LLM-based intelligent generation method and device for equipment maintenance and operation support models

Abstract

This application relates to the interdisciplinary field of artificial intelligence and reliability systems engineering, and provides a method and apparatus for intelligent generation of equipment maintenance and operation support models based on LLM (Limited Linear Modeling). This method combines a vector database to achieve enhanced retrieval and constructs an agent-based framework for generating equipment maintenance and operation support models. Then, for each business component involved in the equipment maintenance and operation support model, structural elements in the agent are generated one by one. Based on the model structure, the results are integrated and merged, ultimately achieving a complete and operational equipment maintenance and operation support model. This method integrates the interactive modeling capabilities of agents with the semantic understanding and logical reasoning capabilities of large language models, realizing the intelligent generation of maintenance and operation support models for complex equipment, and effectively improving the generation efficiency of equipment maintenance and support models.

Description

Technical Field

[0001] This application relates to the interdisciplinary field of artificial intelligence and reliability systems engineering, and in particular to a method and apparatus for intelligent generation of equipment maintenance and operation support models based on LLM. Background Technology

[0002] Equipment maintenance and operation support models, as the core supporting the entire lifecycle management of equipment and the evaluation of support effectiveness, have significant application value in complex systems engineering fields such as military equipment and high-end manufacturing. Traditionally, the construction of equipment maintenance and operation support models mainly relies on domain experts to complete the process based on experience, historical data, and domain specifications, through manual coding, flowchart drawing, or gradually building a logical framework in Model-Based Systems Engineering (MBSE) tools. This approach suffers from problems such as long development cycles, low modeling efficiency, and difficulty in ensuring model consistency. Especially when facing multi-scenario support tasks or dynamically changing support resource constraints, it is difficult to respond promptly and iterate quickly, limiting the timeliness and adaptability of the model in practical applications.

[0003] With the development of artificial intelligence technology, large language models (LLMs) have shown great potential in natural language understanding, knowledge extraction, logical reasoning, and code generation, providing new ideas for the intelligent generation of models. However, when directly using large language models to generate equipment maintenance and support models, the lack of a deep, structured understanding of professional domain knowledge such as equipment maintenance processes, support resource scheduling, and task collaboration mechanisms can easily lead to "illusion" phenomena, generating model outputs that do not conform to physical reality and contain logical contradictions. In addition, the generated results are mostly discrete code fragments or text descriptions, lacking organic connections between modules, making it difficult to automatically integrate them into a simulation model with complete interactive logic and executable functionality. The integrity of the model and the consistency of business rules lack a systematic verification mechanism, making it difficult to directly apply to high-reliability equipment support analysis.

[0004] Therefore, there is an urgent need for an intelligent model construction method that can deeply integrate knowledge of equipment domains, natural semantic understanding, and logical reasoning capabilities. Summary of the Invention

[0005] In view of this, embodiments of this application provide an intelligent generation method and apparatus for equipment maintenance and operation support models based on LLM, in order to solve the problems of low generation efficiency and insufficient matching degree between the model and the business logic of the equipment in the prior art.

[0006] A first aspect of this application provides a method for intelligently generating equipment maintenance and operation support models based on LLM, comprising:

[0007] A logical model of the equipment is constructed using the Model-Based Systems Engineering (MBSE) method, and retrieval-enhanced knowledge is injected into the logical model to obtain an enhanced logical model.

[0008] Based on the enhanced logic model and the equipment's task scenario, LLM (Large Language Model) is used to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents. Attribute-layer features and behavioral-layer features are defined for each entity unit in each intelligent agent. Attribute-layer features are used to characterize the static features of the entity unit, while behavioral-layer features are used to characterize the dynamic interaction logic of the entity unit.

[0009] LLM is used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, thus obtaining the equipment composition model. Based on the equipment composition model, LLM is used to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent. Furthermore, LLM is used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, thus obtaining the task scenario model.

[0010] Based on the task scenario model, the attribute structure and behavior structure of each entity unit in the equipment intelligent agent, maintenance personnel intelligent agent and support resource intelligent agent are updated using LLM based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment intelligent agent, maintenance personnel intelligent agent and support resource intelligent agent, so as to obtain the fault maintenance support behavior model.

[0011] Based on the task scenario model, the attribute structure and behavior structure of the equipment-type intelligent agent are updated using the usage support-related behavior layer features of each entity unit in the equipment-type intelligent agent. The attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent are updated using the usage support-related attribute layer features and usage support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, thus obtaining the usage support behavior model.

[0012] By merging the fault repair support behavior model and the usage support behavior model, we obtain the equipment maintenance and usage support model.

[0013] A second aspect of this application provides an intelligent generation device for equipment maintenance and operation support models based on LLM, comprising:

[0014] The logical model building module is configured to build a logical model of the equipment using the model-based systems engineering (MBSE) method, and inject retrieval-enhanced knowledge into the logical model to obtain an enhanced logical model.

[0015] The definition module is configured to use the Large Language Model (LLM) to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents based on the enhanced logic model and equipment task scenarios. It also defines attribute layer features and behavioral layer features for each entity unit in each intelligent agent. Attribute layer features are used to characterize the static features of entity units, and behavioral layer features are used to characterize the dynamic interaction logic of entity units.

[0016] The equipment composition model and task scenario model construction module is configured to use LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, thereby obtaining the equipment composition model; based on the equipment composition model, LLM is used to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent, and LLM is used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, thereby obtaining the task scenario model;

[0017] The fault repair and support behavior model construction module is configured to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the task scenario model, thereby obtaining the fault repair and support behavior model.

[0018] The support behavior model building module is configured to, based on the task scenario model, use LLM to update the attribute structure and behavior structure of the equipment-type intelligent agent based on the use support-related behavior layer features of each entity unit in the equipment-type intelligent agent, and use LLM to update the attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent based on the use support-related attribute layer features and use support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, so as to obtain the use support behavior model;

[0019] The merging module is configured to merge the fault maintenance support behavior model and the usage support behavior model to obtain the equipment maintenance and usage support model.

[0020] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described method.

[0021] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method.

[0022] The beneficial effects of the embodiments in this application compared with the prior art are:

[0023] This application's embodiments combine vector databases to achieve enhanced retrieval and construct an agent-based framework for generating equipment maintenance and operational support models. Then, for each business component involved in the equipment maintenance and operational support model, structural elements within the agent are generated one by one. Based on the model structure, the results are integrated and merged, ultimately achieving a complete and operational equipment maintenance and operational support model. This method integrates the interactive modeling capabilities of agents with the semantic understanding and logical reasoning capabilities of large language models, enabling intelligent generation of maintenance and operational support models for complex equipment and effectively improving the generation efficiency of equipment maintenance and support models. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a flowchart illustrating an intelligent generation method for equipment maintenance and operation support model based on LLM provided in an embodiment of this application.

[0026] Figure 2 This is an example of a Markdown-formatted text illustration.

[0027] Figure 3 This is a flowchart illustrating another method for intelligently generating equipment maintenance and support models based on LLM, provided in an embodiment of this application.

[0028] Figure 4 This is a schematic diagram of an intelligent generation device for equipment maintenance and operation support model based on LLM provided in an embodiment of this application.

[0029] Figure 5 This is a schematic diagram of the electronic device provided in the embodiments of this application. Detailed Implementation

[0030] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0031] The following will describe in detail, with reference to the accompanying drawings, an intelligent generation method and apparatus for equipment maintenance and operation support model based on LLM according to embodiments of this application.

[0032] As mentioned above, traditional equipment maintenance and operation support models rely on manual work by experts, which is inefficient and has a long development cycle. If LLM is directly introduced into the construction of equipment maintenance and operation support models, the resulting models usually differ significantly from the actual equipment failure logic because LLM lacks a deep understanding of equipment domain knowledge.

[0033] In view of this, this application provides an intelligent generation method for equipment maintenance and operation support models based on LLM (Limited Language Modeling). It combines a vector database to achieve enhanced retrieval and constructs an agent-based framework for generating equipment maintenance and operation support models. Then, for each business component involved in the equipment maintenance and operation support model, structural elements within the agent are generated one by one. Based on the model structure, the results are integrated and merged, ultimately achieving a complete and operational equipment maintenance and operation support model. This method integrates the interactive modeling capabilities of agents with the semantic understanding and logical reasoning capabilities of large language models, enabling intelligent generation of maintenance and operation support models for complex equipment and effectively improving the generation efficiency of equipment maintenance and support models.

[0034] Figure 1 This is a flowchart illustrating an intelligent generation method for equipment maintenance and operation support models based on LLM, provided in an embodiment of this application. Figure 1 As shown, the method includes the following steps:

[0035] In step S101, the logical model of the equipment is constructed using the model-based systems engineering (MBSE) method, and retrieval enhancement knowledge is injected into the logical model to obtain the enhanced logical model.

[0036] In step S102, LLM is used to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents based on the enhanced logical model and equipment task scenario, and attribute layer features and behavior layer features are defined for each entity unit in each intelligent agent.

[0037] Among them, attribute layer features are used to characterize the static features of entity units, and behavior layer features are used to characterize the dynamic interaction logic of entity units.

[0038] In step S103, the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent are updated using LLM based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent to obtain the equipment composition model. Based on the equipment composition model, the attribute structure and behavior structure of each entity unit in the task-type intelligent agent are updated using LLM based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent. Furthermore, the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent are updated using LLM based at least on the behavior layer features of each entity unit in the task-type intelligent agent to obtain the task scenario model.

[0039] In step S104, based on the task scenario model, the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent are updated using LLM based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent, so as to obtain the fault maintenance support behavior model.

[0040] In step S105, based on the task scenario model, the attribute structure and behavior structure of the equipment-type intelligent agent are updated using LLM based on the usage and support-related behavior layer features of each entity unit in the equipment-type intelligent agent. The attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent are also updated using LLM based on the usage and support-related attribute layer features and usage and support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, thus obtaining the usage and support behavior model.

[0041] In step S106, the fault maintenance support behavior model and the usage support behavior model are merged to obtain the equipment maintenance and usage support model.

[0042] In some embodiments of this application, the method may be executed by a server or by a terminal device with certain processing capabilities.

[0043] In some embodiments of this application, the MBSE method can be used to first construct a logical model of the analysis object, and then retrieval enhancement knowledge can be injected into the logical model to obtain an enhanced logical model.

[0044] In some embodiments of this application, LLM can also be used to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents based on the enhanced logical model and equipment task scenario, and attribute layer features and behavioral layer features can be defined for each entity unit in each intelligent agent.

[0045] In some embodiments of this application, LLM can be used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, thereby obtaining an equipment composition model.

[0046] Furthermore, based on the equipment composition model, LLM can be used to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent. LLM can also be used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, thus obtaining the task scenario model.

[0047] Furthermore, based on the task scenario model, the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent are updated using LLM based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent, thus obtaining the fault maintenance support behavior model.

[0048] Based on the task scenario model, the attribute structure and behavior structure of the equipment-type intelligent agent are updated using the usage support-related behavior layer features of each entity unit in the equipment-type intelligent agent. Furthermore, the attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent are updated using the usage support-related attribute layer features and usage support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, thus obtaining the usage support behavior model.

[0049] There are no restrictions on the execution order of the steps for constructing the fault repair support behavior model and the steps for constructing the use support behavior model.

[0050] Finally, the fault repair support behavior model and the usage support behavior model can be merged to obtain the equipment maintenance and usage support model.

[0051] According to the technical solution provided in this application, enhanced retrieval is achieved by combining a vector database, and a framework for generating equipment maintenance and operation support models based on intelligent agents is constructed. Then, for each business component involved in the equipment maintenance and operation support model, structural elements in the intelligent agent are generated one by one, and the results are integrated and merged based on the model structure, ultimately realizing a complete and operable equipment maintenance and operation support model. This method integrates the interactive modeling capabilities of intelligent agents with the semantic understanding and logical reasoning capabilities of large language models, realizing the intelligent generation of maintenance and operation support models for complex equipment, and effectively improving the generation efficiency of equipment maintenance and support models.

[0052] In some embodiments of this application, constructing the logical model of the analysis object using the MBSE method can be done by assuming the overall analysis object system is as follows: The MBSE method is used to divide the analyzed object into multiple logical levels based on its actual hardware architecture, denoted as... , of which each layer It includes a set of components that make up this layer. .in this way, It can represent the first The first in the layer Each component, also known as a physical unit, For the first The total number of components in a layer.

[0053] In some embodiments of this application, injecting retrieval enhancement knowledge into the logical model can be achieved by, based on the equipment's mission scenario, converting non-handwritten format (such as Excel, Word, XML, etc.) documents (e.g., equipment detailed design reports, equipment reliability analysis reports, equipment maintainability analysis reports, etc.) into a structured intermediate text format using an Optical Character Recognition (OCR) engine; subsequently, segmenting the document by physical pages or logical chapters, and extracting semantic tags such as heading levels, paragraph boundaries, and list items based on document layout analysis, ultimately outputting a unified structured intermediate representation format—a Markdown string with semantic tags. This format preserves the original document's logical structure, facilitating subsequent semantic slicing. An exemplary Markdown format text is as follows: Figure 2 As shown.

[0054] To improve the semantic integrity and contextual coherence of knowledge fragments, a sliding window-driven dynamic aggregation slicing mechanism can be used to convert a preset format string into a semantically coherent set of knowledge slices. The specific operation is as follows:

[0055] The Markdown text above is initially segmented according to natural sentence boundaries to obtain a sentence sequence. ; Utilizing pre-trained embedding models (For example, Yinka) will process each statement Mapped to dense vectors Initialize the content of the current slice. ,vector ; for subsequent statements (in Calculate its cosine similarity to the current slice vector:

[0056] ;

[0057] like ,in If the preset similarity threshold is used, then... Merge into the current slice: Otherwise, the current slice Stored as a complete knowledge unit in the slice set, and... Start a new slice from the beginning. This strategy generates a set of semantically cohesive, length-adaptive knowledge slices. and its corresponding set of embedding vectors .

[0058] The obtained knowledge slice set and its embedding vectors can be imported into an efficient vector database (such as FAISS) to build a semantic index that can be retrieved quickly. Subsequently, for each component in the logical model... (in The following knowledge injection process will be executed:

[0059] (1) Construct structured query suggestions based on component identifiers, functional descriptions and interface features. ;

[0060] (2) Using the same embedding model right Encode the query vector;

[0061] (3) Perform an approximate nearest neighbor search in the vector database to recall and match the vectors. Most similar front A slice of knowledge;

[0062] (4) Introduce a cross-coding re-ranking model (such as bce-reranker) to perform refined relevance scoring on the recall results, and select the top scorers. Each slice serves as final contextual evidence. ;

[0063] (5) will be with Original query prompts The words are concatenated to form enhanced prompts, which are used to guide the large language model in subsequent analysis.

[0064] (6) Repeat the above process until knowledge injection of all components is completed.

[0065] Taking the first paragraph of the equipment maintenance and support simulation document as an example, the original text reads: "The simulation scenario simulates the mission execution and fault maintenance of 10 aircraft over a year. Each aircraft has 4 subsystems, and each subsystem consists of 5 Line Replaceable Units (LRUs). Each LRU experiences a random failure during operation according to an exponential probability distribution. The aircraft performs training missions daily, and when a failure occurs, the current mission is interrupted, and a maintenance mission is then performed. There are 10 maintenance personnel on site who can perform fault maintenance on the aircraft. Each fault maintenance requires the replacement of spare parts, with an initial inventory of 20. The working time, sorties, number of failures, and various maintenance information of each aircraft are recorded to evaluate the overall equipment availability and maintenance support efficiency."

[0066] First, the text is divided into 6 sentences based on natural sentence boundaries (using periods as delimiters). Then, a pre-trained embedding model, Yinka, maps each sentence to a 768-dimensional dense vector. A cosine similarity threshold τ = 0.65 is set, and a sliding window-driven dynamic aggregation slicing mechanism is used: the current slice is initialized as sentence 1, and the similarity between subsequent sentences and the average vector of the current slice is calculated sequentially. Sentences 2 (0.74 similarity), 3 (0.71), and 4 (0.68) are all above the threshold and are merged into the current slice; sentence 5 has a similarity of 0.63, below the threshold, triggering the end of the slice and the start of a new slice; sentence 6 has a similarity of 0.66 with sentence 5, so they are merged. Finally, two semantically cohesive, length-adaptive knowledge slices are generated.

[0067] Slice 1: "The simulation scenario simulates the mission execution and fault maintenance of 10 aircraft over a year. Each aircraft has 4 subsystems, and each subsystem consists of 5 LRUs. During operation, each LRU will fail randomly according to an exponential probability distribution. The aircraft performs daily tasks, and when a fault occurs, the current task is interrupted, and then a maintenance task is performed."

[0068] Segment 2: "There are 10 maintenance personnel on site who can perform fault repairs on aircraft. Each fault repair requires the replacement of spare parts, with an initial inventory of 20. Record the working hours, sorties, number of faults, and various maintenance information for each aircraft to evaluate the overall equipment availability and maintenance support efficiency."

[0069] Each slice is assigned its corresponding embedding vector, forming a set of indexable knowledge units.

[0070] Taking the "Spare Parts" component in the logical model as an example, a structured query prompt can be constructed based on its identifier "SPARE-INV-01", functional description "supports LRU fault replacement", and interface features "inventory quantity and consumption rate": "configuration strategies, consumption mechanisms, and impact on task continuity related to spare parts inventory". This prompt, after being encoded by the Yinka embedding model, is used to perform an approximate nearest neighbor search in the FAISS vector database to recall the Top-5 knowledge slices. Then, a cross-coding re-ranking model (such as bce-reranker) is used for refined scoring, selecting the Top-2 slices with the highest scores. The contents of these slices are: "Spare parts need to be replaced for each fault repair, with an initial inventory of 20." and "Record the working hours, sorties, number of faults, and various maintenance information for each aircraft to evaluate overall equipment availability and support effectiveness."

[0071] The above-mentioned slices are concatenated with the original query suggestions to form an enhanced suggestion word input large language model. Based on this, the model generates the following analysis conclusion: "According to the simulation settings, each LRU failure requires the consumption of one spare part, with an initial inventory of 20 parts. In a high-concurrency failure scenario with 10 aircraft, each containing 20 LRUs (4×5), if the total number of failures per year exceeds 20, spare parts will be depleted, leading to maintenance delays and mission interruptions. Therefore, it is recommended to establish a dynamic replenishment mechanism based on failure rate prediction, or to set a minimum safety stock threshold (e.g., ≥10) to maintain the sustainable operation of the support system. The current simulation data can be used to calibrate the effectiveness of the inventory strategy."

[0072] In some embodiments of this application, the functional requirements and business logic of the simulation scenario of the equipment can be analyzed based on the mission scenario of the equipment, and the equipment-type intelligent agent can be developed. Task-based intelligent agents Maintenance personnel-type intelligent agents and resource-protecting intelligent agents This enables the description of various objects involved in the equipment maintenance and operation support model; among them, equipment-type intelligent agents... Used to characterize equipment entities and their constituent entity units; task-oriented intelligent agents Used to characterize the initiation, execution, and management of various tasks (combat, maintenance, etc.); maintenance personnel-type intelligent agents Used to characterize maintenance personnel, etc.; resource-saving intelligent agents. Used to characterize spare parts, fuel, etc. Among them, Indicates the first intelligent agents, Represents the first in the equipment class of intelligent agents Each entity unit.

[0073] Taking an airport equipment maintenance simulation as an example, this simulation scenario simulates the mission execution and fault maintenance of 10 aircraft over a year. Each aircraft has 4 subsystems, and each subsystem consists of 5 LRUs. Each LRU experiences a fault randomly during its operation according to an exponential probability distribution. The aircraft performs two routine tasks, interrupting the current task when a fault occurs, and then performing a maintenance task. There are 10 maintenance personnel on-site who can perform fault repairs on the aircraft. The working time, sorties, number of faults, and various maintenance information for each aircraft are recorded to evaluate the overall equipment availability and maintenance support efficiency.

[0074] The four types of intelligent agents constructed based on the above steps can be:

[0075] Equipment-type intelligent agents, Includes equipment intelligent agents j={1,2,3…,250};

[0076] Task-oriented intelligent agents Includes task-oriented intelligent agents j={1,2,3};

[0077] Maintenance personnel-type intelligent agents, Including maintenance personnel intelligent agents j={1,2,3...,10};

[0078] Protecting resource-based intelligent agents, Includes resource protection intelligent agents , j={1,2,3...,20}.

[0079] Various intelligent agents can be analyzed one by one, targeting the first... In intelligent agents Entity unit Design and define a two-layer structure as follows:

[0080] Attribute layer: Describes the static characteristics of the agent (identifier, name, parameters, variables, functions, descriptions, etc.), defined as: .

[0081] Behavioral layer: describes the dynamic interaction logic of the agent (state transition rules, event response mechanisms, relationships with other agents, etc.), defined as: Each entity unit intelligent agent has a unique identifier. ; It is a set that can be designed according to the specific needs of the equipment to perform the mission, and can include multiple specific parameters (such as failure rate, maintenance rate, etc.). This is a function of the intelligent agent, which can be implemented by writing relevant code; This is a supplementary explanation of intelligent agents; ; The relationships between agents can be represented as follows: ;in, This indicates that other intelligent agents have a relationship with this type of intelligent agent. Indicates the specific impact of the relationship (such as failure, performance degradation, etc.).

[0082] Among them, other entity units are other entity units in this intelligent agent, or other entity units in other intelligent agents.

[0083] In some embodiments of this application, the intelligent agent constructed through the above steps can be stored in the structured data format XML.

[0084] Taking the equipment maintenance and support simulation scenario as an example, the equipment-type intelligent agents in the logical model include equipment such as aircraft, subsystems, and LRUs. Taking the entity unit "Aircraft Intelligent Agent Plane" as an example, its ID is set to 1001, its model is "J01", its intelligent agent description is "the main simulation object, composed of subsystems", and its logical states are defined as "standby state, working state", which is saved in XML format.

[0085] Furthermore, based on the equipment's mission scenarios and requirements, and combined with natural language processing, a verifiable equipment maintenance and operation support checklist can be generated. This checklist can then be used to verify the completeness, consistency, and correctness of the simulation model implementation item by item. Each checklist item can be used... express.

[0086] Taking aircraft malfunction logic as an example, it can be further broken down into multiple checks: checking whether the aircraft has fault information reception and processing logic; checking whether the aircraft has fault state transition logic; and checking whether the aircraft has mission fault interruption logic. This leads to the generation of an equipment maintenance and operation support checklist.

[0087] In some embodiments of this application, the business logic of the equipment maintenance and operation support model can be used to analyze and generate the model in four parts: equipment composition model modeling, task scenario model modeling, fault maintenance support behavior model modeling, and operation support behavior model modeling, as detailed below:

[0088] First, an equipment composition model is constructed. This involves identifying target entity units within the target equipment class of intelligent agents; the target equipment class of intelligent agents can be any type of intelligent agent, and the target entity unit can be any entity unit within that class. Static prompts for the equipment composition model are constructed based on the attribute-level features of the target entity units, guiding the LLM to generate static attributes of the target entity units. Dynamic prompts for the equipment composition model are then constructed based on the behavioral-level features of the target entity units, guiding the LLM to generate dynamic interaction logic for the target entity units. Finally, the static attributes and dynamic interaction logic of each entity unit are embedded into its attribute and behavioral structures to obtain the equipment composition model.

[0089] In other words, targeting equipment-type intelligent agents Each entity unit in Based on its attribute layer characteristics, it can construct guiding prompts using identifiers, names (such as equipment component names), and descriptions as clues. LLM then generates static attribute codes for the intelligent agent that conform to the equipment background and mission requirements, thereby realizing the construction of basic information for the entity unit.

[0090] And, for equipment-type intelligent agents Each entity unit in Based on its behavioral characteristics, and using the relationships with other intelligent agents (such as the relationships between equipment components and the subordinate relationships between units) as clues, guiding prompts are constructed. Based on this, LLM generates dynamic interaction logic code for intelligent agents that conforms to the equipment background and mission requirements, thereby realizing the construction of relationships between entity units.

[0091] Finally, the static attributes and dynamic interaction logic code of the agent generated by LLM are embedded into the attribute structure and behavior structure of each entity unit of the equipment-type agent.

[0092] Taking an equipment maintenance and support simulation scenario as an example, the equipment-type intelligent agents in the logical model include aircraft, subsystems, LRUs, and other equipment. Taking the entity unit "Aircraft Intelligent Agent Plane" as an example, a description is constructed: "Please analyze the simulation scenario, consider the hierarchy and compositional relationships between equipment, and based on the static attributes of the Aircraft Intelligent Agent Plane, construct XML code to implement its static attributes." This description is then concatenated with the static attributes of the entity unit to form a prompt, which is then provided by the larger model. Simultaneously, based on the behavioral logic description of the intelligent agent Plane, prompts are constructed to generate the implementation code for the behavioral logic. Finally, the two parts of the implementation code are extracted into one... <activeobjectclass>Under the tag, merge them into a single XML fragment, which is the implementation code of the intelligent agent Plane.

[0093] Then, a task scenario model can be constructed. The target task unit in the target task class intelligent agent can be determined; the target task class intelligent agent can be any task class intelligent agent, and the target task unit can be any entity unit in the target task class intelligent agent.

[0094] Then, based on the attribute layer features of the target task unit in the equipment composition model, static prompt words for the task scenario model are constructed to guide the LLM to generate static attributes of the target task unit based on the static prompt words of the task scenario model; and based on the behavioral layer features of the target task unit in the equipment composition model, dynamic prompt words for the task scenario model are constructed to guide the LLM to generate dynamic interaction logic of the target task unit based on the dynamic prompt words of the task scenario model.

[0095] At the same time, it is also possible to identify the target entity unit in the target equipment class intelligent agent; the target equipment class intelligent agent is any equipment class intelligent agent, and the target entity unit is any entity unit in the target equipment class intelligent agent.

[0096] Then, based on the behavioral features of the target entity unit in the equipment composition model, supplementary prompts for the first task scenario model are constructed, guiding the LLM to generate supplementary dynamic interaction logic between the target equipment class agent and its corresponding task class agent based on the supplementary prompts for the first task scenario model.

[0097] It can also identify the target personnel entity in the target maintenance personnel class intelligent agent; the target maintenance personnel class intelligent agent is any maintenance personnel class intelligent agent, and the target personnel entity is any entity unit in the target maintenance personnel class intelligent agent.

[0098] Then, based on the behavioral features of the target personnel entity in the equipment composition model, supplementary prompts for the second task scenario model are constructed. This guides the LLM to generate supplementary dynamic interaction logic between the target maintenance personnel agent and its corresponding task agent based on the supplementary prompts for the second task scenario model.

[0099] It can also identify the target resource entity in the target resource protection class intelligent agent; the target resource protection class intelligent agent is any protection resource class intelligent agent, and the target resource entity is any entity unit in the target resource protection class intelligent agent.

[0100] Then, based on the behavioral features of the target resource entity in the equipment composition model, supplementary prompts for the third task scenario model are constructed. This guides the LLM to generate supplementary dynamic interaction logic between the target support resource-type intelligent agent and its corresponding task-type intelligent agent based on the supplementary prompts for the third task scenario model.

[0101] Finally, the static attributes and dynamic interaction logic of each task-type intelligent agent are embedded into the attribute structure and behavior structure of this task unit. The supplementary dynamic interaction logic of each equipment-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this entity unit. The supplementary dynamic interaction logic of each maintenance personnel-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this target personnel entity. The supplementary dynamic interaction logic of each maintenance support resource-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this target resource entity, thus obtaining the task scenario model.

[0102] In other words, for task-oriented intelligent agents Each task unit in Based on its attribute layer characteristics, LLM can construct guiding prompts using identifiers, names (such as training, combat, maintenance, etc.), parameters (such as mission type, mission start or end time, mission duration, etc.), variables (such as constraints, etc.), functions, and descriptions as clues. Based on this, LLM generates static attribute codes of intelligent agents that meet the mission requirements of the equipment, thereby realizing the construction of basic information for various mission units.

[0103] Additionally, based on its behavioral characteristics, state transition rules (such as task start rules, task end rules, etc.) and its relationships with other intelligent agents (such as task and equipment intelligent agents), it can be further analyzed. Maintenance personnel-type intelligent agents and resource-protecting intelligent agents Using the relationships between different tasks as clues, guiding prompts are constructed, and the LLM generates dynamic interaction logic code for intelligent agents that meets the needs of equipment tasks, thereby realizing the construction of relationships between various task units.

[0104] Then, the static attributes and dynamic interaction logic code of the agent generated by LLM are embedded into the attribute structure and behavior structure of the task-type agent, respectively.

[0105] Taking an equipment maintenance and support simulation scenario as an example, the task-type intelligent agents in the logical model include training tasks and maintenance tasks. Taking the training task intelligent agent Mission as an example, first, its ID is set to 2001, its type is "training", the required number of aircraft is 5, the task start time is "10:00", and the end time is "12:00". The agent description is "Task intelligent agent, assigns aircraft to perform tasks". Its behavioral logic "At the task start time, assigns aircraft to perform tasks" is defined as an event and saved in XML format. The description "Please analyze the task scenario, based on the static attributes and behavioral logic of the intelligent agent Mission, and considering its interaction logic with the aircraft intelligent agent, construct and implement the implementation code of Mission" is constructed. After concatenating the description with the attributes and behavioral characteristics of Mission into a prompt, the large model first provides the implementation code of Mission's static attributes; at the same time, the large model provides the implementation code of the behavioral logic. Finally, the two parts of the implementation code are extracted into one. <activeobjectclass>Under the tag, merge them into a single XML fragment, which is the implementation code of the intelligent agent Mission.

[0106] For equipment-type intelligent agents Each entity unit in Based on its behavioral characteristics, guided prompts can be constructed using state transition rules (such as standby to task execution, standby to maintenance, etc.), event response mechanisms (such as executing tasks after receiving task instructions, etc.), and relationships with other intelligent agents. LLM can then generate equipment-type intelligent agents based on these prompts. Task-oriented intelligent agents The dynamic interaction behavior logic supplementary code was then added. This supplementary code was then merged into the entity unit. In the structure.

[0107] Taking the equipment maintenance and support simulation scenario, the entity unit "Aircraft Intelligent Agent Plane" is used as an example. The prompt "Please analyze the task scenario and, based on the existing static attributes and behavioral logic of the Aircraft Intelligent Agent Plane, construct supplementary interaction logic between it and the task intelligent agent" is provided by the large model. Then, the implementation code of the supplementary logic is merged into the Aircraft Intelligent Agent Plane structure.

[0108] Intelligent agents targeting maintenance personnel Each resource entity in Based on its behavioral characteristics, guided prompts can be constructed using state transition rules (such as maintenance personnel going from idle to working), event response mechanisms (such as executing maintenance tasks after receiving task instructions), and relationships with other intelligent agents. The LLM then generates maintenance personnel-type intelligent agents based on these prompts. Task-oriented intelligent agents The dynamic interaction behavior logic supplementary code was then added. This supplementary code was then merged into the entity unit. In the structure.

[0109] Taking an equipment maintenance and support simulation scenario as an example, the logical model contains a maintenance personnel agent called `MaintenacePersonnel`. Using `MaintenacePersonnel` as an example, the model describes "analyzing the task scenario and constructing supplementary logic code related to task execution based on the existing static attributes and behavioral logic of `MaintenacePersonnel`". Combining the attribute and behavioral definitions of `MaintenacePersonnel`, the larger model provides the implementation code for the supplementary logic. This implementation code is then merged into the `MaintenacePersonnel` structure.

[0110] For resource-protecting intelligent agents Each resource entity in Based on its behavioral characteristics, guided prompts can be constructed using state transition rules (such as ensuring resources are not used), event response mechanisms (such as ensuring resource allocation / replenishment), and relationships with other agents. Based on these prompts, LLM can generate resource-secure agents. Task-oriented intelligent agents The dynamic interaction logic logic is supplemented with additional code. Then, this supplementary code is merged into the resource entity. In the structure.

[0111] Taking an equipment maintenance and support simulation scenario as an example, the logical model includes a spare parts agent (SparePart) within the support resource class. Using the SparePart agent as an example, the model describes "analyzing the task scenario and constructing supplementary logic code related to task execution based on the existing static attributes and behavioral logic of the SparePart agent." Combining the attributes and behavioral definitions of the SparePart, the larger model provides the implementation code for the supplementary logic. This implementation code is then merged into the SparePart agent structure.

[0112] In some embodiments of this application, the fault repair and support behavior model can be constructed in the following manner: determining the target entity unit in the target equipment class intelligent agent; the target equipment class intelligent agent is any equipment class intelligent agent, and the target entity unit is any entity unit in the target equipment class intelligent agent.

[0113] Based on the fault-related attribute layer features of the target entity unit, fault-related static supplementary prompt words are constructed to guide LLM to generate fault-related supplementary static attributes of the target entity unit based on the fault-related static supplementary prompt words.

[0114] Based on the fault-related behavioral layer features of the target entity unit, fault-related dynamic supplementary prompts are constructed to guide the LLM to generate fault-related supplementary dynamic interaction logic of the target entity unit.

[0115] Identify the target personnel entity within the target maintenance personnel class of intelligent agents; the target maintenance personnel class of intelligent agents can be any maintenance personnel class of intelligent agents, and the target personnel entity can be any entity unit within the target maintenance personnel class of intelligent agents.

[0116] Using the attribute layer features of the target person entity as clues, static prompt words for the person are constructed, which guides the LLM to generate static attributes of the target person entity based on the static prompt words.

[0117] Based on the behavioral characteristics of the target personnel entity, dynamic prompt words are constructed to guide the LLM to generate dynamic interaction logic for the target personnel entity.

[0118] Identify the target resource entity within the target resource protection class of intelligent agents; the target resource protection class of intelligent agents can be any type of intelligent agent of resource protection, and the target resource entity can be any entity unit within the target resource protection class of intelligent agents.

[0119] Based on the fault-related attribute layer features of the target resource entity, fault-related static prompt words are constructed to guide LLM to generate fault-related static attributes of the target resource entity based on the fault-related static prompt words.

[0120] Based on the fault-related behavioral features of the target resource entity, fault-related dynamic prompt words are constructed to guide the LLM to generate fault-related dynamic interaction logic of the target resource entity.

[0121] By embedding the fault-related supplementary static attributes and fault-related supplementary dynamic interaction logic of entity units in each equipment-type intelligent agent into the attribute structure and behavior structure of that entity unit, the static attributes and dynamic interaction logic of entity units in each maintenance personnel-type intelligent agent into the attribute structure and behavior structure of that entity unit, and the fault-related static attributes and fault-related dynamic interaction logic of entity units in each maintenance support resource-type intelligent agent into the attribute structure and behavior structure of that entity unit, a fault maintenance support behavior model is obtained.

[0122] In other words, when constructing a fault repair and support behavior model, it is necessary to consider equipment-type intelligent agents. Each entity unit in Based on its attribute layer characteristics, and using fault-related parameters (such as Mean Time Between Failures (MTBF) and failure rate), variables (such as remaining time after failure), and functions (such as failure probability distribution function) as clues, guiding prompts can be constructed. The LLM then generates static attribute code for the agent that conforms to the equipment's fault behavior logic, thus constructing fault information for the entity unit. Furthermore, based on its behavioral layer characteristics, and using fault-related state transition rules (such as transition from normal state to fault state), event response mechanisms (from task execution to task termination), and relationships with other agents as clues, guiding prompts can be constructed. The LLM then generates dynamic interaction logic code for the agent that conforms to the equipment's fault generation logic, thus constructing the entity unit's dynamic fault behavior. Finally, the static fault information code and dynamic fault behavior code of the equipment-type agent are supplemented and merged into the entity unit. In the structure.

[0123] Taking the entity unit "Aircraft Agent Plane" within the equipment-type intelligent agent in an equipment maintenance and support simulation scenario as an example, a prompt is constructed: "Please analyze the fault and maintenance support behavior logic. Based on the existing static attributes and behavior logic of the Aircraft Agent Plane, supplement the static attribute code that conforms to the fault and maintenance support logic." The larger model then provides the implementation code. Simultaneously, behavior logic implementation code that conforms to the fault and maintenance support logic interaction is generated. Finally, the two parts of implementation code are merged into the structure of the agent Plane.

[0124] Meanwhile, for intelligent agents such as maintenance personnel Each resource entity in Based on its attribute-level characteristics, using identifiers, names (e.g., maintenance personnel specialties), parameters (e.g., number of maintenance personnel), variables (e.g., number of available maintenance personnel), functions, and descriptions as clues, guiding prompts can be constructed. The LLM then generates static attribute code for the intelligent agent that conforms to the equipment fault maintenance behavior logic, thus constructing the basic fault maintenance information of the resource entity. Furthermore, based on its behavioral-level characteristics, using state transition rules (e.g., transition from idle to maintenance state), event response mechanisms (e.g., maintenance personnel being occupied), and relationships with other intelligent agents as clues, guiding prompts can be constructed. The LLM then generates dynamic interaction logic code for the intelligent agent that conforms to the equipment fault maintenance behavior logic, thus constructing the dynamic fault maintenance behavior of the resource entity. Finally, the static fault maintenance information code and the dynamic fault maintenance behavior code of the maintenance personnel intelligent agent are supplemented and merged into the resource entity. In the structure.

[0125] Taking an equipment maintenance and support simulation scenario as an example, the logical model includes maintenance personnel agents. Taking the maintenance personnel agent `MaintenacePersonnel` as an example, its ID is first set to 3001, its specialty is "Mechatronics," and its behavioral logic "If a maintenance task is detected, proceed to maintenance" is defined as a time event. A description is constructed: "Please analyze the fault and maintenance support behavioral logic. Based on the existing static attributes and behavioral logic of the agent `MaintenacePersonnel`, supplement the static attribute code that conforms to the fault and maintenance support logic." The agent's attribute features and behavioral features are concatenated as prompts. The larger model first provides the implementation code for the static attributes; simultaneously, it supplements the implementation code for the generated behavioral logic. Finally, the two parts of the implementation code are supplemented and merged into the structure of the agent `MaintenacePersonnel`.

[0126] For resource-protecting intelligent agents Each resource entity in Based on its attribute-level characteristics, using identifiers, names (such as spare parts names), parameters (such as spare parts quantity), functions, and descriptions as clues, guiding prompts can be constructed. The LLM then generates static attribute codes for the intelligent agent that conform to equipment maintenance and support behaviors, thus constructing basic fault maintenance information for the resource entity. Furthermore, based on its behavioral-level characteristics, using state transition rules (such as transitioning from unused to used) and relationships with other intelligent agents as clues, guiding prompts can be constructed. The LLM then generates dynamic interaction logic code for the intelligent agent that conforms to equipment fault maintenance behavior logic, thus constructing dynamic fault maintenance behaviors for the resource entity. Finally, the static fault maintenance information code and the dynamic fault maintenance behavior code of the maintenance personnel intelligent agent are supplemented and merged into the resource entity. In the structure.

[0127] Taking an equipment maintenance and support simulation scenario as an example, the support resource-type intelligent agents in the logical model include the spare parts intelligent agent SparePart. Taking the Spare Part intelligent agent SparePart as an example, first, its ID is set to 4001 and its name is "Converter". A description is constructed: "Please analyze the fault and maintenance support logic. Based on the existing static attributes and behavioral logic of the SparePart intelligent agent, supplement the static attribute code that conforms to the fault and maintenance support logic." The intelligent agent's attribute features and behavioral features are concatenated as prompts. The larger model first provides the implementation code for the static attributes; then, it supplements the implementation code for the generated behavioral logic. Finally, the two parts of the implementation code are supplemented and merged into the structure of the SparePart intelligent agent.

[0128] In some embodiments of this application, the safeguard behavior model can be constructed in the following manner: determine the target entity unit in the target equipment class intelligent agent; the target equipment class intelligent agent is any equipment class intelligent agent, and the target entity unit is any entity unit in the target equipment class intelligent agent.

[0129] Using the usage guarantee-related behavioral layer features of the target entity unit as clues, construct the first usage guarantee-related dynamic supplementary prompt words, and guide the LLM to generate the usage guarantee-related supplementary dynamic interaction logic of the target entity unit based on the first usage guarantee-related dynamic supplementary prompt words.

[0130] Identify the target task unit within the target task class intelligent agent; the target task class intelligent agent can be any task class intelligent agent, and the target task unit can be any entity unit within the target task class intelligent agent.

[0131] Using the features of the usage guarantee-related attribute layer of the target task unit as clues, a second usage guarantee-related static supplementary prompt word is constructed, which guides LLM to generate usage guarantee-related supplementary static attributes of the target task unit based on the second usage guarantee-related static supplementary prompt word.

[0132] Using the usage guarantee-related behavioral layer features of the target task unit as clues, a second usage guarantee-related dynamic supplementary prompt is constructed, which guides the LLM to generate usage guarantee-related supplementary dynamic interaction logic of the target task unit based on the second usage guarantee-related dynamic supplementary prompt.

[0133] Identify the target resource entity within the target resource protection class of intelligent agents; the target resource protection class of intelligent agents can be any type of intelligent agent of resource protection, and the target resource entity can be any entity unit within the target resource protection class of intelligent agents.

[0134] Based on the attribute layer features of the target resource entity, static resource prompt words are constructed to guide LLM to generate static attributes of the target resource entity based on the static resource prompt words.

[0135] Based on the behavioral characteristics of the target resource entity, dynamic resource prompts are constructed to guide the LLM to generate dynamic interaction logic for the target resource entity.

[0136] By embedding the supplementary dynamic interaction logic related to the use and maintenance of entity units in each equipment-type intelligent agent into the behavioral structure of that entity unit, by embedding the supplementary static attributes and supplementary dynamic interaction logic related to the use and maintenance of entity units in each task-type intelligent agent into the attribute structure and behavioral structure of that entity unit, and by embedding the static attributes and dynamic interaction logic of entity units in each maintenance support resource-type intelligent agent into the attribute structure and behavioral structure of that entity unit, a fault maintenance support behavior model is obtained.

[0137] In other words, when constructing and using the safeguard behavior model, it is necessary to target equipment-type intelligent agents. Each entity unit in Based on its behavioral characteristics, guided prompts can be constructed using state transition rules (such as transitioning from standby state to operational support state), event response mechanisms (such as increasing equipment fuel level), and relationships with other agents. The LLM then generates dynamic interaction logic code for the agents that conforms to the equipment operational support logic, thus constructing the operational support behavior of the entity unit. Finally, the operational support behavior code of the equipment-type agents is supplemented and merged into the entity unit. In the structure.

[0138] Taking the aircraft agent Plane, which is included in the equipment-type intelligent agents, as an example in the equipment maintenance and support simulation scenario, a prompt is constructed: "Please analyze the equipment use and support behavior logic, consider the dynamic interaction between intelligent agents, and supplement the static attribute code that conforms to the equipment use and support behavior logic based on the existing static attributes and behavior logic of the agent Plane." The large model then provides the implementation code. This implementation code is then merged into the agent Plane structure.

[0139] For task-oriented intelligent agents Each task unit in Based on its attribute layer characteristics, and using identifiers, names (e.g., usage support tasks), parameters (e.g., usage support task cycle, task start / end time, task duration, etc.), variables (e.g., constraints), functions, and descriptions as clues, guiding prompts can be constructed. Based on this, LLM generates static attribute codes for intelligent agents that meet the requirements of equipment usage support tasks, thus realizing the construction of basic information for various task units. Simultaneously, based on its behavioral layer characteristics, and using state transition rules (e.g., task start rules, task end rules, etc.) and relationships with other intelligent agents (e.g., task and equipment-related intelligent agents), guidance prompts can be constructed. Maintenance personnel-type intelligent agents and resource-protecting intelligent agents Using relationships (such as connections) as clues, guiding prompts are constructed. Based on these, the LLM generates dynamic interaction logic code for intelligent agents that meets the needs of equipment use and support tasks, thus realizing the construction of relationships between various task units. Finally, the static attributes and dynamic interaction logic code of the intelligent agents generated by the LLM are embedded into the attribute structure and behavior structure of the task-type intelligent agents, respectively.

[0140] Taking the training task agent Mission within a task-type agent in an equipment maintenance and support simulation scenario as an example, a prompt is constructed: "Please analyze the equipment usage and support behavior logic, consider the dynamic interaction between agents, and supplement the static attribute code that conforms to the equipment usage and support behavior logic based on the existing static attributes and behavior logic of the agent Mission." The large model then provides the implementation code. Simultaneously, the implementation code for generating the behavior logic is also provided. Finally, the two parts of the implementation code are merged into the structure of the agent Mission.

[0141] For resource-protecting intelligent agents Each resource entity in Based on its attribute-level characteristics, the LLM can construct guiding prompts using identifiers, names (e.g., names of resources used), parameters (e.g., quantity of resources used), variables (e.g., remaining quantity of resources), functions, and descriptions as clues. The LLM then generates static attribute codes for the intelligent agent that conform to the equipment's usage support behavior, thus constructing the basic information for the resource entity's usage support. Furthermore, based on its behavioral-level characteristics, the LLM can construct guiding prompts using state transition rules (e.g., the transition from having to not having resources used) and relationships with other intelligent agents (e.g., the increase of variables related to resource usage in the equipment intelligent agent). The LLM then generates dynamic interaction logic codes for the intelligent agent that conform to the equipment's usage support behavior, thus constructing the resource entity's usage support behavior. Finally, the static and dynamic usage support behavior codes generated by the LLM are supplemented and merged into the resource entity. In the structure.

[0142] Taking the SparePart agent, a component of the support resource agent, as an example in an equipment maintenance and support simulation scenario, a prompt is constructed: "Analyze the equipment usage support behavior logic, consider the dynamic interaction between agents, and supplement the static attribute code that conforms to the equipment usage support behavior logic based on the existing static attributes and behavior logic of the SparePart agent." The large model then provides the implementation code. Simultaneously, behavioral logic implementation code that conforms to the equipment usage support behavior logic interaction is generated. Finally, the two parts of implementation code are merged into the structure of the SparePart agent.

[0143] In some embodiments of this application, merging the fault maintenance support behavior model and the usage support behavior model to obtain the equipment maintenance and usage support model may include: determining the target entity unit; the target entity unit is any entity unit in any intelligent agent; merging the attribute structure and behavior structure of the target entity unit in the fault maintenance support behavior model with the attribute structure and behavior structure of the target entity unit in the usage support behavior model to obtain the equipment maintenance and usage support model.

[0144] In some embodiments of this application, after obtaining the equipment maintenance and operation support model, the equipment maintenance and operation support model can be logically verified against the equipment maintenance and operation support checklist, and compensation can be made to the equipment maintenance and operation support model based on the verification results.

[0145] Specifically, this may include: in response to determining, based on the equipment maintenance and operation support checklist, that the target business logic in the equipment maintenance and operation support model has not been implemented, locating the target intelligent agent associated with the target business logic.

[0146] Based on the existing attributes and behavioral context of the target agent, LLM is guided to generate features at the missing attribute layer and features at the missing behavioral layer.

[0147] Add values ​​to the target agent for missing attribute layer features and missing behavior layer features.

[0148] The process involves iteratively executing the checklist for equipment maintenance and operation support to logically verify the equipment maintenance and operation support model. If the target business logic is not realized, the missing attribute layer features and missing behavior layer features are regenerated, and the missing attribute layer features and missing behavior layer features are added to the target intelligent agent. This process continues until the logical verification result indicates that the target business logic has been realized.

[0149] In addition, in response to determining the implementation or functional deficiencies of the target business logic part in the equipment maintenance and operation support model based on the equipment maintenance and operation support checklist, the target intelligent agent associated with the target business logic, and the code implementation fragments in the target intelligent agent.

[0150] While preserving the existing effective logic of the target agent, the LLM is guided to generate supplementary attribute layer features and supplementary behavior layer features, and the existing behavioral logic is expanded or modified.

[0151] Add values ​​to the target agent by supplementing attribute layer features and supplementary behavior layer features.

[0152] The process involves iteratively executing the checklist for equipment maintenance and operation support to logically verify the equipment maintenance and operation support model. When it is determined that the target business logic is partially implemented or the function is incomplete, supplementary attribute layer features and supplementary behavior layer features are generated again, and the supplementary attribute layer features and supplementary behavior layer features are added to the target intelligent agent. This process continues until the logical verification result indicates that the target business logic has been fully implemented.

[0153] In addition, in response to identifying target business logic errors in the equipment maintenance and operation support model based on the equipment maintenance and operation support checklist, the target intelligent agent associated with the target business logic is located.

[0154] The target agent is validated for compliance based on the enhanced retrieval knowledge, and the validation results guide the LLM to update the target agent based on at least the enhanced retrieval knowledge.

[0155] The process involves iteratively executing a checklist to logically verify the equipment maintenance and operation support model, and then performing compliance verification and updating the target agent when an error is detected in the target business logic, until the logic verification result indicates that the target business logic is executed correctly.

[0156] In other words, LLM can be used to logically review the generated equipment maintenance and support model, and to implement categorized compensation for any problems found, ensuring that the model meets business logic requirements. For each inspection item... The logical checks and compensations are as follows:

[0157] Step 1: Logical Gap Compensation. When the LLM determines that a certain business logic is completely unimplemented based on the equipment maintenance and operation support checklist, the following operations are performed in sequence: Locate the agent associated with the logic; Based on the existing attributes and behaviors of the agent, guide the LLM to generate the missing attribute and behavior elements; Add the newly generated elements to the corresponding agent structure; Re-execute the logic check until it passes.

[0158] For example, if the check item is sent as "Check if the aircraft has fault state transitions," the LLM is specified to judge and return one of three errors based on the review results: "Logical blank, logical missing, logical error." When the LLM returns "Logical blank," it further locates the specific agent and, based on the existing code, supplements and generates the missing attributes and behaviors. The LLM locates the aircraft agent Plane, finds that "fault" and state transitions are missing, and provides the implementation code for the "fault" state and state transitions. Then, the implementation code is added to the structure of the aircraft agent Plane, and the check item is re-validated.

[0159] Step 2: Logic Gap Compensation. When the LLM determines that a certain business logic is partially implemented or functionally incomplete, it performs the following operations in sequence: Locate the relevant intelligent agent and specific elements; While retaining the existing valid logic, guide the LLM to generate supplementary attributes or behavioral elements, and expand or modify the existing behavioral logic; Replace the original implementation with the updated elements; Re-verify the integrity of the logic.

[0160] For example, when the LLM returns "Logic Missing," it further locates the specific agent and generates corrected and missing attributes and behaviors based on existing code. The LLM locates the aircraft agent Plane, finds that only the "Fault" state is missing, and provides the implementation code for the "Fault" state. This implementation code is then added to the structure of the aircraft agent Plane; if a "Fault" state already exists, it is directly replaced. Finally, the check item is re-validated.

[0161] Step 3: Logical Error Compensation. When the LLM determines that a certain business logic contains a factual or structural error, it performs the following operations in sequence: Locate the element in the specific erroneous intelligent body; Perform compliance verification on the erroneous element based on the original knowledge definitions injected in Step 1 (such as attribute tuple specifications and behavioral semantic constraints) and the attribute and behavior definitions in Step 2; If the element does not conform to the definition, guide the LLM to regenerate the correct version based on the original knowledge and execute replacement or give a deletion instruction; After all relevant elements are corrected, re-execute the logic check.

[0162] For example, if the LLM returns a "logic error," it further locates the specific agent and element that erred. Based on the existing code, it performs compliance checks on the erroneous element according to the original knowledge definition (such as attribute tuple specifications and behavioral semantic constraints) used when injecting and retrieving augmented knowledge. The LLM locates the aircraft agent Plane and finds that the transition from the "standby" state to the "fault" state is incorrect. The LLM then modifies this to transition from the "working" state to the "fault" state and provides the implementation code. This implementation code replaces the original code, and finally, the logic check is re-executed.

[0163] Figure 3 This is a flowchart illustrating another method for intelligently generating equipment maintenance and operation support models based on LLM, provided in an embodiment of this application. Figure 3 As shown, the method includes the following steps:

[0164] In step one, enhanced knowledge injection can be performed to introduce simulation data information into the equipment's logical model, resulting in an enhanced logical model.

[0165] In step two, an intelligent agent model can be constructed, and a checklist can be generated. This checklist may include an intelligent agent logic checklist and an equipment maintenance and operation support checklist.

[0166] In step three, an LLM model of equipment failure behavior can be constructed based on the built agents and checklists. In one example, the agents can first be validated and compensated using a verifiable agent logic checklist, and then the validated and compensated agents can be used to build an equipment maintenance and operation support model.

[0167] In step four, the logical model of equipment failure and maintenance and operation assurance behavior can be verified and compensated based on LLM and a verifiable equipment maintenance and operation assurance checklist.

[0168] By adopting the technical solution provided in the embodiments of this application, and combining the dynamic interaction advantages of intelligent agent modeling with the logical reasoning capabilities of large language models, efficient and intelligent generation of equipment maintenance and operation support models is achieved.

[0169] All of the above-mentioned optional technical solutions can be combined in any way to form the optional embodiments of this application, and will not be described in detail here.

[0170] The following are embodiments of the apparatus described in this application, which can be used to execute the embodiments of the method described in this application. For details not disclosed in the apparatus embodiments of this application, please refer to the embodiments of the method described in this application.

[0171] Figure 4 This is a schematic diagram of an intelligent generation device for equipment maintenance and operation support model based on LLM, provided in an embodiment of this application. Figure 4 As shown, the device includes:

[0172] The logical model building module 401 is configured to build a logical model of the equipment using the model-based systems engineering (MBSE) method, and inject retrieval enhancement knowledge into the logical model to obtain an enhanced logical model.

[0173] The definition module 402 is configured to use the Large Language Model (LLM) to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents based on the enhanced logic model and the equipment's task scenario, and to define attribute layer features and behavioral layer features for each entity unit in each intelligent agent; the attribute layer features are used to characterize the static features of the entity unit, and the behavioral layer features are used to characterize the dynamic interaction logic of the entity unit.

[0174] The equipment composition model and task scenario model construction module 403 is configured to use LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, thereby obtaining an equipment composition model; based on the equipment composition model, it uses LLM to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent, and uses LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, thereby obtaining a task scenario model.

[0175] The fault maintenance and support behavior model construction module 404 is configured to, based on the task scenario model, use LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent, so as to obtain the fault maintenance and support behavior model.

[0176] The support behavior model construction module 405 is configured to, based on the task scenario model, update the attribute structure and behavior structure of the equipment-type intelligent agent using LLM based on the use support-related behavior layer features of each entity unit in the equipment-type intelligent agent, and update the attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent using LLM based on the use support-related attribute layer features and use support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, thereby obtaining the use support behavior model.

[0177] The merging module 406 is configured to merge the fault maintenance support behavior model and the usage support behavior model to obtain the equipment maintenance and usage support model.

[0178] According to the technical solution provided in this application, enhanced retrieval is achieved by combining a vector database, and a framework for generating equipment maintenance and operation support models based on intelligent agents is constructed. Then, for each business component involved in the equipment maintenance and operation support model, structural elements in the intelligent agent are generated one by one, and the results are integrated and merged based on the model structure, ultimately realizing a complete and operable equipment maintenance and operation support model. This method integrates the interactive modeling capabilities of intelligent agents with the semantic understanding and logical reasoning capabilities of large language models, realizing the intelligent generation of maintenance and operation support models for complex equipment, and effectively improving the generation efficiency of equipment maintenance and support models.

[0179] In some implementations, the equipment composition model is constructed as follows: A target entity unit within a target equipment class intelligent agent is identified; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within that target equipment class intelligent agent; static prompts for the equipment composition model are constructed using the attribute layer features of the target entity unit as clues, guiding the LLM to generate static attributes of the target entity unit based on these static prompts; dynamic prompts for the equipment composition model are constructed using the behavioral layer features of the target entity unit as clues, guiding the LLM to generate dynamic interaction logic of the target entity unit based on these dynamic prompts; the static attributes and dynamic interaction logic of each entity unit are embedded into the attribute structure and behavioral structure of that entity unit to obtain the equipment composition model.

[0180] The task scenario model is constructed as follows: A target task unit is identified within a target task-class intelligent agent; the target task-class intelligent agent can be any task-class intelligent agent, and the target task unit is any entity unit within that target task-class intelligent agent; static prompts for the task scenario model are constructed using the attribute-level features of the target task unit in the equipment composition model as clues, guiding the LLM to generate static attributes of the target task unit based on these static prompts; dynamic prompts for the task scenario model are constructed using the behavioral-level features of the target task unit in the equipment composition model as clues, guiding the LLM to generate dynamic interactions of the target task unit based on these dynamic prompts. Logic; Identify the target entity unit within the target equipment class intelligent agent; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within the target equipment class intelligent agent; Construct supplementary prompts for the first task scenario model using the behavioral layer features of the target entity unit in the equipment composition model as clues, guiding the LLM to generate supplementary dynamic interaction logic between the target equipment class intelligent agent and its corresponding task class intelligent agent based on the supplementary prompts for the first task scenario model; Identify the target personnel entity within the target maintenance personnel class intelligent agent; the target maintenance personnel class intelligent agent can be any maintenance personnel class intelligent agent, and the target personnel entity is any entity within the target maintenance personnel class intelligent agent. The unit; constructs supplementary prompts for the second task scenario model based on the behavioral features of the target personnel entity in the equipment composition model, guiding the LLM to generate supplementary dynamic interaction logic between the target maintenance personnel agent and its corresponding task agent based on the supplementary prompts of the second task scenario model; determines the target resource entity in the target support resource agent; the target support resource agent is any support resource agent, and the target resource entity is any entity unit in the target support resource agent; constructs supplementary prompts for the third task scenario model based on the behavioral features of the target resource entity in the equipment composition model, guiding the LLM to ... The system generates supplementary dynamic interaction logic between the target support resource-type intelligent agent and its corresponding task-type intelligent agent using prompt words. It then embeds the static attributes and dynamic interaction logic of each task-type intelligent agent into the attribute structure and behavior structure of the task unit, the supplementary dynamic interaction logic of each equipment-type intelligent agent and its corresponding task-type intelligent agent into the attribute structure and behavior structure of the entity unit, the supplementary dynamic interaction logic of each maintenance personnel-type intelligent agent and its corresponding task-type intelligent agent into the attribute structure and behavior structure of the target personnel entity, and the supplementary dynamic interaction logic of each maintenance support resource-type intelligent agent and its corresponding task-type intelligent agent into the attribute structure and behavior structure of the target resource entity, thus obtaining the task scenario model.

[0181] In some implementations, the fault repair and support behavior model is constructed as follows: First, a target entity unit is identified within a target equipment-type intelligent agent; the target equipment-type intelligent agent can be any equipment-type intelligent agent, and the target entity unit is any entity unit within that target equipment-type intelligent agent. Second, fault-related static supplementary prompts are constructed using the fault-related attribute layer features of the target entity unit as clues, guiding the LLM to generate fault-related supplementary static attributes of the target entity unit based on these prompts. Third, fault-related dynamic supplementary prompts are constructed using the fault-related behavioral layer features of the target entity unit as clues, guiding the LLM to generate fault-related supplementary dynamic interaction logic for the target entity unit based on these prompts. Finally, a target personnel entity is identified within a target maintenance personnel-type intelligent agent; the target maintenance personnel-type intelligent agent can be any maintenance personnel-type intelligent agent, and the target personnel entity is any entity unit within that target maintenance personnel-type intelligent agent. Fourth, personnel static prompts are constructed using the attribute layer features of the target personnel entity as clues, guiding the LLM to generate static attributes of the target personnel entity based on these prompts. Fifth, personnel dynamic prompts are constructed using the behavioral layer features of the target personnel entity as clues, guiding the LLM to generate dynamic attributes of the target personnel entity based on these prompts. The system generates dynamic interaction logic for the target personnel entity using dynamic prompts; identifies target resource entities within the target resource support agent; the target resource support agent can be any resource support agent, and the target resource entity can be any entity unit within that agent; constructs static prompts for fault-related resources based on the fault-related attribute layer features of the target resource entity, guiding the LLM to generate fault-related static attributes of the target resource entity based on these static prompts; and constructs dynamic prompts for fault-related resources based on the fault-related behavioral layer features of the target resource entity, guiding the LLM to generate fault-related static attributes of the target resource entity based on these static prompts. The LLM generates fault-related dynamic interaction logic for the target resource entity based on the fault-related resource dynamic prompts; it embeds the fault-related supplementary static attributes and fault-related supplementary dynamic interaction logic of the entity units in each equipment-type intelligent agent into the attribute structure and behavior structure of the entity unit; it embeds the static attributes and dynamic interaction logic of the entity units in each maintenance personnel-type intelligent agent into the attribute structure and behavior structure of the entity unit; and it embeds the fault-related static attributes and fault-related dynamic interaction logic of the entity units in each maintenance support resource-type intelligent agent into the attribute structure and behavior structure of the entity unit, thus obtaining the fault maintenance support behavior model.

[0182] In some implementations, the support behavior model is constructed as follows: First, a target entity unit is identified within a target equipment-type intelligent agent; the target equipment-type intelligent agent can be any equipment-type intelligent agent, and the target entity unit is any entity unit within that target equipment-type intelligent agent. Second, a target task unit is identified within a target task-type intelligent agent; the target task-type intelligent agent can be any task-type intelligent agent, and the target task unit is any entity unit within that target task-type intelligent agent. Third, a second use-support related static supplementary prompt is constructed within a target task-type intelligent agent; the target task unit can be any task-type intelligent agent, and the target task unit is any entity unit within that target task-type intelligent agent. Fourth, a second use-support related static supplementary prompt is constructed within a target task-type intelligent agent; the LLM generates use-support related static supplementary attributes of the target task unit based on the second use-support related static supplementary prompt. Finally, a second use-support related dynamic supplementary prompt is constructed within a target task-type intelligent agent, guiding the LLM to generate use-support related static attributes of the target task unit based on the second use-support related static supplementary prompt. The system generates usage support-related supplementary dynamic interaction logic for the target task unit using dynamic supplementary prompts related to support; it identifies the target resource entity in the target support resource class intelligent agent; the target support resource class intelligent agent is any support resource class intelligent agent, and the target resource entity is any entity unit in the target support resource class intelligent agent; it constructs resource static prompts based on the attribute layer features of the target resource entity, guiding the LLM to generate the static attributes of the target resource entity based on the resource static prompts; it constructs resource dynamic prompts based on the behavior layer features of the target resource entity, guiding the LLM to generate the dynamic interaction logic of the target resource entity based on the resource dynamic prompts; it embeds the usage support-related supplementary dynamic interaction logic of the entity unit in each equipment class intelligent agent into the behavior structure of this entity unit, embeds the usage support-related supplementary static attributes and usage support-related supplementary dynamic interaction logic of the entity unit in each task class intelligent agent into the attribute structure and behavior structure of this entity unit, and embeds the static attributes and dynamic interaction logic of the entity unit in each maintenance support resource class intelligent agent into the attribute structure and behavior structure of this entity unit, thus obtaining the usage support behavior model.

[0183] In some implementations, the fault repair support behavior model and the usage support behavior model are merged to obtain the equipment maintenance and usage support model, including: determining a target entity unit; the target entity unit is any entity unit in any intelligent agent; merging the attribute structure and behavior structure of the target entity unit in the fault repair support behavior model with the attribute structure and behavior structure of the target entity unit in the usage support behavior model, respectively, to obtain the equipment maintenance and usage support model.

[0184] In some implementations, after obtaining the enhanced logic model, the method further includes: determining a verifiable agent logic checklist and an equipment maintenance and use support checklist based on the equipment's mission scenario and requirements; after constructing equipment-type agents, mission-type agents, maintenance personnel-type agents, and support resource-type agents, verifying and compensating the equipment-type agents, mission-type agents, maintenance personnel-type agents, and support resource-type agents respectively against the agent logic checklist; and after obtaining the equipment maintenance and use support model, the method further includes: performing logical verification on the equipment maintenance and use support model against the equipment maintenance and use support checklist, and compensating the equipment maintenance and use support model based on the verification results.

[0185] In some implementations, performing logical verification and compensation on the equipment maintenance and operation support model against the equipment maintenance and operation support checklist includes: in response to determining, based on the equipment maintenance and operation support checklist, that the target business logic in the equipment maintenance and operation support model is not implemented, locating the target agent associated with the target business logic; based on the existing attributes and behavioral context of the target agent, guiding the LLM to generate missing attribute layer features and missing behavioral layer features; adding the missing attribute layer features and missing behavioral layer features to the target agent; iteratively executing the steps of logically verifying the equipment maintenance and operation support model against the equipment maintenance and operation support checklist, and regenerating the missing attribute layer features and missing behavioral layer features when it is determined that the target business logic is not implemented, and adding the missing attribute layer features and missing behavioral layer features to the target agent, until the logical verification result indicates that the target business logic has been implemented; and, in response to determining, based on the equipment maintenance and operation support checklist, that the target business logic in the equipment maintenance and operation support model is partially implemented or functionally incomplete, locating the target agent associated with the target business logic, and the code implementation fragment in the target agent; while retaining the existing valid logic of the target agent... The process involves guiding the LLM to generate supplementary attribute layer features and supplementary behavioral layer features, and expanding or modifying existing behavioral logic; adding the supplementary attribute layer features and supplementary behavioral layer features to the target agent; iteratively executing the steps of logically verifying the equipment maintenance and use support model against the equipment maintenance and use support checklist, and generating supplementary attribute layer features and supplementary behavioral layer features again when it is determined that the target business logic is partially implemented or functionally incomplete, and adding the supplementary attribute layer features and supplementary behavioral layer features to the target agent, until the logical verification result indicates that the target business logic has been fully implemented; and, in response to determining that the target business logic in the equipment maintenance and use support model is incorrect based on the equipment maintenance and use support checklist, locating the target agent associated with the target business logic; performing compliance verification on the target agent based on the retrieval enhancement knowledge, and guiding the LLM to update the target agent based on the verification result at least based on the retrieval enhancement knowledge; iteratively executing the steps of logically verifying the equipment maintenance and use support model against the equipment maintenance and use support checklist, and performing compliance verification and updating the target agent again when it is determined that the target business logic is incorrect, until the logical verification result indicates that the target business logic is correctly executed.

[0186] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0187] Figure 5 This is a schematic diagram of the electronic device provided in an embodiment of this application. For example... Figure 5 As shown, the electronic device 5 of this embodiment includes: a processor 501, a memory 502, and a computer program 503 stored in the memory 502 and executable on the processor 501. When the processor 501 executes the computer program 503, it implements the steps in the various method embodiments described above. Alternatively, when the processor 501 executes the computer program 503, it implements the functions of each module / unit in the various device embodiments described above.

[0188] Electronic device 5 can be a desktop computer, laptop, handheld computer, cloud server, or other electronic device. Electronic device 5 may include, but is not limited to, processor 501 and memory 502. Those skilled in the art will understand that... Figure 5 This is merely an example of electronic device 5 and does not constitute a limitation on electronic device 5. It may include more or fewer components than shown, or different components.

[0189] The processor 501 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0190] The memory 502 can be an internal storage unit of the electronic device 5, such as a hard disk or RAM of the electronic device 5. The memory 502 can also be an external storage device of the electronic device 5, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the electronic device 5. The memory 502 can also include both internal and external storage units of the electronic device 5. The memory 502 is used to store computer programs and other programs and data required by the electronic device.

[0191] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0192] If an integrated module / unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program may include computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable medium may include: any entity or device capable of carrying computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0193] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.< / activeobjectclass> < / activeobjectclass>

Claims

1. A method for intelligently generating equipment maintenance and operation support models based on LLM, characterized in that, include: A logical model of the equipment is constructed using the Model-Based Systems Engineering (MBSE) method, and retrieval-enhanced knowledge is injected into the logical model to obtain an enhanced logical model. Using the Large Language Model (LLM), based on the enhanced logical model and the equipment's task scenario, we construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents, and define attribute layer features and behavioral layer features for each entity unit in each intelligent agent. The attribute layer features are used to characterize the static features of the entity unit, and the behavior layer features are used to characterize the dynamic interaction logic of the entity unit. Using LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, an equipment composition model is obtained. Based on the equipment composition model, using LLM to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent, and using LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, a task scenario model is obtained. Based on the task scenario model, the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent are updated using LLM based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent, so as to obtain the fault maintenance support behavior model. Based on the task scenario model, the attribute structure and behavior structure of the equipment-type intelligent agent are updated using LLM based on the usage support-related behavior layer features of each entity unit in the equipment-type intelligent agent. Furthermore, the attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent are updated using LLM based on the usage support-related attribute layer features and usage support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, thus obtaining the usage support behavior model. The fault repair support behavior model and the usage support behavior model are combined to obtain the equipment maintenance and usage support model. The equipment maintenance and operation support model is obtained by merging the fault repair support behavior model and the operation support behavior model, including: Identify the target entity unit; the target entity unit is any entity unit in any intelligent agent; The attribute structure and behavior structure of the target entity unit in the fault maintenance support behavior model are merged with the attribute structure and behavior structure of the target entity unit in the use support behavior model to obtain the equipment maintenance and use support model.

2. The method according to claim 1, characterized in that, The equipment composition model was constructed in the following manner: Identify the target entity unit within the target equipment class intelligent agent; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within the target equipment class intelligent agent; Using the attribute layer features of the target entity unit as clues, static prompt words for the equipment composition model are constructed to guide the LLM to generate static attributes of the target entity unit based on the static prompt words of the equipment composition model; Using the behavioral features of the target entity unit as clues, dynamic prompt words for the equipment composition model are constructed to guide the LLM to generate dynamic interaction logic for the target entity unit based on the dynamic prompt words of the equipment composition model. The static attributes and dynamic interaction logic of each entity unit are embedded into the attribute structure and behavior structure of the entity unit to obtain the equipment composition model. The task scenario model is constructed in the following manner: Identify the target task unit within the target task class intelligent agent; the target task class intelligent agent can be any task class intelligent agent, and the target task unit can be any entity unit within the target task class intelligent agent; Using the attribute layer features of the target task unit in the equipment composition model as clues, static prompt words for the task scenario model are constructed to guide the LLM to generate static attributes of the target task unit based on the static prompt words for the task scenario model. Using the behavioral features of the target task unit in the equipment composition model as clues, dynamic prompt words for the task scenario model are constructed to guide the LLM to generate the dynamic interaction logic of the target task unit based on the dynamic prompt words of the task scenario model. Identify the target entity unit within the target equipment class intelligent agent; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within the target equipment class intelligent agent; Using the behavioral layer features of the target entity unit in the equipment composition model as clues, a first task scenario model supplementary prompt words are constructed to guide the LLM to generate supplementary dynamic interaction logic between the target equipment class intelligent agent and its corresponding task class intelligent agent based on the first task scenario model supplementary prompt words. Determine the target personnel entity within the target maintenance personnel class intelligent agent; the target maintenance personnel class intelligent agent is any maintenance personnel class intelligent agent, and the target personnel entity is any entity unit within the target maintenance personnel class intelligent agent; Using the behavioral features of the target personnel entity in the equipment composition model as clues, construct supplementary prompts for the second task scenario model, and guide the LLM to generate supplementary dynamic interaction logic between the target maintenance personnel agent and its corresponding task agent based on the supplementary prompts for the second task scenario model. Identify the target resource entity within the target resource protection class of intelligent agents; the target resource protection class of intelligent agents can be any resource protection class of intelligent agents, and the target resource entity can be any entity unit within the target resource protection class of intelligent agents; Using the behavioral features of the target resource entity in the equipment composition model as clues, a third task scenario model supplementary prompt words are constructed to guide the LLM to generate supplementary dynamic interaction logic between the target support resource class intelligent agent and its corresponding task class intelligent agent based on the third task scenario model supplementary prompt words. The static attributes and dynamic interaction logic of each task-type intelligent agent are embedded into the attribute structure and behavior structure of this task unit. The supplementary dynamic interaction logic of each equipment-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this entity unit. The supplementary dynamic interaction logic of each maintenance personnel-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this target personnel entity. The supplementary dynamic interaction logic of each maintenance support resource-type intelligent agent and its corresponding task-type intelligent agent is embedded into the attribute structure and behavior structure of this target resource entity, thus obtaining the task scenario model.

3. The method according to claim 1, characterized in that, The fault repair and support behavior model is constructed in the following manner: Identify the target entity unit within the target equipment class intelligent agent; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within the target equipment class intelligent agent; Using the fault-related attribute layer features of the target entity unit as clues, fault-related static supplementary prompt words are constructed to guide the LLM to generate fault-related supplementary static attributes of the target entity unit based on the fault-related static supplementary prompt words; Based on the fault-related behavioral layer features of the target entity unit, fault-related dynamic supplementary prompt words are constructed to guide the LLM to generate fault-related supplementary dynamic interaction logic of the target entity unit based on the fault-related dynamic supplementary prompt words; Determine the target personnel entity within the target maintenance personnel class intelligent agent; the target maintenance personnel class intelligent agent is any maintenance personnel class intelligent agent, and the target personnel entity is any entity unit within the target maintenance personnel class intelligent agent; Using the attribute layer features of the target person entity as clues, construct static prompt words for the person, and guide the LLM to generate static attributes of the target person entity based on the static prompt words for the person; Using the behavioral features of the target person entity as clues, construct dynamic prompt words for the person, and guide the LLM to generate dynamic interaction logic for the target person entity based on the dynamic prompt words; Identify the target resource entity within the target resource protection class of intelligent agents; the target resource protection class of intelligent agents can be any resource protection class of intelligent agents, and the target resource entity can be any entity unit within the target resource protection class of intelligent agents; Based on the fault-related attribute layer features of the target resource entity, fault-related static prompt words are constructed to guide the LLM to generate fault-related static attributes of the target resource entity based on the fault-related static prompt words. Based on the fault-related behavior layer features of the target resource entity, fault-related dynamic prompt words are constructed to guide the LLM to generate fault-related dynamic interaction logic of the target resource entity based on the fault-related dynamic prompt words. The fault-related supplementary static attributes and fault-related supplementary dynamic interaction logic of the entity units in each equipment-type intelligent agent are embedded into the attribute structure and behavior structure of the entity unit. The static attributes and dynamic interaction logic of the entity units in each maintenance personnel-type intelligent agent are embedded into the attribute structure and behavior structure of the entity unit. The fault-related static attributes and fault-related dynamic interaction logic of the entity units in each maintenance support resource-type intelligent agent are embedded into the attribute structure and behavior structure of the entity unit, thus obtaining the fault maintenance support behavior model.

4. The method according to claim 1, characterized in that, The usage protection behavior model is constructed in the following manner: Identify the target entity unit within the target equipment class intelligent agent; the target equipment class intelligent agent can be any equipment class intelligent agent, and the target entity unit is any entity unit within the target equipment class intelligent agent; Using the usage guarantee-related behavioral layer features of the target entity unit as clues, a first usage guarantee-related dynamic supplementary prompt word is constructed, which guides the LLM to generate the usage guarantee-related supplementary dynamic interaction logic of the target entity unit based on the first usage guarantee-related dynamic supplementary prompt word; Identify the target task unit within the target task class intelligent agent; the target task class intelligent agent can be any task class intelligent agent, and the target task unit can be any entity unit within the target task class intelligent agent; Using the features of the usage guarantee-related attribute layer of the target task unit as clues, a second usage guarantee-related static supplementary prompt word is constructed to guide the LLM to generate the usage guarantee-related supplementary static attributes of the target task unit based on the second usage guarantee-related static supplementary prompt word; Using the usage guarantee-related behavioral layer features of the target task unit as clues, construct a second usage guarantee-related dynamic supplementary prompt word, and guide the LLM to generate the usage guarantee-related supplementary dynamic interaction logic of the target task unit based on the second usage guarantee-related dynamic supplementary prompt word; Identify the target resource entity within the target resource protection class of intelligent agents; the target resource protection class of intelligent agents can be any resource protection class of intelligent agents, and the target resource entity can be any entity unit within the target resource protection class of intelligent agents; Based on the attribute layer features of the target resource entity, static resource prompt words are constructed to guide the LLM to generate static attributes of the target resource entity based on the static resource prompt words; Based on the behavioral features of the target resource entity, dynamic resource prompts are constructed to guide the LLM to generate dynamic interaction logic for the target resource entity. The usage support behavior model is obtained by embedding the supplementary dynamic interaction logic related to the use and maintenance of entity units in each equipment-type intelligent agent into the behavior structure of the entity unit, embedding the supplementary static attributes and supplementary dynamic interaction logic related to the use and maintenance of entity units in each task-type intelligent agent into the attribute structure and behavior structure of the entity unit, and embedding the static attributes and dynamic interaction logic of entity units in each maintenance support resource-type intelligent agent into the attribute structure and behavior structure of the entity unit.

5. The method according to claim 1, characterized in that, After obtaining the enhanced logical model, the method further includes: Based on the mission scenarios and requirements of the equipment, a verifiable checklist of intelligent agent logic and a checklist of equipment maintenance and use support are determined. After constructing equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents, the equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents are respectively verified and compensated according to the intelligent agent logic checklist; Furthermore, after obtaining the equipment maintenance and operation support model, the method further includes: The equipment maintenance and operation support model is logically verified against the equipment maintenance and operation support checklist, and compensation is made to the equipment maintenance and operation support model based on the verification results.

6. The method according to claim 1, characterized in that, The equipment maintenance and operation support model is logically validated and compensated against the aforementioned equipment maintenance and operation support checklist, including: In response to the determination based on the equipment maintenance and operation support checklist that the target business logic in the equipment maintenance and operation support model has not been implemented, the target intelligent agent associated with the target business logic is located. Based on the existing attributes and behavioral context of the target agent, guide the LLM to generate features at the missing attribute layer and features at the missing behavioral layer. Add the missing attribute layer features and missing behavior layer features to the target agent; The process involves iteratively executing the checklist for equipment maintenance and operation support to logically verify the equipment maintenance and operation support model. If it is determined that the target business logic has not been implemented, the missing attribute layer features and missing behavior layer features are regenerated, and the missing attribute layer features and missing behavior layer features are added to the target intelligent agent. This process continues until the logical verification result indicates that the target business logic has been implemented. In addition, in response to determining, based on the equipment maintenance and operation support checklist, that the target business logic part of the equipment maintenance and operation support model is partially implemented or functionally incomplete, the target intelligent agent associated with the target business logic, and the code implementation fragment in the target intelligent agent are located; While retaining the existing effective logic of the target agent, the LLM is guided to generate supplementary attribute layer features and supplementary behavior layer features, and the existing behavior logic is expanded or modified. Add the supplementary attribute layer features and supplementary behavior layer features to the target agent; The process involves iteratively executing the checklist for equipment maintenance and operation support to logically verify the equipment maintenance and operation support model. When it is determined that the target business logic is partially implemented or the function is incomplete, supplementary attribute layer features and supplementary behavior layer features are generated again, and the supplementary attribute layer features and supplementary behavior layer features are added to the target intelligent agent. This process continues until the logical verification result indicates that the target business logic has been fully implemented. In addition, in response to determining a target business logic error in the equipment maintenance and operation support model based on the equipment maintenance and operation support checklist, the target intelligent agent associated with the target business logic is located; Based on the enhanced retrieval knowledge, the target agent is subjected to compliance verification, and based on the verification result, the LLM is guided to update the target agent at least based on the enhanced retrieval knowledge. The process involves iteratively executing a checklist to logically verify the equipment maintenance and operation support model, and then performing a compliance verification update on the target intelligent agent when an error is detected in the target business logic, until the logic verification result indicates that the target business logic is executed correctly.

7. An intelligent generation device for equipment maintenance and operation support models based on LLM, characterized in that, include: The logical model building module is configured to build a logical model of the equipment using the model-based systems engineering (MBSE) method, and inject retrieval enhancement knowledge into the logical model to obtain an enhanced logical model. The definition module is configured to use the Large Language Model (LLM) to construct equipment-type intelligent agents, task-type intelligent agents, maintenance personnel-type intelligent agents, and support resource-type intelligent agents based on the enhanced logical model and the equipment's task scenario, and to define attribute layer features and behavioral layer features for each entity unit in each intelligent agent. The attribute layer features are used to characterize the static features of the entity unit, and the behavior layer features are used to characterize the dynamic interaction logic of the entity unit. The equipment composition model and task scenario model construction module is configured to use LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the equipment-type intelligent agent, thereby obtaining the equipment composition model; based on the equipment composition model, LLM is used to update the attribute structure and behavior structure of each entity unit in the task-type intelligent agent based on the attribute layer features and behavior layer features of each entity unit in the task-type intelligent agent, and LLM is used to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, maintenance personnel-type intelligent agent, and support resource-type intelligent agent based at least on the behavior layer features of each entity unit in the task-type intelligent agent, thereby obtaining the task scenario model; The fault repair and support behavior model construction module is configured to, based on the task scenario model, use LLM to update the attribute structure and behavior structure of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent based on the fault-related attribute layer features and fault-related behavior layer features of each entity unit in the equipment-type intelligent agent, the maintenance personnel-type intelligent agent, and the support resource-type intelligent agent, so as to obtain the fault repair and support behavior model. The support behavior model construction module is configured to, based on the task scenario model, use LLM to update the attribute structure and behavior structure of the equipment-type intelligent agent based on the use support-related behavior layer features of each entity unit in the equipment-type intelligent agent, and use LLM to update the attribute structure and behavior structure of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent based on the use support-related attribute layer features and use support-related behavior layer features of each entity unit in the maintenance personnel-type intelligent agent and the support resource-type intelligent agent, so as to obtain the use support behavior model; The merging module is configured to merge the fault repair support behavior model and the usage support behavior model to obtain the equipment maintenance and usage support model. The equipment maintenance and operation support model is obtained by merging the fault repair support behavior model and the operation support behavior model, including: Identify the target entity unit; the target entity unit is any entity unit in any intelligent agent; The attribute structure and behavior structure of the target entity unit in the fault maintenance support behavior model are merged with the attribute structure and behavior structure of the target entity unit in the use support behavior model to obtain the equipment maintenance and use support model.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 6.

Images