Method and device for constructing and recommending equipment system adversarial network of dynamic time sequence event

Abstract

The invention discloses a dynamic time sequence event equipment system adversarial network construction and recommendation method and device, and relates to the field of killing network design, and the method comprises the steps: constructing an initial detection-command and control-strike warning network; a dynamic time sequence event on the battlefield is continuously monitored, when the event type of the dynamic time sequence event is an equipment state event, the initial detection-command-strike warning network is updated, and when the event type of the dynamic time sequence event is a chained event, a closed detection-command-strike link set containing an enemy target is generated; the closed detection-command-strike link set comprises a plurality of killing chains; based on the multi-dimensional evaluation index system, evaluating each killing chain to obtain an evaluation result corresponding to each killing chain; and according to the evaluation results corresponding to all the killing chains, recommending an optimal interception scheme of an air defense and anti-guide interception action. The method overcomes the problems of large calculation dimension and slow response speed of an existing method.

Description

Technical Field

[0001] This application relates to the field of kill web design, and in particular to a method and apparatus for constructing and recommending equipment system adversarial networks based on dynamic temporal events. Background Technology

[0002] The design and construction of existing equipment system countermeasure networks (real-time kill chains / networks) targeting dynamic, time-series events present numerous challenges. Particularly when engaging highly dynamic and uncertain time-sensitive targets, the kill chain construction process differs significantly from traditional warfare methods. Communication delays are unavoidable during information transmission between combat nodes, reducing the timeliness of strikes against high-speed maneuvering targets; complex collaborative control among heterogeneous equipment and insufficient autonomous decision-making capabilities hinder timely responses to rapidly changing battlefield situations; simultaneously, communication links are vulnerable to interference and destruction, and the network warfare environment is complex and volatile. All these factors pose severe challenges to the closure and robust operation of the kill chain.

[0003] On the other hand, the key characteristics of the modern battlefield situation further highlight the severity of the above problems. Combat targets exhibit high-dynamic characteristics such as high speed and high mobility, frequently breaking through defenses with diverse and unpredictable tactics; battlefield information sources are highly heterogeneous, reconnaissance data from detection equipment is diverse and scattered, and operational mission plans are frequently adjusted. At the same time, the unmanned and intelligent levels of enemy equipment have significantly improved, and human-machine collaborative combat modes are becoming increasingly common. These new characteristics make the combat environment more uncertain and adversarial, requiring kill chains / networks to have stronger adaptive and collaborative capabilities to cope with the challenges posed by the rapid appearance and disappearance of targets.

[0004] Therefore, there is an urgent need to build an efficient, dynamically reconfigurable, real-time responsive, and distributed collaborative time-sensitive kill chain / kill network system to achieve rapid kill chain construction and optimization recommendation for multiple types of equipment in dynamic scenarios, overcoming the problems of large computational dimensions and slow response speed of existing methods. Summary of the Invention

[0005] The purpose of this application is to provide a method and apparatus for constructing and recommending equipment system adversarial networks based on dynamic temporal events, overcoming the problems of large computational dimensionality and slow response speed of existing methods.

[0006] To achieve the above objectives, this application provides the following solution:

[0007] Firstly, this application provides a method for constructing and recommending equipment system adversarial networks based on dynamic temporal events, including:

[0008] Construct an initial detection-command and control-strike early warning network; the network nodes in the initial detection-command and control-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system; the links in the initial detection-command and control-strike early warning network represent the communication relationships between the equipment.

[0009] Continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, update the initial detection-command-strike early warning network to obtain the updated detection-command-strike early warning network.

[0010] When the event type of a dynamic time-series event is a chained event, a closed detection-command-strike chain set containing enemy targets is generated; the closed detection-command-strike chain set includes several kill chains; a chained event is when the detection equipment detects the appearance of an enemy target;

[0011] Based on a multi-dimensional evaluation index system, each kill chain is evaluated to obtain the evaluation result corresponding to each kill chain; the evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system.

[0012] Based on the assessment results corresponding to all kill chains, the optimal interception scheme for air defense and anti-missile interception operations is recommended.

[0013] Secondly, this application provides a device for constructing and recommending equipment system adversarial networks based on dynamic temporal events, comprising:

[0014] The initial detection-command-strike early warning network construction module is used to construct the initial detection-command-strike early warning network. The network nodes in the initial detection-command-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system. The links in the initial detection-command-strike early warning network represent the communication relationships between the equipment.

[0015] The Initial Detection-Command-Strike Warning Network Update Module is used to continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, the Initial Detection-Command-Strike Warning Network is updated to obtain the updated Detection-Command-Strike Warning Network.

[0016] The closed detection-command-strike link set generation module is used to generate a closed detection-command-strike link set containing enemy targets when the event type of the dynamic time-series event is a chained event; the closed detection-command-strike link set includes several kill chains; the chained event is when the detection equipment detects the appearance of an enemy target;

[0017] The kill chain assessment module is used to evaluate each kill chain based on a multi-dimensional evaluation index system, and obtain the evaluation result corresponding to each kill chain. The evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system.

[0018] The optimal interception scheme recommendation module is used to recommend the optimal interception scheme for air defense and anti-missile interception operations based on the evaluation results corresponding to all kill chains.

[0019] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to implement the above-mentioned method for constructing and recommending equipment system adversarial networks based on dynamic timing events.

[0020] Fourthly, this application provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the aforementioned method for constructing and recommending equipment system adversarial networks based on dynamic timing events.

[0021] According to the specific embodiments provided in this application, the following technical effects are disclosed:

[0022] This application provides a method and apparatus for constructing and recommending a dynamic temporal event-based equipment system adversarial network. The method constructs an initial detection-command-strike early warning network. Network nodes in the initial detection-command-strike early warning network correspond to detection equipment, command and control equipment, and strike equipment in an air defense and anti-missile system. Links in the initial detection-command-strike early warning network represent the communication relationships between equipment. The method continuously monitors dynamic temporal events on the battlefield. When the event type of a dynamic temporal event is an equipment status event, the initial detection-command-strike early warning network is updated to obtain an updated detection-command-strike early warning network. When the event type of a dynamic temporal event is a chained event, a closed set of detection-command-strike links containing enemy targets is generated. The closed set of detection-command-strike links includes several kill chains. A chained event is when detection equipment detects the appearance of an enemy target. Based on a multi-dimensional evaluation index system, each kill chain is evaluated to obtain the evaluation result corresponding to each kill chain. The evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system. Based on the evaluation results corresponding to all kill chains, the optimal interception scheme for air defense and anti-missile interception operations is recommended. By introducing a pre-deployment of the warning network and an event-driven link reconstruction mechanism, the timeliness of kill chain generation and decision-making when a target appears is greatly improved. The resource allocation and dynamic iteration method of the warning network are optimized to achieve a rapid chain response when a target appears, and the equipment system's ability to adapt to dynamic changes in battlefield nodes is enhanced, so as to achieve the goal of "destroying the target as soon as it appears" at the lowest cost. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is an application environment diagram of a method for constructing and recommending equipment system adversarial networks based on dynamic temporal events, as described in Embodiment 1 of this application.

[0025] Figure 2 This is a flowchart illustrating a method for constructing and recommending a dynamic temporal event-based equipment system adversarial network, as provided in Embodiment 1 of this application.

[0026] Figure 3 This is a schematic diagram of the technical path for a method of constructing and recommending a dynamic temporal event-based equipment system adversarial network, as provided in Embodiment 1 of this application.

[0027] Figure 4 This is a schematic diagram illustrating the specific process of a method for constructing and recommending a dynamic temporal event-based equipment system adversarial network, as provided in Embodiment 1 of this application.

[0028] Figure 5 This is a schematic diagram of the time-sensitive kill network model structure provided in Embodiment 1 of this application. Detailed Implementation

[0029] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] The kill chain is a key concept in modern information warfare, referring to a series of interdependent operational links that complete the target acquisition, location, tracking, aiming, engagement, and damage assessment. With the development of the operational environment and technology, the US military has expanded the static, single kill chain concept into a multi-domain collaborative "kill network," using multiple nodes to form a highly distributed and cross-domain collaborative network structure, enhancing the flexibility and unpredictability of the operational chain. For particularly urgent time-sensitive targets (TSTs), rapid kill chain closure is required, and in recent years, scholars from various countries have conducted extensive research on how to construct high-speed, closed time-sensitive kill chains.

[0031] Overall Trends and Representative Studies: Review studies indicate that the academic community is exploring kill chain construction methods towards intelligence, networking, and adaptability. Yang Song et al. provided a comprehensive review of the current development status of kill chains / kill networks, summarizing domestic and international research progress and proposing suggestions for future development. Based on this, many research teams have proposed their own unique modeling approaches. For example, Zhao Guohong proposed a time-sensitive target classification method and closed-loop evaluation criteria based on the kill chain perspective for constructing time-sensitive target strike systems, used to measure the response speed and completeness of the kill chain. Shao Jun et al. designed an equipment combination selection framework based on kill networks, clarifying for the first time the concept and characteristics of equipment system kill networks, and using a multi-layer network model to build a kill network structure to reflect the characteristics of cross-domain collaborative operations. Yang Zhengzheng et al., based on the "mosaic warfare" concept, constructed a multi-layered kill network model for the land battlefield, including four elements: reconnaissance network, communication network, command network, and strike network, to achieve dynamic coordination of multiple elements in the land warfare domain. These modeling studies focus on revealing the impact of collaborative relationships and network structure on kill chain effectiveness. Their advantage lies in comprehensively depicting the link structure of multi-domain collaborative operations, but their disadvantage is that most of them are conceptual models, and their feasibility needs to be further verified by combining algorithms and simulations.

[0032] Mission-Capability-Equipment Mapping Method: To translate operational mission requirements into specific equipment actions, some scholars have proposed mapping modeling methods for missions, capabilities, and equipment units. For example, Wang Meng et al., based on hypernetwork theory and executable architecture, designed a "target-mission link" model that matches operational missions with multi-domain equipment resources, improving the systemic completeness of mission planning. This method establishes a mapping relationship between mission requirements and equipment functions, achieving a layer-by-layer decomposition from mission to capability and then to equipment. The characteristic of this type of method is that it uses operational capabilities as an intermediary to map mission requirements to specific strike equipment: for example, mapping mission elements such as "detection, tracking, guidance, attack, and assessment" to resources such as detection equipment, command and control systems, and strike equipment ammunition, and constructing a link between mission nodes and equipment resources. Its advantage is that it ensures that each link in the kill chain has corresponding capabilities and equipment support, forming a complete link; however, due to the complexity of the battlefield environment and the large number of heterogeneous data and collaborative constraints among different equipment, maintaining effective mapping in real-time dynamic scenarios remains a challenge.

[0033] Multi-objective optimization and simulation modeling: For the optimal selection of kill chain schemes, researchers introduced multi-objective optimization algorithms to balance indicators such as combat effectiveness and resource consumption. Wan Silai et al. proposed a kill chain modeling and optimization method based on AGE-MOEA. By establishing a mathematical model encompassing three types of equipment—reconnaissance, command and control, and strike—and combining the "OODA loop" decision-making cycle concept, they formed a multi-objective optimization model covering all links in the kill chain. The team designed an adaptive geometric estimation multi-objective evolutionary algorithm (AGE-MOEA) optimization process to solve the kill chain design problem. The optimization objectives include maximizing strike effectiveness, minimizing strike equipment consumption, and minimizing self-survival threat. Simulation simulations verified that the proposed method can simultaneously optimize multiple objectives while ensuring the closure of all target kill chains, obtaining a distributed optimal kill chain scheme. Similarly, Qian Feng et al. designed a manned-unmanned collaborative kill chain engine to overcome the bottleneck of manned / unmanned combat resource fusion scheduling and enhance the compactness and speed of the chain closure; Lu Jiabo et al. proposed an intelligent method for rapid matching of strike equipment and targets for time-sensitive targets, which significantly improves the efficiency of the strike chain while meeting time constraints. The advantages of the above optimization methods lie in quantitative modeling and algorithm solution, which can automatically select the optimal kill chain scheme in a large solution space and consider the link effectiveness evaluation index. Their disadvantages may lie in the high model complexity, the algorithm convergence speed and global optimality depending on parameter settings, and the need for extensive simulation verification.

[0034] Intelligent Algorithm Scheduling and Knowledge Reasoning Methods: To achieve autonomous and rapid construction of kill chains, many teams have introduced artificial intelligence and knowledge engineering techniques. Wan Silai et al. proposed a kill network intelligent design method based on knowledge reasoning, constructing a kill network ontology model and reasoning rule framework, and associating combat missions (such as reconnaissance, decision-making, and strike, corresponding to each stage of the OODA loop) with equipment capabilities. This method defines three types of kill chain effectiveness evaluation indicators: time chain, accuracy chain, and cost chain, to adapt to the rapid generation of kill networks in dynamic battlefields, and verified its effectiveness in an air defense combat case on an adversarial simulation platform. Its advantage is that it uses knowledge graphs and reasoning engines to explicitly characterize the collaborative relationships between equipment (such as communication links, command and control relationships, etc.), achieving automated combination of kill chains. Meanwhile, scholars both domestically and internationally have also used knowledge graphs for kill chain construction: for example, Hu Wei et al. implemented hierarchical knowledge visualization of equipment data, and Wang et al. constructed a land, sea, and air integrated kill chain knowledge graph based on grouping clustering and rule constraints, improving the efficiency and accuracy of manual planning. These intelligent scheduling methods emphasize adaptability and autonomy, enabling real-time adjustment and reorganization of communication links based on battlefield conditions. However, their limitation lies in their high dependence on knowledge completeness. Early military knowledge graph models often depicted combat entity relationships with coarse granularity and lacked fine-grained descriptions of combinatorial relationships, potentially affecting the accuracy of communication link schemes in complex scenarios. Furthermore, in highly dynamic environments such as unmanned swarms, intelligent algorithm scheduling also faces real-time challenges. For example, some research has employed an improved contract network protocol combined with the Hungarian algorithm to achieve rapid task redistribution of damaged nodes in unmanned swarms, thereby improving local adjustment capabilities under emergency missions.

[0035] However, current methods often focus solely on executing unidirectional chained tasks in a fixed sequence, lacking consideration for multi-node collaboration and dynamic link reconfiguration. This results in insufficient overall robustness and flexibility, making it difficult to cope with highly dispersed adversarial scenarios in cross-domain, multi-faceted combat environments. Therefore, under conditions of heterogeneous operational resources and unstable communication, it is essential to decouple equipment functions and support dynamic grouping. This can enhance the system's rapid reconfiguration and real-time response capabilities, enabling the kill chain to close the OODA loop more quickly, achieving continuous global assessment and dynamic adjustment to meet the urgent needs of modern warfare for time-sensitive kill chains / networks.

[0036] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0037] Example 1

[0038] The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events provided in this application can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on other servers. Terminal 102 can send pending requests to server 104. Upon receiving a pending request, server 104 constructs an initial detection-command-strike early warning network, continuously monitoring dynamic temporal events on the battlefield. When the event type of the dynamic temporal event is an equipment status event, the initial detection-command-strike early warning network is updated. When the event type of the dynamic temporal event is a chained event, a closed detection-command-strike link set containing enemy targets is generated. Based on a multi-dimensional evaluation index system, each kill chain is evaluated, obtaining the evaluation result corresponding to each kill chain. Based on the evaluation results corresponding to all kill chains, the optimal interception scheme for air defense and anti-missile interception operations is recommended. Server 104 can feed back the obtained optimal interception scheme to terminal 102. Furthermore, in some embodiments, the method for constructing and recommending equipment system adversarial networks based on dynamic time-series events can also be implemented separately by the server 104 or the terminal 102. For example, the terminal 102 can directly construct and recommend networks for requests to be processed, or the server 104 can obtain requests to be processed from the data storage system and construct and recommend networks for requests to be processed.

[0039] The terminal 102 can be, but is not limited to, various desktop computers, laptops, and tablets. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0040] In one exemplary embodiment, such as Figure 2 , Figure 3 , Figure 4 As shown, a method for constructing and recommending equipment system adversarial networks based on dynamic temporal events is provided. This method is executed by computer devices, specifically by a terminal or server alone, or by both a terminal and a server. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 205.

[0041] Step 201: Construct an initial detection-command-strike early warning network; the network nodes in the initial detection-command-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system, and the links in the initial detection-command-strike early warning network represent the communication relationships between the equipment.

[0042] Step 202: Continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, update the initial detection-command-strike early warning network to obtain the updated detection-command-strike early warning network.

[0043] Step 203: When the event type of the dynamic time-series event is a chained event, a closed detection-command-strike link set containing the enemy target is generated; the closed detection-command-strike link set includes several kill chains; the chained event is when the detection equipment detects the appearance of the enemy target.

[0044] Step 204: Based on the multi-dimensional evaluation index system, evaluate each kill chain to obtain the evaluation result corresponding to each kill chain; the evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system.

[0045] Step 205: Recommend the optimal interception scheme for air defense and anti-missile interception operations based on the evaluation results corresponding to all kill chains.

[0046] Implementing steps 201 to 205 above, based on the kill chain closure principle, and integrating OODA (Observe, Orient, Decide, Act) cycle theory and complex network theory, a dynamic early warning network (ODA network) is innovatively established. This network is assigned weights that consider the coupling of physical distance, communication accuracy, and reliability to form a battlefield early warning network. A hierarchical optimization strategy is adopted to perform local network reconstruction under non-time-sensitive events such as equipment damage or addition, rapidly generating kill chains when enemy targets appear. Simultaneously, a multi-dimensional evaluation system is proposed, calculating key indicators such as communication loss / gain factors, equipment deployment density, damage effectiveness index, and response latency to rank and recommend candidate solutions. The resource allocation and dynamic iteration method of the early warning network are optimized to achieve a chain-like rapid response when targets appear, improving the equipment system's adaptability to dynamic changes in battlefield nodes, and achieving the goal of "destroying targets upon appearance" at minimal cost.

[0047] The construction of the Initial Detection-Command-Strike (ODA) early warning network includes: using graph theory, with detection equipment O, command and control equipment D, and strike equipment A in the air defense and missile defense system as network nodes, and constructing OD links and DA links based on the communication relationships between the equipment; the OD link is the communication link between the detection equipment and the command and control node; the DA link is the communication link between the command and control node and the strike equipment.

[0048] Within the time-sensitive operational area, based on the existing equipment inventory, an initial ODA network, or "early warning network," is pre-constructed, comprising nodes for Observer (O), Decision (D), and Act (A). Equipment with Observer and Decision functions is collectively referred to as O equipment.

[0049] Initial Early Warning Network Construction: First, based on all relevant equipment deployed within the time-sensitive operational mission area, an initial ODA early warning network model is established. It is assumed that the initial ODA early warning network model has x1 detection equipment O, x2 command and control equipment D, and x3 strike equipment A.

[0050] The network topology and edge weights are determined by analyzing the communication interconnections and performance parameters between the various pieces of equipment.

[0051] ①If a certain detection equipment O i It can directly send intelligence to a certain command and control equipment D j Then at node O i With D j Establish an OD link between them and assign an initial weight odw ij These weights constitute the matrix (Calculated according to the aforementioned model).

[0052] ②If a certain command equipment D j Capable of controlling / commanding a certain strike equipment command and control equipment A k Then in D j With A k Establish a DA link between them and assign an initial weight daw ij These weights constitute the matrix (Calculated according to the aforementioned model). OT and AT links do not have any substantial targets at this time, so they can be left out of the surveillance network topology for now.

[0053] Node O represents detection equipment such as radar and early warning aircraft; node D represents command and control equipment such as command and control centers or fire control units; node A represents strike equipment such as air defense missile launchers or anti-aircraft guns; target T is an enemy target introduced as a special node when an enemy situation occurs.

[0054] Based on the information flow and combat actions of the OODA cycle, four types of key links are defined: OD link (communication link between detection equipment and command and control node), DA link (communication link between command and control node and strike equipment), OT link (detection link between detection equipment and target), and AT link (strike link between strike equipment and target). These links correspond to four types of edge weight matrices: and These represent the performance weight matrices of the aforementioned links, where the elements are the communication loss and gain factors between specific pairs of equipment.

[0055] Link weights are assigned based on a model of "physical distance / interaction time × adjustment coefficient" to reflect the comprehensive effectiveness of link transmission delay and reliability. This model reflects the spatiotemporal transmission efficiency and reliability of different links. Physical distance reflects the distance between nodes, and communication / interaction time reflects the delay in information or attack effect transmission. The ratio of these two represents the coverage distance per unit time (which can be considered as link speed or efficiency). This weight is then multiplied by a corresponding adjustment coefficient to incorporate other influencing factors. Link weights include the following four types:

[0056] 1) Link weight (OTW) between detection equipment and enemy target: This is obtained by multiplying the ratio of detection range to signal round-trip time by the detection reliability coefficient and the target detectability adjustment coefficient. It measures the sensor's detection efficiency per unit time. (Detection link weight matrix) This represents the overall effectiveness of detection equipment O in detecting enemy target T. The elements otw in the weight matrix of the detection link... ij The weights can be modeled as:

[0057]

[0058] Among them, OTX ij It is the detection relation coefficient (whether detection equipment i has the ability to detect enemy target j, with a value of 0 or 1, where 0 represents no and 1 represents yes); otd ij To detect the distance between equipment i and enemy target j, OTP ij This refers to detection accuracy or reliability (such as detection probability), with a value range of (0, 1]; ots ij This represents the target's concealment / stealth coefficient (an inverse factor of target detectability); ov i The speed of the detection signal emitted by detection equipment i; ok i This refers to the performance coefficient of detection equipment i (e.g., the remaining resources or operational status adjustment coefficient of the detector). The weight can be understood as a measure of the effective detection output provided by the detection equipment to the enemy target per unit time. The larger the value, the greater the loss or gain in the detection equipment to detect the enemy target, that is, the less "economical" it is.

[0059] 2) Detection-Command Link Weight (ODW): Obtained by multiplying the ratio of communication distance to communication delay by a communication quality coefficient and a compatibility adjustment coefficient, it measures the efficiency of the sensor in transmitting information to the command node. Detection-Command Link Weight Matrix This represents the communication efficiency of transmitting detection information from detection equipment O to command and control node D. The element odw in the detection command link weight matrix... ij The weighting model is as follows:

[0060]

[0061] Among them, odx ij It is the interaction coefficient (whether the detection equipment i has an interaction relationship with the command and control equipment j, with a value of 0 or 1, where 0 represents no interaction and 1 represents interaction); odd ij The distance between detection equipment i and command and control equipment j; odp ij It refers to the quality or accuracy of the communication link between detection equipment i and command and control equipment j (such as link reliability and data accuracy); odv ij It is the signal transmission speed between detection equipment i and command and control equipment j; odc ij This indicates communication interference or incompatibility factors (such as loss of effective bandwidth due to different protocols or environmental interference) between detection equipment i and command and control equipment j; dk j The performance capacity of the command and control equipment j refers to how many detection devices it can establish communication with simultaneously.

[0062] 3) Command and control link weight (DAW) between command and control equipment and strike equipment: This is obtained by multiplying the ratio of command transmission distance to latency by the link reliability coefficient, and is used to measure the timeliness of command and control orders reaching the weapons. Command and Strike Link Weight Matrix This indicates the communication efficiency of command node D in intercepting commands and transmitting them to strike equipment node A. The element 'daw' in the command and strike link weight matrix... ij Weight calculation can be performed using the same methods as... Similar models:

[0063]

[0064] Among them, dax ij It is the interaction coefficient (whether command and control equipment i has an interaction relationship with attack equipment j, with a value of 0 or 1, where 0 represents no interaction and 1 represents interaction); dad ij It is the distance between the commanding equipment i and the striking equipment j, dap ij It refers to the transmission accuracy and reliability between command and control equipment i and strike equipment j; dav ij It refers to the signal transmission speed between command and control equipment i and strike equipment j; dac ij This indicates communication interference or incompatibility between the command and control equipment i and the strike equipment j (such as loss of effective bandwidth due to different protocols or environmental interference); dk j The performance capacity of the command and control equipment j refers to how many strike equipment it can establish communication links with simultaneously.

[0065] 4) Attack Link Weight (ATW) between the weapon and the enemy target: This is obtained by multiplying the ratio of the attack distance to the ballistic flight time by the weapon's hit probability and the operational environment influence coefficient. It is used to measure the weapon's immediate damage efficiency against the target. Attack Link Weight Matrix This represents the effectiveness and efficiency of strike equipment A in striking target node T. The element atw in the strike link weight matrix... ij The weight can be expressed as:

[0066]

[0067] Among them, atx ij It is the strike relationship coefficient (whether the strike equipment i has the capability to strike enemy target j, with a value of 0 or 1, where 0 represents no capability and 1 represents capability); ATP ij It refers to the accuracy or probability of a strike by weapon i against enemy target j; atd ij The distance between the weapon i and the enemy target j; av i For the strike speed of equipment i; atf ij This is the reduction factor of the effect of the environment on the attack equipment (such as the target adopting electronic countermeasures or maneuvering evasion, which improves survivability).

[0068] Constructing an initial detection-command and control-strike early warning network specifically includes: building the initial detection-command and control-strike early warning network based on equipment performance and capacity constraints. Equipment performance and capacity constraints include: the number of targets each detection device can simultaneously track does not exceed its capacity limit; the number of detection information and control / strike units each command and control device can access does not exceed its processing capacity; and the number of commands each strike device can accept or intercept per unit time is limited. Within these constraints, the early warning network structure is optimized to connect as many equipment nodes as possible to the network and cover a predetermined airspace, forming a multi-sensor collaborative early warning and surveillance network.

[0069] The ODA network generated based on existing equipment is optimized after the occurrence of ad hoc events (dynamic timing events), avoiding redundant budgeting and saving costs. Various constraints must be met during network construction:

[0070] Capacity limitations include the maximum number of command and control nodes each detection device can connect to, the maximum number of detection and strike devices each command and control node can connect to, and the maximum number of command channels each strike device can receive. These constraints can be expressed mathematically:

[0071]

[0072] This represents the number of links from all detection equipment O to all command and control equipment D. This represents the number of links from all command and control equipment D to all strike equipment A. If both link numbers are equal, they represent the sum of the maximum performance capacity of the command and control equipment.

[0073]

[0074] Among them, dkj This indicates that all detection equipment O corresponds to a certain command and control equipment d. j The sum of the interactions does not exceed the charge equipment d j Performance capacity; OK i This indicates that all command and control equipment D is related to a certain detection equipment o. i The sum of the interactions between the detection equipment does not exceed the value of the detection equipment. i Performance capacity; ak j This indicates that all the command equipment D is against a certain strike equipment o. i The sum of the interaction relationships does not exceed the value of the strike equipment a. j Performance capacity; dk i This indicates that all attack equipment A is against a certain command and control equipment d. i The sum of the interactions does not exceed the charge equipment d i Performance capacity.

[0075] The number of sensor messages that each command and control node can receive simultaneously and the number of control and strike equipment are also limited; each strike equipment node can only receive one command link per operation, and so on. Under the premise of meeting the constraints, this application prioritizes network coverage of all equipment, that is, to include all available O, D, and A nodes into the connected network as much as possible, thereby maximizing the overall early warning, surveillance, and strike coverage.

[0076] Warning Network Maintenance and Updates: After construction, the warning network enters operational monitoring mode. The system continuously monitors equipment status and enemy situation information. When non-linked events occur (such as equipment failure, reinforcement equipment joining, or changes in communication link quality), the system adjusts the warning network accordingly based on the event type.

[0077] ① Equipment damage or offline: Mark the corresponding node and related edges as unavailable, remove them from the network topology or reduce their weight to infinity (indicating unreachable). At the same time, notify neighboring nodes to re-plan surveillance / communication tasks (e.g., other nearby radars expand their scanning sectors to fill gaps).

[0078] ② Adding or Restoring Equipment: Add new nodes to the network and connect them to appropriate locations based on their communication / detection capabilities. For example, a new radar connects to the nearest command and control equipment node, and a new strike equipment connects to the command center responsible for air defense in the area. Assign initial weights to the new links.

[0079] ③ Status / Environment Changes: If the communication quality of a certain link deteriorates, the communication parameters can be dynamically adjusted and the weights recalculated; if the ammunition for the attack equipment is exhausted, the performance coefficient ak of the corresponding node A can be reduced. j Or there is a possibility of temporarily removing AT.

[0080] ④ Local Optimization: After event handling, local optimization algorithms can be executed. For example, to address coverage gaps in the detection network, the scanning direction of nearby detection equipment can be adjusted or backup detection equipment can be activated; for links with increased communication latency, switching to backup channels can be attempted. Through these local adaptive measures, the surveillance network can maintain efficient connectivity in dynamic environments, laying the foundation for rapid response when the next target appears.

[0081] Resource Consumption and Warning Network Deployment: In the initial construction of the ODA warning network, the deployment coverage and resource consumption of detection equipment need to be considered. A continuous warning and surveillance network is formed through the reasonable distribution of detection equipment. Detection equipment maintains constant communication with its corresponding command and control equipment, and the command and control equipment maintains constant communication with its corresponding strike equipment. These three types of nodes constitute the warning network, achieving a highly interconnected data communication hinge. This is the fundamental guarantee for the subsequent formation of a kill chain; that is, when a time-sensitive target appears, detection equipment and strike equipment can be quickly matched based on the existing network to form a kill chain. Here, resource consumption refers to the detection capability invested by the detection equipment in performing the warning mission (such as continuous scanning of a certain airspace / sector). This application incorporates the coverage effect and resource utilization rate of the detection network into the design, enabling each detection equipment to collaboratively monitor different areas when no target appears, forming an airborne warning grid and improving the overall detection probability. Once a target appears, the existence of the warning network ensures that at least one detection link (OTW) can be quickly established, providing a prerequisite for subsequent links.

[0082] Event monitoring and network updates: Dynamically updating the ODA (Operational Distress Assessment) network based on event type. This step addresses dynamic, time-series events on the battlefield and handles two scenarios:

[0083] 1) Non-chained events (equipment status events): When equipment is added, damaged, or its status changes, the affected nodes and links are locally adjusted or reconstructed to maintain network connectivity and optimal configuration. This includes events such as the addition, damage, relocation, or communication status changes of friendly equipment, which do not directly involve the appearance of enemy targets. When such events are detected, this application performs a local reconstruction of the ODA (Optical Deployment Detection) network: updating the status of relevant nodes or links. For example, if a detection radar is damaged, the O node and its associated OD (Optical Deployment) links are temporarily removed from the network; if a new detection equipment is added, it is added as a new O node, and OD links are established based on connectable command and control nodes. The local reconstruction follows the event-driven principle, adjusting only the affected local network without changing other areas, thereby saving computational overhead and maintaining network stability. In addition, network connectivity constraints and equipment performance capacity constraints must be ensured during the reconstruction process: such as the maximum number of targets that each detection equipment can track, the maximum number of detection and strike equipment that each command node can simultaneously access, and the ammunition / interception capabilities of each strike equipment. Through continuous monitoring and local optimization, the warning network always maintains an optimal or near-optimal structure to cope with sudden threats.

[0084] 2) Chaining Event (Target Appearance): When the detection equipment detects the appearance of enemy target T, the kill chain construction is triggered: Target T is added as a new node to the updated ODA network, connecting to all O nodes that have detected the target and all A nodes capable of engaging the target, forming OT and AT candidate links; utilizing existing OD and DA links, each O node is connected to each A node through at least one D node, generating a closed detection-command-strike link set containing target T. The specific process is as follows: When the appearance of enemy target T is detected (usually this is first detected by an O node in the surveillance network, triggering the event), the rapid kill chain construction process is immediately triggered. First, based on the target's location and attributes, candidate node sets for the detection chain OT and the strike chain AT are selected: that is, determining which O nodes have detected the target and which A nodes can engage the target within their range and firepower. This application utilizes pre-built and updated ODA (Optical Deployment Detection) network information to quickly assess whether a usable command node D exists between the aforementioned candidate detection node O and strike node A to connect them into a complete link. When constructing the warning network, a relationship matrix between detection node O and strike equipment A is introduced. The element oax ij If it is 0, it means the detection equipment is not... i With strike equipment a j There is no available command node D connecting the two; a value of 1 indicates that the detection equipment is not present. i With strike equipment a jThere is an available command node D connecting the two. If a communication path already exists between the detection node O and the strike node A through a command node D (i.e., the OD and DA links are connected), a kill chain T–O–D–A–T is generated immediately. If there are multiple feasible links, the target T is combined with all these links to form a kill network structure. In simple terms, after the target appears, this application does not only generate a single path scheme, but converges them into a chain and a network: it identifies all link combinations that can complete the target detection-strike closure, forming an adversarial network against the target.

[0085] In step 203 above, when the event type of the dynamic time-series event is a chained event, a closed detection-command-strike chain set containing the enemy target is generated, specifically including the following steps 301 to 302.

[0086] Step 301: When the event type of the dynamic time-series event is a chained event, an improved incremental graph algorithm is used to generate a closed detection-command-strike link set containing enemy targets. Specifically, enemy targets are dynamically added as new nodes to the updated detection-command-strike early warning network, connecting all O nodes that have detected enemy targets and all A nodes that can attack enemy targets, forming OT links and AT links, resulting in a further updated detection-command-strike early warning network; O nodes are detection equipment nodes, and A nodes are strike equipment nodes; when no equipment status event occurs, the updated detection-command-strike early warning network is the initial detection-command-strike early warning network.

[0087] Step 302: Search the updated Detection-Command-Strike early warning network for kill chains that originate from an enemy target and return to the enemy target; all kill chains constitute a closed set of Detection-Command-Strike links.

[0088] By employing an incremental graph algorithm, after adding target T as a new node to the existing ODA network, path search is performed only on the newly added nodes and links related to the target, quickly finding the set of closed paths containing the target, without repeating global calculations on the entire network, thereby improving construction efficiency.

[0089] The method for constructing and recommending the equipment system adversarial network for dynamic time-series events also includes: after the enemy target is connected to the updated detection-command-strike early warning network as a new network node, OT links and AT links are added to the detection-command-strike early warning network, where the OT link is the detection link of the detection equipment to the target, and the AT link is the strike link of the strike equipment to the target.

[0090] Chain / Network Construction Method: An improved incremental graph algorithm is used to rapidly construct the aforementioned kill chain / network. The specific process is as follows: Target T is dynamically added to the ODA network as a new node: it is connected to all O nodes that detected it (forming OT edges) and all A nodes that can attack it (forming AT edges). Thus, target T is connected to several O and A nodes in the graph. Using the incremental graph algorithm, only nodes and edges related to T are added to the existing ODA network. Then, a set of closed paths (T→O→D→A→T) starting from T and returning to T is searched in the new graph. Since the incremental addition is very small (relative to the complete network), the search process is very fast, obtaining all possible closed link combinations the instant the target appears. The algorithm simultaneously calculates the comprehensive weight or cost (according to the evaluation metrics below) for each closed path. If a path already exists and is stored in a previous calculation, it does not need to be recalculated, further saving time. Through this incremental construction, this application achieves the effect of "a chain is formed as soon as the target appears."

[0091] This paper will explain the modeling of time-sensitive kill chains from the perspectives of nodes, edges, and edge weights in kill networks, based on network design.

[0092] This embodiment establishes a time-sensitive kill network model (kill network, also known as a detection-command-strike early warning network) G = (V, E, W), where G represents the detection-command-strike early warning network, V represents the nodes in the detection-command-strike early warning network, E represents the edges between nodes in the detection-command-strike early warning network, and W represents the weights of the edges between nodes in the detection-command-strike early warning network (link weights, where an edge is a link formed by network nodes corresponding to the equipment). This embodiment defines interactive gains and losses to replace these weights.

[0093] in:

[0094] V=T∪O∪D∪A∪T'(10);

[0095] and:

[0096] T = [t1, t2, ..., t z ] ‖T‖=z (11);

[0097]

[0098] T' = [t'1, t'2, ..., t'] z ||T'||=z (15);

[0099] The above model indicates that there are five types of network nodes: detection targets T, reconnaissance equipment O, command and control equipment D, strike equipment A, and strike targets T', with quantities of z, x1, x2, x3, and z, respectively.

[0100] Where: E = ET→O ∪E O→D ∪E D→A ∪E A→T' This formula indicates that edges in the network are formed by connecting nodes in only two adjacent columns. Each pair of adjacent columns has an interaction matrix, which is as follows: and Each element OTX ij odx ij dax ij atx ij The value is 0 or 1, in order to Taking a matrix as an example, if odx ij =0 indicates reconnaissance equipment o i and the charge equipment d j There is no interaction between them, meaning they represent reconnaissance equipment. i Nodes and Command Equipment j There are no edge relationships between nodes; conversely, if odx ij =1 indicates reconnaissance equipment. i and the charge equipment d j There is an interactive relationship between them, representing reconnaissance equipment. i Nodes and Command Equipment j There are edges between the nodes.

[0101] Furthermore, there is also a relationship between the detection equipment O and the target T. and The matrices and vectors represent the detection accuracy matrix between the detection equipment and the detection target, the distance matrix between the detection equipment and the detection target, the concealment matrix of the detection target being detected by the detection equipment, the performance capacity limitation vector of the detection equipment, and the signal velocity vector of the detection equipment, respectively.

[0102] There is still a connection between detection equipment O and command and control equipment D. and The matrices and vectors represent the communication accuracy matrix between the detection equipment and the command and control equipment, the distance matrix between the detection equipment and the command and control equipment, the communication speed matrix between the detection equipment and the command and control equipment, the communication compatibility matrix between the detection equipment and the command and control equipment, and the performance capacity limitation vector of the command and control equipment, respectively.

[0103] There is still a gap between command and control equipment D and strike equipment A. and The matrices represent the communication accuracy matrix between command and control equipment and strike equipment, the distance matrix between command and control equipment and strike equipment, the communication speed matrix between command and control equipment and strike equipment, and the communication compatibility matrix between command and control equipment and strike equipment, respectively.

[0104] There is still a connection between the attack equipment A and the target T. and The matrix and vector represent the strike accuracy matrix between the strike equipment and the strike target, the distance matrix between the strike equipment and the strike target, the operational environment influence matrix between the strike equipment and the strike target, and the strike velocity vector of the strike equipment, respectively.

[0105] The edge weights between nodes form several matrices, namely: and The formulas for calculating the elements of each matrix are shown in formulas (1) to (4).

[0106] The time-sensitive kill network model established based on the above mathematical model (i.e., the updated detection-command-strike warning network) is as follows: Figure 5 As shown.

[0107] When the surveillance network detects the presence of enemy target T (usually detected first by a node O, this instantaneous detection event (dynamic timing event) is denoted as E). detect (T)), the system introduces node T into the current ODA network to initiate the kill chain construction process.

[0108] ①Sensor-Fire Matching: Determine the detection set based on the detection results. and firepower The detection set is obtained from the monitoring results of the surveillance network (which detection equipment saw the target). This detection result is obtained through the reconnaissance equipment O within the surveillance network. The firepower set is obtained by determining which strike equipment has range covering the target and currently has interceptor missiles available; this is determined by the command and control equipment D within the surveillance network. Each of these sets may contain multiple elements.

[0109] ② Fast link splicing: For each pair and Check if there is a path from o in the warning network. i After a certain d k to a l The communication path. If it exists, then t j -o i -d k -a l -t j Construct a line that can target t j The potential kill chain. This application, due to the pre-existing OD and DA links, is essentially equivalent to finding all combinations of connections between the detection set and the fire set via command and control nodes. When or When there are multiple members, this step may result in several different ODA combination links;

[0110] ③ Incremental network generation: formally generating the target t j Nodes join the network, taking all the o values ​​obtained in step 1. i With t j Connect them to form an OT edge, and connect all a l With t j Connect the edges to form AT edges. Calculate the weight matrix for each new edge. and (According to the aforementioned formula, using parameters such as detection range, signal round-trip time, and hit probability, only the ODA network exists before the battle; OTW and ATW are only calculated after the target appears.) Now, T is connected to several O and A nodes, and through the existing OD and DA edges, T can form a closed loop. For example, c is a complete kill chain closed loop path;

[0111] ④ Solution set generation: Search the augmented graph for all simple cyclic paths that start from T and return to T, thus obtaining the candidate solution set. Each loop corresponds to a sensor-command-firepower combination scheme. For each candidate link... Extract its specific components, such as t j -o i -d k -a l -t j and the corresponding edge weights {otw ij ,odw ik ,daw kl ,atw lj};otw ij ,odw ik ,daw kl ,atw lj The detection equipment is respectively i -Target t j Weight of the link between them, detection equipment i -Accusation Equipment d k Weight of the link between them, command and control equipment d k -Strike Equipment a l Weight of the link between them, and the attack equipment a l -Target t j The weight of the links between them.

[0112] ⑤ Feasibility Filtering: If a link does not meet the constraints (e.g., the O or A connections of a certain command node exceed capacity, or the time delay is too long and not suitable for the target type), the solution is marked as infeasible and eliminated. Generally, the solutions generated through previous monitoring network optimization are mostly feasible.

[0113] Through the above process, the system generates a set of all possible interception link schemes almost at the same time the target appears. Since incremental graph construction and querying are used, the computational complexity is mainly related to the number of candidate O and A and the existing network paths. The existence of the warning network ensures that most of the connection relationships are ready in advance, so the overall computation is very efficient.

[0114] If the target threat is high, the commander can also choose a multi-chain strategy, that is, not just using the single kill chain ranked first, but simultaneously using multiple chains (which effectively expands the "kill chain" to the concept of a "kill network"). For example, against a hypersonic missile target, two different missile launchers can be simultaneously directed to intercept from different locations to increase the interception probability. Under the framework of this application, this is equivalent to selecting multiple chains from the set of options for parallel execution. The evaluation ranking provided in this application also supports this decision: the commander can determine whether redundant interception with multiple options is needed based on indicators, and the system can coordinate the selected chains in scheduling (such as staggering interception timing).

[0115] The multi-dimensional evaluation index system includes, but is not limited to, communication loss, deployment density, damage performance, and response latency.

[0116] 1) Communication Loss: Measures the overall cost of communication links in a kill chain. For example, it can be represented by the sum of link weights or other functional forms to account for the cost of communication delays and reliability. Low communication loss indicates smooth and efficient intelligence transmission.

[0117] 2) Deployment Density (Average Distance): This measures the spatial density of equipment nodes in the kill chain and can be represented by the average physical distance of the links. A smaller average distance indicates that detection equipment, command and control equipment, and strike equipment are closer together, resulting in a more concentrated deployment, which is beneficial for reducing communication latency and improving coordination efficiency; conversely, excessive distance indicates a sparse network, which may reduce reaction speed. This indicator also reflects the effectiveness of resource pre-positioning strategies.

[0118] 3) Damage Effectiveness Index: This index comprehensively considers damage probability and damage accuracy, reflecting the effectiveness of the method in intercepting targets. Damage probability typically refers to the probability that the strike equipment will hit and destroy the target, while damage accuracy refers to the strike error or hit precision. The overall success rate p can be obtained by multiplying the success rate of sensor detection, the correctness of command and control decisions, and the hit rate of the strike equipment. j That is, it is calculated using the following formula:

[0119]

[0120] in, To determine the probability of successful sensor detection. To determine the probability of making the correct command decision, To determine the probability of hitting the target, the overall damage probability of multiple strikes or multi-link coordination is calculated, while also considering accuracy factors (e.g., reduced accuracy may require more interceptor missiles). This application integrates these factors into a single index for comparing different scenarios; a higher value indicates a greater likelihood of destroying the target.

[0121] 4) Response latency: also known as kill chain closure time, refers to the total time required from the moment a target is detected until the attacking equipment hits it. This includes sensor detection delay, communication transmission delay, command and decision-making time, and munition flight time. In practical calculations, this indicator can be further divided into the aforementioned decision time and execution time. The decision time in this application is significantly compressed due to the use of a pre-built early warning network and incremental algorithms; the execution time is related to the physical time of the link (which can be obtained from the distance / velocity of each link). A shorter response latency indicates a shorter period from target detection to interception, and a more timely air defense response.

[0122] This embodiment uses complex network design methods to construct an adversarial network. Therefore, the weights mentioned above are only used to describe the optimal overall network performance. However, in actual time-sensitive warfare, time metrics must be constantly monitored. Therefore, based on the above link weight definitions, a global timeliness metric is introduced when evaluating each kill chain scheme. The global timeliness metric is the response delay metric, which is the sum of command decision-making time and kill chain closure time. Decision time refers to the time taken by the command and control unit from acquiring target information to selecting and issuing an interception scheme. Link closure time refers to the time taken from the first detection of the target to its destruction (including physical processes such as detection, communication, and flight of attack equipment). By incorporating decision time into the timeliness metric, along with the pure physical interception time, as an optimization objective, the kill chain scheme can achieve the shortest response time while ensuring the damage effect. This application adds the two together as the total reaction time metric, measuring the complete cycle from "target detection to strike completion" as one of the important considerations for optimization. This design highlights the impact of decision efficiency on overall interception timeliness.

[0123] Suppose a candidate kill chain scheme l0 is t1-o1-d2-a5-t1. The system will calculate:

[0124] ① Communication loss factor value, i.e., communication loss index value The calculation formula is:

[0125] ② Deployment Density: Assuming the distance between detection equipment o1 and command and control equipment d2 is 50km, the distance between command and control equipment d2 and strike equipment a5 is 50km, and the distance between strike equipment a5 and target t1 is 20km, then the average physical distance of the three segments on the link is 40km. If the average distance of the other option is 80km, then the latter deployment is obviously dispersed, which is detrimental to timeliness.

[0126] ③ Damage effectiveness index: Assuming the success rate of sensor detection is OTP ij =0.9, hit probability of hitting equipment (ATP) lj =0.8, Command relay reliability (i.e., the probability of correct command decisions) dp k =0.99, then the overall damage probability is approximately 0.9 × 0.99 × 0.8 ≈ 0.712 (about 71.2%). If the accuracy of the striking equipment (e.g., hit deviation) is also very good, then the damage effectiveness index of this scheme can be set at 0.712 (or further calculated in combination with accuracy).

[0127] ④ Response Delay: If the detection delay of O1 is 0.1 seconds, the communication time from O1 to D2 is 0.2 seconds, the decision time of D2 is 1.0 second, the communication time from D2 to A5 is 0.2 seconds, and the missile flight time is 5.0 seconds, then the total closure time is approximately 6.5 seconds. Separating the decision time of 1.0 second, the global timeliness index is 1.0 second for decision + 5.5 seconds for execution = 6.5 seconds. If another scheme has a shorter decision time but a longer flight time, the total time may be similar. This application prefers the scheme with the shorter total time.

[0128] In step 205 above, the optimal interception scheme for air defense and anti-missile interception operations is recommended based on the evaluation results corresponding to all kill chains. Specifically, this includes: using the improved TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) algorithm, the optimal interception scheme for air defense and anti-missile interception operations is recommended based on the evaluation results corresponding to all kill chains. Specifically, this includes the following steps: for each kill chain, the index values ​​of each indicator of the kill chain are normalized to obtain the normalized index values ​​of each indicator of the kill chain; the normalized index values ​​of each indicator of the kill chain are weighted and summed to obtain the comprehensive score corresponding to the kill chain; all kill chains are sorted according to the comprehensive scores of all kill chains to obtain the comprehensive recommendation order result; all kill chains are sorted according to the index values ​​of each indicator to determine the recommendation order result under each indicator; the kill chain in the first order from largest to smallest among the comprehensive recommendation order result and the recommendation order result under each indicator is the optimal interception scheme.

[0129] After obtaining the aforementioned index values ​​for each candidate kill chain, this application introduces a multi-attribute decision-making method to rank and optimize the schemes. The multi-attribute decision-making method can employ an improved TOPSIS algorithm: first, each index is normalized and weighted (this weighting can be flexibly adjusted according to operational needs; for example, response time can be given more weight to high-threat targets); then, the distance of each candidate kill chain scheme relative to the "ideal optimal" and "ideal worst" schemes is calculated using weighted processing to obtain a comprehensive score. Finally, the schemes are ranked according to their scores, and a recommended order is output. The optimal kill chain scheme is selected based on the recommended order as a decision suggestion for air defense and anti-missile interception operations. When necessary, multiple chains can be selected to form a kill network for simultaneous interception, improving the interception success rate. When multiple threats or high-threat targets exist, a combination of schemes that prioritize response time or damage probability can be selected based on weight preferences.

[0130] For example, the system can provide rankings such as "Option A ranks 1st, Option B ranks 2nd..." and the rankings of each option under different single indicators (e.g., Option B has the lowest communication delay, Option A has the highest probability of destruction, etc.), providing commanders with decision-making references (e.g., ...). Figure 3 As shown, different evaluation criteria are selected to rank the schemes for reference. Through this intelligent recommendation mechanism, this application achieves "decision-based strike": when a target appears, the system automatically provides the best interception scheme, and the commander only needs to confirm to quickly execute the interception, without the need for time-consuming manual selection among multiple schemes.

[0131] After the above indicators are calculated, the system enters the decision support phase. Typically, the system comprehensively considers all indicators to provide a comprehensive score and ranking. For example, if target T may have three kill chains, A, B, and C, and the comprehensive scores calculated using the TOPSIS algorithm are 0.85, 0.80, and 0.60 respectively, then the ranking is A>B>C. The system will recommend option A as the preferred interception option. It can also list the performance of each option on individual indicators (e.g., option B has the lowest communication loss, option A has the highest probability of destruction), for the commander's reference in decision-making. If options A and B are not significantly different, the commander can also decide to use both links A and B simultaneously (forming kill network redundancy) to improve the success rate. Regardless of the chosen option, the system ultimately enters the action execution phase (Act) to intercept and strike the target, and the results are fed back into the next OODA loop.

[0132] It should be noted that this application achieves a high degree of automation and intelligent optimization throughout the entire decision-making and recommendation process. Once a target is identified, the system can automatically complete the entire process from detection chain extraction, communication link matching, scheme generation to optimal recommendation, with extremely short decision-making time. This effectively solves the problem of missed opportunities that may arise from traditional manual decision-making, gaining a competitive advantage in time-sensitive operations.

[0133] Application of Complex Network Algorithms: This application extensively utilizes complex network algorithms and graph theory optimization techniques in its implementation. For example, incremental graphs combined with greedy algorithms are used to pre-calculate the shortest paths between nodes in the ODA network, and heuristic algorithms are employed to optimize the detector layout to cover the maximum area. Furthermore, incremental graph algorithms are introduced to handle dynamic nodes, improving the algorithm's efficiency under adversarial events. Simultaneously, the robustness analysis of the network (e.g., the impact of node failure on the overall network) can draw upon node importance measurement methods from complex networks, considering redundant configurations during the early warning network deployment phase to enhance wartime survivability.

[0134] Multi-target cooperative interception: When multiple enemy targets t1, t2,... appear simultaneously on the battlefield, the method in this application can perform multi-target kill chain construction and recommendation in parallel. The warning network will generate a set of candidate links for each target and evaluate and rank them separately. When resources are sufficient, the best solution can be selected for each target for simultaneous interception; if there are resource conflicts (e.g., two targets need to compete for the same strike equipment A), resources can be comprehensively allocated according to the target threat level and solution score to ensure priority interception of high-threat targets. It also supports scheduling optimization when one type of detection equipment detects multiple targets and one type of strike equipment intercepts multiple targets.

[0135] System Implementation: This application can be integrated into the software module of a time-sensitive combat command and control system as an "intelligent interception planning and recommendation" function. By linking real-time radar detection information and strike equipment status data, the above calculations are completed automatically. The output can be a graphical network diagram of interception schemes and textual suggestions. Commanders can also adjust the weights of evaluation indicators or make a final decision on recommended schemes based on practical experience. This system can also continuously improve through machine learning based on combat data, such as adjusting the damage probability estimation based on historical interception results, to more accurately evaluate the effectiveness of the schemes.

[0136] Through the detailed description of the above implementation methods, it is clear how this application enables the rapid construction and optimization recommendation of kill chains involving multiple equipment in dynamic and complex time-sensitive combat scenarios. Experimental simulations show that, compared with traditional methods, this application significantly shortens the time from target appearance to interception, demonstrating superior performance in high-speed target interception and saturation attack defense.

[0137] The method provided in this application is applied to a computerized air defense and anti-missile command system. It can automatically execute steps 201 to 205 above when receiving radar or other detection warning signals, and present the recommended kill chain / network scheme to the commander in graphical and textual form. The system supports manual adjustment of parameter weights and scheme selection, and has continuous learning and optimization capabilities to adapt to different battlefield environments and operational needs.

[0138] This application also provides an application scenario in which the above-mentioned method for constructing and recommending equipment system adversarial networks based on dynamic temporal events is applied. Specifically, the method for constructing and recommending equipment system adversarial networks based on dynamic temporal events provided in this embodiment can be applied in air defense and anti-missile command scenarios. Air defense and anti-missile command scenarios include a request sending stage, an interception scheme recommendation link, and a content distribution stage. Requests to be processed enter the interception scheme recommendation link from the request sending stage, obtain the corresponding optimal interception scheme through human-machine collaboration, and then enter the downstream content distribution stage. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events provided in this embodiment belongs to the interception scheme recommendation link. Specifically, in the process of recommending interception schemes for requests, an initial detection-command-strike early warning network can be constructed to continuously monitor dynamic temporal events on the battlefield. When the event type of the dynamic temporal event is an equipment status event, the initial detection-command-strike early warning network is updated. When the event type of the dynamic temporal event is a chained event, a closed detection-command-strike link set containing enemy targets is generated. Based on a multi-dimensional evaluation index system, each kill chain is evaluated to obtain the evaluation result corresponding to each kill chain. Based on the evaluation results corresponding to all kill chains, the optimal interception scheme for air defense and anti-missile interception operations is recommended.

[0139] Example 2

[0140] Based on the same inventive concept, this application also provides a device for constructing and recommending equipment system adversarial networks for dynamic temporal events, which is used to implement the above-mentioned method for constructing and recommending equipment system adversarial networks for dynamic temporal events. The solution provided by this device is similar to the solution described in the above-described method. Therefore, the specific limitations of one or more embodiments of the device for constructing and recommending equipment system adversarial networks for dynamic temporal events provided below can be found in the limitations of the method for constructing and recommending equipment system adversarial networks for dynamic temporal events described above, and will not be repeated here.

[0141] In one exemplary embodiment, a device for constructing and recommending equipment system adversarial networks based on dynamic temporal events is provided, comprising the following modules:

[0142] The initial detection-command-strike early warning network construction module is used to construct the initial detection-command-strike early warning network. The network nodes in the initial detection-command-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system. The links in the initial detection-command-strike early warning network represent the communication relationships between the equipment.

[0143] The Initial Detection-Command-Strike Warning Network Update Module is used to continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, the Initial Detection-Command-Strike Warning Network is updated to obtain the updated Detection-Command-Strike Warning Network.

[0144] The closed detection-command-strike link set generation module is used to generate a closed detection-command-strike link set containing enemy targets when the event type of the dynamic time sequence event is a chained event; the closed detection-command-strike link set includes several kill chains; the chained event is when the detection equipment detects the appearance of an enemy target.

[0145] The kill chain assessment module is used to evaluate each kill chain based on a multi-dimensional evaluation index system, and obtain the evaluation result corresponding to each kill chain. The evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system.

[0146] The optimal interception scheme recommendation module is used to recommend the optimal interception scheme for air defense and anti-missile interception operations based on the evaluation results corresponding to all kill chains.

[0147] This application, through the above technical solution, can significantly improve the efficiency of kill chain construction and response in time-sensitive combat dynamic environments, specifically in the following ways:

[0148] ① Introducing the concept of a "pre-deployment early warning network": This involves pre-integrating detection, command, and strike units to construct an ODA network, ensuring battlefield resources are in a coordinated and interconnected state. On one hand, it maximizes the utilization of equipment performance capacity (all available equipment plays a role in the network), avoiding repetitive work of temporary connection calculations after a target appears. On the other hand, layered optimization (optimizing the early warning network structure during peacetime and specific links during wartime) reduces the problem scale and improves design efficiency. When a target appears, there is no need to search the entire domain from scratch; only incremental updates to the local area are required, greatly reducing the computational dimensionality and time, and solving the problems of excessive dimensionality and slow response in existing methods.

[0149] ② Dynamic Event-Driven Adaptive Network: Integrating complex network optimization theory, this approach allows the air defense system to rapidly adjust to dynamic battlefield changes. For non-time-sensitive events, event-driven local reconstruction is employed to maintain the continuous optimization of the early warning network; for time-sensitive targets, an improved incremental graph algorithm is used to construct links in real time. In complex and ever-changing battlefield networks, the method described in this application can adapt to equipment additions and subtractions and environmental changes, consistently ensuring the connectivity and efficiency of the detection-command-strike link.

[0150] ③ Innovative Link Weight Model and Evaluation Index Design: A link weight model is constructed by comprehensively considering various factors affecting detection and strike effectiveness, making the evaluation of each link more complete and accurate. In particular, a global timeliness index (decision time + closure time) is introduced to quantify the time efficiency of the kill chain, giving the system a clear direction for optimization of "rapid response" in scheme evaluation. This time index has not been fully valued in traditional methods and is one of the key innovations of this application in improving the success rate of intercepting time-sensitive targets.

[0151] ④ Intelligent Recommendation and Decision-Making Mechanism: Through multi-indicator fusion evaluation and improved decision-making algorithms, multiple candidate kill chain schemes are automatically sorted to achieve the goal of "the chain is formed as soon as the target appears, and the decision is made to strike immediately." When a high-speed target approaches, this application can recommend the best interception scheme the instant the target appears, reducing human decision-making intervention and greatly compressing the decision-making time of the OODA loop. This not only improves the rapid response capability of time-sensitive operations but also reduces the decision-making burden of commanders under high-pressure situations, which has important significance for intelligent warfare.

[0152] In summary, the method for constructing and recommending equipment system adversarial networks under dynamic time-series events provided in this application has outstanding advantages such as dynamic adaptation, multi-equipment collaboration, and rapid closed-loop operation. It can greatly improve the timeliness and reliability of time-sensitive operations and has broad application prospects in informationized battlefield environments.

[0153] Example 3

[0154] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0155] Example 4

[0156] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0157] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0158] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for constructing and recommending equipment system adversarial networks based on dynamic temporal events, characterized in that, The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events includes: Construct an initial detection-command and control-strike early warning network; the network nodes in the initial detection-command and control-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system; the links in the initial detection-command and control-strike early warning network represent the communication relationships between the equipment. Continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, update the initial detection-command-strike early warning network to obtain the updated detection-command-strike early warning network. When the event type of a dynamic time-series event is a chained event, a closed detection-command-strike chain set containing enemy targets is generated; the closed detection-command-strike chain set includes several kill chains; a chained event is when the detection equipment detects the appearance of an enemy target; Based on a multi-dimensional evaluation index system, each kill chain is evaluated to obtain the evaluation result corresponding to each kill chain; the evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system. Based on the assessment results corresponding to all kill chains, the optimal interception scheme for air defense and anti-missile interception operations is recommended.

2. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 1, characterized in that, Building an initial detection-command-strike early warning network includes: Based on graph theory, detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system are used as network nodes. OD links and DA links are constructed based on the communication relationships between the equipment. The OD link is the communication link between the detection equipment and the command and control node; the DA link is the communication link between the command and control node and the strike equipment.

3. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 1, characterized in that, Constructing an initial detection-command-strike early warning network, specifically including: Based on equipment performance and capacity constraints, an initial detection-command and control-strike early warning network is constructed.

4. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 1, characterized in that, The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events also includes: After adding enemy targets as new network nodes to the updated detection-command-strike early warning network, OT links and AT links are added to the detection-command-strike early warning network. The OT link is the detection link of the detection equipment to the target, and the AT link is the strike link of the strike equipment to the target.

5. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 1, characterized in that, The multi-dimensional evaluation index system includes communication loss, deployment density, damage performance, and response latency indicators.

6. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 4, characterized in that, When the event type of a dynamic time-series event is a chained event, a closed detection-command-strike chain set containing enemy targets is generated, specifically including: When the event type of the dynamic time-series event is a chained event, an improved incremental graph algorithm is used to generate a closed detection-command-strike link set containing enemy targets. Specifically, enemy targets are dynamically added as new nodes to the updated detection-command-strike early warning network, connecting all O nodes that have detected enemy targets and all A nodes that can attack enemy targets, forming OT links and AT links, resulting in a further updated detection-command-strike early warning network; O nodes are detection equipment nodes, and A nodes are strike equipment nodes; when no equipment status event occurs, the updated detection-command-strike early warning network is the initial detection-command-strike early warning network. Search for kill chains originating from and returning to enemy targets within the updated detection-command-strike early warning network; all kill chains constitute a closed set of detection-command-strike links.

7. The method for constructing and recommending equipment system adversarial networks based on dynamic temporal events according to claim 1, characterized in that, Based on the evaluation results corresponding to all kill chains, the optimal interception scheme for air defense and missile defense interception operations is recommended. Specifically, this includes: utilizing the improved TOPSIS algorithm, the optimal interception scheme for air defense and missile defense interception operations is recommended based on the evaluation results corresponding to all kill chains. For each kill chain, the index values ​​of each index of the kill chain are normalized to obtain the normalized index values ​​of each index of the kill chain. The comprehensive score corresponding to the kill chain is obtained by weighted summation of the normalized index values ​​of each index in the kill chain. All kill chains are sorted according to their comprehensive scores to obtain a comprehensive recommendation order result; The kill chains are sorted according to their index values ​​under each metric to determine the recommended order for each metric. The kill chain ranked first in descending order of the combined recommended order and the recommended order for each metric is the optimal interception scheme.

8. A device for constructing and recommending equipment system adversarial networks based on dynamic temporal events, characterized in that, The device for constructing and recommending equipment system adversarial networks based on dynamic temporal events includes: The initial detection-command-strike early warning network construction module is used to construct the initial detection-command-strike early warning network. The network nodes in the initial detection-command-strike early warning network correspond to the detection equipment, command and control equipment, and strike equipment in the air defense and anti-missile system. The links in the initial detection-command-strike early warning network represent the communication relationships between the equipment. The Initial Detection-Command-Strike Warning Network Update Module is used to continuously monitor dynamic time-series events on the battlefield. When the event type of the dynamic time-series event is an equipment status event, the Initial Detection-Command-Strike Warning Network is updated to obtain the updated Detection-Command-Strike Warning Network. The closed detection-command-strike link set generation module is used to generate a closed detection-command-strike link set containing enemy targets when the event type of the dynamic time-series event is a chained event; the closed detection-command-strike link set includes several kill chains; the chained event is when the detection equipment detects the appearance of an enemy target; The kill chain assessment module is used to evaluate each kill chain based on a multi-dimensional evaluation index system, and obtain the evaluation result corresponding to each kill chain. The evaluation result includes the index values ​​of each index in the multi-dimensional evaluation index system. The optimal interception scheme recommendation module is used to recommend the optimal interception scheme for air defense and anti-missile interception operations based on the evaluation results corresponding to all kill chains.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the method for constructing and recommending equipment system adversarial networks based on dynamic temporal events as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the method for constructing and recommending equipment system adversarial networks based on dynamic temporal events as described in any one of claims 1-7.

Images