Military target graph expression method

Abstract

The invention relates to a graph expression method for a military target, and the method comprises the following steps: S1, abstracting a military target network into a directed weighted heterogeneous dynamic graph G = (V (t), E (t), W (t)), deeply analyzing the classification basis of the target capability and the target relationship, constructing a target capability system, and achieving the vector expression of a target node and the target relationship; s2, aiming at a military target, constructing a target capability system and carrying out capability analysis to realize vector expression of a target node; s3, target relation features are extracted based on the classification basis of the relation between military targets, and vector expression of the target relation is achieved; s4, in order to accurately reflect the actual situation of the military target network, fusing historical data and real-time information data, dynamically adjusting the feature representation of a target node, constructing a time sequence data matrix of a target changing along with time, and realizing the representation of time sequence data; a military target map model with higher practicability and accuracy is constructed, and powerful support is provided for military command decisions.

Description

Technical Field

[0001] The present invention relates to the technical fields of computer science and military target technology, and particularly to a method for graphically representing military targets. Background Art

[0002] In the modern military field, with the increasingly complex combat environment and the continuous expansion of the combat system, accurately analyzing the relationships between military targets has become crucial. Graphs, as a powerful data structure, with their unique expressive and analytical capabilities, have become a powerful tool for dissecting the relationships between military targets.

[0003] The application of graphs can span multiple fields and effectively capture the complex relationships therein. In the specific scenario of military target analysis, the concept of graphs is cleverly used to construct a military target network. In this network model, each vertex represents a specific military target, which can be a command center, a military base, a combat equipment, or even a combat unit. The edges connecting these vertices are used to represent the intricate relationships between the targets, covering command relationships, support relationships, communication relationships, and coordinated combat relationships, etc. For example, the command relationship between a command center and its subordinate combat units is like an edge in the graph, pointing from the command center to the combat unit, clarifying the direction of command; while the support relationship between a support base and various combat units is reflected by the connection of edges, showing the transmission path of supplies, equipment, etc. Through such an abstract method, the originally intricate relationships between military targets are clearly presented, providing an intuitive and comprehensive perspective for military decision-making.

[0004] Due to the inherently high complexity of the military target network, the target types are rich and diverse, covering multiple fields such as command, combat, support, and intelligence, and the relationships between them are intricate, with both direct associations and indirect influences. In this case, the characteristics of heterogeneous graphs provide a more accurate description tool. The nodes and edges of heterogeneous graphs have multiple types, which highly conforms to the actual situation of the military target network. For example, in a military target network, the nodes can be divided into different types such as command nodes, combat nodes, support nodes, etc., and the edges can also be divided into command edges, communication edges, support edges, etc. Utilizing this characteristic of heterogeneous graphs, the complexity of the military target network can be more accurately characterized. Different types of nodes and edges can carry different attribute information, thus more comprehensively reflecting the relationships between military targets. This not only helps to deeply understand the internal structure of the military target network, but also lays a solid foundation for subsequent analysis using specialized graph neural network methods.

[0005] Based on this, the present invention deeply explores the vector expressions of target nodes and relationships in the construction of military target graphs, fully considering the support for time-series data. Through in-depth analysis of target feature data (including the physical attributes, combat capabilities, strategic values, etc. of targets), associated relationship data (the attributes and intensities of various relationships), and time-series data (information recording the changes of targets and relationships over time), a more practical and accurate military target graph model is constructed. This model can not only accurately reflect the dynamic change process of the military target network, but also provide a solid data foundation and theoretical support for the evaluation of the importance of military targets, facilitating the scientific and precise military command decision-making. Summary of the Invention

[0006] The present invention synthesizes graph theory concepts and abstracts military targets into directed weighted heterogeneous dynamic graphs. This abstract expression fully considers the dynamics, heterogeneity, and weighted nature of the military target network, providing theoretical support for constructing a graph model that accurately reflects the structure of the military target network. Further, through in-depth analysis of target feature data, associated relationship data, and time-series data, a more practical and accurate military target graph model is constructed, aiming to provide strong support for military command decision-making.

[0007] The present invention provides a graph expression method for military targets, including the following steps:

[0008] Step S1: Abstract the military target network into a directed weighted heterogeneous dynamic graph G=(V(t), E(t), M(t)), deeply analyze the classification basis of target capabilities and target relationships, construct a target capability system, realize the vector expression of target nodes and target relationships, and further consider the dynamics of the military target network to construct a time-series data matrix of target changes over time, realizing the expression of time-series data.

[0009] Step S2: For military targets, construct a target capability system and conduct capability analysis to realize the vector expression of target nodes;

[0010] Step S3: Based on the classification basis of the relationships between military targets, extract target relationship features and realize the vector expression of target relationships;

[0011] Step S4: To accurately reflect the actual situation of the military target network, fuse historical data and real-time intelligence data, dynamically adjust the feature representation of target nodes, construct a time-series data matrix of target changes over time, and realize the expression of time-series data.

[0012] The said Step S2 includes the following steps:

[0013] Step S21: Based on the evaluation method from the capability perspective in the US Department of Defense Architecture Framework (DoDAF), divide the target capability system, classify it according to factors such as functional attributes, strategic status, effect, and geographical deployment, integrate it into a multi-level system, and deepen the understanding of the target characteristics and their operating laws.

[0014] Step S22: According to the target functional characteristics, the target capability system is often subdivided into six major capability categories: early warning and reconnaissance, information transmission, command and control, strike and destruction, battlefield protection, and logistics support. The main capability indicators are as follows:

[0015] Early warning and reconnaissance capability: As a key part of the military combat system, it covers anti-submarine detection, sea detection, space-based detection, and air detection. The anti-submarine detection range determines the early warning scope; sea detection is measured by detection range, angle, maximum early warning time, the number and type of tracked targets, and information capacity; space-based detection depends on image resolution and multi-source information reception capabilities; air detection's detection range, angle, maximum early warning time, the number and type of tracked targets, and information capacity affect air defense combat effectiveness.

[0016] Information communication capability: It is the "nerve center" of the military system, measured from four dimensions: communication equipment type (optical fiber, digital microwave, short wave, etc.), support unit type (command agency, reconnaissance node, etc.), the number of support units, and communication station level (core, edge).

[0017] Command and control capability: It dominates military operations and is composed of command agency level (strategic group, campaign corps, etc.), force command and control capability (the number of commanded troops, the number of subordinate agencies, etc.), and deployment location level (from level one to level six).

[0018] Strike and damage capability: It directly determines the outcome of the battle and includes air defense and anti-missile, land strike, anti-ship, anti-submarine, and information suppression capabilities. Among them, air defense and anti-missile capability includes anti-missile combat capability and air defense system combat capability; land strike capability covers air force's land combat capability and ground combat unit's land combat capability; anti-ship capability includes air force's anti-ship combat capability, ground combat unit's anti-ship combat capability, and surface combat unit's anti-ship combat capability; anti-submarine capability involves air force's anti-submarine combat capability and surface combat unit's anti-submarine combat capability (ships + submarines); information suppression capability: It is measured by indicators such as the action distance of electromagnetic interference, the number of platforms, platform mobility, interference power, and target categories that can be interfered.

[0019] Battlefield protection capability: It ensures the survival and continuous combat of military forces and includes defense capabilities (air defense and anti-missile, armor protection, etc.), mobility capabilities (ground, air, sea mobility), and anti-destruction capabilities.

[0020] Support and guarantee capability: It includes energy, ammunition, maintenance, and medical support.

[0021] Step S23: Construct a vector representation of the target capabilities. Using the method based on capability feature extraction, first extract various capability features from the target, and then convert them into numerical vectors. Let the target node vector be According to the above division of the target capability system and the description of each capability index, the mathematical form of the target node vector can be expressed as:

[0022]

[0023] where, V 预警侦察 , V 信息通信 , V 指挥控制 , V 打击毁伤 , V 战场防护 , V 支援保障 is composed of the corresponding indexes. The target node vector describes the capability features of military targets in the form of numerical vectors by integrating the indexes of each capability category.

[0024] Furthermore, the vector expression of military target relationships is to convert the relationships between targets into the form of numerical vectors. Each target can be represented by a vector, and each dimension of the vector can correspond to different features or attributes. The relationships between targets can also be expressed by a vector.

[0025] Therefore, step S3 includes the following steps:

[0026] Step S31: Target relationship classification. According to the classification basis of target relationships, classify the relationships between targets into relationships such as command, support, communication, cooperation, backup, repair, succession, concealment, self-connection, etc.

[0027] Step S32: Target relationship expression. There are zero or more relationships between targets. If there is no connection relationship between targets, it is 0. If there is, it is the level of connection. The level range is 0 - 10 to express the strength of the relationship. The relationship between targets can be expressed as E = [e1, e2, e3, e4, e5, e6, e7, e8, e9].

[0028] Step S33: Construct a vector representation of the target relationships. Using the method based on capability feature extraction, first extract various features from the target, and then convert them into numerical vectors.

[0029] Furthermore, military operations are full of dynamics and uncertainties. The battlefield situation changes rapidly, and the states, capabilities, and mutual relationships of military targets are constantly changing. To accurately reflect the actual situation of the military target network, fuse historical data and real-time intelligence data, dynamically adjust the feature representation of target nodes, and capture the real-time changes of target capabilities.

[0030] Therefore, step S4 includes the following steps:

[0031] Step S41: Clearly define the states. Use "on" to indicate that the target is enabled and "off" to indicate that it is not enabled, which facilitates quick judgment of the target state and provides accurate information for subsequent analysis and military operation planning.

[0032] Step S42: Precisely set the target timestamps. Each target is represented by a vector that contains information such as the target number, timestamp, and status information. For example, for target A, the vector at a certain moment records its number and status at that time point. In this way, the state changes of the target at different times can be tracked to understand its development trend.

[0033] Step S43: Efficiently implement the dynamic update of target capabilities. When the target capabilities change, the original target vector remains unchanged, and a new vector recording the new capabilities is added and marked with a new timestamp. If the target is destroyed, disappears, or reappears, its status is changed in a timely manner and a new timestamp is marked. Through continuous updates, the model can always reflect the actual capabilities of the target.

[0034] Step S44: Accurately determine the edge timestamps: The relationships between military targets are represented by edges, and each edge is also represented by a vector that contains the edge number and timestamp. In this way, the status and changes of the edges can be tracked to help analyze the dynamic changes in the military target network structure.

[0035] Step S45: Carefully perform the dynamic update of the relationships between targets (i.e., edges): Once the relationships between targets change, the original edge vector remains unchanged, and a new edge vector is added and marked with a new timestamp. If a relationship disappears, reappears, or a new relationship is added, the relevant status should be changed in a timely manner and a new timestamp should be marked. This can enable the model to accurately reflect the complex relationship changes between targets.

[0036] Step S46: Intelligently and dynamically adjust the graph: Introduce the concept of a time window and combine it with a graph neural network. Within the time window, continuously monitor the changes in the capabilities and relationships of the targets. Once there are new changes, the graph neural network will adjust the nodes and edges of the graph. For example, if a target's capabilities improve, the importance of the nodes related to it will change; if there are changes in the relationships between targets, the connection status of the edges will also be adjusted. Through this dynamic adjustment, the model can more accurately reflect the changes in the military situation and provide timely and accurate support for military decision-making.

[0037] Step S47: Use the time window to obtain the latest data.

[0038] Target node data target feature vector: For each military target i, at time point t j ∈[t now -Δt,t now , its status and capabilities are represented by the feature vector is described as follows, where represents the k-th attribute value of target i at time point t j .

[0039] Data of target nodes within the time window: In the latest time window, the time-series data of target i is That is, all now satisfying t j -Δt ≤ t now ≤ t . The data edge vector of the edge: The relationship between military targets is represented by an edge. Each edge e pq connects target p and target q and is represented by the vector , where is the edge number, is the timestamp.

[0040] Data of the edge within the time window: In the latest time window, the time-series data of edge e pq is That is, all now satisfying t j -Δt ≤ t now ≤ t . When the relationship between target p and target q changes within the latest time window, a new edge vector appears, recording the new relationship status and timestamp.

[0041] Obtaining the latest target graph structure data Target graph structure: The target graph G = (V, E). In the latest time window, the target graph structure data is where

[0042] Update rule Target node update: If the ability of target i changes at time point t j+1 ∈ [t now -Δt, t now , a new feature vector is added and incorporated into , and at the same time, the feature of node i in the target graph is updated. Edge update: If the relationship between target p and target q changes at time point t j+1 ∈ [t now -Δt, t now , a new edge vector is added and incorporated into , and at the same time, the status of edge e in the target graph pq is updated;

[0043] Temporal Data Representation Method for Dynamic Assessment of Military Targets:

[0044] The dynamic feature vector and the temporal data matrix target core define the time series, the time axis of military operations, used to capture the dynamic evolution of the target state:

[0045] T = {t1, t2, …, t n}, t1 < t2 < … < t n (2-1)

[0046] The target feature vector, an m-dimensional vector, used to describe various attributes of target i at time point t j including firepower intensity, defense ability:

[0047]

[0048] The temporal data matrix, arranging the feature vectors of target i at each time point in chronological order to form a matrix to reflect the evolution of the target's capabilities over time:

[0049]

[0050] The relationship edge core defines the edge vector representation:

[0051]

[0052] The edge connecting targets p and q, ID pq is the unique number of the edge, t j is the timestamp, indicating the k-th relationship attribute between targets p and q, such as communication intensity, degree of coordinated operation, etc. Relationship attribute examples: Indicates the communication link strength, and the value range can be [0,1], where 0 means no communication and 1 means the strongest communication, Indicates the tightness of coordinated operation, quantified according to the frequency and effect of joint operations,

[0053] The dynamic graph model and the real-time update mechanism time window technology definition:

[0054] Δt = t now -t now-Δt (2-5)

[0055] Set a time window length Δt to obtain the latest battlefield data, such as information data screening rules within the current 1 hour:

[0056] Target node:

[0057]

[0058] Relationship edge:

[0059]

[0060] Only retain the valid data of the target nodes and relationship edges within the time window to ensure the real-time performance of the model.

[0061] Dynamic update rule Target node update: New

[0062]

[0063] When the ability of target i changes at the new time point t j+1 Add the corresponding feature vector to the existing data and retain the historical data trajectory.

[0064] Relationship edge update:

[0065]

[0066] When the relationship between targets p and q changes at the new time point t j+1 Add the corresponding edge vector to the existing data to track the dynamic changes of the relationship.

[0067] Graph structure adjustment Dynamic graph model:

[0068] G Δt =(V Δt , E Δt ) V Δt (2-10)

[0069] The set of target nodes existing within the time window (such as undestroyed target nodes), E Δt : The set of relationship edges existing within the time window (such as currently valid communication or cooperation relationship edges). Military significance: When a target is destroyed Remove the node from the node set V Δt ; When a new relationship (such as a communication or cooperation combat relationship) is established, add the corresponding edge to the edge set E Δt .

[0070] An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the method for graph expression of military targets when executing the program.

[0071] A computer-readable storage medium, on which computer instructions are stored, and when the computer instructions are executed by a processor, the method for graph expression of military targets is implemented.

[0072] Compared with the prior art, the advantages of the present invention are as follows:

[0073] 1. Systematic ability modeling

[0074] Based on the DoDAF framework, a target capability system covering six dimensions such as early warning and reconnaissance, information communication, and command and control is constructed. Through 16 - 75 sub - indicators, the multi - dimensional quantitative expression of military targets is realized, breaking through the limitations of traditional single - attribute modeling.

[0075] 2. Multi - relationship dynamic modeling

[0076] Nine types of military target relationships such as command, support, and communication are defined. An innovative 0 - 10 level relationship strength quantification system is introduced to accurately depict complex battlefield relationships. In particular, the introduction of self - connection relationships provides a new mechanism for node self - attention calculation.

[0077] 3. Spatiotemporal dynamic expression

[0078] A time - series data matrix containing timestamps, status information, and feature vectors is constructed. Through time - window technology, real - time screening and updating of battlefield data are realized. The dynamic update mechanism of node capabilities and edge relationships (retaining historical trajectories + adding current states) ensures that the model always reflects the latest battlefield situation.

[0079] 4. Intelligent dynamic adjustment

[0080] Combined with graph neural networks and time - window technology, automatic optimization of the target graph structure is realized. When node capabilities change or relationship strengths are adjusted, the model can automatically update node importance assessment and edge connection status, significantly improving the timeliness of battlefield situation analysis.

[0081] 5. Support for actual combat applications

[0082] The proposed dynamic graph model can be directly mapped to the military command and decision - making process, supporting rapid judgment of target status (on / off status), historical trajectory backtracking, and future trend prediction. Through the integration scheme of electronic devices and storage media, a practical technical architecture is provided for battlefield command information systems.

[0083] 6. Data fusion innovation

[0084] The organic fusion of historical data and real - time intelligence is realized. By means of vector superposition, both the evolution trajectory of target capabilities is retained and the current combat effectiveness is highlighted. This incremental update strategy effectively reduces the computational complexity while ensuring data continuity.

[0085] 7. Military application value

[0086] It provides a unified mathematical framework for military target importance assessment, combat mission planning, strike effect prediction, etc. In particular, the construction of the time - series data matrix provides key technical support for dynamic assessment of combat effectiveness and optimization of command and decision - making.

[0087] This method significantly improves the accuracy and timeliness of military target network analysis through systematic modeling, multi-dimensional relationship quantification, and dynamic update mechanisms, providing an innovative solution for complex battlefield situation awareness in modern warfare. Detailed implementation manners

[0088] The embodiments of the technical solution of the present invention will be described in detail below with reference to the attached tables. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and therefore are only examples and cannot be used to limit the protection scope of the present invention. It should be noted that unless otherwise specified, the technical terms or scientific terms used in this application should have the ordinary meanings understood by those skilled in the art to which the present invention belongs. The present invention will be further described in detail below with reference to the attached tables and specific embodiments.

[0089] Embodiment: A graph expression method for military targets includes the following steps:

[0090] Step S1: Abstract the military target network into a directed weighted heterogeneous dynamic graph G=(V(t), E(t), W(t)), deeply analyze the classification basis of target capabilities and target relationships, construct a target capability system, realize the vector expression of target nodes and target relationships, and further consider the dynamics of the military target network, construct a time-series data matrix of targets changing over time, and realize the expression of time-series data.

[0091] Step S2: For military targets, construct a target capability system and conduct capability analysis to realize the vector expression of target nodes;

[0092] Step S3: Based on the classification basis of the relationships between military targets, extract target relationship features and realize the vector expression of target relationships;

[0093] Step S4: To accurately reflect the actual situation of the military target network, fuse historical data and real-time intelligence data, dynamically adjust the feature representation of target nodes, construct a time-series data matrix of targets changing over time, and realize the expression of time-series data.

[0094] For the specific embodiment:

[0095] Step 1: Construct a target capability system and conduct capability analysis to realize the vector expression of target nodes

[0096] Step (1): Based on the evaluation method from the capability perspective in the Department of Defense Architecture Framework (DoDAF) of the United States, divide the target capability system, classify it according to factors such as functional attributes, strategic status, effect of action, and geographical deployment, and integrate it into a multi-level system to deepen the understanding of target characteristics and their operation rules.

[0097] Step (2): According to the target functional characteristics, the target capability system is often subdivided into six major capability categories: early warning and reconnaissance, information transmission, command and control, strike and destruction, battlefield protection, and logistics support. The main capability indicators are as follows:

[0098] Early warning and reconnaissance capability: As a key part of the military combat system, it covers anti-submarine detection, sea detection, space-based detection, and air detection. The anti-submarine detection range determines the early warning scope; sea detection is measured by detection range, angle, maximum early warning time, the number, type, and information capacity of tracked targets; space-based detection depends on image resolution and multi-source information reception capabilities; the detection range, angle, maximum early warning time, the number, type, and information capacity of air detection affect the air defense combat effectiveness. There are 16 specific indicators, as shown in Table 1.

[0099] Table 1 Early warning and reconnaissance capability indicator table

[0100]

[0101]

[0102] Information communication capability: It is the "nerve center" of the military system, measured from four dimensions: communication equipment type (optical fiber, digital microwave, short wave, etc.), support unit type (command agency, reconnaissance node, etc.), number of support units, and communication station level (core, edge). There are 11 indicators in total, as shown in Table 2.

[0103] Table 2 Information communication capability indicator table

[0104]

[0105]

[0106] Command and control capability: It dominates military operations and is composed of command agency level (strategic group, campaign corps, etc.), force command and control capability (number of commanded troops, number of subordinate agencies, etc.), and deployment location level (from level one to level six). There are 15 indicators in total, as shown in Table 3.

[0107] Table 3 Command and control capability indicator table

[0108]

[0109] Strike and damage capability: It directly determines the outcome of the battle and includes air defense and antimissile, land strike, anti-ship, anti-submarine, and information suppression capabilities. There are 75 indicators in total. Due to the large number of indicators, they will not be presented in the form of a detailed table, and only the indicator composition of the strike and damage capability will be specifically described.

[0110] Among them, the air defense and anti-missile capabilities include anti-missile combat capabilities and air defense system combat capabilities. The anti-missile combat capabilities are determined by indicators such as the maximum interception radius, maximum interception altitude, total number of missile targets that can be struck, warhead power, kill radius, number of launch units (launchers, launch vehicles), interception speed, number of missiles, and interception reaction time; the air defense system combat capabilities are further divided into the air defense combat capabilities of ground combat units, aviation air defense combat capabilities, and naval ship air defense combat capabilities. The indicators of the air defense combat capabilities of ground combat units are similar to some of the indicators of anti-missile combat capabilities; the aviation air defense combat capabilities are determined by the type of aircraft, number of aircraft, endurance, maximum operating radius, maximum weapon range, total ammunition load, maximum number of missiles carried (air-to-air missiles), maximum warhead power (i.e., kill radius), maximum number of air weapons, and maximum number of sorties per day; the indicators of the naval ship air defense combat capabilities also overlap with some of the indicators of anti-missile combat capabilities.

[0111] The land strike capabilities cover the land combat capabilities of aviation and the land combat capabilities of ground combat units. The land combat capabilities of aviation are measured by indicators such as the type of aircraft, number of aircraft, endurance, maximum operating radius, maximum weapon range, total ammunition load, maximum number of missiles carried (air-to-air missiles), maximum warhead power (i.e., kill radius), maximum number of air weapons, maximum number of sorties per day, available forces, and total number of air targets that can be struck; the land combat capabilities of ground combat units are determined by indicators such as range, total number of land targets that can be struck, maximum warhead power (warhead mass), number of launch units (cruise missiles, artillery, tanks, and armored vehicles), and number of missiles.

[0112] The anti-ship capabilities include the anti-ship combat capabilities of aviation, the anti-ship combat capabilities of ground combat units, and the anti-ship combat capabilities of surface combat units. The indicators of the anti-ship combat capabilities of aviation are the same as some of the indicators of the land combat capabilities of aviation, plus the total number of maximum surface targets of available forces; the anti-ship combat capabilities of ground combat units are determined by the flight altitude range, range, firing arc, total number of surface targets that can be struck, maximum warhead power (i.e., kill radius), number of launch units (launchers, launch vehicles), and number of missiles; the indicators of the anti-ship combat capabilities of surface combat units (bases and ships) are similar to those of the anti-ship combat capabilities of ground combat units.

[0113] The anti-submarine capabilities involve the anti-submarine combat capabilities of aviation and the anti-submarine combat capabilities of surface combat units (ships + submarines). The anti-submarine combat capabilities of aviation are measured by indicators such as the number of aircraft, endurance, maximum operating radius, maximum weapon strike range, maximum weapon strike depth, maximum number of missiles carried (depth charges, mines), maximum number of sorties per day, available forces, and total number of surface targets that can be struck; the anti-submarine combat capabilities of surface combat units are determined by indicators such as range, maximum number of underwater targets that can be struck, combat endurance, single-shot kill probability, rapid response ability, and strike depth.

[0114] Information suppression ability: Measured by indicators such as the action distance, number of platforms, platform mobility, interference power, and target categories that can be interfered with of electromagnetic interference.

[0115] Battlefield protection ability: Ensuring the survival and continuous combat of military forces, including defensive capabilities (such as air defense and antimissile, armor protection, etc.), mobility capabilities (ground, air, and sea mobility), and anti-destruction capabilities, with a total of 3 indicators, as shown in Table 4.

[0116] Table 4 Battlefield survival ability indicator table

[0117]

[0118] Support and guarantee ability: The "logistical lifeline" of military operations, including energy, ammunition, maintenance, and medical support, with a total of 4 indicators, as shown in Table 5.

[0119] Table 5 Support and guarantee ability indicator table

[0120]

[0121] Step (3): Construct the vector representation of the target ability. Using the method based on ability feature extraction, first extract various ability features from the target, and then convert them into numerical vectors. Let the target node vector be According to the above division of the target ability system and the description of each ability indicator, the mathematical form of the target node vector can be expressed as:

[0122]

[0123] Among them, V 预警侦察 , V 信息通信 , V 指挥控制 , V 打击毁伤 , V 战场防护 , V 支援保障 Consisted of the corresponding indicators. The target node vector Describes the ability characteristics of the military target in the form of a numerical vector by integrating the indicators of each ability category.

[0124] Step two: Extract the target relationship features and realize the vector expression of the target relationship

[0125] Step (1): Target relationship classification. According to the classification basis of the target relationship, the relationships between targets are divided into relationships such as command, support, communication, cooperation, backup, repair, replacement, concealment, self-connection, etc.

[0126] Command relationship: The command relationship is a hierarchical guiding and controlling connection existing between military targets. The superior military target has the power to issue commands, make decision-making plans, and conduct operation scheduling for the subordinate military target to ensure that military operations proceed in an orderly manner according to the established strategies and tactics.

[0127] Support relationship: The support relationship is reflected in the association of resource supply and support between military targets to maintain combat capabilities and operation sustainability. It includes support in aspects such as materials, equipment, personnel, and technology. The support source target provides necessary conditions for the supported target to ensure its normal operation and task execution.

[0128] Communication relationship: The communication relationship refers to the connection established between military targets to achieve information transmission, exchange, and sharing. Through various communication means and channels, different military targets can communicate intelligence, commands, and combat-related information in a timely manner to ensure the smooth flow of information to support coordinated operations and decision-making.

[0129] Cooperation relationship: The cooperation relationship is the association of mutual cooperation and coordinated actions between military targets to achieve common military tasks or goals. Different targets give play to their respective advantages in combat operations to jointly complete tasks, involving aspects such as action coordination, resource sharing, and task division and cooperation.

[0130] Backup relationship: The backup relationship is the backup alternative connection established between military targets to cope with emergencies or the failure of the main target. When the main target cannot work properly, the backup target can quickly take over its function or task to ensure the continuity and stability of military operations.

[0131] Repair relationship: The repair relationship is the association established for damaged military targets to restore their functions and performance. The target with repair capabilities conducts operations such as maintenance, restoration, and reconstruction on the damaged target to make it resume use as soon as possible or return to a better combat state.

[0132] Replacement relationship: The replacement relationship is that during military operations, when a military target cannot continue to perform its tasks for various reasons, another target can seamlessly take over its tasks and responsibilities and continue to advance the process of military operations to ensure that the tasks are not interrupted.

[0133] Concealment relationship: The concealment relationship is a special association formed between military targets to achieve their own concealment, protection, and reduction of the probability of being discovered by the enemy. By taking camouflage, concealment measures, and coordinated concealment actions, military targets are difficult to be detected under enemy reconnaissance, improving their survival capabilities and combat surprise.

[0134] Self-connection relationship: The self-connection relationship is the connection established between the various components or functional modules within a military target itself to achieve self-management, coordination, and optimization. It enables the military target to conduct information transmission, resource allocation, and function adjustment internally to maintain its stable operation and high performance. The model uses self-connection to calculate edge attention.

[0135] Step (2): Expression of target relationships. There are zero or more relationships between targets. If there is no connection relationship between targets, it is 0; if there is, it is the level of the connection. The level ranges from 0 to 10 to express the strength of the relationship. The relationship between targets can be expressed as E = [e1, e2, e3, e4, e5, e6, e7, e8, e9], as shown in Table 6 specifically.

[0136] Table 6 Target Relationship Table

[0137]

[0138]

[0139] Step 3: Integrate historical data and real-time intelligence data, dynamically adjust the feature representation of target nodes, construct a time-series data matrix of targets changing over time, and achieve the expression of time-series data

[0140] Step (1): Clearly define the status, use "on" to indicate that the target is enabled, and "off" to indicate that it is not enabled, which is convenient for quickly judging the target status and providing accurate information for subsequent analysis and military operation planning.

[0141] Step (2): Accurately set the target timestamp. Each target is represented by a vector, which contains information such as the target number, timestamp, and status. For example, for target A, the vector at a certain moment records its number and status at this time point. In this way, the status changes of the target at different times can be tracked, and its development trend can be understood.

[0142] Step (3): Efficiently implement the dynamic update of target capabilities. When the target capabilities change, the original target vector remains unchanged, and a new vector recording the new capabilities is added and marked with a new timestamp. If the target is destroyed, disappears, or reappears, its status is changed in a timely manner and a new timestamp is marked. Through continuous updates, the model can always reflect the actual capabilities of the target.

[0143] Step (4): Accurately determine the edge timestamp: The relationship between military targets is represented by edges, and each edge is also represented by a vector, which contains the edge number and timestamp. In this way, the status and changes of the edges can be tracked, helping to analyze the dynamic changes in the network structure of military targets.

[0144] Step (5): Conduct a detailed dynamic update of the relationships between targets (i.e., edges): Once the relationship between targets changes, the original edge vector remains unchanged, a new edge vector is added and a new timestamp is marked. If a relationship disappears, reappears, or a new relationship is added, the relevant status should be changed in a timely manner and a new timestamp should be marked. This enables the model to accurately reflect the complex relationship changes between targets.

[0145] Step (6): Intelligently and dynamically adjust the graph: Introduce the concept of a time window and combine it with a graph neural network. Within the time window, continuously monitor the changes in the capabilities and relationships of targets. Once there are new changes, the graph neural network will adjust the nodes and edges of the graph. For example, if the capability of a target improves, the importance of the nodes related to it will change; if there are changes in the relationships between targets, the connection status of the edges will also be adjusted. Through this dynamic adjustment, the model can more accurately reflect the changes in the military situation and provide timely and accurate support for military decision-making.

[0146] Step (7): Use the time window to obtain the latest data.

[0147] Target node data - target feature vector: For each military target i, at time point t j ∈[t now -Δt, t now , its status and capabilities are described by the feature vector where represents the k-th attribute value of target i at time point t j .

[0148] Target node data within the time window: In the latest time window, the time-series data of target i is That is, all now t - Δt ≤ t j ≤ t now of the set. Edge data - edge vector: The relationship between military targets is represented by an edge. Each edge e pq connects target p and target q and is represented by the vector where is the edge number, is the timestamp.

[0149] Edge data within the time window: In the latest time window, the time-series data of edge e pq is That is, all now t - Δt ≤ t j ≤ t now of the set. When the relationship between target p and target q changes within the latest time window, there will be a new edge vector Appear, record the new relationship status and timestamp.

[0150] Obtain the latest target graph structure data. Target graph structure: The target graph G=(V, E). Within the latest time window, the target graph structure data is where

[0151] Update rule. Target node update: If the ability of target i changes at time point t j+1 ∈[t now -Δt, t now , a new feature vector is added and added to , and at the same time, the feature of node i in the target graph is updated. Edge update: If the relationship between target p and target q changes at time point t j+1 ∈[t now -Δt, t now , a new edge vector is added and added to , and at the same time, the status of edge e in the target graph pq is updated.

[0152] Finally, it should be noted that: The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit them; Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: They can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; And these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the description of the present invention.

Claims

1. A graphical representation method for military targets, characterized in that, The method includes the following steps: Step S1: Abstract the military target network into a directed weighted heterogeneous dynamic graph G=(V(t), E(t), W(t)), deeply analyze the classification basis of target capabilities and target relationships, construct a target capability system, realize the vector expression of target nodes and target relationships, and further consider the dynamics of the military target network, construct a time-series data matrix of targets changing over time, and realize the expression of time-series data. Step S2: For military targets, construct a target capability system and conduct capability analysis to realize the vector expression of target nodes. Step S3: Based on the classification basis of the relationships between military targets, extract target relationship features and realize the vector expression of target relationships. Step S4: To accurately reflect the actual situation of the military target network, fuse historical data and real-time intelligence data, dynamically adjust the feature representation of target nodes, construct a time-series data matrix of targets changing over time, and realize the expression of time-series data.

2. The graphical representation method of military targets according to claim 1, wherein Step S2 includes the following steps: Step S21: Based on the evaluation method from the capability perspective in the Department of Defense Architecture Framework (DoDAF) of the United States, divide the target capability system, classify it according to functional attributes, strategic status, effect of action, and geographical deployment factors, and integrate it into a multi-level system to deepen the understanding of target characteristics and their operation rules. Step S22: According to the functional characteristics of the target, the target capability system is often subdivided into six major capability categories: early warning and reconnaissance, information transmission, command and control, strike and destruction, battlefield protection, and logistics support. The main capability indicators are as follows: Early warning and reconnaissance capability, information communication capability, command and control capability, strike and damage capability, battlefield protection capability, support and guarantee capability. Step S23: Construct a vector representation of the target capabilities. Using a method based on capability feature extraction, first extract various capability features from the target, and then convert them into numerical vectors. Let the target node vector be According to the above division of the target capability system and the description of each capability index, the mathematical form of the target node vector can be expressed as: Among them, V 预警侦察 , V 信息通信 , V 指挥控制 , V 打击毁伤 , V 战场防护 , V 支援保障 is composed of corresponding indicators, and the target node vector describes the ability characteristics of military targets in the form of a numerical vector by integrating the indicators of each ability category.

3. The graphical representation method of military targets according to claim 1, characterized in that The vector expression of military target relationships is to transform the relationships between targets into the form of numerical vectors. Each target is represented by a vector, and each dimension of the vector corresponds to different features or attributes. The relationships between targets are expressed by a vector. Step S3 includes the following steps: Step S31: Target relationship classification. According to the classification basis of target relationships, divide the relationships between targets into relationships such as command, support, communication, cooperation, backup, repair, succession, concealment, self-connection, etc. Step S32: Target relationship expression. There are zero or more relationships between targets. If there is no connection relationship between targets, it is 0. If there is, it is the connection level, and the level range is 0-10 to express the strength of the relationship. The relationship between targets is expressed as E=[e1, e2, e3, e4, e5, e6, e7, e8, e9]. Step S33: Construct the vector representation of target relationships. Adopt the method of extracting based on capability features, first extract various features from the targets, and then transform them into numerical vectors.

4. The graphical representation method of military targets according to claim 1, wherein Step S4 includes the following steps: Step S41: Clearly define the status. Use "on" to indicate that the target is enabled and "off" to indicate that it is not enabled, which is convenient for quickly judging the target status and providing accurate information for subsequent analysis and military operation planning. Step S42: Accurately set the target timestamp. Each target is represented by a vector, and the vector contains the target number, timestamp, and status information. Step S43: Efficiently implement the dynamic update of the target capabilities. When the target capabilities change, the original target vector remains unchanged, and a new vector recording the new capabilities is added, along with a new timestamp. Step S44: Accurately determine the edge timestamps. The relationships between military targets are represented by edges, and each edge is also represented by a vector, which contains the edge number and the timestamp. Step S45: Carefully perform the dynamic update of the relationships between targets (i.e., edges). Once the relationships between targets change, the original edge vector remains unchanged, a new edge vector is added and marked with a new timestamp. If a relationship disappears, reappears, or a new relationship is added, the relevant status should be changed in a timely manner and a new timestamp should be marked. Step S46: Intelligently and dynamically adjust the graph. Introduce the concept of a time window and combine it with a graph neural network. Within the time window, continuously monitor the changes in the capabilities and relationships of the targets. Once there are new changes, the graph neural network will adjust the nodes and edges of the graph. Step S47: Use the time window to obtain the latest data. Target node data target feature vector: For each military target i, at time point t j ∈[t now -Δt, t now , its state and capabilities are described by the feature vector , where represents the k-th attribute value of target i at time point t j . Target node data within the time window: In the latest time window, the time-series data of target i is That is, all those satisfying t now -Δt ≤ t j ≤ t now of set, edge data edge vector: The relationship between military targets is represented by an edge. Each edge e pq connects target p and target q, and is represented by the vector where is the edge number, is the timestamp. Data of the edge within the time window: In the latest time window, the time series data of edge e pq is That is, all satisfying t now -Δt ≤ t j ≤ t now of the set. When the relationship between target p and target q changes within the latest time window, a new edge vector appears, recording the new relationship status and timestamp Obtain the latest target graph structure data. Target graph structure: Target graph G = (V, E). Within the latest time window, the target graph structure data is Wherein Update rule: Target node update: If the ability of target i changes at time point t j+1 ∈[t now -Δt,t now , a new feature vector is added and added to . At the same time, the feature of node i in the target graph is updated. Edge update: If the relationship between target p and target q changes at time point t j+1 ∈[t now -Δt,t now , a new edge vector is added and added to . At the same time, the status of edge e in the target graph pq is updated. Temporal data expression method for dynamic assessment of military targets. Dynamic feature vectors and temporal data matrices. The target core defines the time series, the time axis of military operations, which is used to capture the dynamic evolution of the target states. T = {t1, t2, …, t n}, t1 < t2 < … < t n (2-1) The target feature vector, an m-dimensional vector, is used to describe the various attributes of target i at time point t, including firepower intensity and defense ability: j ​ Temporal data matrix. Arrange the feature vectors of target i at each time point in chronological order to form a matrix to reflect the evolution of the target capabilities over time. Relationship edge core defines the edge vector representation. The edge connecting targets p and q, ID pq is the unique number of the edge, t j is the timestamp, represents the k-th relationship attribute between targets p and q. Example of relationship attributes: represents the communication link strength, and the value range can be [0, 1]. 0 means no communication, and 1 means the strongest communication. represents the degree of closeness of joint operations, which is quantified according to the frequency and effectiveness of joint operations. Dynamic graph model and real-time update mechanism. Definition of the time window technology. Δt = t now -t now-Δt (2 - 5) Set a time window length Δt to obtain the latest battlefield data, such as the information data screening rules within the current 1 hour. Target node. Relationship edge. Only retain the valid data of the target nodes and relationship edges within the time window to ensure the real-time performance of the model. Dynamic update rule. Target node update: Add When the ability of target i changes at the new time point t j+1 When a change occurs, add the corresponding feature vector to the existing data and retain the historical data trajectory Relationship edge update. When the relationship between the targets p and q changes at the new time point t j+1 a corresponding edge vector is newly added and incorporated into the existing data to track the dynamic changes of the relationship Graph structure adjustment. Dynamic graph model. G Δt = (V Δt , E Δt ) V Δt (2 - 10) Set of target nodes existing within the time window, E Δt : Set of relational edges existing within the time window, Military significance: When a target is destroyed Remove the node from the node set V Δt When a new relationship is established, add the corresponding edge to the edge set E Δt Add the corresponding edge.

5. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that: When the processor executes the program, it implements the graph expression method of military targets as described in any one of claims 1 to 4 above.

6. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instruction is executed by the processor, it implements the graph expression method of military targets as described in any one of claims 1 - 4.

Images