing combat operations such as precision strikes.

[0003] How to discover the key targets of the enemy in the dynamic combat system for strikes, so as to both strike important targets and paralyze the combat system, is an important research topic for improving combat effectiveness. Traditional evaluation methods are to conduct qualitative and quantitative analysis based on expert experience, experimental data and combat cases, establish a threat degree model of the target, so as to realize the value evaluation of the target. However, the general effectiveness and quantification degree of such evaluation methods are insufficient. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method for evaluating target importance based on complex networks, effectively avoiding the defects of insufficient effectiveness and quantification degree in the prior art for evaluating combat targets.

[0005] To overcome the deficiencies in the prior art, the present invention provides a solution for a method for evaluating target importance based on complex networks, specifically as follows:

[0006] A method for evaluating target importance based on complex networks, comprising:

[0007] Step 1: Modeling combat capabilities, that is, quantifying the combat capabilities of the target according to the target characteristics;

[0008] Step 2: Modeling the network topology structure of the combat system, that is, quantifying the role and status of the target in the network according to the combat network topology structure;

[0009] Step 3: Expanding the target combat capability model, that is, evaluating the target according to the combat intention, and converting the strike intention into the capabilities that the target to be struck has and the role and status of the target in the combat network;

[0010] Step 4: Giving specific evaluation scores, that is, setting target constraints and solving.

[0011] Further, the specific steps of Step 1 include:

[0012] The analytic hierarchy process is used to divide the levels of the target system. The specific method is to divide a series of individual targets with the same or similar properties into multiple system levels according to rules such as the functional characteristics, status and role, and location distribution of the targets.

[0013] Further, step 2 specifically includes:

[0014] For each target node, its role in the combat network system can be described by network parameters such as degree centrality, closeness centrality, clustering coefficient, betweenness centrality, and K-Shell core degree.

[0015] Further, the degree centrality D Ci is calculated as shown in the following formula (1):

[0016]

[0017] k i represents the number of existing edges connected to node i, where i is a positive integer and node i is the i-th node;

[0018] N - 1 represents the number of nodes except node i, and N represents the total number of nodes.

[0019] Further, the closeness centrality C Ci is calculated as shown in the following formula (2):

[0020]

[0021] N - 1 represents the number of nodes except node i;

[0022] d vi represents the average shortest distance of a node.

[0023] Further, the betweenness centrality B Ci is calculated as shown in the following formula (3):

[0024]

[0025] represents the number of paths connecting node s and node t, passing through node i, and being the shortest path, where s and t are both positive integers, node s is the s-th node, and node t is the t-th node;

[0026] g st represents the number of shortest paths connecting s and t;

[0027] After normalization, formula (4) can be obtained:

[0028]

[0029] Further, the K-Shell core degree is the K-Shell2 core degree, and the K-Shell2 core degree is calculated by the method of layer-by-layer peeling: First, directly peel off the edge nodes. Since the minimum degree of the network is 1, the K-Shell2 core degree of the peeled-off nodes is 1. Set a variable total and record total = 1. Then peel off the second layer. The minimum degree is still 1, and record total = 1 + 1 = 2, that is, the K-Shell2 core degree of the peeled-off nodes is 2. Then peel off the third layer. The minimum degree is 2, and record total = 2 + 2 = 4, that is, the K-Shell2 core degree of the peeled-off nodes is 4. Finally, the fourth layer can be completely peeled off. The minimum degree is 3, and record total = 4 + 3 = 7, that is, the K-Shell2 core degree of the remaining nodes is 7.

[0030] Further, step 3 specifically includes:

[0031] Combined with the quantitative model of combat capabilities, the capabilities of the targets in the combat network are expanded using degree centrality, closeness centrality, betweenness centrality, and K-Shell2 core degree. Combined with the capability matrix M of the targets, the augmented matrix M+ of the capabilities of the targets is obtained.

[0032] Further, step 4 specifically includes:

[0033] Based on the requirements constraints for modeling, set the number of target strikes as x, and represent the quantitative capabilities of each target as a row vector M k , represent the combat intention as a column vector W, F = M k W to obtain the quantitative score obtained by evaluating the kth target. Then, according to formula (5), take the maximum value of the sum of the evaluation scores of x capability indicators from all targets:

[0034]

[0035] The top x targets with the highest evaluation scores of the overall capability indicators are the solutions sought, where k is a positive integer.

[0036] The beneficial effects of the present invention are:

[0037] The present invention divides a series of individual targets with the same or similar properties into multiple system levels according to rules such as the functional characteristics, status and role, and location distribution of the targets, facilitating better research and understanding of the characteristics and operation laws of combat targets. Additionally, by focusing on joint operation command under informationized conditions, based on the properties of network topology, the K-Shell2 core degree is defined, and a quantization method for the combat capabilities of network nodes is designed, taking into account both the capabilities of the targets themselves and their status and role in the combat network. Combining with the combat intention, the targets are effectively evaluated. When the combat intention changes, the weight values of the capabilities change accordingly, and the ranking of the targets also changes. In practical applications, the connections between targets change dynamically, and the computational complexity increases significantly with the increase in the number of network nodes, the number of target attacks, and the computational amount of overall performance indicators. This effectively avoids the deficiencies in the effectiveness and quantization degree of the existing methods for evaluating combat targets. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a structural diagram of the quantization model for the combat capabilities of the targets of the present invention.

[0039] Figure 2 It is a schematic diagram of the detailed table of the capabilities of a certain airport target of the present invention.

[0040] Figure 3 It is a schematic diagram of the K-shell core degree algorithm.

[0041] Figure 4 It is a schematic diagram of the K-shell2 core degree algorithm of the present invention.

[0042] Figure 5 It is an extended structural diagram of the quantization of the target capabilities of the present invention.

[0043] Figure 6 It is a schematic diagram of the distribution structure of the target system of Party B in the embodiment of the present invention.

[0044] Figure 7 It is a list of the overall capability values of the targets of Party B in the embodiment of the present invention.

[0045] Figure 8 It is an assignment table of the weight relationship of the capability items in the embodiment of the present invention.

[0046] Figure 9 It is a quantization scoring table of the targets of Party B in the embodiment of the present invention.

[0047] Figure 10 It is the comparison intention of the evaluation of the importance of the targets of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] The present invention will be further described below in conjunction with the drawings and embodiments.

[0049] As Figures 1 - 10 shown, the method for evaluating the importance of a target based on a complex network includes:

[0050] Step 1: Operational capability modeling, that is, quantifying the operational capabilities of the target according to the target characteristics;

[0051] Step 2: Modeling the network topology structure of the combat system, that is, quantifying the role and status of the target in the network according to the combat network topology structure;

[0052] Step 3: Expansion of the target operational capability model, that is, evaluating the target according to the combat intention, and converting the strike intention into the capabilities that the target to be struck should have and the role and status of the target in the combat network;

[0053] Step 4: Giving specific evaluation scores, that is, setting target constraints and solving.

[0054] Furthermore, the specific content of Step 1 includes:

[0055] Based on the research on the characteristics of the combat system under complex network conditions, analyze various nodes with complex functions and relationships in the combat system, and construct a target system through means such as target characteristic selection, system level division, and network structure association. Use the analytic hierarchy process to divide the target system levels. The specific method is to divide a series of single targets with the same or similar properties into multiple system levels according to the rules of target functional characteristics, status and role, and location distribution, so as to better study and grasp the target characteristics and operation rules. For example, according to the target functional characteristics, the target system can usually be divided into six categories: reconnaissance and early warning system, command and control system, air defense and antimissile system, fire strike system, information attack and defense system, and integrated support system.

[0056] For example, based on the constructed target system, conduct capability analysis, use the analytic hierarchy process to divide the target capabilities into six major capabilities: reconnaissance and early warning, command and control, air defense and antimissile, fire strike, information attack and defense, and integrated support for quantification, and establish a target operational capability quantification model as Figure 1 shown.

[0057] At the same time, the combat target is composed of sub-targets, and the combat capabilities of the combat target are composed of the combat capabilities of the sub-targets. Taking an airport as an example, this target is composed of sub-targets such as radar stations, communication stations, fighter jets, and runways. Through target decomposition, as Figure 2 shown, the capabilities of this target include air reconnaissance capability, air strike capability, command and control capability, and communication capability, etc.

[0058] Quantitatively assign values to the target capability refinement table, and the capability vector V of this target can be obtained as V = [m 1 , m 2 , …, m n, so as to finally obtain the ability matrix M of all targets in the combat network:

[0059]

[0060] Among them, m 11 , m 21 ...m p1 are the elements of m1, m 12 , m 22 ...m p2 are the elements of m2, m 1n , m 2n ...m pn are the elements of m2. Both p and n are positive integers. Each element in the ability vector V is the ability of the corresponding single target, and each ability of each target constitutes the elements of the ability matrix M.

[0061] For example, the relationship between the target and the ability can be shown in Table 1:

[0062] Table 1

[0063]

[0064] Furthermore, the specific steps of step 2 include:

[0065] The combat system network can also be represented by the graph G=(V, E). V represents the nodes, representing each combat target in the combat network. E represents the directed or undirected edges between the nodes, which not only describes the relationships such as command, support, communication, and cooperation between the nodes, but also can express the dynamic relationships between the flexible combat targets.

[0066] For each target node, its role in the combat network system can be described by network parameters such as degree centrality, closeness centrality, clustering coefficient, betweenness centrality, and K-Shell core degree.

[0067] Furthermore, degree is the most basic concept among the single-node attributes. The degree of a node is the number of other nodes connected to this node, which is a simple but very important value in network analysis. The degree centrality D Ci is calculated as shown in the following formula (1):

[0068]

[0069] k i represents the number of existing edges connected to node i. i is a positive integer, and node i is the i-th node. D Ci is the degree centrality of the ci-th node, and the value of ci is equal to i;

[0070] N - 1 represents the number of nodes except node i, and N represents the total number of nodes.

[0071] Furthermore, the distance between two points in the combat network is defined as the number of edges included in the shortest path connecting the two points. The average path length of the combat network refers to the average distance between all node pairs in the combat network, which reflects the separation degree between nodes in the combat network and the global characteristics of the network. The closeness centrality C Ci The calculation formula is as shown in formula (2) below:

[0072]

[0073] N - 1 represents the number of nodes except node i;

[0074] d vi represents the average shortest distance of a node.

[0075] Furthermore, betweenness includes node betweenness and edge betweenness. Node betweenness refers to the proportion of the number of shortest paths passing through the node in the entire network, and edge betweenness refers to the proportion of the number of shortest paths passing through the edge in the entire network. Betweenness reflects the role and influence of the corresponding node or edge in the entire network and has strong practical significance. For example, in a transportation network, the probability of congestion on roads with higher betweenness is very high; in a communication network, the channels with higher betweenness have the highest usage rate; in a power network, transmission lines and nodes with higher betweenness are prone to danger. The betweenness centrality B Ci The calculation formula is as shown in formula (3) below:

[0076]

[0077] represents the number of paths connecting node s and node t, passing through node i, and being the shortest path. Both s and t are positive integers. Node s is the s - th node, and node t is the t - th node;

[0078] g st represents the number of shortest paths connecting s and t;

[0079] After normalization, formula (4) can be obtained:

[0080]

[0081] Furthermore, Kitsak proposed applying the K - core decomposition algorithm in complex networks in 2010 and found that the K - core decomposition method based on network topology can obtain an evaluation index that more accurately describes the importance of nodes than degree centrality and betweenness centrality, namely the K - shell core degree index. The K - shell decomposition method gives a relatively coarse - grained division of the importance of nodes, and its basic idea is as Figure 2As shown in the figure, the nodes in the network are divided into three layers through K-shell decomposition, corresponding to ks = 1, 2, and 3 respectively. The specific operation is as follows: Just like peeling an eggshell, first define the K-shell value of the edge nodes as 1, and then enter the core of the network layer by layer towards the inside. In the first step, all nodes with a degree of 1 in the network are peeled off, and the edges of these nodes are removed. Then, check whether there are still nodes with a degree of 1 among the remaining nodes. If there are, continue to peel off the nodes until the degrees of all the remaining nodes are greater than 1. The K-shell values of these peeled-off nodes are 1, that is, these nodes are all in the layer with ks value of 1; in the next step, nodes with a degree less than or equal to k (k is an integer, k≥2) and their connecting edges are peeled off in turn until all nodes have corresponding ks values. The K-Shell core degree is the K-Shell2 core degree. As Figure 4 shown, the K-Shell2 core degree is calculated by the method of peeling layer by layer: First, directly peel off the edge nodes. Since the minimum degree of the network is 1, the K-Shell2 core degree of the peeled-off nodes is 1. Set a variable total and record total = 1; then peel off the second layer. The minimum degree is still 1, and record total = 1 + 1 = 2, that is, the K-Shell2 core degree of the peeled-off nodes is 2; then peel off the third layer. The minimum degree is 2, and record total = 2 + 2 = 4, that is, the K-Shell2 core degree of the peeled-off nodes is 4; finally, the fourth layer can be completely peeled off. The minimum degree is 3, and record total = 4 + 3 = 7, that is, the K-Shell2 core degree of the remaining nodes is 7.

[0082] Further, the specific steps of step 3 include:

[0083] Combined with the quantitative model of combat capabilities, the capabilities of the targets in the combat network are expanded using degree centrality, closeness centrality, betweenness centrality, and K-Shell2 core degree. Combined with the capability matrix M of the targets, the augmented matrix M+ of the capabilities of the targets is obtained.

[0084] Obtaining the augmented matrix M+ of the capabilities of the targets is to add the capabilities and effects of the combat network on the basis of the capability matrix M of the targets.

[0085] An example of the augmented matrix M+ of the capabilities of the targets is shown in Table 2:

[0086] Table 2

[0087]

[0088] Further, the specific steps of step 4 include:

[0089] The number of targets that can be struck in each operation is limited, and its combat purpose is to maximize the degree of decline in the capabilities of the combat system. Based on this requirement constraint, a model is established. Set the number of target strikes as x, and represent the quantitative capabilities of each target as a row vector M k, the combat intention is represented as a column vector W, F = M k W obtains the quantization score obtained from the k-th target evaluation, and then takes the maximum value of the sum of the evaluation scores of x ability indicators from all targets according to formula (5):

[0090]

[0091] W represents the weight value of each ability. For example, the air reconnaissance ability accounts for 0.1, the sea reconnaissance ability accounts for 0.2,..., the degree centrality accounts for 0.1, and the KShell2 core degree accounts for 0.1. The sum of the ability values multiplied by the weight values is the score of a target.

[0092] The top x targets with the highest evaluation scores of the overall ability indicators are the solutions, where k is a positive integer.

[0093] The following further illustrates the present invention with specific embodiments:

[0094] Party A needs to strike the targets of Party B with air and missile defense capabilities. Assume the target distribution of Party B is as Figure 6 shown, where targets 1-4 are early warning nodes, targets 5-8 are command and control nodes, targets 9-14 are firepower nodes, and targets 15-24 are communication nodes.

[0095] According to the target classification characteristics and the role status in the combat network, determine the target list ability values of Party B by the method of quantization assignment; according to the combat network topology structure, calculate the network topology ability values of the targets, specifically including degree centrality, closeness centrality, betweenness centrality and KShell2 index; after merging the ability indicators, generate the overall ability value list of the targets as Figure 7 shown.

[0096] According to the combat intention, the ability weight relationship is as Figure 8 shown.

[0097] Construct the combat intention vector W = [1, 1, 1, 1, 0, 0, 0, 0.1] according to the ability item weight relationship assignment table, construct the augmented matrix M based on the target overall ability value list, and through matrix operation Score = MW, the quantization score result of the target can be obtained, as Figure 9 shown.

[0098] Under the guidance of the idea of system disruption, the selection of key targets emphasizes quality over quantity. Under the premise of limited strike costs, following the principle of disrupting the connection of the enemy from command nodes to combat units and from sensors to launchers, key nodes of the enemy's combat system are carefully selected for focused strikes to achieve the goal of destroying nodes, severing links, breaking the network, and paralyzing the battlefield. According to the score ranking, the top 6 targets for strikes are: 8, 14, 13, 15, 19, 20. It can be seen that target No. 8 has the highest level with a command ability of 2 and ranks first in the evaluation score; target No. 14 has the strongest air combat ability and ranks second in the evaluation score; target No. 13 has the second-strongest air combat ability and ranks third in the evaluation score; nodes No. 15, 19, and 20 are in important positions in the combat network and rank high in the strike evaluation.

[0099] It can be seen from Figure 10 that the analytic hierarchy process only gives an evaluation based on the combat ability of the node itself and does not consider the role in the combat system. When the target ability values are close, it is difficult to distinguish. It is also impossible to evaluate the target from the perspective of destroying the combat system.

[0100] By focusing on joint operation command under information-based conditions, the present invention defines the K-Shell2 core degree according to the network topology properties and designs a quantification method for the combat ability of network nodes, taking into account both the ability of the target itself and the status and role of the target in the combat network. Combining with the combat intention, the target is effectively evaluated. When the combat intention changes, the weight value of the ability changes accordingly, and the ranking of the target also changes. In practical applications, the connections between targets will change dynamically, and the computational complexity increases significantly with the increase in the number of network nodes, the number of target disruptions, and the computational volume of the overall performance indicators.

[0101] In modern warfare, when selecting strike targets, it is necessary to consider both the ability of the target itself and the status and role of the target in the combat network system. Based on the analytic hierarchy process, the present invention proposes a quantification method for the hierarchical classification of the abilities of each target under the combat system; based on complex network theory, it defines the K-Shell2 core degree and establishes an importance evaluation model for battlefield targets; with the maximization of the strike target ability as the constraint condition, it solves the key nodes of the combat system, providing theoretical guidance and computational support for target analysis and selection.

[0102] The present invention has been described by way of examples above. It is obvious to those skilled in the art that the present disclosure is not limited to the above-described embodiments, and various changes, alterations, and substitutions can be made without departing from the scope of the present invention.

Claims

1. A method for evaluating the importance of a target based on complex networks, characterized in that, it includes: Step 1: Operational capability modeling, that is, quantifying the operational capabilities of the target according to the target characteristics; Step 2: Modeling the network topology structure of the combat system, that is, quantifying the role and status of the target in the network according to the combat network topology structure; Step 3: Expansion of the target operational capability model, that is, evaluating the target according to the combat intention, and transforming the strike intention into the capabilities that the target to be struck possesses and the role and status of the target in the combat network; Step 4: Giving specific evaluation scores, that is, setting target constraints and solving; The specific content of the said Step 1 includes: Using the analytic hierarchy process to divide the target system levels. The specific method is to divide a series of single targets with the same or similar properties into multiple system levels according to the rules of the functional characteristics, status and role, and location distribution of the targets; The specific content of the said Step 2 includes: For each target node, its role in the combat network system can be described by network parameters such as degree centrality, closeness centrality, clustering coefficient, betweenness centrality, and K-Shell core degree; Degree centrality D Ci The calculation formula is as shown in the following formula (1): k i represents the number of existing edges connected to node i, where i is a positive integer and node i is the i-th node; N - 1 represents the number of nodes except node i, and N represents the number of all nodes; Closeness centrality C Ci The calculation formula is shown in the following formula (2): N - 1 represents the number of nodes except node i; d vi represents the average shortest distance of a node; Betweenness centrality B Ci is calculated as shown in the following formula (3): Denotes the number of paths that connect node s and node t, pass through node i, and are the shortest paths. Both s and t are positive integers. Node s is the s-th node, and node t is the t-th node; g st represents the number of the shortest paths connecting s and t; After normalization, the formula (4) can be obtained: The K-Shell core degree is the K-Shell2 core degree, and the K-Shell2 core degree is calculated by the method of layer-by-layer peeling: First, directly peel off the edge nodes. Since the minimum degree of the network is 1, the K-Shell2 core degree of the peeled-off nodes is 1, and set a variable total, record total = 1; then peel off the second layer, the minimum degree is still 1, record total = 1 + 1 = 2, that is, the K-Shell2 core degree of the peeled-off nodes is 2; then peel off the third layer, the minimum degree is 2, record total = 2 + 2 = 4, that is, the K-Shell2 core degree of the peeled-off nodes is 4; finally, the fourth layer can be completely peeled off, the minimum degree is 3, record total = 4 + 3 = 7, that is, the K-Shell2 core degree of the remaining nodes is 7.

2. The method for evaluating the importance of a target based on complex networks according to claim 1, characterized in that, the specific content of the said Step 3 includes: Combining the quantification model of operational capabilities, expanding the capabilities of the target in the combat network with degree centrality, closeness centrality, betweenness centrality, and K-Shell2 core degree, and combining with the capability matrix M of the target to obtain the augmented matrix M+ of the target's capabilities.

3. The method for evaluating the importance of a target based on complex networks according to claim 1, characterized in that, the specific content of the said Step 4 includes: Modeling based on demand constraints, setting the target strike quantity as x, and representing the quantification ability of each target as a row vector M k , representing the combat intention as a column vector W, F = M k W to obtain the quantified score obtained from the evaluation of the k-th target. Then, according to formula (5), take the maximum value of the sum of the evaluation scores of x ability indicators from all targets: The top x targets with the highest evaluation scores of the overall capability indicators are the solutions sought, where k is a positive integer.

Images