LVC intelligent planning system and method based on intelligent test resource interaction relation model

Abstract

The invention provides an LVC intelligent planning system and method based on an intelligent test resource interaction relation model. The method comprises the steps of obtaining a test task document; determining a test resource demand according to the test task document; obtaining test resources matched with the test resource demand from an infrastructure layer; performing intelligent planning processing on the test resources through a functional layer to generate a test resource configuration scheme; generating and outputting the test scheme file in an application layer; wherein the infrastructure layer comprises a resource database and a knowledge graph database of a plurality of discrete test target ranges, the functional layer comprises a test task definition module and an intelligent planning module, and the application layer comprises a test scheme generation module and a recommendation module. According to the intelligent planning method based on the knowledge graph, the efficiency of resource allocation and scheme generation in a complex test environment is remarkably improved, and an effective solution is provided for management and utilization of large-scale heterogeneous test resources.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of information technology, and in particular to an LVC intelligent planning system and method based on an intelligent test resource interaction relationship model. BACKGROUND

[0002] The joint test system can carry out cross-domain virtual-real integration joint tests by combining real, virtual and constructed resources (Live Virtual Constructive, LVC). The joint test system can make up for the shortcomings of single test ranges in carrying out certain test tasks, flexibly combine and integrate all test ranges, facilities and equipment, give full play to the advantages of each test range resource, avoid repeated construction and waste of resources, and greatly improve the efficiency of joint test.

[0003] In a complex test environment, how to efficiently manage and utilize large-scale heterogeneous test resources has become a technical problem to be solved. The traditional method is difficult to cope with the integration of multiple sub-test range resources, resulting in low resource utilization and low efficiency of test scheme development. At the same time, the complex correlation and influence mechanism between test factors are difficult to accurately grasp, which restricts the scientificity and effectiveness of the test scheme. In addition, different types of test tasks have different requirements for resource allocation and process arrangement, and how to realize intelligent generation and personalized recommendation of test schemes also faces challenges. The core of these problems is the lack of a unified knowledge representation and reasoning framework, which cannot effectively integrate and utilize the professional knowledge and experience scattered in various test fields. How to build a knowledge system that can integrate multi-source heterogeneous data, depict the correlation of test resources, and support intelligent planning and decision-making is the key to solving the above problems. This technical problem involves knowledge representation, data fusion, relationship reasoning and other technical fields, and needs innovation and breakthrough in system architecture, algorithm model, application implementation and other aspects. Only by establishing a comprehensive knowledge graph covering test resources, factor relationships and task processes can intelligent resource management and scheme generation in a complex test environment be truly realized. SUMMARY

[0004] The purpose of the present application is to solve the problem of intelligent resource management and scheme generation in a complex test environment in the prior art, and to provide an LVC intelligent planning system and method based on an intelligent test resource interaction relationship model, which significantly improves the efficiency of resource allocation and scheme generation in a complex test environment.

[0005] The present application adopts the following technical solutions:

[0006] On the one hand, the present application provides an LVC intelligent planning system based on an intelligent test resource interaction relationship model, comprising:

[0007] An infrastructure layer includes a resource database and a knowledge graph database of a plurality of sub-test ranges;

[0008] A function layer includes a test task definition module and an intelligent planning scheme module;

[0009] An application layer includes a test scheme generation module and a recommendation module;

[0010] The test resource matching the test resource requirement is obtained from the infrastructure layer, the test resource is intelligently planned by the function layer to generate a test resource configuration scheme, and a test scheme file is generated and output in the application layer.

[0011] Further, the infrastructure layer includes a test resource sharing library, a test data sharing library, a test resource model library, a test resource database, and a knowledge graph database; the test resource sharing library stores sharable resources of each sub-test range; the test data sharing library stores historical test data; the test resource model library stores test resource models; the test resource database stores test resource parameters; and the knowledge graph database stores relationship data between test resources.

[0012] Further, the function layer includes a test task requirement definition module, a test logical target range design module, a resource loading module, an intelligent planning scheme module, and a knowledge graph module; the test task requirement definition module parses test task requirements according to a test task document; the test logical target range design module selects target test resources from the test resource sharing library, the test resource model library, and the test resource database according to the test task requirements, and defines the interaction relationship between the target test resources; the resource loading module manages the state of the target test resources; the intelligent planning scheme module generates a test resource configuration scheme according to the interaction relationship between the target test resources and the state of the test resources; and the knowledge graph module constructs a test resource relationship graph according to the test resource configuration scheme.

[0013] Further, the test logical target range design module defines the interaction relationship between test resources, specifically including: obtaining a test resource list from the test resource sharing library, the test resource model library, and the test resource database; selecting target test resources from the test resource list according to test task requirements; defining the logical attribution relationship of the target test resources in a visual manner; configuring node allocation information of the target test resources; and generating a logical target range definition file for defining the interaction relationship between the target test resources.

[0014] Further, the intelligent planning scheme module generates a test resource configuration scheme, comprising: constructing a test factor knowledge graph; determining resource integration rules according to the test factor knowledge graph; optimizing test factor influence relationships; automatically generating a test resource configuration scheme according to the resource integration rules, test factor influence relationships, interaction relationships between target test resources, and states of the test resources; wherein the test factor knowledge graph contains test resource information and interaction relationships thereof.

[0015] Further, the application layer comprises: a test task access module and a test scheme automatic generation module; the test task access module receives test task input and generates a test task document; the test scheme automatic generation module generates a test scheme according to the test task document and the test resource relationship graph; and the test scheme recommendation module optimizes the test scheme based on constraint conditions.

[0016] Further, the test scheme automatic generation module generates a test scheme by the following method: importing the test task document into the test task demand definition module of the functional layer and obtaining a corresponding test resource relationship graph; extracting target test resources from the test resource relationship graph; planning a test task flow; configuring a test data acquisition scheme; defining a test information display interface; and generating a test scheme file containing test resource configuration, interaction relationships, and a task flow according to the target resources and the test task flow.

[0017] In another aspect, the present application also provides a planning method of the LVC intelligent planning system based on the intelligent test resource interaction relationship model, which is characterized in that the method comprises:

[0018] obtaining a test task document;

[0019] determining test resource requirements according to the test task document;

[0020] obtaining test resources matching the test resource requirements from the infrastructure layer;

[0021] performing intelligent planning processing on the test resources through the functional layer to generate a test resource configuration scheme;

[0022] generating and outputting a test scheme file in the application layer.

[0023] Further, the intelligent planning processing on the test resources through the functional layer to generate a test resource scheme comprises: constructing a test factor knowledge graph; determining resource integration rules according to the test factor knowledge graph; optimizing test factor influence relationships; and automatically generating a test resource configuration scheme according to the resource integration rules, test factor influence relationships, interaction relationships between target test resources, and states of the test resources.

[0024] The technical scheme provided by the embodiment of the present application can include the following beneficial effects:

[0025] The application discloses an intelligent test scheme generation method, and realizes efficient management and intelligent planning of test resources through construction of a multi-layer architecture. The infrastructure layer integrates resources and knowledge of multiple sub-test ranges, the function layer is responsible for test task definition and intelligent planning, and the application layer generates and recommends test schemes. The core innovation is to construct a test factor knowledge graph, extract and fuse knowledge entities from multi-source heterogeneous data, and build the association relationship between test resources. Based on the knowledge graph, the application can automatically determine resource integration rules, optimize test factor influence relationship, and generate test schemes containing resource configuration, interaction relationship and task flow. The intelligent planning method based on the knowledge graph significantly improves the efficiency of resource configuration and scheme generation in a complex test environment, and provides an effective solution for management and utilization of large-scale heterogeneous test resources. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 The figure shows a flowchart of an LVC intelligent planning method based on an intelligent test resource interaction relationship model. DETAILED DESCRIPTION

[0027] In order to further understand the content of the present application, the present application will be described in detail in conjunction with the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the related application, and not to limit the application. In addition, it should be noted that, in order to facilitate description, only the parts related to the application are shown in the drawings.

[0028] The embodiment of the present application is an LVC intelligent planning system based on an intelligent test resource interaction relationship model, which comprises:

[0029] The infrastructure layer comprises resource databases and knowledge graph databases of multiple sub-test ranges;

[0030] The function layer comprises test task definition module and intelligent planning scheme module;

[0031] The application layer comprises test scheme generation module and recommendation module;

[0032] The test resource matching the test resource requirement is obtained from the infrastructure layer, the test resource is intelligently planned and processed through the function layer, and the test resource configuration scheme is generated. The test scheme file is generated and output in the application layer.

[0033] In one specific embodiment, the infrastructure layer comprises:

[0034] a test resource sharing library, a test data sharing library, a test resource model library, a test resource database, and a knowledge graph database;

[0035] The test resource sharing library stores sharable resources of each sub-test range.

[0036] The test data sharing library stores historical test data.

[0037] The test resource model library stores test resource models.

[0038] The test resource database stores test resource parameters.

[0039] The knowledge graph database stores relationship data between test resources.

[0040] In one specific embodiment, the functional layer comprises:

[0041] a test task requirement definition module, a test logical range design module, a resource loading module, an intelligent planning scheme module, and a knowledge graph module.

[0042] The test task requirement definition module parses test task requirements according to a test task document.

[0043] The test logical range design module selects target test resources from the test resource sharing library, the test resource model library, and the test resource database according to the test task requirements, and defines interaction relationships between the target test resources.

[0044] The resource loading module manages the states of the target test resources.

[0045] The intelligent planning scheme module generates a test resource configuration scheme according to the interaction relationships between the target test resources and the states of the test resources.

[0046] The knowledge graph module constructs a test resource relationship graph according to the test resource configuration scheme.

[0047] In one specific embodiment, the test logical range design module defines interaction relationships between test resources, specifically comprising:

[0048] obtaining a test resource list from the test resource sharing library, the test resource model library, and the test resource database;

[0049] selecting target test resources from the test resource list according to test task requirements;

[0050] defining logical attribution relationships of the target test resources in a visual manner;

[0051] configuring node allocation information of the target test resources;

[0052] generating a logical target range definition file for defining the interaction relationship between target test resources.

[0053] In one embodiment, the intelligent planning scheme module generates a test resource configuration scheme, including:

[0054] constructing a test factor knowledge graph;

[0055] determining resource integration rules according to the test factor knowledge graph;

[0056] optimizing test factor influence relationships;

[0057] automatically generating a test resource configuration scheme according to the resource integration rules, test factor influence relationships, interaction relationships between target test resources, and states of test resources;

[0058] The test factor knowledge graph contains information about test resources and their interaction relationships.

[0059] In one embodiment, the application layer includes:

[0060] a test task access module and a test scheme automatic generation module;

[0061] The test task access module receives test task input and generates a test task document.

[0062] The test scheme automatic generation module generates a test scheme according to the test task document and the test resource relationship graph.

[0063] The test scheme recommendation module optimizes the test scheme based on constraint conditions.

[0064] In one embodiment, the test scheme automatic generation module generates a test scheme, including:

[0065] Importing the test task document into the test task demand definition module of the functional layer and obtaining the corresponding test resource relationship graph;

[0066] Extracting target test resources from the test resource relationship graph;

[0067] Planning a test task flow;

[0068] Configuring a test data collection scheme;

[0069] Defining a test information display interface;

[0070] Generating a test scheme file containing test resource configuration, interaction relationships, and task flow according to the target resources and test task flow.

[0071] As Figure 1 The embodiment of the LVC intelligent planning method based on the intelligent test resource interaction relationship model can specifically include:

[0072] In step S101, a test task document is obtained; and test resource requirements are determined according to the test task document.

[0073] The process of obtaining the test task document involves receiving a structured task description file from a test command system or a task issuing party. The document usually contains elements such as test type (such as electronic countermeasures, fire strike), red and blue confrontation rules, scenario background (such as island seizure battle), key evaluation indicators (hit rate ≥ 90%, response time ≤ 2 seconds), etc. Taking a certain air defense missile test as an example, the task document will clearly specify that the radar needs to verify the multi-target tracking capability in a complex electromagnetic environment, and requires processing more than 8 air targets at the same time, with a false signal suppression ratio of 20 dB. The document is stored in XML format, including <testobjective>Node defines core indicators, <scenarioparameters>The node describes the battlefield environment parameters. The system extracts these structured data through the parsing engine, and automatically identifies that 3 phased array radars and 2 electronic jamming stations are needed as core resources. When determining the test resource requirements according to the test task document, the system will start the knowledge graph matching engine. The knowledge graph has pre-stored 200 test resource entities and 400 attribute relationships, such as the association of radar resources with attributes such as "detection distance" and "anti-jamming level". When the task requirement "detects stealth targets" is identified, the system automatically filters radar models with detection capability of RCS≤0.1㎡. For the requirement of "multi-platform cooperative tracking", the knowledge graph recommends a radar networking scheme with data link interworking protocol. The requirement determination process is divided into three levels of matching: the first level of matching is based on resource function tags (such as "electronic countermeasure type-jammer"), the second level of matching verifies performance parameters (transmit power≥100kW), and the third level of matching checks resource status (available period coincides with test time). In a certain test scheme generation, the system parses the "need to simulate saturated attack" requirement from the task document, and automatically combines 4 rocket simulators and 1 set of track generation system to meet the technical requirement of generating 32 incoming tracks at the same time. The constraint satisfaction problem (CSP) solving technology is introduced in the test resource requirement determination link. The system establishes a variable set corresponding to the selected resources, and the domain value range comes from the resource library instance, and the constraint condition comes from the task document. For example, a networked test requires that the time delay of each node be ≤50ms, and the system will exclude communication links with transmission delay >30ms. For the special requirement of "need to support dynamic joining", the command and control system with hot plug interface is automatically filtered. This process generates a feasibility matrix of resource combination, and in a certain test scheme design, the system evaluates 6 radar combination schemes, and finally selects a scheme of 3 radars of different frequency bands to form complementary coverage, meeting the requirement of seamless coverage of 300km monitoring area in the task document. The test resource requirement analysis includes time and space dimension verification. The time axis verification ensures that the resource available period covers the test period, and because a satellite reconnaissance resource can only provide a 2-hour window in a certain test, the system automatically adjusts the test process to be implemented in 3 stages. The space verification is completed through the GIS engine, and when the task document requires "cross-war zone joint test", the system automatically filters out 8 target range resources distributed in three war zones and calculates the data transmission route. For special environment requirements (such as sea test), land-based fixed radar stations are excluded and mobile platforms on ships are selected. When generating a certain test scheme, the system detects that the electromagnetic environment level required by the task document does not match the qualifications of a certain target range, and automatically recommends using an alternative site with 10V / m strong field test capability. The requirement determination process includes an intelligent conflict detection mechanism. When the task document requires "high-precision positioning" and "strong electromagnetic interference" at the same time, the system detects that the GPS type resource is not applicable, and instead recommends a Beidou anti-jamming positioning device. Resource combination conflict detection uses rule-based reasoning, such as identifying that a certain type of jammer and fire control radar exist in the same frequency band interference, and automatically inserting a filter configuration or adjusting the frequency point allocation scheme.During the generation of a certain test scheme, the system found that the 4 target aircraft required by the document conflicted with the available airspace capacity, and intelligently suggested an alternative scheme of 2 target aircrafts for multi-flight.

[0074] Step S102, obtaining test resources matching the test resource requirements from the infrastructure layer; intelligently planning the test resources through the functional layer to generate a test scheme file.

[0075] 1. When obtaining matching test resources from the infrastructure layer, the system first parses the requirement parameters in the test task document (such as target range type, equipment accuracy, data sampling rate, etc.), and screens out discrete target range resources that meet the conditions through semantic association retrieval of the knowledge graph database. For example, a certain missile test needs to meet the conditions of a range of 500 kilometers and a dynamic target tracking accuracy of 0.1 meters per second. The system automatically matches the radar array of A target range (accuracy of 0.08 meters per second) and the trajectory measurement system of B target range (coverage of 600 kilometers) from the test resource sharing library, and simultaneously calls the historical calibration data of similar scenarios in the test data sharing library as auxiliary reference. 2. The intelligent planning process of the functional layer adopts a multi-level decision mechanism: the task requirement definition module decomposes the test target into three sub-tasks of "environment simulation-equipment deployment-data acquisition", the logical target range design module analyzes the interaction relationship of each resource through the knowledge graph, for example, it constructs an interaction link of "radar tracking→ trajectory feedback→ damage assessment" for the above-mentioned missile test, and the resource loading module dynamically allocates the IP ports of No. 3 radar station of A target range and the mobile measurement vehicle of B target range to ensure real-time data intercommunication. 3. When generating the test scheme file, the intelligent planning scheme module optimizes based on the constraint conditions (such as time window, resource conflict rate), for example, in the scenario of multiple test tasks at the same period, the target range equipment with high reuse rate (a certain weather simulation equipment with a utilization rate of 90%) is preferentially scheduled, and finally an XML format scheme file containing resource topology graph, time sequence Gantt chart and data flow protocol is output, which includes more than 300 parameters such as the starting time of each node (such as 120 seconds in advance for radar preheating) and data packet size (single transmission limit of 2GB). 4. The knowledge graph module continuously updates the resource state, for example, when a certain device of C target range suddenly fails, the system immediately triggers dynamic re-planning, calls the performance simulation data of the replacement device from the test resource model library (delay increases by 15 milliseconds but meets the error tolerance), and marks the emergency switching process in the test scheme to ensure that the reliability of the scheme is improved by more than 40%. 5. The visualization construction link compares the old scheme (such as a certain anti-missile test in 2019) with a similarity of 85% through the scheme recommendation function of the application layer, automatically highlights the difference items (such as the newly added infrared imaging requirement), assists the planner to quickly confirm the key configurations, and compresses the 72 hours required for traditional manual planning to 4 hours.

[0076] Step S103, outputting the test scheme file at the application layer; wherein the infrastructure layer comprises a resource database and a knowledge graph database of multiple sub-test ranges, the function layer comprises a test task definition module and an intelligent planning module, and the application layer comprises a test scheme generation module and a recommendation module.

[0077] Implementation method of test plan file output. The test plan generation module of the application layer generates a structured test plan file by integrating the resource configuration results of the intelligent planning module of the functional layer. In specific implementation, the system encapsulates test resource node allocation table, interaction relationship topology, task timing logic and other elements into XML or JSON format files. For example, in a certain missile cooperative interception test, the output file contains the coordinates of 3 radar stations (longitude 116.5°, latitude 39.9°, etc.), the flight path parameters of 2 simulated target aircrafts (altitude 8000 meters, speed 2.5 Mach), and the data link interaction frequency band (C band). The recommendation module matches historical similar cases based on the knowledge graph, such as preferentially recommending the "distributed data fusion" scheme with a confidence of 92%, as it has the highest success rate in the past 20 similar tests. The coordination mechanism of the multi-target range resource library of the infrastructure layer. The test resource databases of various discrete target ranges achieve data synchronization through standardized interfaces (such as RESTful API). For example, the jammer parameter library of a certain electronic countermeasure range is associated with the receiver sensitivity database of another range, forming a joint knowledge graph. When the functional layer initiates a resource query, the system automatically retrieves more than 5000 device attribute records across target ranges and filters out 8 types of compatible signal generators and 3 types of analyzers. The knowledge graph database realizes relationship reasoning through Neo4j graph storage, such as the triple relationship chain of "Radar A-anti-jamming mode B-environmental condition C", supporting subsequent intelligent planning. Constraint processing flow of the task definition module of the functional layer. This module analyzes 5 types of core constraints in the test task document: time window (such as 2024Q3), resource type (must include 2 sets of satellite navigation simulators), performance indicator (positioning error <1 meter), cost upper limit (5 million yuan), and safety requirement (electromagnetic radiation level III). Through constraint satisfaction algorithm, the task topology graph is generated, for example, in a certain networking test, 3 communication devices with insufficient bandwidth are automatically excluded, and 4 candidate devices that meet the delay <10ms are retained. Dynamic optimization method of the intelligent planning module. Based on the reinforcement learning framework, the system optimizes the resource deployment scheme in 100 iterations, for example, the patrol path of a certain early warning aircraft is reduced from 6 waypoints in the initial scheme to 4, while maintaining the coverage rate ≥95%. When planning, the dynamic nature of resources is considered, such as when the weather database of a certain target range is updated to warn of heavy rain, the optical observation device is automatically replaced with a radar imaging device. Multidimensional evaluation of the application layer scheme recommendation. The recommendation module constructs an evaluation matrix containing 3 dimensions: historical similarity (weight 40%), resource utilization rate (weight 35%), and cost coefficient (weight 25%). For example, for a certain reconnaissance test, the system recommends a "multi-spectral + SAR satellite combination" after comparing 3 schemes, as it has an average data acquisition completeness of 98.7% in similar tasks from 2019 to 2023, which is 12% higher than the single satellite scheme.

[0078] The infrastructure layer includes a test resource sharing library, a test data sharing library, a test resource model library, a test resource database, and a knowledge graph database; the test resource sharing library stores sharable resources of each sub-test range; the test data sharing library stores historical test data; the test resource model library stores test resource models; the test resource database stores test resource parameters; and the knowledge graph database stores relationship data between test resources.

[0079] The test resource sharing library stores the shareable resources of each sub-test target range, and the core is to break the information barrier between physical target ranges. For example, a joint test involves the radar systems of three target ranges, the X-band radar of A target range, the phased array radar of B target range, and the electronic countermeasure equipment of C target range are all abstracted as standardized resource objects, containing device ID, frequency range (such as X-band 8-12GHz), available time period, etc. Through the resource matching algorithm, when the test task needs to call X-band radar and electronic countermeasure equipment at the same time, the system automatically filters the idle resources of A target range and C target range and establishes the space-time synchronization relationship, avoiding the resource conflicts caused by traditional manual coordination. The test data sharing library realizes experience reuse by structurally storing historical test data. For example, the radar echo data (sampling rate 1MHz, signal-to-noise ratio 30dB), environmental parameters (wind speed 5m / s, temperature 25℃) and interception results (hit rate 92%) of a certain missile interception test are marked as the "interception test-2023” data set. When a new test needs to simulate a similar scenario, the system automatically recommends this data set as a reference, and through the data comparison module, it prompts the deviation threshold of the current test parameters and historical data (such as triggering a warning when the wind speed difference exceeds 3m / s), thereby improving the scientificity of test design. The test resource model library uses parameterized modeling technology to describe the behavior characteristics of resources. For example, a certain unmanned aerial vehicle model is defined as a three-dimensional motion model (maximum speed 300m / s, climb rate 20m / s), a load model (optical lens focal length 500mm), and a communication model (data link bandwidth 10Mbps). In the logical target range design, when the test needs to simulate a UAV cluster reconnaissance task, the system automatically calls the model and deduces its interaction trajectory with the ground radar based on the dynamics equation, verifying whether the reconnaissance path meets the task requirements. The test resource database stores the static parameters of devices in a relational structure. For example, a certain satellite test resource record contains orbit parameters (perigee 200km, inclination 45°), load parameters (infrared resolution 0.1K) and state parameters (on-orbit life remaining 5 years). When planning a satellite cooperative observation task, the system quickly filters the satellite resources that meet the inclination range (40°-50°) and have a resolution higher than 0.2K through SQL query, ensuring the technical feasibility of the observation scheme. The knowledge graph database constructs a multi-dimensional association network between resources. For example, a certain electronic warfare equipment is associated with the jamming frequency band (2-18GHz), the countermeasure target (a certain radar ID), and the historical countermeasure effect (success rate 85%). When designing a complex electromagnetic environment test, the system automatically associates the affected radar list through graph reasoning and recommends the combination of interference resources that need to be called synchronously, forming an intelligent decision chain based on association rules.

[0080] Step S105, the functional layer, comprising: test task requirement definition module, test logical range design module, resource loading module, intelligent planning scheme module and knowledge graph module;The test task requirement definition module analyzes the requirement according to test task document;The test logical range design module defines the interaction relationship between test resources;The resource loading module manages the state of test resources;The intelligent planning scheme module generates test resource configuration scheme;The knowledge graph module constructs test resource relationship graph. Further, the test logical range design module comprises: obtaining test resource list;According to the test task requirement, the target resource is selected from the test resource list;The logical belonging relationship of the target resource is defined in a visual manner;The node allocation information of the target resource is configured;The logical range definition file is generated.

[0081] 1The core function of the test task requirement definition module is to convert the test task document into structured requirements.

[0082] For example, when a task document describing “test the interception performance of a new missile in a complex electromagnetic environment” is input, the module extracts key elements through natural language processing techniques: test object (missile model H-9), environmental conditions (electromagnetic intensity ≥ 50 dBm), evaluation indicators (interception success rate ≥ 90%). This process uses semantic parsing algorithms to map unstructured text into machine-readable parameters such as “test type = performance verification, resource type = electromagnetic interference equipment + target missile launch system, constraint conditions = frequency band 2.4 GHz ± 5%”. This structured conversion eliminates ambiguity in manual interpretation and provides standardized input for subsequent modules.2The key to the test logic target range design module is to establish dynamic interaction between resources. Taking “radar detection system” and “electronic countermeasure equipment” as target resources as an example, the module first matches the physical interface standards of the two (such as IEEE 1522 protocol) through resource list matching, then defines the interference relationship link “electronic countermeasure equipment → radar detection system” through a drag-and-drop visual interface, and configures node parameters: interference signal intensity dynamic range is 30-60 dBm, action period is 200 ms interval. The final generated logic target range definition file contains topology structure diagram and timing constraint table. This virtual target range construction method improves the efficiency of physical joint test by more than 70%.3The resource loading module manages the whole life cycle of test resources through a state machine model. For example, the loading process of a certain type of unmanned aerial vehicle target includes: checking the flight control software version (≥ V2.3) in the initialization stage, monitoring the GPS positioning error (< 5 m) in the ready stage, and recording the fuel consumption rate (0.8 L / min ± 5%) in real time in the execution stage. The module dynamically synchronizes the resource state to the central database, and when it detects that the transmission power of a certain radar deviates from the set value (nominal value 20 kW, actual value 18.5 kW), it automatically triggers the calibration process. This closed-loop management ensures that test resources are always under control.4The intelligent planning scheme module uses a knowledge graph-based reasoning algorithm to generate resource configuration schemes. When constructing an anti-missile test scheme, the module analyzes the correlation rules between “early warning radar detection distance” and “interceptor reaction time” in historical data (such as a 0.3-second reduction in reaction time for every 50 km increase in detection distance), and automatically matches X-band radars (detection distance ≥ 300 km) and Y-12 interception systems (reaction time ≤ 2.5 seconds) that meet the conditions. The scheme output includes 23 parameters such as resource deployment coordinates (longitude error < 0.001°) and communication link bandwidth allocation (≥ 100 Mbps), reducing the time for manual planning by 85%.5The knowledge graph module constructs a test resource relationship network through multi-source data fusion. In specific implementation, it extracts device attributes (such as the frequency range 1-18 GHz of a jammer) from the test resource database, collaboration records (12 times of confrontation test with S-type radar) from the shared library, and simulation results (92% success probability of interfering QPSK signals) from the model library, and constructs a knowledge graph containing 387 nodes and 1524 edges.When dealing with the new task of "multi-target identification in complex electromagnetic environment", the atlas can automatically recommend radar combinations with co-site interference experience (A-type + B-type combination historical conflict rate < 3%).

[0083] The application layer includes a test task access module, a test scheme automatic generation module, and a test scheme recommendation module. The test task access module receives test task input. The test scheme automatic generation module generates a test scheme according to a knowledge graph. The test scheme recommendation module optimizes the test scheme based on constraint conditions. Further, the test scheme automatic generation module includes obtaining test resource information, determining resource interaction relationships according to test task requirements, planning test task processes, configuring test data acquisition schemes, defining test information display interfaces, and generating a test scheme file containing resource configurations, interaction relationships, and task processes.

[0084] When the test task access module receives test task input, it needs to parse key parameters in the task document, such as the test target "verify the detection accuracy of a certain type of radar in a complex electromagnetic environment". The task type, resource requirements, environmental conditions, and other structured data need to be extracted. The parsing process uses natural language processing technology to identify constraint conditions such as "detection accuracy ≥ 90%" and "frequency band 2-18 GHz" and converts them into standardized fields that the system can process. The technical value of this module lies in the automated parsing of unstructured task descriptions, reducing manual input errors. The core of the knowledge graph construction module for establishing a task-based knowledge graph is to associate test resource entities with task requirements. For example, for the above radar test, the graph nodes include "radar A", "jammer B", and "darkroom C" resource entities, and the edge relationships are labeled as "radar A supports frequency band 1-20 GHz" and "darkroom C can simulate 5 types of electromagnetic environments". Historical data from the test resource database, such as "the historical test data confidence of radar A in the 2-10 GHz frequency band is 95%", need to be integrated to form a weighted knowledge network. This module achieves resource matching through semantic reasoning, such as automatically excluding jammer models that are not compatible with the frequency band.

[0085] In the test scheme automatic generation module, the test resource information needs to be dynamically queried from the shared library of the infrastructure layer. For example, three radars that meet the frequency band requirements are selected according to the task requirements, and the interface protocol document is called to verify the communication compatibility. The resource interaction relationship determination link needs to establish a topology structure, such as "radar A→jammer B" needs to be configured with a 1 ms synchronization time delay, and the parameter comes from the historical average value of the device response time recorded in the knowledge graph. The task flow planning adopts the directed acyclic graph algorithm, and the steps such as "initialization→jammer loading→data acquisition" are sorted according to the dependency relationship to avoid resource conflicts. The test data acquisition scheme configuration needs to be combined with the sensor characteristics. For example, to measure the radar detection accuracy, the sampling rate of the calibration device D matched with the radar A is queried in the knowledge graph, and the data packet size and transmission interval are set accordingly. The information display interface definition is based on the rules of human-machine engineering, such as placing the key indicator "false alarm rate" at the top of the visualization panel, and the layout template comes from the pre-defined "electronic warfare test UI specification" in the model library. The constraint condition optimization of the test scheme recommendation module includes multi-objective trade-off. For example, in two generated schemes, scheme 1 uses a high-precision but scarce "darkroom E", and scheme 2 uses an available but 5% less accurate "darkroom F". The system recommends scheme 1 according to the resource priority rules in the knowledge graph (such as "war task priority guarantee accuracy"), and displays the time delay and cost difference of the two schemes through a visual comparison chart. The module realizes dynamic decision-making through a rule engine, for example, when the task urgency is higher than the threshold, the resource level limit is automatically relaxed. The specific implementation case of the resource configuration link: to meet the "multi-target synchronous tracking" requirement, the module derives from the knowledge graph that at least 4 data processing nodes are needed, and calculates the maximum allocatable target number according to "the throughput of node M is 800 MB / s". The node allocation adopts the greedy algorithm, which preferentially deploys high-load tasks to the node with the largest performance margin, which can improve the resource utilization rate by 22%. The technical details of the interaction relationship planning are reflected in the time slot allocation. For example, when "radar A" and "electronic countermeasure device B" need to share bandwidth, the module divides the 0.1 ms level time slice according to the device communication duty cycle recorded in the knowledge graph, and the precision comes from the standard deviation of the device switching delay measured in the historical test. The task flow visualization construction depends on the pre-defined symbol system. For example, red arrows represent electromagnetic interference links, and dashed boxes mark temporary loading resources, which are rendered according to the IEEE standard graphics library. The system automatically generates a Gantt chart to display the time consumption of each stage, and the estimated time length of the "environment calibration stage" comes from the average value of the same type of task in the knowledge graph ± 15% floating interval.

[0086] The intelligent planning process comprises: constructing a test factor knowledge graph; determining resource integration rules according to the knowledge graph; optimizing test factor influence relations; and automatically generating a test scheme file; wherein the test factor knowledge graph contains test resource information and interaction relations thereof. Further, the construction of the test factor knowledge graph comprises: extracting knowledge entities from multi-source heterogeneous data; performing fusion processing on the knowledge entities; storing the fused knowledge entities into a graph database; building association relations among test resources; and forming a knowledge graph containing test resource attributes and relations.

[0087] The method for constructing a test factor knowledge graph comprises extracting knowledge entities from multi-source heterogeneous data.

[0088] For example, the test resource database of a certain range contains structured data such as radar models and frequency range, while the test report document may contain unstructured device coordination records. Through natural language processing technology, the entity relationship "Radar A has interference relationship with electronic countermeasure device B" is extracted from the document, combined with the device parameters in the database, to form a "entity-attribute-relationship" triple. This multi-source extraction can solve the data island problem and ensure that the knowledge graph covers the static attributes and dynamic interaction characteristics of the test resources. The specific operation of fusing knowledge entities includes: when different ranges record the same type of radar as "X-band radar" and "frequency range 8-12 GHz radar" respectively, the ontology mapping technology is used to identify them as the same entity, and a standardized name "model XX-X band radar" is established. For conflicting data such as a device being labeled as "available" and "under maintenance" at the same time, a timestamp weighting algorithm is used to preferentially adopt the latest state data. This processing makes the fragmented knowledge scattered in various systems form a unified expression, providing a consistent data foundation for subsequent intelligent planning. When storing the fused knowledge entities into a graph database, an attribute graph model is used. For example, using the Neo4j database as an example, the node types include "test device", "test subject", etc., and the radar node contains attributes such as "peak power: 100kW". The relationship types include "interference relationship", "communication protocol", etc., and the relationship is attached with quantitative attributes such as "interference strength: -30dB". This storage method can more intuitively express the complex network relationship between resources than traditional relational databases, and improve the efficiency of subsequent graph traversal. Building the association between test resources requires domain rules. For example, by analyzing historical test data, it is found that when the blue electronic jamming power exceeds the red radar receiver sensitivity by 20dB, an "strong interference" relationship is automatically established; or according to the equipment interface protocol document, a "data intercommunication" relationship is established between devices supporting the 1553B bus. This objective index-based association establishment method is more accurate than manual experience in reflecting the real interaction possibility between resources. After forming a complete knowledge graph, typical application scenarios include: when a user queries "resource combination suitable for testing radar anti-jamming performance", the system can quickly locate the device combination with interference relationship through the graph, and recommend the standard test process according to the "test subject-device type" association between nodes. This structured knowledge representation can reveal the implicit relationship between resources better than traditional keyword search, providing a reasoning basis for intelligent planning. When determining the resource integration rules based on the knowledge graph, the topological structure of the graph needs to be analyzed. For example, when the test scheme needs to include the electronic countermeasure subject, the system automatically triggers the rule: if the participating radar and jammer have a "same frequency band" relationship, a spectrum monitoring device must be loaded; if historical data show that the "task success rate" of a certain combination is less than 80%, a warning prompt is generated. These rules are not static presets, but are dynamically optimized by continuously analyzing the statistical characteristics of entity relationships in the graph.For example, by calculating the node centrality, it is found that a certain type of GPS simulator is associated with 80% of the navigation test subjects, and it is marked as a key resource; or through community discovery algorithm, it is found that the coupling degree between radar test resource group and photoelectric test resource group is only 0.2, and it is suggested to test in parallel. This quantitative analysis can find the resource coordination mode that is difficult for human to detect. The process of automatically generating test scheme file is the visualization application of knowledge graph. According to the graph path of "test task-resource requirement-constraint condition", the system automatically combines the required resources, such as selecting the radar equipment that meets the requirements of "distance resolution ≤1m" and "anti-interference level ≥3 level", and generates the equipment connection topology graph according to the interface relationship. In the finally output XML scheme file, the configuration parameters of each resource node are derived from the attribute data in the graph, ensuring the executability of the scheme.

[0089] The above merely describes preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.< / scenarioparameters> < / testobjective>

Claims

1. An LVC intelligent planning system based on an intelligent experimental resource interaction relationship model, characterized in that, The system includes: The infrastructure layer includes resource databases and knowledge graph databases for multiple discrete test ranges; The functional layer includes an experimental task definition module and an intelligent planning scheme module; The application layer includes a test plan generation module and a recommendation module; The system obtains test resources that match the test resource requirements from the infrastructure layer; it performs intelligent planning processing on the test resources through the functional layer to generate a test resource configuration scheme; and it generates and outputs a test scheme file at the application layer.

2. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 1, characterized in that, The infrastructure layer includes: Experimental resource sharing library, experimental data sharing library, experimental resource model library, experimental resource database, and knowledge graph database; The test resource sharing library stores the shareable resources of each discrete test range; The shared experimental data library stores historical experimental data; The experimental resource model library stores experimental resource models; The experimental resource database stores experimental resource parameters; The knowledge graph database stores the relationship data between experimental resources.

3. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 2, characterized in that, The functional layer includes: The test task requirements definition module, test logic range design module, resource loading module, intelligent planning scheme module, and knowledge graph module are all included. The test task requirement definition module parses the test task requirements based on the test task document; The test logic target range design module selects target test resources from the test resource sharing library, test resource model library, and test resource database according to the test task requirements, and defines the interaction relationships between the target test resources. The resource loading module manages the status of the target experimental resources; The intelligent planning module generates an experimental resource configuration scheme based on the interaction relationships between the target experimental resources and the status of the experimental resources; The knowledge graph module constructs an experimental resource relationship graph based on the experimental resource configuration scheme.

4. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 3, characterized in that, The test logic range design module defines the interaction relationships between test resources, specifically including: Obtain the list of experimental resources from the experimental resource sharing library, experimental resource model library, and experimental resource database; Select the target test resource from the test resource list according to the test task requirements; Define the logical ownership relationship of the target experimental resources in a visual manner; Configure the node allocation information for the target experimental resources; Generate a logical target range definition file to define the interaction relationships between target test resources.

5. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 3, characterized in that, The intelligent planning module generates an experimental resource allocation scheme, including: Construct a knowledge graph of experimental factors; Resource integration rules are determined based on the knowledge graph of experimental factors; Optimize the influence relationships of experimental factors; Based on the resource integration rules, the influence relationship of experimental factors, the interaction relationship between target experimental resources, and the status of experimental resources, an experimental resource configuration scheme is automatically generated. The experimental factor knowledge graph contains information on participating resources and their interaction relationships.

6. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 3, characterized in that, The application layer includes: Test task access module, test plan automatic generation module; The test task access module receives test task input and generates test task documents; The test plan automatic generation module generates a test plan based on the test task document and the test resource relationship map; The test scheme recommendation module optimizes test schemes based on constraints.

7. The LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in claim 6, characterized in that, The method for generating test plans by the automatic test plan generation module includes: Import the test task document into the test task requirement definition module of the functional layer, and obtain the corresponding test resource relationship graph; Extract the target experimental resources from the experimental resource relationship graph; Planning the experimental task process; Configure the test data acquisition plan; Define the test information display interface; Based on the target resources and test task flow, generate a test plan file that includes test resource configuration, interaction relationships, and task flow.

8. A planning method for an LVC intelligent planning system based on an intelligent experimental resource interaction relationship model as described in any one of claims 1-7, characterized in that, The method includes: Obtain the test task document; Determine the experimental resource requirements based on the experimental task document; Obtain test resources that match the test resource requirements from the infrastructure layer; The experimental resources are intelligently planned and processed through the functional layer to generate an experimental resource configuration scheme. Generate and output test plan documents at the application layer.

9. The planning method of the LVC intelligent planning system based on the intelligent experimental resource interaction relationship model as described in claim 8, characterized in that, The step of intelligently planning and processing the experimental resources through the functional layer to generate an experimental resource scheme includes: constructing an experimental factor knowledge graph; determining resource integration rules based on the experimental factor knowledge graph; optimizing the influence relationships of experimental factors; and automatically generating an experimental resource configuration scheme based on the resource integration rules, the influence relationships of experimental factors, the interaction relationships between target experimental resources, and the status of experimental resources.

Images