Natural language requirement-based automatic reconfiguration method for equipment system-of-systems model

Abstract

This invention belongs to the field of systems engineering and provides an automatic reconstruction method for equipment system composition models based on natural language requirements. Existing technologies rely on manual experience and fixed rules for configuring equipment system composition models, resulting in low design efficiency and difficulty in handling interface compatibility and cascading failure risks in dynamic battlefield environments. This invention defines an equipment system composition model reconstruction dictionary and uses templates to parse natural language requirements. Based on an equipment system database and a domain meta-model library, it generates preliminary solutions and employs a discrete multi-objective particle swarm optimization algorithm to find optimal solutions under multiple objectives such as weight, cost, interface compatibility, dynamic behavior compliance, and system topological resilience, ultimately generating a reconstructed equipment system composition model. This invention achieves automated and intelligent reconstruction of equipment system composition models, improving design efficiency and solution robustness.

Description

Technical Field

[0001] This invention belongs to the fields of systems engineering and architecture engineering, and specifically relates to an automatic reconstruction method for equipment system composition models based on natural language requirements. Background Technology

[0002] Model-based systems engineering (MBSE) uses digital models as its core and runs through the entire operational system design process, including operational system design, requirements demonstration, and digital prototype design, which can effectively improve the development efficiency of operational systems.

[0003] In the iterative design of combat systems, in order to cope with the changing battlefield situation and mission requirements, it is necessary to design different combat system composition models to meet the capability requirements of the combat system. As a complex technical system, the modeling of the equipment system is to digitally and structurally express the complex real equipment and their interrelationships. This is crucial for achieving the scientific design, efficient coordination, dynamic evaluation and rapid iteration of the equipment system.

[0004] This process involves the reconstruction of the equipment system composition model. The speed of this reconstruction determines the equipment's physical and tactical indicators' ability to quickly adapt to battlefield evolution, thus significantly impacting the overall effectiveness of the equipment system. Traditional methods for reconstructing equipment system composition models primarily rely on human experience and manual configuration, resulting in low efficiency, poor reliability, and a high risk of configuration errors. Furthermore, manual configuration easily overlooks the physical interface compatibility (such as power type, communication protocol, and data format) and energy supply constraints between equipment, leading to technical malfunctions such as communication interruptions, overloads, and response delays in subsequent actual operation. Therefore, there is an urgent need for an automated method to achieve adaptive reconstruction of the equipment system composition model.

[0005] Analysis of existing methods reveals the following main drawbacks: There is a lack of methods for reconstructing equipment system composition models. Existing technologies are mostly focused on top-level design methods for combat systems, system confrontation simulation, and effectiveness assessment, but have not yet proposed systematic technical means for the reconstruction process of equipment system composition models. In particular, when facing changes in system capability requirements or adjustments in combat scenarios, there is a lack of methods to support the rapid reconstruction of equipment system composition models, making it difficult to iterate system designs in a timely manner.

[0006] Lacking intelligent optimization and adaptive capabilities, existing methods for reconstructing equipment system configuration models primarily rely on human experience and fixed rules, lacking intelligent optimization mechanisms and adaptive capabilities. When faced with complex and ever-changing combat mission requirements and dynamic battlefield environments, existing methods struggle to automatically generate optimal system configuration schemes, often requiring multiple manual iterations and adjustments. This is inefficient, time-consuming, and easily influenced by the subjective factors of designers, making it difficult to guarantee the stability and rationality of the system design results.

[0007] Existing model reconstruction methods lack the ability to process refactoring requests from engineers. They primarily rely on structured data or predefined parameters, lacking the ability to parse and process unstructured descriptions. When users input operational mission requirements or equipment configuration constraints in natural language, existing methods cannot effectively identify and transform them, resulting in inflexible requirement expression and limiting the practicality and ease of use of system composition model reconstruction. Summary of the Invention

[0008] The purpose of this invention is to provide an automatic reconstruction method for equipment system composition models based on natural language requirements. This method solves the technical problems of low design efficiency, difficulty in responding to dynamic battlefield requirements, and easy occurrence of interface incompatibility and cascading failures caused by the reliance on manual experience and fixed rules in the reconstruction of existing equipment system composition models.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows: The method for automatic reconstruction of equipment system composition models based on natural language requirements includes the following steps: Step S1: Organize the equipment system database to form a structured form containing equipment name, type, combat and technical indicators and indicator values. The combat and technical indicators include static indicators, interface attributes and dynamic behavior functions that change with the environment. Step S2: Construct a domain meta-model library based on the equipment system database. The domain meta-model library includes a graph meta-model, an object meta-model, and a relation meta-model, wherein the object meta-model encapsulates the static attribute group, interface attribute group, and dynamic behavior attribute group of the equipment. Step S3: Define a dictionary for identifying reconstruction requirement text, the dictionary containing the parts of speech and content of equipment type, performance indicators, interface terms and dynamic behavior terms; Step S4: Define a template library for reconstructing equipment composition models. The template library shall include at least: equipment system type declaration template, equipment system overall target template, equipment addition constraint template, equipment system overall performance constraint template, equipment quantity constraint template, interface compatibility constraint template, and dynamic behavior constraint template. Step S5: Construct an equipment system composition model library based on the domain meta-model library, and instantiate an equipment object meta-model containing specific attribute values; Step S6: Receive the structured natural language reconstruction requirement statement input by the engineer, and use the dictionary and template library to perform word segmentation, part-of-speech tagging, block segmentation and information extraction to obtain equipment system type declaration information, overall target information, added constraint information, overall performance constraint information, quantity constraint information, interface attribute information and dynamic behavior information; Step S7: Based on the extracted information, match equipment instances that meet performance constraints, interface compatibility constraints, and dynamic behavior constraints from the equipment system composition model library to generate a preliminary equipment system composition model scheme. Step S8: Construct an objective function based on the overall objective information, construct constraint conditions based on the overall performance constraint information, quantity constraint information, interface attribute information and dynamic behavior information, and optimize the preliminary scheme using the discrete multi-objective particle swarm optimization algorithm to obtain the Pareto optimal solution set that satisfies the multi-objective trade-off; Step S9: Based on the scheme selected from the Pareto optimal solution set, match the equipment data in the equipment system composition model library to generate the reconstructed equipment system composition model.

[0010] Furthermore, in step S2, the attributes of the object meta-model are stored in a nested structure, and the interface attribute group includes at least: power supply type enumeration, communication protocol enumeration, data format enumeration, physical size value and installation method enumeration; the dynamic behavior attribute group includes at least: performance degradation function identifier, function parameter value and applicable environment range.

[0011] Furthermore, the template in step S4 is constructed based on a predefined part-of-speech sequence and is used to map unstructured text into structured information blocks extracted in step S6.

[0012] Furthermore, step S6 further includes: performing syntactic pattern matching and information extraction using regular expression rules, wherein the rules define extraction patterns for indicator names, comparison relationships, and thresholds.

[0013] Furthermore, the multi-objective particle swarm algorithm in step S8 adopts mixed integer encoding, the position vector includes equipment type index, model index and quantity, and multiple constraints are handled through the Deb constraint dominance principle.

[0014] Furthermore, the objective function in step S8 includes: Total weight minimization objective: ,in For the first The weight of each piece of equipment; Total cost minimization objective: ,in For the first The cost of each piece of equipment; The goal is to maximize interface compatibility. ,in For the first The interface compatibility coefficient of each piece of equipment; The goal of maximizing dynamic behavioral compliance: ,in For the first The compliance coefficient of dynamic behavior of each piece of equipment.

[0015] Furthermore, the interface compatibility coefficient Based on equipment The compatibility of the power interface, communication protocol, data format, and physical dimensions with other equipment or carriers within the system is calculated, specifically as the arithmetic mean of the compatibility scores for each item.

[0016] Furthermore, the dynamic behavior compliance coefficient According to equipment The inherent performance degradation function is used to calculate the degree to which its output value meets the demand threshold under specified environmental variables. The calculation formula is a piecewise function.

[0017] Furthermore, the constraints in steps S7 and S8 include: Performance constraints: such as minimum quantity of equipment type, minimum detection range, and minimum kill probability; Interface constraints: such as compatibility of power interface types, communication protocol versions, and data format standards; Dynamic behavior constraints: such as the performance parameters of the equipment under specified environmental conditions (such as rainfall intensity, target speed) not being lower than the preset threshold.

[0018] Furthermore, the reconstructed equipment system model generated in step S9 consists of equipment components that are optimized schemes automatically selected and combined from the candidate pool by an algorithm, under the premise of satisfying all constraints.

[0019] Furthermore, in step S8, the multi-objective optimization algorithm incorporates a system topological resilience assessment based on complex network graph theory when calculating the objective function, specifically as follows: Construct a system functional coupling weight connection matrix Among them, the source equipment node To the target equipment node weight The calculation formula is:

[0020] in, For interface matching factors, when node With nodes When the interface is fully compatible ,otherwise ; destination node Peak power consumption For the source node Maximum power supply capacity; For data exchange bandwidth between nodes, This represents the system's maximum communication bandwidth. To control the latency of command interaction; , , These are the weighting coefficients for energy flow, data flow, and control flow, respectively. ; Based on the connection matrix Calculate the Laplace matrix And extract its algebraic connectivity. ; Combining dynamic behavioral compliance coefficients Calculate the probability of cascade failure. ; Calculate the topological resilience index of the system and will This is set as the fifth optimization objective.

[0021] Furthermore, in each iteration of the multi-objective optimization algorithm, if the topological toughness index of a certain particle is... Below the preset safe survival threshold Then, an exponential penalty term is introduced for the degree of constraint violation of the particle:

[0022] in, To penalize the gain constant, this mechanism forces the particle swarm to evolve towards a system topology with distributed redundancy, thus avoiding equipment combination schemes that produce single-point vulnerabilities.

[0023] Compared with the prior art, the present invention has the following beneficial effects: This invention establishes a standardized equipment system database and domain meta-model library, and defines a reconstruction requirement dictionary and template library. It automatically parses unstructured natural language requirements input by engineers into structured constraint information, fundamentally solving the technical problems of inconsistencies in requirement understanding, low configuration efficiency, and susceptibility to errors caused by reliance on human experience in existing technologies. Building upon this, the invention introduces a discrete multi-objective particle swarm optimization algorithm. Using total weight, total cost, interface compatibility, dynamic behavior compliance, and system topological resilience index as optimization objectives, it automatically seeks optimal solutions for the initially generated equipment system composition model under the joint constraints of performance, interface, and dynamic behavior. This technique frees the equipment system composition model reconstruction process from dependence on manual iterative adjustments, significantly improving the response speed and rationality of equipment system design schemes to dynamic battlefield requirements. Furthermore, this invention incorporates complex network graph theory into the optimization mechanism. By constructing a connection matrix based on functional coupling weights and calculating algebraic connectivity and cascade failure probability, it can quantitatively assess and proactively avoid the risk of system cascade failure caused by the performance degradation of a single node. The algorithm-driven model for equipment system configuration converges towards a distributed defense architecture with redundant data links, balanced energy load, and high resilience, effectively enhancing the survivability and mission reliability of the equipment system in harsh battlefield environments. In summary, this invention provides a fully automated reconfiguration method from requirements analysis and model matching to multi-objective resilience optimization, achieving substantial progress in improving design efficiency, ensuring the scientific validity of the solution, and enhancing system robustness. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of the method flow described in this invention.

[0026] Figure 2 This is a schematic diagram of the method framework described in this invention.

[0027] Figure 3 This invention provides a data form for the equipment system.

[0028] Figure 4 This invention provides a model library for the composition of equipment systems.

[0029] Figure 5 The preliminary equipment system configuration model form of this invention.

[0030] Figure 6 The reconstructed equipment system configuration model of this invention. Detailed Implementation

[0031] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0032] The following is in conjunction with the appendix Figures 1-6 The embodiments of the present invention will be described in detail below.

[0033] Example 1: This example discloses an automatic reconstruction method for equipment system composition model based on natural language requirements. This example further elaborates on the complete process from equipment database sorting, requirement analysis, model matching to multi-objective optimization solution.

[0034] The specific steps are as follows: Step S1: Organize the equipment system database; By combining the existing, under-development, and newly launched equipment in this field, an equipment system database was compiled and structured forms were created, with the form format as follows: Figure 3 As shown, it contains 8 headers, namely serial number, equipment name, equipment type, equipment status, equipment combat skills indicators, indicator type, measurement, and indicator value.

[0035] The equipment name corresponds to the name of the specific model of the equipment; the equipment type corresponds to the category of the equipment model, such as radar detection equipment, optoelectronic detection equipment, radio jamming equipment, laser weapons, etc. Equipment of the same category contains the same equipment tactical and technical indicators; the equipment status corresponds to four values: in service, under development, newly launched, and newly developed; the equipment tactical and technical indicators correspond to the core performance indicators of the equipment, such as detection range, working time, kill probability, interface data, etc.; the indicator type corresponds to upper limit type, lower limit type, and center point type; the measurement corresponds to the unit corresponding to the equipment tactical and technical indicators; and the indicator value corresponds to the actual parameter value of the equipment indicator. The equipment interface data includes: power interface type (e.g., 28V DC, 220V AC), communication protocol version (e.g., 1553B, Ethernet, CAN bus), data format standard (e.g., XML Schema, 1553B message format, CAN data frame), physical dimensions (length, width, height, mounting holes), peak power consumption, average power consumption, and maximum power supply capacity. In addition to assigning real numbers, the index values ​​can also store mathematical functions or tables showing how the performance parameters of each piece of equipment change with the environment, such as the attenuation curve of radar detection range with visibility and rainfall intensity:

[0036] in, The actual detection distance is in meters. The reference detection distance under standard conditions is expressed in meters. The attenuation coefficient is determined by the inherent characteristics of the equipment. Rainfall intensity is expressed in millimeters per hour (mm / h).

[0037] The weapon's kill probability as a function of target velocity:

[0038] in, The kill probability has a value range of [0,1]. The steepness coefficient is the control curve at... The slope nearby; The target speed is expressed in meters per second (m / s). The half-effective velocity is the target velocity when the kill probability is 0.5, and the unit is meters per second (m / s).

[0039] This data is stored in a structured format for later optimization and retrieval.

[0040] Step S2: Construct a domain meta-model library for the equipment system composition model; Based on the equipment system database, a domain meta-model library for the equipment system composition model is constructed, including a graph meta-model, an object meta-model, and a relation meta-model. The graph meta-model serves as the top-level framework, defining the overall topological structure of the equipment system composition model (e.g., "equipment system" as the root object, containing several "equipment type" objects, and the "containment" relationships between them). The object meta-model describes the individual / type attributes of equipment and is divided into two categories: System-level object meta-model: "Equipment System" object meta-model (describes system-level attributes, such as name, task type, etc.); Equipment Types and Object Meta-Models: Each type of equipment (such as radar detection equipment, optoelectronic detection equipment) corresponds to an object meta-model, encapsulating the static attribute groups (such as detection range, working time), interface attribute groups (power supply, communication, etc.), and dynamic behavior attribute groups (performance variation functions with the environment) of that type of equipment; the relational meta-model is used to describe the relationships between equipment (such as "containment" relationship, indicating that "equipment system" contains a certain type of "equipment object meta-model"), and the two ends of the relationship can be connected to object meta-models. This domain meta-model library contains one graph meta-model: the equipment system composition graph meta-model, which contains several object meta-models, including the "equipment system" object meta-model (this object meta-model contains a "name" attribute, and the attribute value type is string) and the same number of object meta-models as the equipment types in step S1 (that is, each type of equipment corresponds to one object meta-model; taking radar detection equipment as an example, this equipment type corresponds to the "radar detection equipment" object meta-model, and this object meta-model contains static indicator attribute groups, such as "detection range" attribute, "working time" attribute, "weight" attribute, "power consumption" attribute, etc., and the attribute value type is string); Interface attribute groups, such as power supply type (enumeration), communication protocol (enumeration), data format (enumeration), physical dimensions (numerical value), installation method (enumeration), etc.; dynamic behavior attribute groups, such as performance degradation function identifier, function parameter value, applicable environment range, etc., are referred to below as the object meta-models of various equipment types. The attribute groups of the object meta-models are stored in a nested structure within the meta-model, providing a data foundation for subsequent compatibility checks. This meta-model contains one type of relation meta-model, namely the "contains" relation meta-model, which contains a "name" attribute. The starting end of the "contains" relation meta-model is connected to the "equipment system" object meta-model, and the ending end is connected to the object meta-models of various equipment types. The attributes of the object meta-models are stored in a nested structure (example structure as follows): "Equipment Type Name": "Radar Detection Equipment", "Static Attribute Group": {"Detection Range": "String (or value, e.g., "50km")", "Operating Time": "String (or value, e.g., "8h")", "Weight": "String (or value, e.g., "20kg")", "Power Consumption": "String (or value, e.g., "100W")"}, "Interface Attribute Group": {"Power Supply Type": "Enumeration (e.g., "28V DC" "220V AC")", "Communication Protocol": "Enumeration (e.g., "1553B" "Ethernet")", "Data Format": "Enumeration (e.g., "XML Schema" "1553B Message Format" "CAN Data Frame")", "Physical Dimensions": "Value (Length × Width × Height, unit mm, e.g., "300×200×150")", "Installation Method": "Enumeration (e.g., "Vehicle-mounted" "Back-mounted" "Fixed")", "Peak Power Consumption": "Value (unit W)", "Maximum Power Supply Capacity": "Value (unit W)"}, "Dynamic Behavior Attribute Group": {"Performance Attenuation Function Identifier": "String (e.g., "radar_rain_attenuation")", "Function Parameter Value": "Value Set (e.g., α parameter value of radar attenuation function "0.05")", "Applicable Environment Range": "String (e.g., "Visibility ≥ 500m, Rainfall Intensity ≤ 10mm / h")"} Step S3: Define the equipment system composition model reconstruction requirement dictionary; The purpose of defining the model reconstruction dictionary is to provide the ability to identify professional terms in the requirements for reconstructing equipment system composition models, which helps to accurately match and understand the information in the text of the requirements for reconstructing equipment system composition models.

[0041] The definition of each word in the dictionary consists of two parts: word content and part of speech.

[0042] The dictionary definition includes (in {}, the content before the colon is the word, and the content after the colon is the part of speech): {new equipment system: new structure}, {should: should}, {possess: possess}, {max: max}, {min: min}, {more than: more than}, {more than or equal to: more than}, {less than: less than}, {less than or equal to: less than}, {add: add}, {equipment: equipment}, {type: type}, {over: over}, {below: below}, {equipment type: equipment type}, {contain: contain}. Interface-related terms: {power interface: power interface}, {communication protocol: communication protocol}, {data format: data format}, {installation method: installation method}.

[0043] Dynamic behavior related terms: {performance attenuation}, {environments sensitivity}, {kill probability correction}.

[0044] Step S4: Define the equipment composition model reconstruction requirement template library; The purpose of defining the equipment composition model refactoring requirement template library is to provide specific statement templates, allowing engineers to input equipment composition model refactoring requirements based on the corresponding templates. The equipment composition model refactoring requirement templates mainly include: equipment system type declaration template, equipment system overall objective template, equipment addition constraint template, equipment system overall performance constraint template, equipment quantity constraint template, interface compatibility constraint template, and dynamic behavior constraint template.

[0045] Equipment System Type Declaration Template: This template allows engineers to input structured statements and declare the types of equipment included in a new equipment system. The template is based on parts of speech.<new structure> <should> <contain>…, …。 An example is: The new equipment system should include radar detection equipment, laser weapons, machine gun equipment, conventional firepower weapons, and grenade launching equipment.

[0046] Equipment system overall objective template: The purpose of this template is to allow engineers to input structured statements and convert them into optimization objectives for the mathematical model in the algorithm.

[0047] The template is created based on parts of speech:<new structure> . <should> <possess><max / min>Examples include: a new equipment system should have the minimum weight, a new equipment system should have the minimum cost, etc.

[0048] Equipment Constraint Template: In the new equipment system composition model, it may be necessary to add some equipment that was not included in the previous model. This template allows engineers to input structured statements, extract the corresponding equipment information from the equipment system database, and encapsulate it into an object meta-model instance to add to the equipment system composition model. The template is created based on parts of speech.<new structure> … <should> <add><equipment type>… <equipment>An example is: The new equipment system should include high-power microwave combat weapons of the microwave weapon type.

[0049] Equipment System Overall Performance Constraint Template: This template allows engineers to input structured statements and convert them into constraints for the mathematical model in the algorithm. The template is created based on parts of speech.<new structure> <possess> … <should><more than / less than>Examples include: the new equipment system should have a detection range greater than 500 meters, and the new equipment system should have an operating time greater than 50 hours, etc.

[0050] Equipment Quantity Constraint Template: In new equipment system composition models, in addition to constraints on equipment performance indicators, there may also be constraints on the quantity of certain types of equipment. This template allows engineers to input structured statements and convert them into constraints for the mathematical model in the algorithm. The template is created based on parts of speech.<new structure> <should> <possess> … <type><over / below>…Example: The new equipment system should have more than three types of radar detection equipment, etc.

[0051] Interface Compatibility Constraint Template: This template allows engineers to input structured statements to constrain interface compatibility between equipment. The template is created based on parts of speech.<new structure> <should> <contain> <equipmenttype>…<power interface> …For example: The new equipment system should include radar detection equipment, and the power interface should be 28V DC.

[0052] Dynamic Behavior Constraint Template: This template allows engineers to input structured statements to constrain the performance of equipment under specific environments. The template is created based on parts of speech.<new structure> <possess> <equipmenttype>…<performance attenuation> …For example: The radar detection equipment of the new equipment system should have a detection range attenuation function that satisfies… ,in .

[0053] Step S5: Construct an equipment system composition model library; Based on the equipment system composition model, a domain meta-model library is constructed, which includes an instantiated "equipment system" object meta-model, a "containment" relation meta-model, and meta-models of various types of equipment objects.

[0054] by Figure 2 For example, instantiate one "Equipment System" object meta-model with the "Name" attribute set to "Individual Soldier Anti-Unequipped System"; instantiate one "Electro-optical Detection Equipment" object meta-model, two "Radar Detection Equipment" object meta-models, one "Laser Weapon" object meta-model, one "Machine-to-Machine Equipment" object meta-model, four "Conventional Firepower Weapons" object meta-models, one "Interceptor Missile Launching Equipment" object meta-model, and one "Power Supply Vehicle" object meta-model. Instantiate 11 "Containment" relation meta-models.

[0055] An example of the "Photoelectric Detection Equipment" object meta-model is: Anti-tracking Turntable. The Anti-tracking Turntable has the following attributes: "Name" = Anti-tracking Turntable, "Detection Distance" = 800 meters, "Weight" = 800 kilograms, "Cost" = 400,000 yuan, "Probability of Kill" = 0.9, and "Power" = 100 watts. Its interface attributes are: Power Supply Type "28V DC", Communication Protocol "Ethernet", Data Format "XML", Peak Power Consumption 100 watts, Maximum Power Supply Capacity 0 watts. Its dynamic behavior attribute is: None (assuming its performance does not change significantly with the environment).

[0056] Examples of the object meta-model for "Radar Detection Equipment" include: The UAV detection radar system has the following attributes: "Name" - UAV Detection Radar System; "Detection Range" - 750 meters; "Weight" - 22 kg; "Cost" - 60,000 yuan; "Operating Time" - 15 hours; and "Power" - 80 watts. Its interface attributes are: Power Supply Type - 28V DC; Communication Protocol - Ethernet; Data Format - XML; Peak Power Consumption - 80 watts; Maximum Power Supply Capacity - 0 watts. Its dynamic behavior attributes include: Performance Attenuation Function Identifier - "radar_rain_attenuation"; and parameters... .

[0057] The area array radar detection system has the following attributes: "Name" - Area Array Radar Detection System; "Detection Range" - 1200 meters; "Weight" - 75 kg; "Cost" - 120,000 yuan; "Operating Time" - 100 hours; and "Power" - 150 watts. Its interface attributes are: Power Supply Type - 28V DC; Communication Protocol - Ethernet; Data Format - MIL-STD-1553B; Peak Power Consumption - 150 watts; Maximum Power Supply Capacity - 0 watts. Its dynamic behavior attributes include: Performance Attenuation Function Identifier - "radar_rain_attenuation"; and parameters... .

[0058] An example of the "laser weapon" object meta-model is: a certain type of anti-drone laser weapon. The "Name" attribute of this anti-drone laser weapon is "certain type of anti-drone laser weapon," the "Detection Range" attribute is 1000 meters, the "Weight" attribute is 1000 kilograms, the "Cost" attribute is 300,000 yuan, the "Kill Probability" attribute is 0.83, and the "Power" attribute is 400 watts. Its interface attributes are: power supply type "220V AC," communication protocol "1553B," data format "CAN data frame," peak power consumption 400 watts, and maximum power supply capacity 0 watts. Its dynamic behavior attributes are: kill probability function identifier "kill_prob_velocity," and parameters... , .

[0059] An example of the "Machine-to-Machine Equipment" object meta-model is: Collision Anti-Drone System. The Collision Anti-Drone System has the following attributes: "Name" = Collision Anti-Drone System, "Detection Range" = 900 meters, "Weight" = 10 kg, "Cost" = 150,000 yuan, "Probability of Kill" = 0.75, and "Operating Time" = 0.5 hours. Its interface attributes are: Power Supply Type "24V DC", Communication Protocol "CAN Bus", Data Format "CAN Data Frame", Peak Power Consumption 120 watts, Maximum Power Supply Capacity 0 watts. Its dynamic behavior attributes are: None.

[0060] Examples of the "conventional firepower weapon" object metamodel include: The lightweight anti-drone remote control weapon station has the following attributes: "Name" - Lightweight anti-drone remote control weapon station; "Detection range" - 1000 meters; "Weight" - 75 kg; "Cost" - 200,000 yuan; and "Probability of kill" - 0.85. Peak power consumption is 200 watts.

[0061] The "Name" attribute of the man-portable air defense system is "Man-portable air defense system", the "Detection range" attribute is 1500 meters, the "Weight" attribute is 200 kg, the "Cost" attribute is 1,300,000 yuan, and the "Probability of Leakage" attribute is 0.9. The peak power consumption is 300 watts.

[0062] The anti-drone gun has the following attributes: "Name" = anti-drone gun, "Detection Range" = 500 meters, "Weight" = 6.5 kg, "Cost" = 20,000 yuan, and "Probability of Kill" = 0.7. Peak power consumption is 50 watts.

[0063] The "Name" attribute of a certain type of grenade launcher is "Type of grenade launcher", the "Detection Range" attribute is 400 meters, the "Weight" attribute is 4.5 kg, the "Cost" attribute is 30,000 yuan, and the "Probability of Lethality" attribute is 0.65. The peak power consumption is 0 watts.

[0064] An example of the "Interceptor Launching Equipment" object metamodel is: Directional Anti-Non-Interceptor Launcher. The "Name" attribute of the Directional Anti-Non-Interceptor Launcher is "Directional Anti-Non-Interceptor Launcher," the "Detection Range" attribute is 850 meters, the "Weight" attribute is 120 kilograms, the "Cost" attribute is 180,000 yuan, and the "Probability of Kill" attribute is 0.88. Its interface attributes are: Power Supply Type "28V DC," Communication Protocol "Ethernet," Data Format "XML," Peak Power Consumption 150 watts, Maximum Power Supply Capacity 0 watts. Its dynamic behavior attributes are: None.

[0065] An example of the "Power Supply Vehicle" object metamodel is: Field Power Supply Vehicle. The Field Power Supply Vehicle's "Name" attribute is "Field Power Supply Vehicle", its "Weight" attribute is 500 kg, and its "Cost" attribute is 50,000 yuan. Its interface attributes are: Power supply type "220V AC", maximum power supply capacity 600 watts, and peak power consumption 100 watts.

[0066] Among them, the "Name" attribute value of all 12 "Contains" relational meta-model instances is "Contains", and the starting point of all 12 "Contains" relational meta-model instances is connected to "Individual Soldier Anti-Equipment System", and the terminal is connected to the above twelve equipment instances respectively.

[0067] Step S6: Processing of requirements statements for reconstructing the equipment system composition model; The purpose of this step is to reconstruct textual requirements based on the equipment system composition model input by the engineer. Through word segmentation and part-of-speech tagging, block segmentation, and requirement information extraction, the system can effectively extract and identify the engineer's reconstruction requirements and provide input for subsequent reconstruction operations. In this case, the reconstruction requirements include equipment system type declaration requirements, overall equipment system objective requirements, equipment addition constraint requirements, overall equipment system performance constraint requirements, equipment quantity constraint requirements, interface attribute requirements, and dynamic behavior requirements.

[0068] The specific input requirement statement is as follows: The new equipment system should include radar detection equipment, laser weapons, machine gun equipment, conventional firepower weapons, and grenade launching equipment; The new equipment system should have a minimum weight. The new equipment system should have the lowest possible cost; The new equipment system should include more high-power microwave warfare weapons and equipment of the microwave weapon type. The new equipment system should have a detection range greater than 700 meters; The new equipment system should have an attack range of more than 500 meters; The new equipment system should have more than three types of radar detection devices; The new equipment system should include radar detection equipment, and the power interface should be 28V DC. The radar detection equipment in the new equipment system should have a detection range attenuation function that satisfies the following requirements. ,in .

[0069] The following details the specific steps for processing requirement statements in the equipment system composition model reconstruction: Step S6.1: Word segmentation and part-of-speech tagging; The purpose of this step is to accurately break down the structured natural language text input by engineers into independent word units and assign corresponding part-of-speech tags to each word unit, providing a foundation for subsequent segmentation and semantic analysis. The specific operational flow of this step is as follows: First, the statements required for the reconstruction of the equipment system composition model are preprocessed, removing punctuation marks and common stop words. This simplifies the text structure, facilitating subsequent word segmentation, part-of-speech tagging, and efficient processing of the overall statements. Then, using a predefined equipment system composition model reconstruction dictionary, the text is finely segmented and tagged with parts of speech, thus providing high-quality data support for subsequent natural language processing tasks.

[0070] Step S6.1.1: Syntactic pattern matching and information extraction; To enhance the ability to handle different input variations, this step extracts information based on regular expression rules defined in the template. For example, for performance constraints, the rules are defined as follows: (Equipment system|System)([e00-fa5]+)(Greater than|Less than|Not lower than|Not more than|Greater than or equal to|Less than or equal to)((.)?)([e00-fa5]+); If the input text is "System detection distance is not less than 800 meters", then the extracted index name is "detection distance", the relation is "not less than", the threshold is 800, and the unit is "meter".

[0071] If the unit is inconsistent with the unit stored in the equipment attribute, it will be automatically converted (e.g., meters to kilometers).

[0072] Step S6.2: Segmentation and Information Extraction; This step first segments the word segmentation results from the previous step into blocks based on the corresponding templates in the template library, dividing the equipment system composition model reconstruction requirement text into independent information blocks. Then, based on the block segmentation results, the information required for model reconstruction is extracted. This information will be input into the equipment system composition model reconstruction system. The information types include seven categories: equipment system type declaration information, overall equipment system objective information, equipment addition constraint information, overall equipment system performance constraint information, equipment quantity constraint information, interface attribute information, and dynamic behavior information.

[0073] The information extraction results in the case are as follows: Equipment system type declaration information: {New equipment system, radar detection equipment, laser weapons, machine gun equipment, conventional firepower weapons, grenade launching equipment}; Overall equipment system objectives: {New equipment system, minimum weight}, {New equipment system, minimum cost}; Equipment constraint information added: {New equipment system, high-power microwave combat weapon system}; Overall performance constraints of the equipment system: {New equipment system, detection range, greater than 700 meters}, {New equipment system, attack range, greater than 500 meters}; Equipment quantity constraint information: {New equipment system, 3 or more, radar detection equipment}; Interface attribute information: {New equipment system, radar detection equipment, power interface, 28V DC}; Dynamic behavior information: {New equipment system, radar detection equipment, performance degradation function,} , }

[0074] Step S7: Match the equipment system composition model library and generate a preliminary equipment system composition model scheme. Step S7.1: Match the equipment system to form a model library; The purpose of matching the equipment system composition model library is to select equipment object meta-model instances that meet performance constraints, interface compatibility constraints, and dynamic behavior constraints from the equipment system type declaration information extracted in step S6.2, based on the equipment system composition model library, so as to provide a candidate object pool for subsequent scheme generation.

[0075] S7.1.1 Performance constraint matching; Based on the "equipment system type" and "performance index requirements" (such as detection range, kill probability, etc.) extracted in step S6.2, instances whose performance attribute values ​​meet the constraints are matched from the "equipment type object meta-model" in the model library. For example, if the requirement is "radar detection range ≥ 50km", then instances with a "detection range" attribute value ≥ 50km in the "radar detection equipment" object meta-model are selected; if the requirement is "weapon kill probability ≥ 0.8", then instances with a "kill probability" attribute value ≥ 0.8 in the "laser weapon" object meta-model are selected.

[0076] S7.1.2 Interface compatibility constraint matching; Based on the "interface requirements" (such as power interface type, communication protocol, data format, etc.) extracted in step S6.2, instances with fully compatible interface attribute groups are matched from the "equipment type object metamodel" in the model library. The matching logic needs to cover the following dimensions: Power Interface: Filter instances whose "Power Interface Type" matches the requirements (or is compatible, such as "28V DC" is compatible with "24V DC" which requires a DC-DC conversion module, but no conversion is preset here as the initial matching).

[0077] Communication Protocol: Filter instances whose "communication protocol version" matches the requirements.

[0078] Data Format: Filter instances whose "data format standard" matches the requirements.

[0079] Physical Dimensions and Installation Methods: Filter for instances where the "physical dimensions" (length, width, height, mounting hole positions) are compatible with the space of the carrier (e.g., armored vehicles, ships) and the "installation method" (e.g., wall-mounted, embedded) is compatible with the carrier structure. Example: If the requirement is "new equipment system adopts 28V DC power supply, 1553B communication protocol, XML data format", then filter for equipment instances with "power interface type = 28V DC", "communication protocol version = 1553B", and "data format standard = XML".

[0080] S7.1.3 Dynamic behavior constraint matching; Based on the "dynamic behavior requirements" extracted in step S6.2 (such as "detection range attenuation rate with rainfall intensity ≤ 20%" and "kill probability varies with target speed range ≤ 0.1"), instances that satisfy the constraints are matched from the "equipment type object meta-model" in the model library. The matching logic needs to combine attributes such as "performance attenuation function identifier", "function parameter value", and "applicable environment range". If the requirement is "Radar detection range attenuation rate ≤ 20% when rainfall intensity is 10mm / h", then filter out the "Performance Attenuation Function Identifier =" in the "Dynamic Behavior Attribute Group". "and (because ) instances; If the requirement is "weapon kill probability ≥ 0.8 when target speed is 500m / s", then filter out the "kill probability function identifier =" in the "dynamic behavior attribute group". "and" , The parameter combination makes exist Examples of "time ≥ 0.8".

[0081] Step S7.2: Generate a preliminary equipment system composition model scheme; The purpose of generating a preliminary equipment system composition model is to convert and generate an editable form containing interface parameters and dynamic behavior parameters based on the performance-compliant, interface-compatible, and dynamic behavior-compliant equipment instances matched in step S7.1 and the equipment constraint information extracted in step S6.2. Figure 5 As shown, the form displays the equipment type, equipment name, attribute name, and attribute value for each piece of equipment. The equipment type originates from the name of each equipment object meta-model instance; the equipment name originates from the "Name" attribute value of each equipment object meta-model instance; the attribute name originates from the attribute name of each equipment object meta-model instance; and the attribute value originates from the corresponding attribute value of each equipment object meta-model instance (engineers can modify attribute values ​​as needed). For newly added equipment, engineers need to define the equipment's attribute values. After completing the above operations, a preliminary equipment system composition model scheme is obtained. Attribute values ​​are divided into static indicator attributes, interface attributes, and dynamic behavior attributes. Static indicator attributes represent the core performance indicators of the equipment (such as detection range, operating time, and kill probability), and their attribute values ​​originate from the "Performance Attribute Value" of the equipment type object meta-model. Interface attributes represent the interface-related parameters of the equipment (power interface, communication protocol, data format, physical dimensions, and installation method), and their attribute values ​​originate from the "Interface Attribute Group" of the equipment type object meta-model. Dynamic behavior attributes represent the dynamic behavior-related parameters of the equipment (performance decay function identifier, function parameter value, and applicable environment range), and their attribute values ​​originate from the "Dynamic Behavior Attribute Group" of the equipment type object meta-model.

[0082] Step S8: Multi-objective optimization algorithm processing; The algorithm processing is a multi-objective, multi-constraint optimization process. Its purpose is to optimize and solve the preliminary equipment system composition model scheme based on the objective function defined by the overall objective information of the equipment system, the constraints defined by the overall performance constraints of the equipment system, and the constraints defined by the equipment quantity constraints. This yields an equipment system composition model scheme that meets the current engineer's needs and system capability requirements, providing a foundation for the generation of subsequent equipment system composition models. The algorithm used in this processing is the multi-objective particle swarm optimization algorithm, which supports solving multi-objective programming problems. The algorithm treats each piece of equipment in the equipment system composition model as a node, and all equipment in the model constitutes a node set. Based on the objective function and constraints, a mathematical model is constructed, and each node in the node set is adaptively selected or discarded to finally obtain the optimal solution set.

[0083] S8.1 Mathematical modeling of optimization problems; Each piece of equipment in the equipment system composition model is abstracted as a decision variable node, and all equipment constitutes a node set. ( (Total number of equipment). The optimization objective and constraints are defined as follows: S8.1.1 Optimization Objective; Minimize total weight:

[0084] in, For the first The weight of a piece of equipment is derived from the "weight" attribute of the object's metamodel, and the unit is kilograms.

[0085] Minimize total cost:

[0086] in, For the first The cost of a piece of equipment comes from the "cost" attribute of the object metamodel, and the unit is yuan.

[0087] Maximizing interface compatibility: Defining interface compatibility coefficients , indicating equipment The interface compatibility with other equipment within the system, taking values ​​[0,1], where 1 indicates full compatibility. Therefore:

[0088] The interface compatibility coefficient The value of is dynamically calculated in each generation of particle fitness evaluation in step S8, based on the equipment combination scheme represented by the current particle. Interface compatibility with other equipment or designated carriers within the solution.

[0089] The compatibility calculation logic is as follows: Power interface compatibility score: Establish the system power supply topology and define the set of "power supply equipment" nodes (specified by the user or based on preset rules, such as equipment types containing the keyword "power supply"). For equipment... If its "power interface type" matches the interface type of any power supply equipment in the system (voltage level tolerance ±10%, current capacity greater than the equipment) If the peak power consumption is 1, then the score is 1; otherwise, the score is 0. If equipped with... If it is a power supply device, the score is 1 by default.

[0090] Communication protocol compatibility score: A preset protocol compatibility mapping table is used (e.g., 1553B bus and Ethernet are compatible after conversion via a gateway; conversion latency ≤ 1ms is included in performance constraints). If equipped with... If the protocol is directly consistent with the communication hub (specified by the user) or exists in the mapping table, the score is 1; otherwise, the score is 0.

[0091] Data format compatibility score: If equipment If the "data format standard" is consistent with the format of the data interaction partners in the system or can be converted by a lossless converter, the score is 1; otherwise, the score is 0.

[0092] Physical size and installation method compatibility score: If equipped If the "physical dimensions" parameter is compatible with the reserved space constraints of the carrier or other related equipment within the system, and the "installation method" parameter is consistent with the mechanical interface standard of the mounting base, then this item scores 1; if either dimension fails to meet the assembly space or mechanical interface requirements, this item scores 0. This score is included as an arithmetic mean. Through the objective function To maximize the performance, the driving algorithm automatically eliminates equipment solutions that do not meet physical compatibility standards, thereby achieving hard constraints on physical size and installation method.

[0093] Ultimately, the interface compatibility coefficient .

[0094] Maximizing Dynamic Behavioral Compliance: Defining Dynamic Behavioral Compliance Coefficient , indicating equipment The performance parameter represents the degree to which the requirements are met under environmental changes, taking values ​​[0,1], where 1 indicates full compliance. Therefore: ; The dynamic behavior compliance coefficient The value of is determined during the constraint check phase in step S8 by reading the equipment. The pre-defined inherent performance decay function identifier and parameter value in the object meta-model are substituted into the environmental variable values ​​specified in the dynamic behavior constraints extracted in step S6.2 for real-time calculation, rather than being written with fixed values ​​during modeling in step S5.

[0095] The compliance calculation steps are as follows: For equipment Let the performance degradation function in its dynamic behavior attribute group be... ,in For equipment inherent parameter vectors (such as radar's) , ), For environmental variables (such as rainfall intensity) Target speed Demand constraints are specified in a standard environment. Lower performance Threshold should be met .but:

[0096] in, Calculations are performed in real time based on the function expression stored in step S1. For example, for radar detection equipment, That is .

[0097] S8.1.2 Constraints; Performance constraints: Radar detection equipment quantity constraints: ( (The set of nodes for all radar detection devices). Detection capability constraint: For all equipment with a "detection range" parameter Its actual detection distance rice( (Calculated from a dynamic behavior model) Damage Capability Constraint: Applicable to all equipment with an "Attack Range" parameter. Its kill probability ( (Calculated from a dynamic behavior model).

[0098] Interface constraints: Power interface compatibility constraints: The "power supply type" attribute of each electrical equipment in the system must be compatible with the "220V AC" provided by the field power vehicle or the interface type of other power supply nodes in the system. Incompatible combinations will be excluded from the scheme. Communication protocol compatibility constraints: The "communication protocol" attribute of each piece of equipment in the system must support the protocol conversion mapping table preset by the communication center to ensure smooth information flow; Data format compatibility constraint: The "data format" attribute of each piece of equipment in the system must be consistent with the format of the equipment whose data is being exchanged, or consistent after lossless conversion. Physical size and installation method compatibility constraints: The "physical size" and "installation method" attributes of each piece of equipment in the system must meet the preset constraints of carrier space and installation base.

[0099] Dynamic behavioral constraints: Environmental attenuation constraints for radar detection equipment: under rainfall intensity of In typical mission environments, radar detection equipment relies on its inherent attenuation function The calculated actual detection distance still needs to meet the following requirements. Rice, of which The baseline detection distance under standard conditions. The attenuation coefficient is... This refers to the intensity of rainfall.

[0100] Among them, dynamic behavior constraints refer to ensuring that the performance parameters of the equipment, calculated based on its inherent performance attenuation function, are not lower than a preset threshold under specified environmental conditions (such as rainfall intensity and target velocity). For example, the radar under rainfall intensity... The actual detection distance below meters, or the speed of the weapon at the target. The probability of killing .

[0101] S8.2 Main steps of the multi-objective particle swarm algorithm; Initialize the particle swarm: Each particle represents an equipment system configuration scheme (i.e., a node set). A specific example, including the selection, quantity, interface configuration, and dynamic behavior parameters of each piece of equipment. During initialization, it is randomly generated. Individual particles (population size) A value of 50-100 is recommended. Particle encoding uses a mixed integer vector. ,in For equipment type index (classes 1 to T). This refers to the index (integer value) of the specific model under this type in the candidate pool in step S7.1. This represents the quantity (positive integer) of equipment of this model. After the speed update, discrete variables are rounded to the nearest integer and limited to a valid range. To prevent illegal solutions, a repair operation is performed before calculating the fitness: if the selected model is outside the candidate pool, it is mapped to the nearest valid index.

[0102] Calculate fitness value: for each particle Calculate the fitness values ​​of its four objective functions. It then determines whether all constraints (performance, interface, and dynamic behavior constraints) are satisfied. Constraint handling follows the Deb constraint dominance principle: when comparing two particles, if one satisfies a constraint while the other does not, the one that satisfies the constraint dominates; if neither is satisfied, the one with the lower degree of constraint violation dominates. The degree of violation is defined as the normalized weighted sum of all constraint violations, such as the percentage of incompatible interfaces.

[0103] Update particle velocity and position: Particles are sorted using non-dominated sorting and crowding distance, and the non-dominated solution set (Pareto front) of the current swarm is selected as the globally optimal guide set. Then, according to the velocity update formula for particle swarm optimization, the velocity and position of each particle are adjusted.

[0104] in: For particles In the The velocity vector of the generation; For particles In the The position vector of the substitute; The inertia weight is set to 0.4~0.9 to balance global exploration and local development. , The learning factor is set to 2.0 to control the impact of individual optimality on global optimality. , A random number in the interval [0,1]. For particles In the The best historical position of the era; For the first The globally optimal position of the generation (selected from the non-dominated solution set).

[0105] Iteration Termination and Solution Set Selection: Repeat steps (b) and (c) until the preset number of iterations is reached (100-200 iterations are recommended) or the fitness converges (the change in the objective function value is less than the threshold). Finally, from the historical best positions of all particles, the non-dominated solution set (Pareto front) is selected as the candidate optimization scheme for the equipment system composition model.

[0106] S8.3 Output and Application of Optimization Results After optimization, the Pareto optimal solution set is output (containing multiple feasible solutions with trade-offs). Each solution satisfies the following: minimizing total weight and total cost; maximizing interface compatibility and dynamic behavior compliance; and strictly meeting performance constraints (detection and kill capabilities). Engineers can select a solution from the Pareto solution set as needed (e.g., choosing the option prioritizing cost control). , For smaller solutions that focus on interfaces / dynamic behavior, choose... , (The larger plan).

[0107] Step S9: Generate the reconstructed equipment system composition model; The purpose of this step is to match the equipment in the equipment system composition model library with the optimal node set obtained by the algorithm, obtain all the data of these equipment, including equipment type, equipment name, attribute value set, position, size, etc., construct a new blank equipment system composition model, and instantiate the object meta-model corresponding to these equipment according to the equipment type.

[0108] like Figure 6 As shown, the final equipment system composition model includes: optoelectronic detection equipment: anti-unmanned aerial vehicle (UAV) tracking turntable; radar detection equipment: UAV detection radar system and area array radar detection system; laser weapon: a certain type of anti-UAV laser weapon; interceptor missile launching equipment: directional anti-unmanned aerial vehicle (UAV) interceptor missile launching device; and conventional firepower weapons: lightweight anti-UAV remote control weapon station and man-portable air defense system.

[0109] Among them, a certain type of grenade launcher (detection range 400 meters) and an anti-drone gun (detection range 500 meters) were automatically filtered out by the algorithm because they did not meet the performance constraint of a detection range ≥ 700 meters. Although the collision anti-drone system met the detection range requirement, its kill probability was 0.75, which did not meet the constraint condition. Furthermore, its power supply interface is 24V DC, which is incompatible with the 28V DC of most equipment in the system (requiring an additional conversion module, increasing costs), and therefore it was eliminated during the optimization process. It was replaced by the "directional anti-interceptor missile launcher" in the candidate pool, which has a kill probability of 0.88, a 28V DC power supply interface, better compatibility with other equipment in the system, and a higher overall target value within an acceptable range of total weight and cost.

[0110] Example 2: Based on Example 1, this example further introduces a system topology toughness assessment mechanism to defend against the risk of cascading failures in dynamic battlefield environments.

[0111] In the dynamic battlefield environment of modern high intensity, strong electromagnetic interference, and extreme weather, the equipment nodes in the equipment system composition model do not exist in isolation. The energy supply links in the physical space and the data communication links in the logical space create a deep coupling dependency within the system.

[0112] The performance degradation of a single equipment node in harsh environments (such as a sharp drop in radar detection range due to heavy rain) can easily trigger cascading failures of adjacent decision-making or strike nodes along highly coupled weighted links (such as the fire control system being unable to meet the closed-loop control cycle requirements due to interruption of target data stream input from front-end sensors or data refresh rate, resulting in timeout invalidation of the calculated interceptor missile launch parameters, ultimately causing a miss). This ultimately leads to a topological collapse of the overall effectiveness of the equipment system.

[0113] To avoid technical failures in the system's operation and achieve automatic defense during the design reconfiguration phase, a system topological resilience assessment mechanism based on complex network graph theory is established in step S8 of the multi-objective optimization algorithm. This mechanism is defined as the fifth optimization objective of the multi-objective particle swarm optimization algorithm, and its specific execution is as follows: Step 1: Digital feature extraction and weight matrix construction of physical coupling attributes; In each generation of the multi-objective particle swarm optimization algorithm, the decoded value for the current particle position includes... For each piece of equipment, the system analyzes the "interface attribute group" and "static attribute group" of the metamodel of each equipment object in the node set, extracts the functional coupling strength representing physical dependencies, and then constructs the system connection matrix. .

[0114] System connection matrix In the middle, source equipment node To the target equipment node Functional coupling weight The calculation formula is as follows:

[0115] in: For nodes To the node Functional coupling weights; As the interface matching factor, when the source equipment node The power supply type or communication protocol and the target equipment node The value is 1 when fully compatible, and 0 when there is conversion loss or incompatibility. The preset energy flow weight allocation coefficient; Equipment node for the purpose The peak power consumption attribute value; For source equipment nodes Maximum power supply capacity attribute value; Assign coefficients to the preset data stream weights; For nodes With nodes The actual data exchange bandwidth between them is determined based on the corresponding communication protocol; This is the highest communication protocol bandwidth benchmark used in the equipment system configuration model; Assign the preset control flow weights; For nodes With nodes The limit of low-level interaction latency for instruction transmission between them.

[0116] The above weighting coefficients satisfy the normalization constraint: The peak power consumption, communication protocol, and other parameters mentioned above are all directly derived from the interface attribute group of the object meta-model defined in step S2 of Example 1.

[0117] Step 2: Dimensionality reduction of the system's Laplace matrix and quantification of topological resilience; Obtain the complete connection matrix Then, the system extracts the degree matrix that characterizes the topological robustness of the node set. Degree matrix diagonal elements All off-diagonal elements are zero. Based on this, the Laplace matrix of this equipment system scheme is constructed as follows:

[0118] in: The Laplace matrix is ​​used to characterize the network structure of the system. This is the degree matrix derived from the network connectivity status; This is the functional coupling weight connection matrix generated in step one.

[0119] Laplace matrix Perform eigenvalue decomposition to obtain its eigenvalue set. Sort the values ​​in the set from smallest to largest and extract the second eigenvalue, which is defined as the algebraic connectivity. Algebraic connectivity The magnitude of this value, at the physical level, strictly maps the algebraic topological baseline for maintaining the flow of information and energy between the remaining nodes after the equipment system has been damaged and removed due to enemy fire.

[0120] Step 3: Evolution of Cascade Failure Probability Driven by Environmental Variables; Beyond simple structural analysis, the system further introduces the destructive effects of dynamic battlefield environment disturbances on the topology network. This is combined with the dynamic behavior compliance coefficient calculated based on the "dynamic behavior attribute group" in step S8.1.1 of Example 1. Calculate the propagation probability of errors and failures in the system network:

[0121] in: The overall cascading failure probability faced by the current equipment system solution; This represents the total number of equipment in the current system configuration model. For the first The compliance coefficient of dynamic behavior of a piece of equipment under a specified extreme environment; This is the sum of the functional coupling weights that the equipment outputs to other equipment within the system.

[0122] This physical model reveals a high-risk phenomenon in systems engineering: if the equipment in the core hub position (with a very large sum of output weights) is extremely sensitive to the environment (with a very low compliance coefficient), then once the performance of this node deteriorates due to environmental interference, the risk of failure will spread exponentially along the highly dependent links throughout the entire network.

[0123] Calculate the topological resilience index of the final system that integrates algebraic topology and environmental attenuation propagation laws:

[0124] in: It is the topological toughness index of the system; The algebraic connectivity obtained from the calculation in step two; These are preset penalty control scaling parameters; This represents the probability of cascading failure.

[0125] Step 4: Algorithm fitness reshaping and numerical implementation simulation; In the fitness calculation and constraint handling steps of the multi-objective particle swarm optimization algorithm, maximizing the topological toughness index of the system is set as the fifth optimization objective: Additionally, a new system topology resilience constraint is introduced: for any feasible equipment system configuration scheme, its system topology resilience index... It must be greater than or equal to the preset safe survival threshold. .

[0126] An exponential penalty defense mechanism is employed during implementation, targeting any system whose topology resilience index falls below the safety survivability threshold. The particle scheme is subjected to a disadvantageous guidance:

[0127] in: To impose a penalty gain term on schemes that violate the topological resilience baseline; The constant of the basic penalty operator; The preset system topology resilience safety survival threshold; This is the currently calculated topological toughness index of the system.

[0128] This penalty will be factored into the overall constraint violation rate of the particle, forcing infeasible solutions to be eliminated in the evolution process.

[0129] In Example 1 Figure 5 and Figure 6 The implementation simulation is carried out using the corresponding individual soldier anti-unmanned equipment system as an example. When the user's input requirements include the formation of an air defense strike chain, the multi-target particle swarm algorithm may generate "Solution A" in the early iterations: select an array radar detection system to provide target guidance for a certain type of anti-UAV laser weapon, and share a 220V AC field power vehicle with a peak capacity of 600 watts.

[0130] In this scheme, the power consumption of the area array radar detection system is 150 watts, and the power consumption of a certain type of anti-drone laser weapon is 400 watts. The energy function coupling weight of the two is... Approaching the limit ( In extreme rainfall environments, the performance degradation parameters of area array radar detection systems cause a sharp decrease in detection range. (The coefficient drops sharply), and this failure state is instantly transmitted to a certain type of anti-drone laser weapon through a high-weight link, causing the system to calculate the cascade failure probability of "Scheme A". Extremely high, system topological toughness index It plummeted to 0.12, triggering an exponential penalty and being removed from the Pareto non-dominated solution set.

[0131] Through automatic optimization and evolution by the algorithm, the final reconstructed "Solution B" is output: a combination of an unmanned aerial vehicle (UAV) detection radar system and an area array radar detection system, coupled with a directional anti-interceptor missile launcher. The power consumption of the two radars is 80 watts and 150 watts respectively, and energy flow and data flow are decoupled and distributed load is achieved between them and the directional anti-interceptor missile launcher. The system calculates the algebraic connectivity of this topology. Significantly improved, and interface matching factor The system reaches its full value without requiring an external transformer module. The topological toughness index of "Scheme B" is then calculated. It climbed to 0.86, perfectly surpassing the safety threshold. .

[0132] This process forces the system design to converge toward a physical defense architecture with "redundant data links", "energy load balancing" and "high resilience connectivity", avoiding system integration vulnerability caused by unilaterally pursuing cost or extreme performance of individual components, and completely making up for the shortcomings of model reconstruction in dynamic environments.

[0133] This invention provides a standardized way to express the reconstruction requirements of equipment system composition models. Existing methods for reconstructing equipment system composition models lack unified standards for expressing requirements. Engineers' reconstruction requirements are often submitted in unstructured text, leading to inconsistent understanding and ambiguity. This patent provides a standardized way to express reconstruction requirements by defining an equipment system composition model reconstruction dictionary and a reconstruction requirement template. Through word segmentation, part-of-speech tagging, and sentence processing technologies, it provides unified semantic standards for the terminology used in the reconstruction process, enabling accurate parsing of engineers' textual reconstruction requirements and effectively improving the efficiency, reliability, and adaptability of equipment system composition model reconstruction.

[0134] This invention optimizes equipment system configuration models based on multi-objective optimization algorithms. Existing equipment system optimization typically relies on manual experience or local adjustments to single indicators, lacking systematic and intelligent algorithmic support. This makes it difficult to effectively address the optimization needs of multiple objectives and constraints in complex and dynamically changing combat environments. This invention introduces a discrete particle swarm optimization algorithm, combining objective functions and constraints, to automatically optimize and solve equipment system configurations, rapidly generating superior equipment system configuration schemes that meet multiple tactical and technical indicator constraints.

[0135] This invention improves the efficiency of equipment system design. Existing methods for reconstructing equipment system models generally rely on manual experience and configuration, which is time-consuming and labor-intensive, severely impacting the development efficiency of equipment systems. This invention establishes an automated reconstruction and optimization mechanism for equipment system models, enabling intelligent processing of reconstruction requirement analysis, model optimization, and scheme generation. This method can quickly adjust and reconstruct equipment system models when battlefield environments and mission requirements change, significantly improving the efficiency and accuracy of overall equipment system design.

[0136] This invention further incorporates complex network graph theory into the optimization process. By constructing a system connection matrix based on physical coupling weights and calculating the system topology resilience index, it can quantitatively assess the system's resilience to node failures or environmental disturbances. The optimization process forces the system configuration to evolve towards a distributed defense architecture with redundant links, load balancing, and high algebraic connectivity, avoiding system vulnerability caused by pursuing extreme values ​​for a single metric.

[0137] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the invention.

[0138] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.< / equipmenttype> < / possess> < / equipmenttype> < / contain> < / should> < / type> < / possess> < / should> < / should> < / possess> < / equipment> < / add> < / should> < / possess> < / should> < / contain> < / should>

Claims

1. An automatic reconstruction method for equipment system composition models based on natural language requirements, characterized in that, Includes the following steps: Step S1: Organize the equipment system database to form a structured form containing equipment name, type, combat and technical indicators and indicator values. The combat and technical indicators include static indicators, interface attributes and dynamic behavior functions that change with the environment. Step S2: Construct a domain meta-model library based on the equipment system database. The domain meta-model library includes a graph meta-model, an object meta-model, and a relation meta-model, wherein the object meta-model encapsulates the static attribute group, interface attribute group, and dynamic behavior attribute group of the equipment. Step S3: Define a dictionary for identifying reconstruction requirement text, the dictionary containing the parts of speech and content of equipment type, performance indicators, interface terms and dynamic behavior terms; Step S4: Define a template library for reconstructing equipment composition models. The template library shall include at least: equipment system type declaration template, equipment system overall target template, equipment addition constraint template, equipment system overall performance constraint template, equipment quantity constraint template, interface compatibility constraint template, and dynamic behavior constraint template. Step S5: Construct an equipment system composition model library based on the domain meta-model library, and instantiate an equipment object meta-model containing specific attribute values; Step S6: Receive the structured natural language reconstruction requirement statement input by the engineer, and use the dictionary and template library to perform word segmentation, part-of-speech tagging, block segmentation and information extraction to obtain equipment system type declaration information, overall target information, added constraint information, overall performance constraint information, quantity constraint information, interface attribute information and dynamic behavior information; Step S7: Based on the extracted information, match equipment instances that meet performance constraints, interface compatibility constraints, and dynamic behavior constraints from the equipment system composition model library to generate a preliminary equipment system composition model scheme. Step S8: Construct an objective function based on the overall objective information, construct constraint conditions based on the overall performance constraint information, quantity constraint information, interface attribute information and dynamic behavior information, and optimize the preliminary scheme using the discrete multi-objective particle swarm optimization algorithm to obtain the Pareto optimal solution set that satisfies the multi-objective trade-off; Step S9: Based on the scheme selected from the Pareto optimal solution set, match the equipment data in the equipment system composition model library to generate the reconstructed equipment system composition model.

2. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 1, characterized in that, In step S2, the attributes of the object meta-model are stored in a nested structure. The interface attribute group includes at least: power supply type enumeration, communication protocol enumeration, data format enumeration, physical size value and installation method enumeration; the dynamic behavior attribute group includes at least: performance degradation function identifier, function parameter value and applicable environment range.

3. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 1, characterized in that, The template in step S4 is constructed based on a predefined part-of-speech sequence and is used to map unstructured text into structured information blocks extracted in step S6.

4. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 1, characterized in that, Step S6 further includes: performing syntactic pattern matching and information extraction using regular expression rules, wherein the rules define extraction patterns for indicator names, comparison relationships, and thresholds.

5. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 1, characterized in that, The multi-objective particle swarm algorithm in step S8 uses mixed integer encoding. The position vector contains equipment type index, model index and quantity, and multiple constraints are handled through the Deb constraint dominance principle.

6. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 1, characterized in that, The objective function in step S8 includes: Total weight minimization objective: ,in For the first The weight of the equipment; Total cost minimization objective: ,in For the first The cost of each piece of equipment; The goal is to maximize interface compatibility. ,in For the first The interface compatibility coefficient of each piece of equipment; The goal of maximizing dynamic behavioral compliance: ,in For the first The compliance coefficient of dynamic behavior of each piece of equipment.

7. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 6, characterized in that, The interface compatibility coefficient Based on equipment The compatibility of the power interface, communication protocol, data format, and physical dimensions with other equipment or carriers within the system is calculated, specifically as the arithmetic mean of the compatibility scores for each item.

8. The automatic reconstruction method for equipment system composition model based on natural language requirements according to claim 6, characterized in that, The dynamic behavior compliance coefficient According to equipment The inherent performance degradation function is used to calculate the degree to which its output value meets the demand threshold under specified environmental variables. The calculation formula is a piecewise function.

9. The method for automatic reconstruction of equipment system composition model based on natural language requirements according to claim 1, characterized in that, The constraints in steps S7 and S8 include: Performance constraints: such as minimum quantity of equipment type, minimum detection range, and minimum kill probability; Interface constraints: such as compatibility of power interface types, communication protocol versions, and data format standards; Dynamic behavior constraints: such as the performance parameters of the equipment under specified environmental conditions (such as rainfall intensity, target speed) not being lower than the preset threshold.

10. The method for automatic reconstruction of equipment system composition model based on natural language requirements according to claim 1, characterized in that, The reconstructed equipment system model generated in step S9 consists of equipment components that are optimized schemes automatically selected and combined from the candidate pool by an algorithm under the premise of satisfying all constraints.

Images