Automatic reliability modeling system, method and equipment

Through a large language model, the fault knowledge is extracted from multiple data sources and formed a knowledge graph to identify and link fault tree events, the problem of insufficient efficiency and accuracy of automatic construction of fault tree models of large complex system is solved, and a fast and accurate fault tree is achieved. Modeling.

CN119026682APending Publication Date: 2024-11-26ANHUI HUCHAO TECHNOLOGY CO LTD
View PDF 0 Cites 4 Cited by

Patent Information

Application Number
CN202411194342.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-29
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing technology lacks intelligent, automated and humanized modeling methods in the automatic construction of fault tree models for large and complex systems, resulting in inefficient modeling efficiency and insufficient model accuracy, and traditional reliability automatic modeling equipment is not portable and consumed. Time-consuming.

Method used

Using a reliability automatic modeling system including knowledge acquisition and generation module, fault knowledge base module, knowledge analysis module and fault tree generation module, fault knowledge extracts fault knowledge from documents, web pages, audio and video carriers through large language models to form knowledge The graph is used to identify and link the top, intermediate and bottom events of the fault tree to form a system fault tree model.

Benefits of technology

Automatically establishing a fault tree model is realized, which improves modeling speed and accuracy, reduces manual labeling workload, broadens the source of fault knowledge, and is suitable for reliability analysis of complex systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119026682A_ABST
    Figure CN119026682A_ABST
Patent Text Reader

Abstract

The invention provides an automatic reliability modeling system, method and equipment, and belongs to the technical field of reliability analys.The automatic reliability modeling system comprises a knowledge collection and generation module, a fault knowledge base module, a knowledge analysis module and a fault tree generation module.The automatic reliability modeling system comprises the steps that firstly, fault knowledge is extracted from documents, webpages and audio and video carriers through the knowledge collection and generation module; the information is stored in a fault knowledge base module in the form of knowledge graph entities, relationships and attributes, and then a top event, a middle event and a bottom event of a fault tree and a logical relationship among the top event, the middle event and the bottom event are identified from the fault knowledge base module through a knowledge analysis module; and finally, linking the recognized top event, middle event and bottom event according to a certain logical relationship through a fault tree generation module to form a system fault tree model. According to the method, the reliability model of the analyzed object can be automatically established, automatic fault tree modeling is realized through a set of equipment comprising a camera, a microphone, a large model server, a data memory and a computing processor, and a quick modeling tool is provided for reliability analysis of a complex system; and a reliability engineer is helped to design, evaluate and manage the system reliability.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention mainly relates to the technical field of reliability analysis, and in particular to a reliability automatic modeling system, method and equipment. Background Art

[0002] Fault tree is a commonly used reliability analysis model. Usually, reliability engineers perform functional decomposition and failure cause analysis of the system based on existing experience, and then manually establish a fault tree model. In the field of reliability analysis, fault tree modeling is a tedious task that is highly dependent on expert knowledge and experience. It is very time-consuming and not easy to automate. It not only reduces the efficiency of people carrying out reliability analysis work, but also increases the technical threshold for non-professionals to carry out reliability analysis work. The fault tree model is one of the more complex models in the field of reliability analysis. Especially for large and complex systems (such as nuclear power plants), it may even take several man-years to establish a complete fault tree model. For this reason, researchers hope to develop a computer-aided modeling method. With the development of artificial intelligence technology, conducting automatic modeling research based on artificial intelligence methods has always been a key problem and research hotspot in this field.

[0003] For example, Comparative Document 1 (Zhang Peng, et al. Development of Fault Tree Automatic Modeling Software Based on Visio Platform. "Industrial Control Computer". 2015, Vol. 8 (Issue 1)) published a fault tree automatic modeling method based on Visio file format. This method only uses the Visio file format interpretation method, and can be better compatible with the Visio file format in terms of fault tree modeling software functions; Comparative Document 2 (CN114281590A, 2022.04.05; CN115809348A, 2023.03.17) published a fault tree automatic generation method based on the bill of materials (BOM) list and FMEA structure of the system to be analyzed. It combines the BOM list and FMEA structure. As an information source for fault tree modeling, it has achieved auxiliary modeling to a certain extent, but it is highly dependent on two structured information: BOM list and FMEA structure; Comparative document 3 (CN114443856A, 2022.05.06) announced a knowledge graph construction method based on fault tree images, which is a method for identifying fault tree images and establishing a knowledge graph, and did not propose a reverse modeling method from knowledge graph to fault tree; other comparative documents (CN115718472A, 2023.02.28; CN115828173A, 2023.03.21) announced a fault diagnosis method based on knowledge graph, and did not involve a method for automatically constructing a fault tree model.

[0004] It can be seen that the automatic construction of fault tree models for large and complex systems currently still lacks a more intelligent, automated, and humanized modeling method. It is necessary to be able to automatically extract knowledge from massive data and automatically construct fault trees based on this knowledge. In addition, the source of knowledge needs to be more generalized and humanized, close to human natural language, which is very important for improving modeling efficiency and model accuracy. However, existing methods and systems cannot meet the above requirements well.

[0005] At the same time, the currently commonly used reliability automatic modeling equipment is a computer. For extremely large and complex systems, a workstation or even a supercomputer is required. It is not cheap and portable, which is not conducive to the large-scale promotion of reliability analysis technology. In addition, before modeling the system, traditional reliability modelers need to prepare a large amount of written and electronic data, and then combine these data with their own experience to model. The whole process is time-consuming and laborious, and there is a lack of an intelligent and convenient reliability automatic modeling equipment. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a reliability automatic modeling system, method and equipment in view of the deficiencies in the prior art.

[0007] The technical solution of the present invention to solve the above technical problems is as follows:

[0008] A reliability automatic modeling system includes a knowledge acquisition and generation module, a fault knowledge base module, a knowledge analysis module, and a fault tree generation module:

[0009] The knowledge collection and generation module is used to extract fault knowledge from documents, web pages, audio and video carriers;

[0010] The fault knowledge base module is used to store fault knowledge in the form of knowledge graph entities, relationships and attributes;

[0011] The knowledge parsing module is used to identify the top event, the middle event and the bottom event of the fault tree from the fault knowledge base module, and the logical relationship between them;

[0011] The fault tree generation module is used to link the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

[0012] The knowledge acquisition and generation module includes a general large model, a model fine-tuning unit, and a retrieval generation unit;

[0013] The general large models include but are not limited to Llama series versions and Qwen series versions;

[0014] The model fine-tuning unit has the functions of setting retrieval and generation weight factors, selecting data sets, and self-evolving according to user input information;

[0015] The retrieval generation unit has the function of retrieving knowledge from the data storage device, and integrates and processes the retrieved content and the generated content according to the retrieval and generation weight factors set by the model fine-tuning unit.

[0016] The knowledge collection and generation module can automatically fill in knowledge and automatically update knowledge through a large model. The document types include TXT, XML, Word, EXCEL, PPT, and PDF. In addition to the unstructured data in the documents, web pages, and audio and video carriers, the knowledge source also supports directly importing fault knowledge from structured databases.

[0017] The fault knowledge base module adopts a graph database, including a data management system, a general knowledge graph, and a professional knowledge graph;

[0018] The data management system includes but is not limited to Neo4j, Amazon Neptune, TigerGraph, ArangoDB, NebulaGraph;

[0019] The general knowledge graph has the function of importing public graph data and the function of self-evolving according to user input information;

[0020] The professional knowledge graph has the function of importing user-defined data and the function of self-evolving according to user input information.

[0021] Another technical solution of the present invention to solve the above technical problem is as follows: a reliability automatic modeling method, comprising the following steps:

[0022] S1, firstly extract fault knowledge from documents, web pages, audio and video carriers through a large model, and store it in the fault knowledge base in the form of knowledge graph entities, relationships and attributes;

[0023] S2, then identifying the top event, the middle event and the bottom event of the fault tree, and the logical relationship between them from the fault knowledge base through the knowledge parser;

[0024] S3, finally, the fault tree generator links the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

[0025] The step S2 specifically includes the following steps:

[0026] S21, searching the corresponding entity from the fault knowledge base according to the name of the object to be analyzed, and taking its failure as the top event T of the fault tree01 (The top event is the fault tree at level 0, and the number of nodes is 1);

[0027] S22, search for all intermediate events or bottom events that lead to the top event based on the entity and the relationships of "belonging to the system component" and "causing system failure" (m1 is the number of fault tree nodes at level 1), determine the logical relationship between the above intermediate events or bottom events according to the category of the “causing system failure” relationship, as the logic gate type of the top event;

[0028] S23, for Repeat step S22 to find the intermediate event or bottom event that causes the above intermediate event to occur. (m2 is the number of fault tree nodes at the second level) and the logic gate types of all intermediate events until all intermediate events find the bottom events that cause them to occur. The last level of fault tree nodes is recorded as (m n is the number of fault tree nodes at level n);

[0029] S24, searching the fault knowledge base for the "failure probability" attribute of all entities corresponding to the bottom events as the probability parameter of the bottom events.

[0030] The step S3 specifically includes the following steps:

[0031] S31, generating the 0th layer node T of the fault tree according to the top event obtained by the knowledge parser 01 ;

[0032] S32, generating the ITE structure corresponding to T01 according to the top event logic gate type, intermediate event, bottom event and probability parameters of the bottom event obtained by the knowledge parser

[0033] S33, will The intermediate event in is marked as G 11 ,G 12 ,…G 1k According to the intermediate event logic gate type, intermediate event, bottom event and probability parameters of bottom event obtained by the knowledge parser, G is generated in sequence. 11 ,G 12 ,…G 1k Corresponding ITE structure Where i = 1, 2…, k, and t0 = 0,

[0034] S34, will The bottom event probability parameter in is converted into the failure probability of the bottom event;

[0035] S35 repeats steps S33 and S34 until all intermediate events are converted into corresponding ITE structures, and all bottom event probability parameters are converted into failure probabilities of bottom events;

[0036] S36 links the above ITE structures into a complete fault tree, including top events and their logic gate types, middle events and their logic gate types, bottom events and their failure probability information.

[0037] The step S36 specifically includes the following steps:

[0038] S361, record the s ITE structures as Where j = 1, 2…, s;

[0039] S362, first find the ITE structure corresponding to the top event Generate a fault tree subtree, G1 is the father node of the subtree, is the logic gate of the father node of the subtree, is the child node of the subtree;

[0040] S363, according to the method of hierarchical traversal, traverse in sequence For all intermediate events in G, for each intermediate event G j Find the corresponding ITE structure And generate a fault tree subtree according to the method in step S362, and replace the child nodes of the fault tree subtree in step S362 with the above fault tree subtree G j ;

[0041] S364, repeat step S363 until all intermediate events G j The fault tree subtrees generated by the corresponding I TE structure replace the G in the original fault tree. j until.

[0042] Another technical solution of the present invention to solve the above technical problem is as follows: a reliability automatic modeling device, characterized in that it includes a camera, a microphone, a large model server, a data storage device, and a computing processor;

[0043] The camera is used to obtain images of paper documents;

[0044] The microphone is used to obtain audio information carrier;

[0045] The large model server deploys the knowledge collection and generation module described in claim 1 to generate fault knowledge information;

[0046] The data storage device deploys the fault knowledge base module according to claim 1, and is used to store fault knowledge information;

[0047] The computing processor deploys the knowledge parsing module and the fault tree generating module as described in claim 1, and is used to parse the fault tree node information from the data storage device and generate a fault tree structure.

[0048] The computing processor is connected to the large model server and the data storage device via a self-organizing network.

[0049] The beneficial effects of the present invention are:

[0050] (1) The present invention can automatically establish a fault tree model of the object being analyzed, and can provide a quick modeling tool for reliability analysis of complex systems. The fault tree modeling is performed based on the hierarchical linking of the ITE structure, thereby improving the speed and reliability of modeling.

[0052] (2) By automatically building fault knowledge graphs in batches through large language models, not only can the workload of manual labeling be reduced, but also "new knowledge" beyond the scope of expert experience and cognition can be "emerged" through large language models, thereby improving the accuracy of the model.

[0053] (3) Based on the existing function of recognizing text information, by adding paper document image recognition and audio information carrier voice recognition equipment, the fault knowledge base is established and updated in a way that is closest to human natural language, which greatly broadens the source of fault knowledge and helps to further improve the accuracy of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A module block diagram of the reliability automatic modeling system provided by an embodiment of the present invention;

[0055] Figure 2 A schematic diagram of a process flow of an automatic reliability modeling method provided by an embodiment of the present invention;

[0056] Figure 3 A component composition diagram of the reliability automatic modeling device provided by an embodiment of the present invention;

[0057] Figure 4 A structural diagram of a fault knowledge graph provided by an embodiment of the present invention;

[0058] Figure 5 A fault tree diagram automatically constructed based on a knowledge graph provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0059] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention. Example

[0060] like Figure 1 As shown, a reliability automatic modeling system includes a knowledge acquisition and generation module, a fault knowledge base module, a knowledge analysis module, and a fault tree generation module:

[0061] The knowledge collection and generation module is used to extract fault knowledge from documents, web pages, audio and video carriers;

[0062] The fault knowledge base module is used to store fault knowledge in the form of knowledge graph entities, relationships and attributes;

[0063] The knowledge parsing module is used to identify the top event, the middle event and the bottom event of the fault tree from the fault knowledge base module, and the logical relationship between them;

[0064] The fault tree generation module is used to link the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

[0065] The knowledge acquisition and generation module includes a general large model, a model fine-tuning unit, and a retrieval generation unit;

[0066] The general large models include but are not limited to Llama series versions and Qwen series versions;

[0067] The model fine-tuning unit has the functions of setting retrieval and generation weight factors, selecting data sets, and self-evolving according to user input information;

[0068] The retrieval generation unit has the function of retrieving knowledge from the data storage device, and integrates and processes the retrieved content and the generated content according to the retrieval and generation weight factors set by the model fine-tuning unit.

[0069] The knowledge collection and generation module can automatically fill in knowledge and automatically update knowledge through a large model. The document types include TXT, XML, Word, EXCEL, PPT, and PDF. In addition to the unstructured data in the documents, web pages, and audio and video carriers, the knowledge source also supports directly importing fault knowledge from structured databases.

[0070] The fault knowledge base module adopts a graph database, including a data management system, a general knowledge graph, and a professional knowledge graph;

[0071] The data management system includes but is not limited to Neo4j, Amazon Neptune, TigerGraph, ArangoDB, NebulaGraph;

[0072] The general knowledge graph has the function of importing public graph data and the function of self-evolving according to user input information;

[0073] The professional knowledge graph has the function of importing user-defined data and the function of self-evolving according to user input information. Example

[0074] like Figure 2 As shown, a reliability automatic modeling method includes the following steps:

[0075] S1, firstly extract fault knowledge from documents, web pages, audio and video carriers through a large model, and store it in the fault knowledge base in the form of knowledge graph entities, relationships and attributes;

[0076] S2, then identifying the top event, the middle event and the bottom event of the fault tree, and the logical relationship between them from the fault knowledge base through the knowledge parser;

[0077] S3, finally, the fault tree generator links the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

[0078] The step S2 specifically includes the following steps:

[0079] S21, searching the corresponding entity from the fault knowledge base according to the name of the object to be analyzed, and taking its failure as the top event T of the fault tree 01 (The top event is the fault tree at level 0, and the number of nodes is 1);

[0080] S22, search for all intermediate events or bottom events that lead to the top event based on the entity and the relationships of "belonging to the system component" and "causing system failure" (m1 is the number of fault tree nodes at level 1), determine the logical relationship between the above intermediate events or bottom events according to the category of the “causing system failure” relationship, as the logic gate type of the top event;

[0081] S23, for Repeat step S22 to find the intermediate event or bottom event that causes the above intermediate event to occur. (m2 is the number of fault tree nodes at the second level) and the logic gate types of all intermediate events until all intermediate events find the bottom events that cause them to occur. The last level of fault tree nodes is recorded as (m n is the number of fault tree nodes at level n);

[0082] S24, searching the fault knowledge base for the "failure probability" attribute of all entities corresponding to the bottom events as the probability parameter of the bottom events.

[0083] The step S3 specifically includes the following steps:

[0084] S31, generating the 0th layer node T of the fault tree according to the top event obtained by the knowledge parser 01 ;

[0085] S32, generating the ITE structure corresponding to T01 according to the top event logic gate type, intermediate event, bottom event and probability parameters of the bottom event obtained by the knowledge parser

[0086] S33, will The intermediate event in is marked as G 11 ,G 12 ,…G 1k According to the intermediate event logic gate type, intermediate event, bottom event and probability parameters of bottom event obtained by the knowledge parser, G is generated in sequence. 11 ,G 12 ,…G 1k Corresponding ITE structure Where i = 1, 2…, k, and t0 = 0,

[0087] S34, will The bottom event probability parameter in is converted into the failure probability of the bottom event;

[0088] S35 repeats steps S33 and S34 until all intermediate events are converted into corresponding ITE structures, and all bottom event probability parameters are converted into failure probabilities of bottom events;

[0089] S36 links the above ITE structures into a complete fault tree, including top events and their logic gate types, middle events and their logic gate types, bottom events and their failure probability information.

[0090] The step S36 specifically includes the following steps:

[0091] S361, record the s ITE structures as Where j = 1, 2…, s;

[0092] S362, first find the ITE structure corresponding to the top event Generate a fault tree subtree, G1 is the father node of the subtree, is the logic gate of the father node of the subtree, is the child node of the subtree;

[0093] S363, according to the method of hierarchical traversal, traverse in sequence For all intermediate events in G, for each intermediate event G j Find the corresponding ITE structure And generate a fault tree subtree according to the method in step S362, and replace the child nodes of the fault tree subtree in step S362 with the above fault tree subtree G j ;

[0094] S364, repeat step S363 until all intermediate events G j The fault tree subtrees generated by the corresponding I TE structure replace the G in the original fault tree. j until.

[0095] In the above embodiment, some steps of the reliability automatic modeling method can be exchanged in order or have additional steps added. For example, S24 can be performed after S23 is fully executed, or it can be performed when S23 traverses each intermediate event. Because the executed tasks are relatively independent, adding one or more steps will not affect the final result, but will only affect the modeling efficiency. Example

[0096] A reliability automatic modeling device, characterized by comprising a camera, a microphone, a large model server, a data storage device, and a computing processor;

[0097] The camera is used to obtain images of paper documents;

[0098] The microphone is used to obtain audio information carrier;

[0099] The large model server deploys the knowledge collection and generation module described in claim 1 to generate fault knowledge information;

[0100] The data storage device deploys the fault knowledge base module according to claim 1, and is used to store fault knowledge information;

[0101] The computing processor deploys the knowledge parsing module and the fault tree generating module as described in claim 1, and is used to parse the fault tree node information from the data storage device and generate a fault tree structure.

[0102] The computing processor is connected to the large model server and the data storage device via a self-organizing network.

[0103] In the above embodiment, the camera in the reliability automatic modeling device can be replaced with a scanner, or a scanner can be added; the data storage device can be integrated into the large model server, as long as the large model server has a storage unit, but it increases the cost and is not conducive to expansion, and does not affect the function of the device. The main solution provides the most scalable solution; the general knowledge graph and professional knowledge graph on the data storage device can be merged, or divided into multiple finer general knowledge graph subgraphs and professional knowledge graph subgraphs.

[0104] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0105] It should be noted that, in this article, the terms "comprises", "includes" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or apparatus that includes a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or apparatus.

[0106] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0107] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0108] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0109] If the integrated unit is implemented in the form of 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, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0110] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A reliability automatic modeling system, characterized in that: Including knowledge acquisition and generation module, fault knowledge base module, knowledge analysis module, fault tree generation module: The knowledge collection and generation module is used to extract fault knowledge from documents, web pages, audio and video carriers; the fault knowledge base module is used to store fault knowledge in the form of knowledge graph entities, relationships and attributes; The knowledge parsing module is used to identify the top event, the middle event and the bottom event of the fault tree from the fault knowledge base module, and the logical relationship between them; The fault tree generation module is used to link the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

2. The reliability automatic modeling system according to claim 1, characterized in that: The knowledge acquisition and generation module includes a general large model, a model fine-tuning unit, and a retrieval generation unit; The general large models include but are not limited to Llama series versions and Qwen series versions; The model fine-tuning unit has the functions of setting retrieval and generation weight factors, selecting data sets, and self-evolving according to user input information; The retrieval generation unit has the function of retrieving knowledge from the data storage device, and integrates and processes the retrieved content and the generated content according to the retrieval and generation weight factors set by the model fine-tuning unit.

3. The reliability automatic modeling system according to claim 1 and claim 2, characterized in that: The knowledge collection and generation module can automatically fill in knowledge and automatically update knowledge through a large model. The document types include TXT, XML, Word, EXCEL, PPT, and PDF. In addition to the unstructured data in the documents, web pages, and audio and video carriers, the knowledge source also supports directly importing fault knowledge from structured databases.

4. The reliability automatic modeling system according to claim 1, characterized in that: The fault knowledge base module adopts a graph database, including a data management system, a general knowledge graph, and a professional knowledge graph; The data management system includes but is not limited to Neo4j, Amazon Neptune, TigerGraph, ArangoDB, NebulaGraph; The general knowledge graph has the function of importing public graph data and the function of self-evolving according to user input information; The professional knowledge graph has the function of importing user-defined data and the function of self-evolving according to user input information.

5. A reliability automatic modeling method, characterized in that: The steps include: S1, extracting fault knowledge from documents, web pages, audio and video carriers through a large model, and storing it in the fault knowledge base in the form of knowledge graph entities, relationships and attributes; S2, identifying the top event, the middle event and the bottom event of the fault tree, and the logical relationship between them from the fault knowledge base through the knowledge parser; S3, the fault tree generator links the identified top events, intermediate events and bottom events according to certain logical relationships to form a system fault tree model.

6. The reliability automatic modeling method according to claim 5, characterized in that: The step S2 specifically includes the following steps: S21, searching the corresponding entity from the fault knowledge base according to the name of the object to be analyzed, and taking its failure as the top event T of the fault tree 01 (The top event is the fault tree at level 0, and the number of nodes is 1); S22, based on the entity and the "belongs to the system component" and "causes system failure" relationships, jointly search for all intermediate events or bottom events that lead to the top event (m1 is the number of fault tree nodes at level 1), determine the logical relationship between the above intermediate events or bottom events according to the category of the "causing system failure" relationship, as the logic gate type of the top event; S23, for Repeat step S22 to find the intermediate event or bottom event that causes the above intermediate event to occur. (m2 is the number of fault tree nodes at the second level) and the logic gate types of all intermediate events until all intermediate events find the bottom events that cause them to occur. The last level of fault tree nodes is recorded as (m n is the number of fault tree nodes at level n); S24, searching the fault knowledge base for the "failure probability" attribute of all entities corresponding to the bottom events as the probability parameter of the bottom events.

7. The reliability automatic modeling method according to claim 5, characterized in that: The step S3 specifically includes the following steps: S31, generating the 0th layer node T of the fault tree according to the top event obtained by the knowledge parser 01 ; S32, generating the ITE structure corresponding to T01 according to the top event logic gate type, intermediate event, bottom event and probability parameters of the bottom event obtained by the knowledge parser S33, will The intermediate event in is marked as G 11 ,G 12 ,…G 1k According to the intermediate event logic gate type, intermediate event, bottom event and probability parameters of bottom event obtained by the knowledge parser, G is generated in sequence. 11 ,G 12 ,…G 1k Corresponding ITE structure Where i = 1, 2…, k, and t0 = 0, S34, will The bottom event probability parameter in is converted into the failure probability of the bottom event; S35 repeats steps S33 and S34 until all intermediate events are converted into corresponding ITE structures, and all bottom event probability parameters are converted into failure probabilities of bottom events; S36 links the above ITE structures into a complete fault tree, including top events and their logic gate types, middle events and their logic gate types, bottom events and their failure probability information.

8. The reliability automatic modeling method according to claim 7, characterized in that: The step S36 specifically includes the following steps: S361, record the s ITE structures as S362, first find the ITE structure corresponding to the top event Generate a fault tree subtree, G1 is the father node of the subtree, is the logic gate of the father node of the subtree, is the child node of the subtree; S363, according to the method of hierarchical traversal, traverse in sequence For all intermediate events in G, for each intermediate event G j Find the corresponding ITE structure And generate a fault tree subtree according to the method in step S362, and replace the child nodes of the fault tree subtree in step S362 with the above fault tree subtree G in j ; S364, repeat step S363 until all intermediate events G j The corresponding fault tree subtrees generated by the ITE structure replace the G in the original fault tree. j until.

9. A reliability automatic modeling device, characterized in that: It includes cameras, microphones, large model servers, data storage, and computing processors; The camera is used to obtain images of paper documents; The microphone is used to obtain audio information carrier; The large model server deploys the knowledge collection and generation module described in claim 1 to generate fault knowledge information; The data storage device deploys the fault knowledge base module described in claim 1 to store fault knowledge information; the computing processor deploys the knowledge parsing module and fault tree generation module described in claim 1 to parse fault tree node information from the data storage device and generate a fault tree structure.

10. The reliability automatic modeling device according to claim 9, characterized in that: The computing processor is connected to the large model server and the data storage device via a self-organizing network.

Citation Information

Cited By

  • Complex production line reliability analysis method based on combination of fault operation and maintenance knowledge graph and fault tree

    CN119903430A

  • Large model courseware automatic generation method based on LangChain framework

    CN120951977A

  • Fault tree intelligent modeling method and device based on MBSE-LLM

    CN121809543A

  • Fault tree intelligent modeling method and device based on mbse-llm

    CN121809543B