Large language model-based scenario generation method and system

Abstract

The invention discloses a scenario generation method and system based on a large language model, and relates to the technical field of computers. By designing the scenario generation system comprising a scenario generation subsystem, a scenario setting subsystem, a scenario deduction subsystem and a scenario management subsystem, based on a large language model, a process of quickly generating a simulation scenario in a manner of retrieval enhancement generation or combat scenario text conversion is realized; moreover, the system has the functions of scenario editing, scenario deduction, scenario management and the like, achieves a workflow closed loop of intelligent generation, editing management, deduction inspection and modular storage, and improves the scenario generation efficiency and the intelligent level.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a scenario generation method and system based on a large language model. Background Technology

[0002] Simulation scenarios are the fundamental basis and an important component of equipment combat simulation. They primarily describe a synthetic environment, including the behavior and activities of entities such as specific battlefield environments, weapons and equipment, and participating forces. They provide the equipment combat simulation system with initialization data and a series of pre-set actions and commands. Simulation scenarios encompass a wide range of content; without the assistance of relevant tools, generating scenarios manually would be an extremely arduous and complex task.

[0003] Currently, researchers are developing different scenario generation systems for various combat scenarios to shorten scenario generation time and reduce costs. However, due to technical limitations, the intelligence level of these systems is not high. There is an urgent need for a scenario generation system and method to improve generation efficiency and intelligence.

[0004] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a scenario generation method and system based on a large language model.

[0006] To achieve the above objectives, the present invention provides a scenario generation system based on a large language model. The system includes a scenario generation subsystem, a scenario setting subsystem, a scenario deduction subsystem, and a scenario management subsystem. The multiple subsystems interact with each other through a network data interface.

[0007] The scenario generation subsystem includes scenario retrieval and enhanced generation functions and / or combat scenario text conversion and generation functions. It is used to analyze user needs obtained through human-computer interaction based on a large language model and generate simulation scenarios that meet the user needs. The scenario retrieval and enhanced generation function is used to achieve accurate matching between user needs and simulation scenario data by relying on the scenario database through the retrieval and enhanced generation system. The combat scenario text conversion and generation function includes combat scenario text parsing, scenario key information extraction, scenario knowledge graph construction, and knowledge graph-scenario template mapping sub-functions.

[0008] The scenario setting subsystem has military standard plotting and attribute parameter setting functions, which are used to support users in editing and modifying the simulation scenarios generated by the scenario generation subsystem.

[0009] The scenario simulation subsystem is used to support users in performing simulation simulations and verifications of the generated scenarios.

[0010] The scenario management subsystem has functions for creating, opening, deleting, and saving simulation scenarios, which are used to support users in managing simulation scenarios.

[0011] In one embodiment of the present invention, the scenario enhancement retrieval generation function is used to perform multi-level feature retrieval of the scenario database based on the user's intent using a semantic matching algorithm, and to obtain scenario knowledge data by performing knowledge distillation and demand alignment on the retrieval results using dynamic prompt construction technology. The scenario knowledge data is then deeply integrated with the user's intent through a structured prompt template to obtain an enhanced input for a large language model. Based on the large language model and the enhanced input, a simulation scenario that meets the user's needs is generated. The user's intent is obtained by parsing the user's needs based on the large language model.

[0012] In one embodiment of the present invention, the user requirement is a combat scenario text input by the user. The combat scenario text conversion and generation function is used to parse the combat scenario text using a large language model, extract the core information of the combat scenario, and construct a knowledge graph of battlefield environment ontology, combat mission ontology, weapon and equipment ontology, and combat mission ontology based on the core information of the combat scenario. Through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is realized to generate a simulation scenario file that meets the user requirement. The core information of the combat scenario includes, but is not limited to, battlefield environment, combat forces, weapon and equipment, combat mission, and combat rules.

[0013] In one embodiment of the present invention, the scenario setting subsystem is used to respond to a user's selection operation on any simulation scenario generated by the scenario generation subsystem, convert the simulation scenario into a DOM tree structure with multi-layer semantic annotation through a large language model, set relevant parameters according to the user's settings, and generate an edited simulation scenario based on the set parameters; the large language model is connected to a geographic information system, a situation display system, and a scenario simulation tool through an API interface, and the scenario simulation subsystem is used to call the geographic information system for map loading and editing, call the situation display system for simulation model visualization deployment, call the timer of the scenario simulation tool to set the simulation time step, and simulate the simulation scenario after completion.

[0014] In one embodiment of the present invention, the scenario management subsystem is used to decompose the simulated scenario into multiple standardized reusable modules, and store the multiple standardized reusable modules in a scenario database. The multiple standardized reusable modules include, but are not limited to, battlefield environment, combat forces, weaponry, and combat mission modules. Correspondingly, the scenario generation subsystem responds by matching any standardized reusable module according to user requirements, and generates a simulation scenario that meets the user's requirements based on the standardized reusable module.

[0015] In one embodiment of the present invention, the scenario generation system based on a large language model is used to receive user operation data and combat scenario data input by the user through a human-computer interaction interface. The user operation data and combat scenario data include, but are not limited to, relevant parameters of user requirements or settings. The scenario setting subsystem is used to receive simulation scenario data input by the scenario generation subsystem and the scenario management subsystem, as well as scenario verification data and simulation scenario data input by the scenario deduction subsystem. The scenario deduction subsystem is used to receive simulation scenario data and simulation force models input by the scenario generation subsystem and the scenario setting subsystem. The scenario management subsystem is used to receive scenario verification data and simulation scenario data input by the scenario deduction subsystem.

[0016] In one embodiment of the present invention, the scenario generation system based on a large language model includes a basic layer, a technical layer, a functional layer, an interface layer, and an application layer.

[0017] The foundation layer consists of a large language model, application tools, related databases, and resource libraries.

[0018] The technology layer is used to provide related technologies for generating, editing, and deducing simulation scenarios based on large language models;

[0019] The functional layer includes scenario generation, setup, simulation, and management functions that are assembled on demand based on the basic operating environment, the foundation layer, and the technical layer.

[0020] The interface layer provides a human-computer interaction interface, which supports users in performing scenario generation, setting, deduction, and management functions.

[0021] The application layer is used to achieve interconnection and interoperability between the scenario generation system based on the large language model and the equipment combat simulation system through the network interface, and to transmit the generated scenario to the equipment combat simulation system.

[0022] To achieve the above objectives, this invention provides a scenario generation method based on a large language model, comprising:

[0023] Collect user needs based on the human-computer interaction interface;

[0024] Based on a large language model, the user needs are analyzed and simulation scenarios that meet the user needs are generated by using scenario retrieval enhancement generation function or combat scenario text conversion generation.

[0025] In response to the user's selection of any simulation scenario generated by the scenario generation subsystem, the generated simulation scenario is edited and modified based on the user's settings.

[0026] The edited simulation scenario is then extrapolated to obtain a simulation scenario that meets the user's requirements.

[0027] In one embodiment of the present invention, the step of analyzing user needs and generating simulation scenarios that meet those needs based on a large language model, through scenario retrieval enhancement generation or combat scenario text conversion generation, includes any one of the following:

[0028] A semantic matching algorithm is used to perform multi-level feature retrieval of the scenario database based on user intent. Dynamic prompting construction technology is employed to perform knowledge distillation and demand alignment on the retrieval results to obtain scenario knowledge data. This scenario knowledge data is then deeply integrated with the user intent using a structured prompt template to obtain an enhanced input for a large language model. Based on the large language model and the enhanced input, a simulation scenario that meets the user's needs is generated. The user intent is obtained by parsing the user's needs using the large language model. Alternatively, the large language model is used to parse the operational scenario text, extracting core operational scenario information. Based on this core operational information, a knowledge graph is constructed, encompassing battlefield environment, operational mission, weapon and equipment, and operational mission ontology. Through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is achieved, generating a simulation scenario file that meets the user's needs. The user's needs are the operational scenario text input by the user, and the core operational scenario information includes, but is not limited to, battlefield environment, operational forces, weapon and equipment, operational mission, and operational rules.

[0029] In one embodiment of the present invention, the step of editing and modifying the generated simulation scenario based on the user's settings, in response to the user's selection operation of any simulation scenario generated by the scenario generation subsystem, includes:

[0030] In response to the user's selection of any simulation scenario generated by the scenario generation subsystem, the simulation scenario is converted into a DOM tree structure with multi-layer semantic annotation through a large language model. According to the user's settings, relevant parameters are set, and based on the set parameters, an edited simulation scenario is generated. The large language model is connected to the geographic information system, the situation display system, and the scenario inference tool through an API interface.

[0031] Use a geographic information system to load and edit maps;

[0032] The situation demonstration system is invoked to visualize and deploy the simulation model.

[0033] Call the timer of the scenario simulation tool to set the simulation time step;

[0034] Simulation and deduction scenarios.

[0035] In one embodiment of the present invention, the method further includes:

[0036] The simulation scenario, after simulation and deduction, is divided into multiple standardized reusable modules, and these modules are stored in the scenario database. These modules include, but are not limited to, battlefield environment, combat forces, weapons and equipment, and combat mission modules.

[0037] In one embodiment of the present invention, the method further includes:

[0038] In response to matching any standardized reusable module according to user needs, a simulation scenario that meets user needs is generated based on the standardized reusable module.

[0039] To achieve the above objectives, the present invention provides an electronic device comprising one or more processors and one or more memories, wherein the one or more memories store at least one computer program, and the at least one computer program is loaded by the one or more processors to perform the above-described scenario generation system based on a large language model, or to execute the above-described scenario generation method based on a large language model.

[0040] To achieve the above objectives, the present invention provides a computer-readable storage medium storing at least one computer program, wherein the at least one computer program is loaded by a processor to form the scenario generation system based on a large language model, or to execute the scenario generation method based on a large language model.

[0041] This invention designs a scenario generation system comprising a scenario generation subsystem, a scenario setting subsystem, a scenario simulation subsystem, and a scenario management subsystem. Based on a large language model, it realizes the rapid generation of simulation scenarios through retrieval-enhanced generation or text conversion of combat scenarios. The system also has functions such as scenario editing, scenario simulation, and scenario management, realizing a closed-loop workflow of "intelligent generation - editing management - simulation verification - modular storage," thereby improving the efficiency and intelligence level of scenario generation. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the implementation environment of a scenario generation method based on a large language model according to an embodiment of the present invention;

[0044] Figure 2 This is a schematic diagram of a scenario generation system based on a large language model according to an embodiment of the present invention;

[0045] Figure 3 This is a schematic diagram of a scenario generation system architecture based on a large language model according to an embodiment of the present invention;

[0046] Figure 4 This is a schematic diagram of a hypothetical intelligent generation system according to an embodiment of the present invention;

[0047] Figure 5 This is a schematic diagram of the interaction relationship of a hypothetical intelligent generation system according to an embodiment of the present invention;

[0048] Figure 6 This is a schematic diagram of a hypothetical intelligent generation system deployment mode according to an embodiment of the present invention;

[0049] Figure 7 This is a schematic diagram of the workflow of a hypothetical intelligent generation system according to an embodiment of the present invention;

[0050] Figure 8 This is a schematic diagram illustrating a hypothetical search enhancement generation method according to an embodiment of the present invention;

[0051] Figure 9 This is a schematic diagram of a method for generating operational scenario text according to an embodiment of the present invention;

[0052] Figure 10 This is a schematic diagram of a hypothetical setting function design according to an embodiment of the present invention;

[0053] Figure 11 This is a schematic diagram of a scenario deduction function design according to an embodiment of the present invention;

[0054] Figure 12 This is a schematic diagram of a hypothetical management function design according to an embodiment of the present invention;

[0055] Figure 13This is a schematic diagram of a retrieval enhancement generation system Prompt template according to an embodiment of the present invention;

[0056] Figure 14 This is a schematic diagram of a hypothetical search enhancement generation example according to an embodiment of the present invention;

[0057] Figure 15 This is a schematic diagram of a key information extraction example according to an embodiment of the present invention;

[0058] Figure 16 This is a schematic diagram of an example of converting and generating combat scenario text according to an embodiment of the present invention;

[0059] Figure 17 This is a schematic diagram of a scenario deduction process according to an embodiment of the present invention;

[0060] Figure 18 This is a schematic diagram of a scenario generation method based on a large language model according to an embodiment of the present invention;

[0061] Figure 19 This is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention;

[0062] Figure 20 This is a schematic diagram of the structure of a terminal according to an embodiment of the present invention;

[0063] Figure 21 This is a schematic diagram of the structure of a server according to an embodiment of the present invention. Detailed Implementation

[0064] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings, but it should be understood that the scope of protection of the present invention is not limited to the specific embodiments.

[0065] Unless otherwise expressly stated, throughout the specification and claims, the term "comprising" or its variations such as "including" or "comprises" shall be understood to include the stated elements or components without excluding other elements or other components.

[0066] like Figure 1 The diagram illustrates an implementation environment for a scenario generation method based on a large language model according to a preferred embodiment of the present invention. This implementation environment includes a terminal 101, or it may include a terminal 101 and a scenario generation platform 102 based on a large language model. The terminal 101 is connected to the scenario generation platform 102 based on a large language model via a wireless network or a wired network.

[0067] Terminal 101 is at least one of a smartphone, game console, desktop computer, tablet computer, e-book reader, MP3 (Moving Picture Experts Group Audio Layer III) player, or MP4 (Moving Picture Experts Group Audio Layer IV) player. Terminal 101 has installed and runs applications that support scenario generation based on a large language model; for example, the application is a system application, instant messaging application, news push application, shopping application, online video application, or social application.

[0068] For example, the terminal 101 performs this task independently, and also receives data services from the scenario generation platform 102 based on a large language model. This embodiment of the invention does not limit the scope of the invention.

[0069] The scenario generation platform 102 based on a large language model includes at least one of a single server, multiple servers, a cloud computing platform, and a virtualization center. The scenario generation platform 102 provides background services for applications supporting scenario generation based on large language models. Optionally, the scenario generation platform 102 undertakes the primary processing work, and the terminal 101 undertakes secondary processing work; or, the scenario generation platform 102 undertakes secondary processing work, and the terminal 101 undertakes the primary processing work; or, the scenario generation platform 102 or the terminal 101 each undertakes processing work independently. Alternatively, the scenario generation platform 102 and the terminal 101 collaborate using a distributed computing architecture.

[0070] Optionally, the scenario generation platform 102 based on a large language model includes at least one server 1021 and a database 1022. The database 1022 is used to store data. In this embodiment of the invention, the database 1022 stores template examples to provide data services for at least one server 1021.

[0071] A server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. A terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these.

[0072] Those skilled in the art will know that the number of terminals 101 and servers 1021 may be more or less. For example, there may be only one terminal 101 or server 1021, or there may be dozens or hundreds of terminals 101 or servers 1021, or even more. The embodiments of the present invention do not limit the number of terminals or servers or the type of devices.

[0073] like Figure 2 The diagram shown is a schematic of a scenario generation system based on a large language model according to a preferred embodiment of the present invention. The system includes a scenario generation subsystem, a scenario setting subsystem, a scenario deduction subsystem, and a scenario management subsystem, with multiple subsystems exchanging information through a network data interface.

[0074] The scenario generation subsystem includes scenario retrieval and enhanced generation functions and / or combat scenario text conversion and generation functions. It is used to analyze user needs obtained through human-computer interaction based on a large language model and generate simulation scenarios that meet the user needs. The scenario retrieval and enhanced generation function is used to achieve accurate matching between user needs and simulation scenario data by relying on the scenario database through the retrieval and enhanced generation system. The combat scenario text conversion and generation function includes combat scenario text parsing, scenario key information extraction, scenario knowledge graph construction, and knowledge graph-scenario template mapping sub-functions.

[0075] The scenario setting subsystem has military standard plotting and attribute parameter setting functions, which are used to support users in editing and modifying the simulation scenarios generated by the scenario generation subsystem.

[0076] The scenario simulation subsystem is used to support users in performing simulation simulations and verifications of the generated scenarios.

[0077] The scenario management subsystem has functions for creating, opening, deleting, and saving simulation scenarios, which are used to support users in managing simulation scenarios.

[0078] This invention designs a scenario generation system comprising a scenario generation subsystem, a scenario setting subsystem, a scenario simulation subsystem, and a scenario management subsystem. Based on a large language model, it realizes the process of rapidly generating simulation scenarios through retrieval-enhanced generation or text conversion of combat scenarios. The system also has functions such as scenario editing, scenario simulation, and scenario management, realizing a closed-loop workflow of "intelligent generation - editing management - simulation verification - modular storage," which greatly improves the efficiency and intelligence of scenario generation.

[0079] In one embodiment of the present invention, the scenario enhancement retrieval generation function is used to perform multi-level feature retrieval of the scenario database based on the user's intent using a semantic matching algorithm, and to obtain scenario knowledge data by performing knowledge distillation and demand alignment on the retrieval results using dynamic prompt construction technology. The scenario knowledge data is then deeply integrated with the user's intent through a structured prompt template to obtain an enhanced input for a large language model. Based on the large language model and the enhanced input, a simulation scenario that meets the user's needs is generated. The user's intent is obtained by parsing the user's needs based on the large language model.

[0080] In one embodiment of the present invention, the user requirement is a combat scenario text input by the user. The combat scenario text conversion and generation function is used to parse the combat scenario text using a large language model, extract the core information of the combat scenario, and construct a knowledge graph of battlefield environment ontology, combat mission ontology, weapon and equipment ontology, and combat mission ontology based on the core information of the combat scenario. Through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is realized to generate a simulation scenario file that meets the user requirement. The core information of the combat scenario includes, but is not limited to, battlefield environment, combat forces, weapon and equipment, combat mission, and combat rules.

[0081] In one embodiment of the present invention, the scenario setting subsystem is used to respond to a user's selection operation on any simulation scenario generated by the scenario generation subsystem, convert the simulation scenario into a DOM tree structure with multi-layer semantic annotation through a large language model, set relevant parameters according to the user's settings, and generate an edited simulation scenario based on the set parameters; the large language model is connected to a geographic information system, a situation display system, and a scenario simulation tool through an API interface, and the scenario simulation subsystem is used to call the geographic information system for map loading and editing, call the situation display system for simulation model visualization deployment, call the timer of the scenario simulation tool to set the simulation time step, and simulate the simulation scenario after completion.

[0082] In one embodiment of the present invention, the scenario management subsystem is used to decompose the simulated scenario into multiple standardized reusable modules, and store the multiple standardized reusable modules in a scenario database. The multiple standardized reusable modules include, but are not limited to, battlefield environment, combat forces, weaponry, and combat mission modules. Correspondingly, the scenario generation subsystem responds by matching any standardized reusable module according to user requirements, and generates a simulation scenario that meets the user's requirements based on the standardized reusable module.

[0083] In one embodiment of the present invention, the scenario generation system based on a large language model is used to receive user operation data and combat scenario data input by the user through a human-computer interaction interface. The user operation data and combat scenario data include, but are not limited to, relevant parameters of user requirements or settings. The scenario setting subsystem is used to receive simulation scenario data input by the scenario generation subsystem and the scenario management subsystem, as well as scenario verification data and simulation scenario data input by the scenario deduction subsystem. The scenario deduction subsystem is used to receive simulation scenario data and simulation force models input by the scenario generation subsystem and the scenario setting subsystem. The scenario management subsystem is used to receive scenario verification data and simulation scenario data input by the scenario deduction subsystem.

[0084] In one embodiment of the present invention, the scenario generation system based on a large language model includes a basic layer, a technical layer, a functional layer, an interface layer, and an application layer.

[0085] The foundation layer consists of a large language model, application tools, related databases, and resource libraries.

[0086] The technology layer is used to provide related technologies for generating, editing, and deducing simulation scenarios based on large language models;

[0087] The functional layer includes scenario generation, setup, simulation, and management functions that are assembled on demand based on the basic operating environment, the foundation layer, and the technical layer.

[0088] The interface layer provides a human-computer interaction interface, which supports users in performing scenario generation, setting, deduction, and management functions.

[0089] The application layer is used to achieve interconnection and interoperability between the scenario generation system based on the large language model and the equipment combat simulation system through the network interface, and to transmit the generated scenario to the equipment combat simulation system.

[0090] The following is through Figure 3 The illustrated embodiment provides a more detailed description of the scenario generation system based on a large language model. Specifically, as shown... Figure 3 The diagram shown is a schematic of a scenario generation system architecture based on a large language model according to a preferred embodiment of the present invention.

[0091] To achieve intelligent generation of combat simulation scenarios for army equipment, this invention designs an intelligent generation system for combat simulation scenarios based on a large language model (i.e., a scenario generation system based on a large language model, hereinafter referred to as the intelligent scenario generation system). This system can quickly generate simulation scenarios through scenario retrieval enhancement generation or combat scenario text conversion, and has functions such as scenario editing, scenario simulation, and scenario management, realizing a closed-loop workflow of "intelligent generation - editing management - simulation verification - modular storage".

[0092] The proposed intelligent generation system architecture adopts a layered architecture, mainly divided into five layers: foundation layer, technology layer, functional layer, interface layer, and application layer.

[0093] The foundation layer consists of a large language model, application tools, related databases, and resource libraries, providing the large language model, application tool software, and related data resources necessary for the application system to run.

[0094] Technical layer: mainly related technologies for generating, editing, and deducing simulation scenarios based on large language models.

[0095] Functional layer: Based on the basic operating environment, foundational layer and technical layer, the system provides scenario generation, setting, simulation and management functions as needed.

[0096] Interface layer: Human-computer interaction interface, supporting users to realize scenario generation, setting, simulation and management functions.

[0097] Application layer: Through network interface, the intelligent scenario generation system and the equipment combat simulation system are interconnected, and the generated scenario is transmitted to the equipment combat simulation system.

[0098] The intelligent scenario generation system primarily completes the visual design and simulation verification of scenarios before equipment combat simulation. Based on logical relevance and different user operation time stages, it can be divided into a scenario generation subsystem, a scenario setting subsystem, a scenario simulation subsystem, and a scenario management subsystem. For example... Figure 4 As shown.

[0099] The scenario generation subsystem is primarily used to generate the simulation scenarios required by the user. This subsystem has two functions: scenario retrieval and enhanced generation, and operational scenario text conversion and generation. The operational scenario text conversion and generation function is further subdivided into four sub-functions: operational scenario text parsing, scenario key information extraction, scenario knowledge graph construction, and knowledge graph-scenario template mapping.

[0100] The scenario setting subsystem primarily supports users in editing and modifying simulation scenarios generated by the scenario generation subsystem until they meet the requirements of the current simulation task. This subsystem has two functions: military insignia plotting and attribute parameter setting. The military insignia plotting function is further subdivided into three sub-functions: electronic map loading, electronic map editing, and military insignia selection and plotting. The attribute parameter setting function is further subdivided into five sub-functions: battlefield environment parameter setting, combat entity parameter setting, combat mission parameter setting, and combat interaction parameter setting.

[0101] Scenario simulation subsystem: Supports users in performing simulations and verifications on the generated scenarios. The subsystem has two functions: timer setting and situation playback.

[0102] Scenario Management Subsystem: Supports user management of simulation scenarios. The subsystem has four functions: creating a new simulation scenario, opening a simulation scenario, deleting a simulation scenario, and saving a simulation scenario.

[0103] In one embodiment of the present invention, such as Figure 5 As shown, the interactive relationships of the intelligent scenario generation system are mainly divided into user-system information interaction and internal system information interaction. User-system information interaction supports human-computer interaction, receiving user operation data and combat scenario data, allowing users to perform simulation scenario generation, setting, deduction, and management. Internal system information interaction is used for scenario generation, scenario setting, scenario deduction, and information interaction between subsystems.

[0104] For information interaction between users and the system, the main method is the human-computer interaction interface. The system receives user operation data and combat scenario data, and supports users in performing operations such as simulation scenario generation, setting, simulation, and management. Details of the interaction information are shown in Table 1 below.

[0105] Table 1 Details of User-System Interaction Information

[0106]

[0107] For internal information exchange within the system, the scenario generation, scenario setting, scenario simulation, and scenario management subsystems mainly exchange information through network data interfaces. Details of the exchange information are shown in Table 2 below.

[0108] Table 2 Details of Internal System Interaction Information

[0109]

[0110] Regarding the deployment mode of the proposed intelligent generation system, in one embodiment of the present invention, the deployment mode can be referred to the above. Figure 1The illustrated embodiment, i.e., the envisioned intelligent generation system, can adopt a layered model consisting of a foundation layer, a technology layer, a functional layer, an interface layer, and an application layer.

[0111] In another embodiment of the invention, see [link to relevant documentation]. Figure 6 The diagram shows the server / browser mode, also known as the B / S mode. This means that the envisioned intelligent generation system can adopt the B / S mode.

[0112] The server is used to deploy software services, while the client provides interfaces to these software services. Users access the application services provided by the server through the client. The server and the client (the user terminal) are connected via a core switch.

[0113] The server-side primarily consists of five types of servers: application servers, used to deploy applications; large language models, used to support the implementation of the system's main functions; data resource servers, used to deploy database and model library services; storage servers, containing multiple specific storage units for storing desired data; and backup servers, used to back up relevant data. These five types of servers are located on the same network segment and connected using fiber optic switches to ensure efficient and fast data transmission between servers.

[0114] The terminals may be distributed across various departments or centrally deployed in a few dedicated rooms, providing access to simulation scenario generation, simulation scenario setting, simulation scenario derivation, and simulation scenario management services.

[0115] In some embodiments of this application, the intelligent scenario generation system supports creating new simulation scenario files, setting simulation scenario files, deriving simulation scenario files, and saving simulation scenario files. Specifically, the workflow of the intelligent scenario generation system is as follows: Figure 7 As shown, it can include the following four steps.

[0116] Step 1: Create a new simulation scenario file.

[0117] The user can choose either scenario retrieval enhancement generation or combat scenario text conversion to generate the simulation scenario. If the scenario meets the user's expectations, proceed to step 3. If not, proceed to step 2.

[0118] Step 2: Set up the simulation scenario file.

[0119] The simulation scenario is analyzed using a large language model. Users set parameters such as battlefield environment, combat entities, combat missions, combat actions, and combat interactions as required. After setting, proceed to step 3.

[0120] Step 3: Simulation scenario file.

[0121] If there are errors in the simulation, return to step 2 until the required simulation scenario is generated.

[0122] Step 4: Save the simulation scenario file.

[0123] Save the simulated scenario blocks that have been derived and verified to the designated location.

[0124] The intelligent scenario generation system is classified according to its functions, including scenario generation subsystem, scenario setting subsystem, scenario simulation subsystem, and scenario management subsystem.

[0125] The scenario generation subsystem uses semantic matching algorithms to perform multi-level feature retrieval on the scenario database, employs dynamic prompting construction technology to perform knowledge distillation and demand alignment on the retrieval results, and deeply integrates scenario knowledge data with user intent through structured prompt templates to generate scenario combat simulation scenarios that meet simulation requirements.

[0126] The scenario generation subsystem uses a large language model to parse combat scenario texts, extract multiple core elements of equipment combat, construct corresponding knowledge graphs based on these core elements, and then associate and map each knowledge graph with a scenario template to generate simulation scenario files.

[0127] The scenario setting subsystem is used to receive the XML simulation scenario in the selected operation, convert the XML simulation scenario into a DOM tree structure with multi-layer semantic annotation, and set parameters such as battlefield environment, combat entities, combat tasks and combat interactions according to the equipment combat simulation mission requirements, using military standard plotting, parameter input and other means. The DOM tree document is then converted into an XML Schema document through a large language model to generate the required simulation scenario.

[0128] The scenario simulation subsystem is used to visualize geographic information systems, situation display systems, and scenario simulation tools connected through API interfaces using a large language model. It provides feedback based on user geographic information loading operations, electronic map editing operations, and time step setting operations, and performs simulation scenario simulation based on the determined information.

[0129] The scenario management subsystem is used to save, query, and delete simulation scenarios.

[0130] Similarly, the functions of the intelligent scenario generation system include scenario generation, scenario setting, scenario simulation, and scenario management. Each function will be described in detail below.

[0131] Function 1: Scenario Generation

[0132] Specifically, the scenario enhancement retrieval generation method mainly achieves accurate matching of combat simulation scenarios from the scenario database through a scenario enhancement generation system. Specifically, firstly, a semantic matching algorithm is used to perform multi-level feature retrieval on the scenario database. Then, dynamic prompting construction technology is used to perform knowledge distillation and requirement alignment on the retrieval results. Finally, a structured prompt template is used to deeply integrate scenario knowledge data with user intent, forming an enhanced input for a large-scale language model, thereby generating customized scenarios that meet simulation requirements. The specific design is as follows... Figure 8 As shown.

[0133] Specifically, the method for generating operational scenario text first uses a large language model to parse the text, extracting core elements such as the battlefield environment, combat forces, weaponry, combat missions, and operational rules. Based on this, knowledge graphs are formed for the battlefield environment ontology, combat mission ontology, weaponry ontology, and combat mission ontology. Finally, through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is achieved, generating XML simulation scenario files, thus realizing the automated conversion from unstructured text to an executable simulation model. The specific design is as follows... Figure 9 As shown.

[0134] Function 2: Preset Settings

[0135] The large language model transforms the user-selected XML simulation scenario into a DOM tree structure with multi-layered semantic annotations, achieving intelligent mapping from XML elements to operational semantics. Then, based on the equipment's operational simulation mission requirements, the user sets parameters such as battlefield environment, combat entities, combat tasks, and combat interactions using methods like military standard plotting and parameter input. Finally, the large language model converts the DOM tree document into an XML Schema document, thereby generating the required simulation scenario. The specific design is as follows... Figure 10 As shown.

[0136] Function 3: Scenario Deduction

[0137] The large language model connects to a Geographic Information System (GIS), a situation display system, and a scenario simulation tool via an API interface. Users load and edit electronic maps using the GIS, visualize and deploy the simulation model using the situation display system, and set the simulation time step using the timer in the scenario simulation tool to begin the simulation scenario. During the simulation, the system performs the simulation according to the set time step. The simulation situation remains unchanged until the next time step is reached; when the next time step is reached, the simulation situation changes. This process continues until the simulation ends. The specific design is as follows... Figure 11 As shown.

[0138] Function 4: Scenario Management

[0139] The scenario management system manages the entire process of saving, querying, and deleting simulation scenarios. Scenarios that have undergone simulation and meet user requirements are broken down into standardized, reusable modules such as battlefield environment, combat forces, weaponry, and combat missions by the retrieval and enhancement generation system, and stored in the scenario resource library. Specific design details are as follows... Figure 12 As shown.

[0140] The structure and functions of the intelligent generation system have been described in detail above. In a specific embodiment of the present invention, a specific application example is provided.

[0141] The system environment construction includes an explanation of the construction of the large language model and the retrieval enhancement generation system.

[0142] (1) Large Language Model

[0143] The intelligent generation system is based on the Ollam framework, with a finely tuned Deepseek-R1-70B as the base model. This model is based on a hybrid expert architecture (MoE), multi-head latent attention mechanism (MLA), and reinforcement learning algorithm (GRPO), and has advantages such as high efficiency, low cost, open source ecosystem, and support for Chinese context.

[0144] The specific fine-tuning steps are as follows:

[0145] ① Constructing a fine-tuned dataset. The dataset is derived from publicly available military documents, historical combat scenario texts, and equipment combat simulation corpora, and after cleaning and enhancement, it contains nearly 200,000 entries.

[0146] ② The LoRA algorithm was fine-tuned and distributed training was performed on 4 NVIDIA RTX 4090s. The optimization was achieved through gradient accumulation (steps=4) and mixed precision (FP16), with a total time consumption of 200 GPU hours.

[0147] ③ Merge the original model with the LoRA adapter to obtain the final fine-tuned model.

[0148] (2) Search Enhancement Generation System

[0149] The retrieval enhancement generation system uses LangChain to construct the core logic of RAG. FastAPI provides a RESTful API interface as the backend, supporting interaction with external systems. The template library is deployed offline using the Milvus vector database, containing over 2000 standardized XML templates, categorized and stored according to combat type and scale hierarchy. Nomic-Embed-TXT encodes user queries and template text into 768-dimensional vectors, combining a hybrid retrieval strategy to improve accuracy. Deepseek-R1-70B is integrated through LangChain, generating simulation scenario text conforming to XML Schema based on Prompt templates, and ensuring correct formatting through a post-processing module. The Prompt template is as follows: Figure 13 As shown.

[0150] The scenario generation capability of this application instance was verified, and the verification process and results are as follows.

[0151] Using a tactical-level urban offensive operation as a backdrop, the intelligent scenario generation system automatically generates XML simulation scenarios through two methods: scenario retrieval-enhanced generation and operational scenario text conversion generation.

[0152] 1. Validation of the scenario-based search enhancement generation method

[0153] Users submit their scenario generation requirements, and the intelligent scenario generation system uses a large language model and a retrieval-enhanced generation system to generate the required scenarios for the user, such as... Figure 14 As shown.

[0154] 2. Verification of the method for converting and generating combat scenario text

[0155] The intelligent scenario generation system reads the urban offensive operation scenario text provided by the user and automatically converts it into a simulation scenario. Key information extracted from the operational scenario text using a large language model is shown in Table 3 below. An example demonstration is provided. Figure 15 As shown.

[0156] Table 3 Key Information Extraction List

[0157]

[0158] Based on the aforementioned key information, knowledge graphs for battlefield environment, combat forces, weaponry, and combat missions are constructed. These knowledge graphs load combat data and map it to a scenario template to generate simulation scenarios, such as... Figure 16 As shown.

[0159] 3. Scenario simulation verification

[0160] The scenario simulation subsystem performs simulations and verifications on scenarios generated by both the scenario retrieval enhancement generation method and the operational scenario text conversion generation method. The simulation process for scenarios generated by the operational scenario text conversion generation method is as follows: Figure 17 As shown.

[0161] 4. Verification Conclusion

[0162] Test results demonstrate that the intelligent scenario generation system can generate simulation scenarios through scenario retrieval enhancement generation and combat scenario text conversion generation.

[0163] For the scenario generation efficiency verification, the intelligent scenario generation system generated 10 tactical-level equipment combat simulation scenarios using the two generation methods mentioned above, and compared the scenario generation efficiency with that of a certain type of armored equipment combat simulation scenario generation system. The results are shown in Table 4.

[0164] Table 4 Comparison of Expected Generation Efficiency

[0165]

[0166] The results of the generation efficiency comparison experiment show that the average generation time and number of manual interventions for the combat scenario text conversion generation method are less than those for the scenario retrieval enhancement generation method and the traditional scenario generation method.

[0167] For scenario reusability verification, generally speaking, since large language models are specifically trained and fine-tuned for scenario generation tasks in specific scenarios, they can meet the needs of scenario generation tasks in specific scenarios. If the large language model has not been previously trained and fine-tuned for scenario generation tasks in other scenarios, can the large language model achieve cross-scenario reuse of scenario modules? If so, what is the reuse rate? Does cross-scenario scenario generation (scenario module reuse + manual editing and adjustment) have any advantages over traditional scenario generation systems in terms of scenario generation time? To this end, the following experiment is designed:

[0168] ① A standardized database is pre-built containing the following modules: tactical-level urban offensive battlefield environment module (10 types), combat force module (15 types), weapon and equipment model (25 types), and combat mission module (15 types);

[0169] ② Test scenarios: Two types of combat scenarios were selected: tactical town defense and island / reef seizure and control;

[0170] ③ Test data volume: 20 scenarios are generated for each type of test scenario. The intelligent scenario generation system and the combat simulation scenario generation system for a certain type of armored equipment each generate 10 scenarios. Among them, the intelligent scenario generation system adopts the scenario retrieval enhancement generation method.

[0171] ④ Testing process: The intelligent simulation system uses tactical-level urban offensive battlefield environment modules, combat force modules, weapon and equipment models, and combat mission modules to be reused in two test scenarios: urban defense and key point seizure and control; a certain type of armored equipment combat simulation scenario generation system generates two types of test scenario equipment combat simulation scenarios.

[0172] ⑤ Test metrics: Module reuse rate and scenario generation time

[0173]

[0174] P represents the module reuse rate, α represents the number of manually adjusted target parameters, and β represents the total number of target parameters. The test statistics are shown in Table 5.

[0175] Table 5 Comparison of the reusability of the intended modules

[0176]

[0177] The comparison results of scenario module reusability show that by combining modular retrieval-enhanced generation technology with an intelligent scenario generation system, scenario modules can be reused across scenarios. If the task scenarios are highly similar, the reuse rate of the scenario modules is higher, the time spent on manual adjustments is less, and the scenario generation time is also reduced. Conversely, if the task scenarios are lowly similar, the reuse rate of the scenario modules is lower, the time spent on manual adjustments is more, and the scenario generation time increases, but it is still less than the time required for traditional manual scenario generation.

[0178] The verification conclusions are as follows.

[0179] (1) The intelligent scenario generation system can generate simulation scenarios through scenario retrieval enhancement generation method and combat scenario text conversion generation method.

[0180] (2) For scenario generation tasks in specific scenarios, the combat scenario text conversion generation method has a better effect;

[0181] (3) For scenario generation tasks across scenarios, a scenario retrieval-enhanced generation method can be selected;

[0182] (4) Search-enhanced generation technology can significantly improve the efficiency and flexibility of scenario generation, while providing automated support for scenario modular reuse.

[0183] This invention designs a scenario generation system comprising a scenario generation subsystem, a scenario setting subsystem, a scenario simulation subsystem, and a scenario management subsystem. Based on a large language model, it realizes the rapid generation of simulation scenarios through retrieval-enhanced generation or text conversion of combat scenarios. The system also has functions such as scenario editing, scenario simulation, and scenario management, realizing a closed-loop workflow of "intelligent generation - editing management - simulation verification - modular storage," thereby improving the efficiency and intelligence level of scenario generation.

[0184] like Figure 18 The diagram illustrates a scenario generation method based on a large language model according to a preferred embodiment of the present invention. This embodiment includes the following steps.

[0185] S1801. Electronic devices collect user needs based on human-computer interaction interfaces.

[0186] S1802. The electronic device, based on a large language model, analyzes the user needs and generates simulation scenarios that meet the user needs by means of scenario retrieval enhancement generation function or combat scenario text conversion generation.

[0187] S1803. In response to the user's selection operation on any simulation scenario generated by the scenario generation subsystem, the electronic device edits and modifies the generated simulation scenario based on the user's settings.

[0188] S1804. The electronic device performs simulations on the edited simulation scenario to obtain a simulation scenario that meets the user's requirements.

[0189] In one embodiment of the present invention, the step of analyzing user needs and generating simulation scenarios that meet those needs based on a large language model, through scenario retrieval enhancement generation or combat scenario text conversion generation, includes any one of the following:

[0190] A semantic matching algorithm is used to perform multi-level feature retrieval of the scenario database based on user intent. Dynamic prompting construction technology is employed to perform knowledge distillation and demand alignment on the retrieval results to obtain scenario knowledge data. This scenario knowledge data is then deeply integrated with the user intent using a structured prompt template to obtain an enhanced input for a large language model. Based on the large language model and the enhanced input, a simulation scenario that meets the user's needs is generated. The user intent is obtained by parsing the user's needs using the large language model. Alternatively, the large language model is used to parse the operational scenario text, extracting core operational scenario information. Based on this core operational information, a knowledge graph is constructed, encompassing battlefield environment, operational mission, weapon and equipment, and operational mission ontology. Through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is achieved, generating a simulation scenario file that meets the user's needs. The user's needs are the operational scenario text input by the user, and the core operational scenario information includes, but is not limited to, battlefield environment, operational forces, weapon and equipment, operational mission, and operational rules.

[0191] In one embodiment of the present invention, the step of editing and modifying the generated simulation scenario based on the user's settings, in response to the user's selection operation of any simulation scenario generated by the scenario generation subsystem, includes:

[0192] In response to the user's selection of any simulation scenario generated by the scenario generation subsystem, the simulation scenario is converted into a DOM tree structure with multi-layer semantic annotation through a large language model. According to the user's settings, relevant parameters are set, and based on the set parameters, an edited simulation scenario is generated. The large language model is connected to the geographic information system, the situation display system, and the scenario inference tool through an API interface.

[0193] Calling the Geographic Information System for map loading and editing; calling the Situation Demonstration System for simulation model visualization and deployment;

[0194] Call the timer of the scenario simulation tool to set the simulation time step;

[0195] Simulation and deduction scenarios.

[0196] In one embodiment of the present invention, the method further includes:

[0197] The simulation scenario, after simulation and deduction, is divided into multiple standardized reusable modules, and these modules are stored in the scenario database. These modules include, but are not limited to, battlefield environment, combat forces, weapons and equipment, and combat mission modules.

[0198] In one embodiment of the present invention, the method further includes:

[0199] In response to matching any standardized reusable module according to user needs, a simulation scenario that meets user needs is generated based on the standardized reusable module.

[0200] Figure 19 This is a schematic diagram of an electronic device 1900 provided in an embodiment of the present invention. The electronic device 1900 can vary considerably depending on its configuration or performance. It can include one or more Central Processing Units (CPUs) 1901 and one or more memories 1902. The memories 1902 store at least one computer program. This at least one computer program is loaded by the processor 1901 and executed to implement the scenario generation method based on the large language model provided in the various method embodiments. The electronic device can also include other components for implementing device functions. For example, the electronic device can also have wired or wireless network interfaces and input / output interfaces for input and output. Further details of this embodiment are not elaborated upon here.

[0201] The electronic device in the above method embodiments can be implemented as a terminal. For example, Figure 20 This is a structural block diagram of a terminal provided in an embodiment of the present invention. The terminal 2000 can be a portable mobile terminal, such as a smartphone, tablet computer, MP3 (Moving Picture Experts Group Audio Layer III) player, MP4 (Moving Picture Experts Group Audio Layer IV) player, laptop computer, or desktop computer. The terminal 2000 may also be referred to as a user device, portable terminal, laptop terminal, desktop terminal, or other names.

[0202] Typically, terminal 2000 includes a processor 2001 and a memory 2002.

[0203] Processor 2001 may include one or more processing cores, such as a quad-core processor, an octa-core processor, etc. Processor 2001 may be implemented using at least one hardware form selected from DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). Processor 2001 may also include a main processor and a coprocessor. The main processor, also known as a CPU (Central Processing Unit), is used to process data in the wake-up state; the coprocessor is a low-power processor used to process data in the standby state. In some embodiments, processor 2001 may integrate a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content required to be displayed on the screen. In some embodiments, processor 2001 may also include an AI (Artificial Intelligence) processor, which is used to handle computational operations related to machine learning.

[0204] The memory 2002 may include one or more computer-readable storage media, which may be non-transitory. The memory 2002 may also include high-speed random access memory and non-volatile memory, such as one or more disk storage devices or flash memory devices. In some embodiments, the non-transitory computer-readable storage media in the memory 2002 is used to store at least one instruction, which is executed by the processor 2001 to implement the scenario generation method based on a large language model provided in the embodiments of the present invention or the scenario generation system based on a large language model described above.

[0205] In some embodiments, the terminal 2000 may also optionally include a peripheral device interface 2003 and at least one peripheral device. The processor 2001, memory 2002, and peripheral device interface 2003 can be connected via a bus or signal line. Each peripheral device can be connected to the peripheral device interface 2003 via a bus, signal line, or circuit board. Specifically, the peripheral device includes at least one of the following: a radio frequency circuit 2004, a display screen 2005, a camera assembly 2006, an audio circuit 2007, a positioning assembly 2008, and a power supply 2009.

[0206] Peripheral device interface 2003 can be used to connect at least one I / O (Input / Output) related peripheral device to processor 2001 and memory 2002. In some embodiments, processor 2001, memory 2002 and peripheral device interface 2003 are integrated on the same chip or circuit board; in some other embodiments, any one or two of processor 2001, memory 2002 and peripheral device interface 2003 can be implemented on separate chips or circuit boards, which is not limited in this embodiment.

[0207] The radio frequency (RF) circuit 2004 is used to receive and transmit RF (Radio Frequency) signals, also known as electromagnetic signals. The RF circuit 2004 communicates with communication networks and other communication devices via electromagnetic signals. The RF circuit 2004 converts electrical signals into electromagnetic signals for transmission, or converts received electromagnetic signals into electrical signals. Optionally, the RF circuit 2004 includes: an antenna system, an RF transceiver, one or more amplifiers, a tuner, an oscillator, a digital signal processor, a codec chipset, a user identity module card, etc. The RF circuit 2004 can communicate with other terminals through at least one wireless communication protocol. This wireless communication protocol includes, but is not limited to: the World Wide Web, metropolitan area networks, intranets, various generations of mobile communication networks (2G, 3G, 4G, and 5G), wireless local area networks, and / or WiFi (Wireless Fidelity) networks. In some embodiments, the RF circuit 2004 may also include circuitry related to NFC (Near Field Communication), which is not limited in this respect.

[0208] Display screen 2005 is used to display a UI (User Interface). This UI may include graphics, text, icons, video, and any combination thereof. When display screen 2005 is a touch display screen, it also has the ability to collect touch signals on or above its surface. These touch signals can be input as control signals to processor 2001 for processing. In this case, display screen 2005 can also be used to provide virtual buttons and / or a virtual keyboard, also known as soft buttons and / or a soft keyboard. In some embodiments, there may be one display screen 2005, disposed on the front panel of terminal 2000; in other embodiments, there may be at least two display screens, disposed on different surfaces of terminal 2000 or in a folded design; in still other embodiments, display screen 2005 may be a flexible display screen, disposed on a curved or folded surface of terminal 2000. Furthermore, display screen 2005 may also be configured as a non-rectangular, irregular shape, i.e., a non-rectangular screen. The display screen 2005 can be made of materials such as LCD (Liquid Crystal Display) and OLED (Organic Light-Emitting Diode).

[0209] The camera assembly 2006 is used to acquire images or videos. Optionally, the camera assembly 2006 includes a front-facing camera and a rear-facing camera. Typically, the front-facing camera is located on the front panel of the terminal, and the rear-facing camera is located on the back of the terminal. In some embodiments, there are at least two rear-facing cameras, which are any one of a main camera, a depth-sensing camera, a wide-angle camera, and a telephoto camera, to achieve background blurring by fusion of the main camera and the depth-sensing camera, panoramic shooting by fusion of the main camera and the wide-angle camera, VR (Virtual Reality) shooting, or other fusion shooting functions. In some embodiments, the camera assembly 2006 may also include a flash. The flash can be a single-color temperature flash or a dual-color temperature flash. A dual-color temperature flash refers to a combination of a warm-light flash and a cool-light flash, which can be used for light compensation at different color temperatures.

[0210] The audio circuit 2007 may include a microphone and a speaker. The microphone is used to collect sound waves from the user and the environment, converting them into electrical signals that are input to the processor 2001 for processing, or to the radio frequency circuit 2004 for voice communication. For stereo sound acquisition or noise reduction purposes, multiple microphones may be used, each positioned at a different location on the terminal 2000. The microphone may also be an array microphone or an omnidirectional microphone. The speaker is used to convert electrical signals from the processor 2001 or the radio frequency circuit 2004 into sound waves. The speaker may be a conventional diaphragm speaker or a piezoelectric ceramic speaker. When the speaker is a piezoelectric ceramic speaker, it can convert electrical signals not only into audible sound waves but also into inaudible sound waves for purposes such as distance measurement. In some embodiments, the audio circuit 2007 may also include a headphone jack.

[0211] The positioning component 2008 is used to locate the current geographical location of the terminal 2000 in order to enable navigation or LBS (Location Based Service). The positioning component 2008 can be a positioning component based on the US GPS (Global Positioning System), China's BeiDou system, or Russia's Galileo system.

[0212] Power supply 2009 is used to power the various components in terminal 2000. Power supply 2009 can be AC ​​power, DC power, a disposable battery, or a rechargeable battery. When power supply 2009 includes a rechargeable battery, the rechargeable battery can be a wired rechargeable battery or a wireless rechargeable battery. A wired rechargeable battery is a battery that is charged via a wired line, and a wireless rechargeable battery is a battery that is charged via a wireless coil. The rechargeable battery can also be used to support fast charging technology.

[0213] In some embodiments, the terminal 2000 further includes one or more sensors 2010. The one or more sensors 2010 include, but are not limited to: an accelerometer 2011, a gyroscope 2012, a pressure sensor 2013, a fingerprint sensor 2014, an optical sensor 2015, and a proximity sensor 2016.

[0214] Accelerometer 2011 can detect the magnitude of acceleration along the three coordinate axes of a coordinate system established by terminal 2000. For example, accelerometer 2011 can be used to detect the components of gravitational acceleration along the three coordinate axes. Processor 2001 can control display screen 2005 to display the user interface in either a landscape or portrait view based on the gravitational acceleration signal acquired by accelerometer 2011. Accelerometer 2011 can also be used for games or for acquiring user motion data.

[0215] The gyroscope sensor 2012 can detect the orientation and rotation angle of the terminal 2000. The gyroscope sensor 2012, in conjunction with the accelerometer sensor 2011, can collect the user's 3D movements on the terminal 2000. Based on the data collected by the gyroscope sensor 2012, the processor 2001 can perform the following functions: motion sensing (e.g., changing the UI based on the user's tilt), image stabilization during shooting, game control, and inertial navigation.

[0216] The pressure sensor 2013 can be installed on the side bezel of the terminal 2000 and / or on the lower layer of the display screen 2005. When the pressure sensor 2013 is installed on the side bezel of the terminal 2000, it can detect the user's grip signal on the terminal 2000, and the processor 2001 can perform left / right hand recognition or quick operation based on the grip signal collected by the pressure sensor 2013. When the pressure sensor 2013 is installed on the lower layer of the display screen 2005, the processor 2001 can control the operable controls on the UI interface based on the user's pressure operation on the display screen 2005. The operable controls include at least one of button controls, scroll bar controls, icon controls, and menu controls.

[0217] The fingerprint sensor 2014 is used to collect the user's fingerprint. The processor 2001 identifies the user's identity based on the fingerprint collected by the fingerprint sensor 2014, or the fingerprint sensor 2014 identifies the user's identity based on the collected fingerprint. When the user's identity is identified as trusted, the processor 2001 authorizes the user to perform relevant sensitive operations, including unlocking the screen, viewing encrypted information, downloading software, making payments, and changing settings. The fingerprint sensor 2014 can be located on the front, back, or side of the terminal 2000. When the terminal 2000 has physical buttons or a manufacturer's logo, the fingerprint sensor 2014 can be integrated with the physical buttons or manufacturer's logo.

[0218] An optical sensor 2015 is used to collect ambient light intensity. In one embodiment, a processor 2001 can control the display brightness of a display screen 2005 based on the ambient light intensity collected by the optical sensor 2015. Specifically, when the ambient light intensity is high, the display brightness of the display screen 2005 is increased; when the ambient light intensity is low, the display brightness of the display screen 2005 is decreased. In another embodiment, the processor 2001 can also dynamically adjust the shooting parameters of a camera assembly 2006 based on the ambient light intensity collected by the optical sensor 2015.

[0219] The proximity sensor 2016, also known as a distance sensor, is typically installed on the front panel of the terminal 2000. The proximity sensor 2016 is used to detect the distance between the user and the front of the terminal 2000. In one embodiment, when the proximity sensor 2016 detects that the distance between the user and the front of the terminal 2000 is gradually decreasing, the processor 2001 controls the display screen 2005 to switch from a screen-on state to a screen-off state; when the proximity sensor 2016 detects that the distance between the user and the front of the terminal 2000 is gradually increasing, the processor 2001 controls the display screen 2005 to switch from a screen-off state to a screen-on state.

[0220] Those skilled in the art will understand that Figure 20 The structure shown does not constitute a limitation on the terminal 2000, and may include more or fewer components than shown, or combine certain components, or use different component arrangements.

[0221] The electronic device in the above method embodiments can be implemented as a server. For example, Figure 21 This is a schematic diagram of a server structure provided in an embodiment of the present invention. The server 2100 can vary considerably depending on its configuration or performance. It can include one or more Central Processing Units (CPUs) 2101 and one or more memories 2102. The memories 2102 store at least one computer program. This at least one computer program is loaded by the processor 2101 and executed to implement the scenario generation method based on the large language model provided in the various method embodiments. Of course, the server can also have wired or wireless network interfaces and input / output interfaces for input and output. The server can also include other components for implementing device functions, which will not be elaborated here.

[0222] In an exemplary embodiment, a computer-readable storage medium is also provided, such as a memory including at least one computer program, which is executable by a processor to perform the scenario generation method based on a large language model in the above embodiments or to implement the scenario generation system based on a large language model. For example, the computer-readable storage medium can be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, and optical data storage device, etc.

[0223] In an exemplary embodiment, a computer program product or computer program is also provided, the computer program product or computer program comprising one or more lines of program code stored in a computer-readable storage medium. One or more processors of an electronic device are capable of reading the one or more lines of program code from the computer-readable storage medium, and the one or more processors execute the one or more lines of program code, enabling the electronic device to perform the functions of the above-described scenario generation method or the above-described scenario generation system based on a large language model.

[0224] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.

Claims

1. A scenario generation system based on a large language model, characterized in that, The system includes a scenario generation subsystem, a scenario setting subsystem, a scenario simulation subsystem, and a scenario management subsystem. The multiple subsystems interact with each other through network data interfaces. The scenario generation subsystem includes scenario retrieval and enhanced generation functions and / or combat scenario text conversion and generation functions. It is used to analyze user needs obtained through human-computer interaction based on a large language model and generate simulation scenarios that meet the user needs. The scenario retrieval and enhanced generation function is used to achieve accurate matching between user needs and simulation scenario data by relying on the scenario database through the retrieval and enhanced generation system. The combat scenario text conversion and generation function includes combat scenario text parsing, scenario key information extraction, scenario knowledge graph construction, and knowledge graph-scenario template mapping sub-functions. The scenario setting subsystem has military standard plotting and attribute parameter setting functions, which are used to support users in editing and modifying the simulation scenarios generated by the scenario generation subsystem. The scenario simulation subsystem is used to support users in performing simulation simulations and verifications of the generated scenarios. The scenario management subsystem has functions for creating, opening, deleting, and saving simulation scenarios, which are used to support users in managing simulation scenarios.

2. The scenario generation system based on a large language model as described in claim 1, characterized in that, The scenario enhancement retrieval generation function uses a semantic matching algorithm to perform multi-level feature retrieval on the scenario database based on user intent. It employs dynamic prompting construction technology to perform knowledge distillation and demand alignment on the retrieval results to obtain scenario knowledge data. Through structured prompt templates, it deeply integrates the scenario knowledge data with the user intent to obtain an enhanced input for a large language model. Based on the large language model and the enhanced input, it generates a simulation scenario that meets the user's needs. The user intent is obtained by parsing the user's needs based on the large language model.

3. The scenario generation system based on a large language model as described in claim 1, characterized in that, The user requirement refers to the operational scenario text input by the user. The operational scenario text conversion and generation function is used to parse the operational scenario text using a large language model, extract the core information of the operational scenario, and construct a knowledge graph of battlefield environment ontology, operational mission ontology, weapon and equipment ontology, and operational mission ontology based on the core operational information. Through a retrieval-enhanced generation system, intelligent association mapping between the scenario knowledge graph and the scenario template is realized to generate a simulation scenario file that meets the user requirement. The core information of the operational scenario includes, but is not limited to, battlefield environment, combat forces, weapon and equipment, operational mission, and operational rules.

4. The scenario generation system based on a large language model as described in claim 1, characterized in that, The scenario setting subsystem is used to respond to the user's selection operation on any simulation scenario generated by the scenario generation subsystem. It converts the simulation scenario into a DOM tree structure with multi-layer semantic annotation through a large language model, sets relevant parameters according to the user's setting operation, and generates an edited simulation scenario based on the set relevant parameters. The large language model connects to the geographic information system, the situation display system, and the scenario simulation tool through an API interface. The scenario simulation subsystem is used to call the geographic information system for map loading and editing, call the situation display system for simulation model visualization deployment, call the timer of the scenario simulation tool to set the simulation time step, and simulate the scenario after completion.

5. The scenario generation system based on a large language model as described in claim 1, characterized in that, The scenario management subsystem is used to break down the simulated scenario into multiple standardized reusable modules, and store these modules in a scenario database. These standardized reusable modules include, but are not limited to, modules for battlefield environment, combat forces, weaponry, and combat missions. Correspondingly, the scenario generation subsystem responds by matching any standardized reusable module according to user requirements and generating a simulation scenario that meets the user's needs based on that standardized reusable module.

6. The scenario generation system based on a large language model as described in claim 1, characterized in that, The scenario generation system based on a large language model is used to receive user operation data and combat scenario data input by the user through a human-computer interaction interface. The user operation data and combat scenario data include, but are not limited to, relevant parameters of user requirements or settings. The scenario setting subsystem is used to receive simulation scenario data input by the scenario generation subsystem and the scenario management subsystem, as well as scenario verification data and simulation scenario data input by the scenario extrapolation subsystem. The scenario simulation subsystem is used to receive simulation scenario data and simulation force models input by the scenario generation subsystem and the scenario setting subsystem; the scenario management subsystem is used to receive scenario verification data and simulation scenario data input by the scenario simulation subsystem.

7. The scenario generation system based on a large language model as described in claim 1, characterized in that, The scenario generation system based on a large language model includes a basic layer, a technical layer, a functional layer, a user interface layer, and an application layer. The foundation layer consists of a large language model, application tools, related databases, and resource libraries. The technology layer is used to provide related technologies for generating, editing, and deducing simulation scenarios based on large language models; The functional layer includes scenario generation, setup, simulation and management functions that are assembled on demand based on the basic operating environment, the foundation layer and the technical layer. The interface layer provides a human-computer interaction interface, which supports users in performing scenario generation, setting, deduction, and management functions. The application layer is used to achieve interconnection and interoperability between the scenario generation system based on the large language model and the equipment combat simulation system through the network interface, and to transmit the generated scenario to the equipment combat simulation system.

8. A scenario generation method based on a large language model, characterized in that, include: Collect user needs based on the human-computer interaction interface; Based on a large language model, the user needs are analyzed and simulation scenarios that meet the user needs are generated by using scenario retrieval enhancement generation function or combat scenario text conversion generation. In response to the user's selection of any simulation scenario generated by the scenario generation subsystem, the generated simulation scenario is edited and modified based on the user's settings. The edited simulation scenario is then extrapolated to obtain a simulation scenario that meets the user's requirements.

9. An electronic device, characterized in that, The electronic device includes one or more processors and one or more memories, wherein the one or more memories store at least one computer program, and the at least one computer program is loaded by the one or more processors. The scenario generation system based on a large language model as described in any one of claims 1 to 7, or the scenario generation method based on a large language model as described in claim 8.

10. A computer-readable storage medium, characterized in that, The storage medium stores at least one computer program, which is loaded by a processor into the scenario generation method system based on a large language model as described in any one of claims 1 to 7, or executes the scenario generation method based on a large language model as described in claim 8.

Images