Cloud simulation deduction platform and method

Abstract

The invention relates to the technical field of simulation deduction, in particular to a cloud simulation deduction platform and method, and the platform comprises a core simulation engine sub-platform which is used for carrying out the online detection of a simulation node based on MQTT, and controlling a simulator plug-in to carry out the frame data updating based on a step length; the matched tool chain sub-platform is used for obtaining a standard digital prototype according to a model embedding framework and managing digital prototype information and mirror image files based on model management and mirror image management, and the platform control subsystem is used for receiving detection information sent by the core simulation engine sub-platform, so that an engine is hooked with the standard digital prototype meeting a preset rule and is used for performing simulation on the standard digital prototype. Carrying out simulation deduction by adopting a platform simulation deduction mode, a multi-platform joint simulation deduction mode or a multi-platform parallel simulation deduction mode; the blue army model system sub-platform is used for carrying out blue army model system management based on equipment categories, and carrying out command and control model configuration management and tactical rule setting. According to the invention, the transmission efficiency and accuracy can be improved, and the deduction delay can be reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of simulation deduction, and particularly relates to a cloud simulation deduction platform and method. BACKGROUND

[0002] With the rapid development of cloud computing and distributed simulation technology, cloud simulation confrontation deduction systems are widely used in the fields of tactical training, network security testing, emergency command, etc. The existing technology mostly adopts a centralized architecture based on virtualization technology, and realizes the confrontation deduction process through pre-setting a rule library and static resource allocation. In a typical implementation, the participants access the central server through remote terminals, and the system performs linear deduction based on preloaded environment data or threat models.

[0003] However, the existing technology still has defects: 1. The traditional architecture adopts a fixed resource configuration mode, and when dealing with multi-dimensional concurrent confrontation deduction scene changes, it cannot realize the elastic scaling of computing resources, resulting in a large increase in deduction delay rate. Especially in high-complexity confrontation deduction, the physical resource bottleneck causes the problem of distortion of key deduction nodes.

[0004] 2. The existing system relies on a serial data processing mode of the central server, and when the number of participating nodes is too large or the simulation step is too small, there is a state synchronization error accumulation phenomenon. A typical manifestation is the strategy response lag in multi-participant collaborative deduction.

[0005] 3. The existing technology relies on a pre-set rule library for threat behavior generation, and lacks a dynamic strategy evolution mechanism based on deep reinforcement learning. In the process of continuous confrontation deduction, the attack strategy pattern is highly predictable, which significantly reduces the tactical reference value of the deduction result.

[0006] 4. The existing technology adopts a single virtualization technology stack, and it is difficult to realize the construction of a hybrid deduction environment across cross-platform architectures. In a complex deduction scene containing semi-physical terminals and cloud native nodes, the device simulation fidelity is significantly reduced. SUMMARY

[0007] The purpose of the present application is to provide a cloud simulation deduction platform and method, which solves the problems of large increase in deduction delay rate, state synchronization error accumulation, lack of dynamic strategy evolution mechanism based on deep reinforcement learning, and significant reduction in device simulation fidelity in the existing technology.

[0008] To achieve the above purpose, the present application provides a cloud simulation deduction platform, comprising: The core simulation engine sub-platform is used for online detection of a simulation node based on MQTT, and sends detection information to the platform control subsystem, performs simulator communication based on a DDS protocol, controls frame data update of a simulator plug-in based on a step length, and performs simulation simulation; The matched tool chain sub-platform is used for obtaining a standard digital mockup according to a model embedding framework, performing digital mockup information and image file management based on model management and image management, constructing a simulation scene based on scenario editing and test design, setting simulation parallel resource information based on parallel simulation service, and performing data review analysis; The platform control subsystem is used for receiving the detection information sent by the core simulation engine sub-platform, making the engine connect a standard digital mockup that meets a preset rule, performing simulation deduction in a platform simulation deduction mode, a multi-platform joint simulation deduction mode or a multi-platform parallel simulation deduction mode, and performing intelligent simulation deduction auxiliary support; The blue army model system sub-platform is used for managing a blue army model system based on equipment categories, and performing command and control model configuration management and tactical rule setting.

[0009] In some embodiments of the present application, the core simulation engine sub-platform comprises: A clock management module is used for controlling simulation deduction step length and data information synchronization of nodes in the engine based on a time advancing manner; A node management module is used for checking and verifying before the simulation engine is online, controlling the simulation engine to allocate multi-node resources of a simulation model, and performing real-time parallel simulation; A simulator management module is used for connecting a digital mockup program and an interface program based on a preset API standard process, and making digital mockups and simulation engines under the same standard interconnect; A basic data type specification module is used for configuring and editing visual instruction data, event data and task data structure information based on preset user relative basis and structure body basis.

[0010] In some embodiments of the present application, the core simulation engine sub-platform further comprises: A plug-in and patch module is used for secondary development based on a test verification platform SDK, and provides a simulator plug-in and an engine patch; A scenario analysis module is used for reading a scenario configuration after the engine is started, and performing initialization; A data synchronization module is used for publishing and subscribing distributed messages of a simulation process by using a message middleware, exchanging and cooperating information between components of a simulation system, making nodes in the simulation engine advance according to time nodes and events, and adjusting time state information of each time node based on a time synchronization mechanism.

[0011] In some embodiments of the present application, the matched tool chain sub-platform comprises: a model management module for generating a C++ template project based on a model attribute foundation configuration and performing a model access process; a scenario editing and test design module for performing red-blue side deployment, task setting, event setting, and task rule setting, and performing Monte Carlo, test round, and test complexity design of a basic scenario; a model embedding framework module for providing a standardized model integration framework template program, wherein the standardized model integration framework template program comprises a communication interface, a node mounting interface, an online-offline detection interface, and a time synchronization interface; a parallel simulation service module for adopting a micro-service structure architecture to split a system into a plurality of functional modules, wherein the plurality of functional modules at least comprise a model framework template management module, a heterogeneous model management module, an agent management module, a test management module, a comparison hall management module, a situation visualization service module, and an account management module; a model / image repository management module for sending a generated model image startup container to a server and performing encapsulation management of a simulation application; a review analysis module for performing statistics, analysis, and playback based on stored simulation data.

[0012] In some embodiments of the present application, the platform control subsystem comprises: a platform master module for adding a blue side model, AI master hooking, situation display control hooking, simulation adjudication, simulation control, and data recording based on a core deduction engine.

[0013] a situation display control module for performing a simulation situation visualization service, wherein the simulation situation visualization service at least comprises data filtering and simulation instruction injection; an AI master module for realizing access to a python node based on an AISDK and performing reinforcement learning and AI model calling; an interface adaptation module for providing digital mockup master hooking requirements under multiple protocols.

[0014] In some embodiments of the present application, the blue army model system sub-platform comprises: a blue army model configuration module for configuring a blue army model, wherein the blue army model at least comprises an equipment function model, a sensor function model, an attack command model, a warning command model, and an interception command model; an equipment type module for integrating, storing, and calling various equipment information and providing equipment services; A charge type module is configured to trigger a model state change charge, wherein the model state change charge includes a formation instruction, an assembly instruction, a path instruction, a launch instruction, a detection instruction, and an interference instruction. A tactical rule module is configured to configure tactical rules based on interaction logic and tactical behaviors between different entities, and different scenarios and tactical requirements.

[0015] In some embodiments of the present application, a cloud simulation deduction method is also disclosed, comprising the following steps: S1, start the core engine subsystem, the clock management module is initialized, the initial simulation deduction step is set, the node management module detects the online of the simulation engine, the multi-node resource allocation of the simulation model is configured, the simulator management module connects the digital prototype program and the interface program based on the API standard and process specified by the interface adaptation environment, the basic data type specification module loads the user-defined basic data type structure definition, the plug-in and patch module loads the simulator plug-in and engine patch based on the secondary development of the test verification platform SDK, the scenario analysis module reads the configuration in the scenario, initializes, and the data synchronization module establishes the distribution message publishing and subscribing mechanism through the message middleware; S2, start the supporting tool chain sub-platform, the model management module automatically generates a C++ template project containing the core deduction engine based on the model attribute basic configuration, the scenario editing and test design module performs red and blue side deployment, task setting, event setting, task rule setting, and Monte Carlo, test round, test complexity design based on the basic scenario, the model embedding framework module provides a standardized model integration framework template program, specifies the standardized interface for simulation engine interaction, and the parallel simulation service module develops multiple function simulations based on the B / S architecture, and the model / image repository management module starts a container for publishing to a server through the generated model image, to realize automatic packaging management of the simulation application; S3, start the platform control subsystem, the platform master module adds the blue side model based on the core deduction engine, hangs the AI master, hangs the situation display control, performs simulation adjudication, simulation control, and data recording, the situation display control module realizes simulation situation visualization services, the AI master module realizes access to the python node based on the AISDK, and the interface adaptation module provides digital prototype master hanging requirements under multiple protocols; S4, the blue army model configuration module configures the blue army model, the equipment type module integrates, stores, and calls various equipment information, and provides equipment services for the blue army model system sub-platform, the charge type module triggers the model state change charge, and the tactical rule module configures tactical rules based on interaction logic and tactical behaviors between different entities, and different scenarios and tactical requirements.

[0016] The present application has the following advantages and beneficial effects compared with the prior art: 1. Through the elastic resource scheduling algorithm, the dynamic adaptation of computing resources is realized. In the face of multi-dimensional concurrent confrontation deduction scene changes, the allocation of computing resources can be automatically adjusted according to the real-time load situation, avoiding the problems of resource shortage or excess in the traditional fixed resource configuration mode, thereby effectively reducing the deduction delay.

[0017] 2. The multi-level clock synchronization algorithm and the incremental state transmission technology are adopted, which can effectively solve the state synchronization error accumulation phenomenon. The multi-level clock synchronization algorithm can ensure that the time reference of each node is highly consistent, and the incremental state transmission technology only transmits the change part of the state, reducing the data transmission amount and improving the transmission efficiency and accuracy, so that in large-scale node cooperative deduction, the clock error and strategy response lag time can be kept low, avoiding the accumulation of state synchronization error.

[0018] 3. Through the online learning mechanism, the deduction rule self-evolution is realized. The system can automatically adjust and optimize the strategy library according to the feedback information in the deduction process, so that the system strategy library is automatically iterated and updated every 24 hours, thereby enhancing the adaptability of the system to different environments and confrontation scenes, effectively improving the environmental adaptability compared with traditional systems, and further making up for the defects of the lack of dynamic strategy evolution mechanism in traditional technologies.

[0019] 4. A cross-architecture hybrid simulation platform is constructed, and the heterogeneous resource abstraction layer is used to realize the unified access and high-fidelity simulation of cross-platform and semi-physical terminals. This platform can support mixed networking of devices with different architectures, solving the problem of difficult construction of cross-platform hybrid deduction environment in traditional systems using a single virtualization technology stack, thereby maintaining high device simulation fidelity in a complex deduction scene containing semi-physical terminals and cloud-native nodes.

[0020] The technical solutions of the present application will be described in further detail below with the help of the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 It is a structure schematic diagram of a cloud simulation deduction platform in an embodiment of the present application; Figure 2 It is a step schematic diagram of a cloud simulation deduction method in an embodiment of the present application; Figure 3 It is a schematic diagram of a scene assumption construction process in an embodiment of the present application; Figure 4 It is a schematic diagram of an entity configuration function in an embodiment of the present application; Figure 5 It is a schematic diagram of environment parameter setting in an embodiment of the present application; Figure 6 It is a schematic diagram of a tactical rule design module in an embodiment of the present application; Figure 7 The figure is a schematic diagram of the anti-test design process in the embodiment of the present application. DETAILED DESCRIPTION

[0022] In the description of the present application, it should be noted that the terms "upper", "lower", "inner", "outer" and the like indicate the orientation or positional relationship shown in the drawings, or the orientation or positional relationship in which the product of the present application is usually placed, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In the description of the present application, it should be noted that, unless otherwise explicitly specified and limited, the terms "arrangement", "installation", "connection" should be understood broadly, for example, it can be fixedly connected, or it can be detachably connected, or integrally connected, it can be mechanically connected, or it can be electrically connected, it can be directly connected, or it can be indirectly connected through an intermediate medium, or it can be the communication between two elements inside. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0023] The embodiments of the present application will be described in detail below with reference to the drawings.

[0024] As Figure 1 shown, the present application provides a cloud simulation deduction platform, comprising: A core simulation engine sub-platform, configured to perform online detection of simulation nodes based on MQTT, and send detection information to a platform control subsystem, perform simulator communication based on a DDS protocol, and perform frame data update based on step control of a simulator plug-in, while performing simulation simulation; A supporting tool chain sub-platform, configured to obtain a standard digital mockup according to a model embedding framework, perform digital mockup information and image file management based on model management and image management, build a simulation scene based on scenario editing and test design, set simulation parallel resource information based on parallel simulation service, and perform data review analysis; The platform control subsystem is configured to receive the detection information sent by the core simulation engine sub-platform, so that the engine is connected to a standard digital mockup that meets the preset rules, and performs simulation deduction in a platform simulation deduction, multi-platform joint simulation deduction or multi-platform parallel simulation deduction mode, and performs intelligent simulation deduction auxiliary support; A blue army model system sub-platform, configured to perform blue army model system management based on equipment categories, and perform command and control model configuration management and tactical rule setting.

[0025] The advantages and beneficial effects of the present application relative to the prior art are: 1. Through the elastic resource scheduling algorithm, the dynamic adaptation of computing resources is realized. In the face of multi-dimensional concurrent confrontation deduction scene changes, the allocation of computing resources can be automatically adjusted according to the real-time load situation, avoiding the problems of resource shortage or excess in the traditional fixed resource configuration mode, thereby effectively reducing the deduction delay.

[0026] 2. The multi-level clock synchronization algorithm and the incremental state transmission technology can effectively solve the state synchronization error accumulation phenomenon. The multi-level clock synchronization algorithm can ensure that the time reference of each node is highly consistent, and the incremental state transmission technology only transmits the change part of the state, reducing the data transmission amount and improving the transmission efficiency and accuracy, so that in large-scale node cooperative deduction, the clock error and strategy response lag time can be kept low, avoiding the accumulation of state synchronization error.

[0027] 3. Through the online learning mechanism, the deduction rule is evolved automatically. The system can automatically adjust and optimize the strategy library according to the feedback information in the deduction process, so that the system strategy library is automatically iterated and updated every 24 hours, thereby enhancing the adaptability of the system to different environments and confrontation scenes, effectively improving the environmental adaptability compared with traditional systems, and further making up for the defects of the lack of dynamic strategy evolution mechanism in traditional technologies.

[0028] 4. A cross-architecture hybrid simulation platform is built, and the heterogeneous resource abstraction layer is used to realize the unified access and high-fidelity simulation of cross-platform and semi-physical terminals. This platform can support mixed networking of devices with different architectures, solving the problem of difficult construction of cross-platform hybrid deduction environment in traditional systems using a single virtualization technology stack, thereby maintaining high device simulation fidelity in a composite deduction scene containing semi-physical terminals and cloud-native nodes.

[0029] In some embodiments of the present application, the core simulation engine sub-platform comprises: A clock management module for controlling the simulation deduction step and the data information synchronization of each node in the engine based on a time advancement method; A node management module for checking and verifying before the simulation engine goes online, controlling the simulation engine to allocate multi-node resources of the simulation model, and performing real-time parallel simulation; A simulator management module for connecting the digital mockup program and the interface program based on the preset API standard process, and allowing the digital mockup and the simulation engine under the same standard to be connected to each other; A basic data type specification module for configuring and editing the visual instruction data, event data, and task data structure information based on the preset user relative basis and structure body basis.

[0030] In some embodiments of the present application, the core simulation engine sub-platform further comprises: Plug-ins and patch modules for secondary development based on the test verification platform SDK, and provide simulator plug-ins and engine patches; Scenario analysis module, for reading the configuration of the scenario after the engine starts, and initializing; Data synchronization module, for publishing and subscribing to distributed messages of the simulation process using message middleware, exchanging information and cooperating between components of the simulation system, advancing each node in the simulation engine according to time nodes and events, and adjusting the time state information of each time node based on a time synchronization mechanism.

[0031] In some embodiments of the present application, the supporting tool chain sub-platform includes: Model management module, for generating C++ template projects based on model attribute basic configuration, and performing model access process; Scenario editing and test design module, for red and blue side deployment, task setting, event setting, and task rule setting, and Monte Carlo, test round, and test complexity design of basic scenarios; Model embedding framework module, for providing a standardized model integration framework template program, wherein the standardized model integration framework template program includes a communication interface, a node mounting interface, an online / offline detection interface, and a time synchronization interface; Parallel simulation service module, for splitting the system into multiple functional modules using a microservice architecture, wherein the multiple functional modules include at least a model framework template management module, a heterogeneous model management module, an agent management module, a test management module, a comparison hall management module, a situation visualization service module, and an account management module; Model / image repository management module, for sending the generated model image startup container to the server, and performing encapsulation management of the simulation application; Debriefing analysis module, for performing statistics, analysis, and playback based on stored simulation data.

[0032] In some embodiments of the present application, the platform control subsystem includes: Platform master module, for adding blue side models, AI master hooking, situation display control hooking, simulation adjudication, simulation control, and data recording based on the core deduction engine.

[0033] Situation display control module, for performing simulation situation visualization services, wherein the simulation situation visualization services include at least data filtering and simulation instruction injection; AI master module, for implementing access to python nodes based on AISDK, and performing reinforcement learning and AI model calling; Interface adaptation module, for providing digital mockup master hooking requirements under multiple protocols.

[0034] In some embodiments of the present application, the blue army model system sub-platform comprises: a blue army model configuration module configured to configure a blue army model, wherein the blue army model comprises at least an equipment function model, a sensor function model, an attack command model, an early warning command model, and an interception command model; an equipment type module configured to integrate, store and call various equipment information, and provide equipment services; a command type module configured to trigger model state change commands, wherein the model state change commands comprise group formation commands, assembly commands, path commands, launch commands, detection commands and interference commands; a tactical rule module configured to configure tactical rules based on interaction logic and tactical behavior between different entities, different scenarios and tactical requirements.

[0035] In some embodiments of the present application, as shown in Figure 2 a cloud simulation deduction method is also disclosed, comprising the following steps: S1, starting the core engine subsystem, initializing the clock management module, setting the initial simulation deduction step, detecting the online of the simulation engine by the node management module, configuring the multi-node resource allocation of the simulation model, based on the API standard and process specified by the interface adaptation environment, the simulator management module connects the digital prototype program and the interface program, the basic data type specification module loads the user-defined basic data type structure definition, the plug-in and patch module loads the simulator plug-in and engine patch based on the secondary development of the test verification platform SDK, the scenario analysis module reads the configuration in the scenario, performs initialization, and the data synchronization module establishes the distribution message publishing and subscribing mechanism through the message middleware; S2, starting the supporting tool chain sub-platform, based on the model attribute basic configuration, the model management module automatically generates a C++ template project containing the core deduction engine, the scenario editing and test design module performs red and blue side deployment, task setting, event setting, task rule setting, and Monte Carlo, test round, test complexity design based on the basic scenario, the model embedding framework module provides a standardized model integration framework template program, specifies the standardized interface for simulation engine interaction, and the parallel simulation service module develops multiple function simulations based on the B / S architecture, the model / image repository management module starts the container for publishing to the server through the generated model image, and realizes automatic packaging management of the simulation application; S3, start the platform control subsystem, the platform master module adds the blue side model based on the core deduction engine, AI master hanging, situation display control hanging, simulation arbitration, simulation control and data recording, the situation display control module realizes the simulation situation visualization service, the AI master module realizes the access with the python node based on AISDK, and the interface adaptation module provides the digital mockup master hanging demand under multiple protocols; S4, the blue army model configuration module configures the blue army model, the equipment type module integrates, stores and calls various equipment information, and provides equipment services for the blue army model system sub-platform, the command and control type module triggers the model state change command and control, and the tactical rule module configures the tactical rules based on the interaction logic and tactical behavior between different entities, different scenes and tactical requirements.

[0036] Compared with the prior art, the present application has the following advantages and positive effects: 1. Super large scale parallel simulation capability: the present application supports not less than 1000 parallel function level simulation scenes by optimizing the task scheduling algorithm and distributed computing architecture, breaking through the scene capacity limitation of traditional simulation systems caused by resource competition. Dynamic resource allocation and load balancing technology is adopted, which can simultaneously carry high complexity, multi-dimensional scene parallel deduction, significantly shorten the multi-task cooperative simulation cycle, and improve the design verification efficiency of complex system.

[0037] 2. High concurrency data interaction performance: based on the improved publish-subscribe communication mechanism and multi-channel data compression technology, the present application realizes not less than 1000 data interaction nodes, solves the delay and packet loss problem caused by data congestion in traditional simulation systems. Through priority queue management and bandwidth dynamic allocation strategy, the real-time transmission of key instructions is ensured, which provides stable data interaction support for large-scale joint tactical simulation.

[0038] 3. Intelligent simulation and intelligent agent dynamic cooperation technology (algorithm / agent): the present application realizes the seamless two-way integration of intelligent agent model and simulation system through the innovative AISDK standardized interface adaptation system and simulation engine dynamic hanging technology, and constructs the intelligent simulation training closed loop based on AI master. Through the cross-platform adaptation ability of AISDK interface, the rapid access of multi-modal intelligent agent (reinforcement learning / decision tree / neural network, etc.) is supported, and the problem of poor coupling of traditional simulation system and AI algorithm is solved; the simulation platform generates high-fidelity environment data stream in real time, provides a dynamic countermeasure training field for reinforcement learning algorithm, and through the closed loop mechanism of situation feedback-strategy optimization-retraining, the convergence speed of attack and defense strategy is significantly improved.

[0039] 4. Millisecond simulation accuracy control: through time slice round robin scheduling algorithm and hardware acceleration technology, the invention supports a minimum simulation step of 1ms of the propulsion period, meets the real-time requirements of high dynamic scenes. Compared with the traditional second-level step simulation system, the invention can reduce the dynamic response delay by two orders of magnitude, and significantly improve the transient process restoration accuracy of the simulation result.

[0040] 5. The invention solves the limitations of traditional simulation systems in scene capacity, data throughput, target processing, time accuracy and scene coverage, etc. through technical innovation, and provides an efficient and accurate technical means for the design optimization, equipment demonstration and tactical deduction of complex systems, which has significant application value and industrialization prospect.

[0041] The embodiments of the present application will be described in detail below with specific examples.

[0042] The Jiutian Yunjie distributed simulation platform is an intelligent cloud platform for large-scale distributed system deduction simulation and decision training, which has a stable core deduction engine, a rich tool chain APP and a high-performance cloud native base. The platform uses B / S front-end and back-end separation technology to realize the mode from offline configuration management to configuration file upload to the cloud, and then the private cloud provides comprehensive simulation services to the outside.

[0043] The cloud runs on a server with a Linux system (ubuntu20.04 / Centos7): Docker is used to realize the containerized management of the simulation engine / simulation platform: K8s is used to realize the arrangement and control of distributed simulation computing resources: SpringBoot micro-service architecture is used to realize the external simulation comprehensive service: Cesium is used to realize the visualization of three-dimensional earth, simulation situation and simulation special effects: Offline client implementation: Based on the scenario editing and experimental design platform software, the red and blue parties are deployed, the tasks / events / rules are set, the experiment is configured and the scenario file is generated and uploaded to the cloud; Access the comprehensive simulation service website through the browser, and the simulation process has full-flow functions: Using the Jiutian Yunjie distributed simulation platform, a large-scale intelligent training or intelligent body comparison can be completed in three simple steps: Based on the model management software, the equipment model is introduced, the AISDK development package is used for intelligent body association and design; Based on the scenario editing and experimental design software, the comparison / training experiment is defined; The cloud native service automatically schedules the computing resources for the training / comparison task.

[0044] The distributed simulation platform of Jiutian Yunjie adopts micro-service architecture, and splits the system into different service modules according to functions. According to function points, the services can be divided into services based on the c++ core engine, Java services based on the SpringCloud framework, and python services based on the Flask framework.

[0045] The scenario editing and test design software: the scenario configuration realizes the whole process of the task scenario, generates a scenario file, and stores it in the test verification platform. The specific process is as Figure 3 shown.

[0046] The scenario configuration function of the test verification platform provides functions of "new", "open", "save", "delete", "modify" and "update". In the scenario configuration interface, the new function supports related information input in the scenario configuration operation interface.

[0047] As Figure 4 shown, the entity adding function in the scenario editing can realize the deployment of the equipment model on the map. The specific operation includes functions of "add", "edit", "copy", "delete" and "attribute information" editing of the equipment model entity. In the scenario editing interface, entity adding, entity attribute configuration, environment configuration and running rule configuration can be performed, and visual drag operation can be performed on the map, and icon operation can be performed. The entity model adding adopts the map operation mode to realize the model adding and deployment, selects the party and model type, sets the position on the map after selecting the model, realizes the model adding and deployment, and the model parameters of the newly added model have default values.

[0048] For the equipment entity participating in the deduction simulation, the parameter editing function of the equipment entity is provided, such as DD parameter, radar parameter, satellite device parameter, etc. After editing, the equipment entity parameter configuration file can be formed. The deployed model supports separate parameter editing. After modifying the parameters of the model to be edited in the editing interface, it is stored in the data management module.

[0049] The environment configuration editing is as follows Figure 5 To meet the needs of providing similar environment information in the specific confrontation scene as the real scene, the global or local environment setting function is provided in the scenario editing, and the atmospheric, electromagnetic and sea state parameter setting is supported.

[0050] The environmental parameter design function supports the addition, modification and deletion of environmental parameters. The atmospheric environmental parameter list includes serial number, area, time, weather, light and cloud data. The added data can be modified and deleted. The area selection can be local or global. The local area selection function supports selecting an area on a map, and the global environmental selection function supports global environmental data. The time selection function supports selecting start and end times. The weather can be selected as rain or snow.

[0051] The next step is to design tactical rules. As shown in the following table, the tactical rules that can be configured in the system include instructions for the execution of events or tasks based on certain judgment conditions. These judgment conditions can be composed of various process data, numerical values, events, distance judgments, etc. Therefore, human knowledge conversion is required before simulation and deduction. The settings of self-defined tactical rules are achieved through the settings of launch instructions, detection instructions and escort instructions. In the simulation process, the rules are triggered by triggers. After model processing, the specified methods are sent to the model for execution of specific actions or tasks. Figure 6

[0052] The tactical rule design mainly realizes the interactive data setting between equipment models, including event setting and task setting.

[0053] Among them, the event setting can configure the equipment model and support the setting of model attribute parameters, including the launch position and movement position of the red side launch vehicle, the detection radar range of the blue side, the initial position, sailing speed, trajectory point coordinates and the type and number of equipment hanging on the blue side ship. The task setting can realize the data interaction between models, such as the related settings of detection tasks and launch tasks.

[0054] The next step is to configure and edit the running rules and store the assumptions. The test and verification platform supports the running rule configuration of the simulation process. According to the complexity of the scene process, it supports task step setting in specific fields and whole-process scenarios. It can also design simulation mode, simulation times, single simulation time and other parameters. After the assumption editing is completed, the data content in the data management module is saved to the assumption file. The assumption file is described in an xml or json format file.

[0055] ​Next, master-slave node configuration is performed, the master-slave node management realizes allocation of equipment deployed in the current scene to different nodes, supports subsequent multi-node distributed simulation, and after node editing, node information is uploaded to the server. The main functions include node setting and allocation of entity models. The functions of adding, modifying, deleting and the like are supported, the node setting function realizes multi-node configuration, and the equipment model of the scene is allocated to each node. The node list displays the hierarchical relationship of the nodes, the node information displays the related information of the selected node, including the node name and the entity model information set that has been allocated, the allocation of entity models is to allocate the equipment deployed in the current scene to different nodes, and the operation of deleting the allocated entity models is supported.

[0056] When designing the confrontation experiment, the edited scenario file is configured with test parameters and uploaded to the test verification platform function, and the specific process is as shown in Figure 7 The simulation test design terminal mainly realizes setting of scenario scheme, test parameter information, deployment information, navigation parameters and radar parameters of simulation test data, realizes initial input of deduction simulation, and the edited scenario file needs to generate a test, a new test is added in the test design function module, the user customizes and changes the scenario round in the parameter configuration list, and batch generation of scenarios is realized. Then, the confrontation test is uploaded, the current test is uploaded to the training simulation platform for simulation deduction.

[0057] Next, test management is performed, the test management realizes operations of adding, deleting, editing and uploading of a test set. The test design list interface supports selection of a test, and the uploaded test is uploaded to the comprehensive situation display and control module for visual display of simulation deduction.

[0058] In the present application, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs. If there is any inconsistency, the meaning described in the present specification or the meaning derived from the content described in the present specification shall prevail. In addition, the terms used herein are only for the purpose of describing the embodiments of the present application and are not intended to limit the present application.

[0059] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application and not to limit them, although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can still be modified or replaced by equivalents, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present application.

Claims

1. A cloud simulation deduction platform, characterized by: include: The core simulation engine sub-platform is used to perform online detection of simulation nodes based on MQTT, send detection information to the platform control subsystem, communicate with the simulator based on the DDS protocol, and update frame data based on the step size control simulator plug-in while performing simulation. A supporting tool chain sub-platform is used to obtain standard digital prototypes based on the model embedding framework, manage digital prototype information and image files based on model management and image management, build simulation scenarios based on scenario editing and experimental design, set simulation parallel resource information based on parallel simulation services, and perform data review and analysis; The platform control subsystem is used to receive detection information sent by the core simulation engine sub-platform, so that the engine can be connected to a standard digital prototype that meets preset rules, and to perform simulation deduction using a platform simulation deduction, a multi-platform joint simulation deduction, or a multi-platform parallel simulation deduction mode, and to provide intelligent simulation deduction auxiliary support; The Blue Force Model System sub-platform is used to manage the Blue Force Model System based on equipment categories, and to perform command and control model configuration management and tactical rule settings.

2. A cloud simulation deduction platform according to claim 1, characterized in that: The core simulation engine sub-platform includes: The clock management module is used to control the simulation step size based on the time advancement method and synchronize data information among various nodes in the engine; The node management module is used to check and verify the simulation engine before it goes online, control the simulation engine to allocate multi-node resources of the simulation model, and perform parallel simulation in real time; The simulator management module is used to connect the digital prototype program and the interface program based on the preset API standard process, and to connect the digital prototype and simulation engine under the same standard to each other; The basic data type specification module is used to configure and edit the visual instruction data, event data and task data structure information based on the preset user relative basis and structure basis.

3. A cloud simulation deduction platform according to claim 2, characterized in that: The core simulation engine sub-platform also includes: Plug-in and patch module, used for secondary development based on the test verification platform SDK, and provides simulator plug-ins and engine patches; The scenario parsing module is used to read the scenario configuration and initialize it after the engine starts; The data synchronization module is used to publish and subscribe to distributed messages of the simulation process using message middleware, exchange information and collaborate between the various components of the simulation system, so that each node in the simulation engine advances according to time nodes and events, and adjusts the time status information of each time node based on the time synchronization mechanism.

4. A cloud simulation deduction platform according to claim 3, characterized in that: The supporting tool chain sub-platform includes: The model management module is used to generate C++ template projects based on the basic configuration of model attributes and perform model access processes; The scenario editing and experiment design module is used to deploy the red and blue teams, set tasks, events, and task rules, and perform Monte Carlo simulations, experiment rounds, and experiment complexity design for basic scenarios. The model embedding framework module is used to provide a standardized model integration framework template program, wherein the standardized model integration framework template program includes a communication interface, a node mounting interface, an online and offline detection interface, and a time synchronization interface; A parallel simulation service module is used to split the system into multiple functional modules using a microservices structure architecture. The multiple functional modules include at least a model framework template management module, a heterogeneous model management module, an agent management module, a test management module, a comparison and measurement hall management module, a situation visualization service module, and an account management module. The model / image repository management module is used to send the generated model image startup container to the server and perform packaging management of simulation applications; The replay analysis module is used to perform statistics, analysis, and playback based on stored simulation data.

5. A cloud simulation deduction platform according to claim 4, characterized in that: The platform control subsystem includes: The platform's main control module is used to add blue team models, connect AI main control, connect situation display and control, perform simulation adjudication, simulation control, and data recording based on the core simulation engine; A situation display and control module, configured to provide a simulation situation visualization service, wherein the simulation situation visualization service includes at least data filtering and simulation instruction injection; AI master control module, used to connect to Python nodes based on AISDK and perform reinforcement learning and AI model calls; The interface adapter module is used to meet the requirements of digital prototype master control connection under multiple protocols.

6. A cloud simulation deduction platform according to claim 5, characterized in that: The Blue Army model system sub-platform includes: A blue force model configuration module is used to configure a blue force model, wherein the blue force model includes at least an equipment function model, a sensor function model, an attack control model, an early warning control model, and an interception control model; Equipment type module, used to integrate, store and call various equipment information and provide equipment services; An accusation type module is used to trigger a model state change accusation, wherein the model state change accusation includes a grouping instruction, an assembly instruction, a path instruction, a launch instruction, a detection instruction, and an interference instruction; The tactical rule module is used to configure tactical rules based on the interaction logic and tactical behavior between different entities, different scenarios and tactical requirements.

7. A cloud simulation deduction method, characterized in that: The cloud simulation deduction method is applied to the cloud simulation deduction platform according to any one of claims 1 to 6, comprising the following steps: S1. Start the core engine subsystem, initialize the clock management module, set the initial simulation deduction step size, the node management module detects the simulation engine online, configures the multi-node resource allocation of the simulation model, the simulator management module connects the digital prototype program and the interface program based on the API standards and processes specified by the interface adaptation environment, the basic data type specification module loads the user-defined basic data type structure definition, the plug-in and patch module loads the simulator plug-in and engine patch developed based on the test verification platform SDK, the scenario analysis module reads the configuration in the scenario and performs initialization, and the data synchronization module establishes a distributed message publishing and subscription mechanism through the message middleware; S2. Start the supporting tool chain sub-platform. The model management module automatically generates a C++ template project containing the core deduction engine based on the basic configuration of model attributes. The scenario editing and experiment design module performs red and blue team deployment, task setting, event setting, task rule setting, and Monte Carlo, experiment rounds, and experiment complexity design based on basic scenarios. The model embedding framework module provides a standardized model integration framework template program and specifies the standardized interface for simulation engine interaction. The parallel simulation service module develops multiple functional simulations based on the B / S architecture. The model / image warehouse management module starts the container by generating the model image and publishes it to the server, realizing the automatic packaging management of simulation applications. S3. Start the platform control subsystem. The platform master control module adds the blue team model, connects the AI ​​master control, connects the situation display and control, performs simulation adjudication, simulation control, and data recording based on the core deduction engine. The situation display and control module implements simulation situation visualization services. The AI ​​master control module connects to the Python node based on the AISDK. At the same time, the interface adapter module provides the digital prototype master control connection requirements under multiple protocols. S4. The Blue Army Model Configuration Module configures the Blue Army Model. The Equipment Type Module integrates, stores, and calls various types of equipment information, and provides equipment services for the Blue Army Model System Sub-Platform. The Command and Control Type Module triggers the command and control of model status changes. The Tactical Rule Module configures tactical rules based on the interaction logic and tactical behaviors between different entities, different scenarios, and tactical requirements.

Images