Subway communication professional equipment fault auxiliary diagnosis system based on knowledge graph

Abstract

The invention discloses a subway communication professional equipment fault auxiliary diagnosis system based on a knowledge graph, and is applied to the technical field of rail transit intelligent operation and maintenance. Comprising a knowledge extraction module, a knowledge storage module, a fault acquisition module, a fault diagnosis module and a human-computer interaction interface module, the knowledge extraction module extracts knowledge required by fault diagnosis from the database; the knowledge storage module stores the knowledge extracted by the knowledge extraction module into a graph database; the fault alarm acquisition module automatically acquires an alarm reported after the equipment side has a fault in real time through a network mode; the fault diagnosis module is used for generating troubleshooting steps, fault reasons and solution measure information according to the fault organization; and the man-machine interaction interface module provides an interface for a user to check fault information conveniently. According to the method, intelligent and automatic equipment fault auxiliary diagnosis of the subway communication professional equipment is realized by constructing the knowledge graph, and the diagnosis efficiency and accuracy are improved.

Description

Technical Field

[0001] The present invention relates to the field of intelligent operation and maintenance technology for rail transit, and more specifically to a knowledge graph-based auxiliary diagnosis system for subway communication professional equipment faults. Background Art

[0002] With the rapid development of subway transportation, subway communication systems play an important role in ensuring the safety and punctuality of trains. However, subway communication equipment is of various types, complex in structure, and frequently fails. In the traditional subway communication equipment fault diagnosis, professional technicians are usually relied on to manually inspect and diagnose the equipment. This method is time-consuming, labor-intensive, and has subjective problems, and often has low diagnostic efficiency and low accuracy. Therefore, how to provide a subway communication equipment fault auxiliary diagnosis system based on knowledge graph is a problem that technicians in this field urgently need to solve. Summary of the invention

[0003] In view of this, the present invention provides a knowledge graph-based auxiliary fault diagnosis system for subway communication professional equipment to realize intelligent and automated auxiliary fault diagnosis of subway communication professional equipment.

[0004] In order to achieve the above object, the present invention provides the following technical solutions:

[0005] A knowledge graph-based subway communication equipment fault auxiliary diagnosis system, including a knowledge extraction module, a knowledge storage module, a fault collection module, a fault diagnosis module and a human-computer interaction interface module;

[0006] The knowledge extraction module is used to extract the knowledge required for fault diagnosis from the database;

[0007] The knowledge storage module includes a graph database program and a database writing program, which is used to store the knowledge extracted by the knowledge extraction module into the graph database;

[0008] The fault alarm collection module collects alarms reported by the equipment side in real time and automatically through the network.

[0009] The fault diagnosis module is used to generate troubleshooting steps, fault causes, and solution information based on the fault organization;

[0010] The human-computer interaction interface module provides an interface for users to view fault information.

[0011] Optionally, the database includes equipment information lists, fault work order data, and ledgers provided by the subway operating unit, equipment maintenance manuals provided by equipment manufacturers, and operation and maintenance experience documents provided by experts.

[0012] Optionally, write big model prompt words according to the characteristics of the document, use the big model to extract knowledge from the extracted text information, increase manual review to ensure the accuracy of the data, and put it into the warehouse after confirmation. The system also supports manual configuration of some fault causes to achieve automatic detection by associating collectible indicators.

[0013] Optionally, the knowledge storage module stores the knowledge in a graph database based on the knowledge graph schema. Specifically: for each piece of fault knowledge data, the knowledge storage module uses the SentenceTransformers library to separately calculate the embedding vector for each node in the knowledge graph schema, and uses the embedding vector to calculate the cosine similarity of the fields in the document as the semantic text similarity, and merges the nodes in the existing nodes whose semantic text similarity is greater than a preset threshold with the knowledge data.

[0014] Optionally, the fault collection module receives device fault information in real time through a standard protocol or a manufacturer's private protocol.

[0015] Optionally, the user triggers the diagnosis by clicking a diagnosis operation in the question-and-answer mode or on the fault page on the human-computer interaction interface module.

[0016] Optionally, the user clicks a diagnosis operation on the fault page, and the fault diagnosis module performs fault diagnosis, specifically:

[0017] The fault collection module sends the fault alarm information and related basic information selected by the user to the fault diagnosis module, including line, site, equipment, alarm code, and subsystem data;

[0018] The fault diagnosis module searches for the corresponding device node and device type node in the graph database based on the basic information; matches the alarm type entity based on the device information and fault information;

[0019] Find the relevant fault cause entity according to the alarm type, locate the corresponding fault node according to the fault cause entity, add it to the fault node list, sort the fault nodes according to the occurrence rate attribute, and mark the node with the highest occurrence rate as the most common fault;

[0020] Arrange the list of fault nodes and find out the troubleshooting steps and solutions for the corresponding fault causes;

[0021] The above information is displayed in the order of fault nodes on the human-computer interaction interface, and the user performs positioning operations based on the information displayed on the page;

[0022] The system provides feedback on whether the actual fault has been resolved through the solution, and the statistical occurrence rate provides data support for subsequent diagnosis.

[0023] Optionally, the user triggers diagnosis through the question-and-answer mode on the fault page, and the fault diagnosis module performs fault diagnosis, specifically:

[0024] The user inputs the fault diagnosis question he wants to ask on the human-computer interaction interface;

[0025] After receiving the request, the fault diagnosis module processes the input question text and identifies whether the question intent is a fault diagnosis question by combining keyword recognition and pattern matching. If so, it proceeds to the next step;

[0026] Extract the subject in the fault diagnosis problem, calculate the similarity with the embedded vector calculated by the knowledge access module to determine whether it contains specific equipment information and alarm information. If so, perform fault diagnosis and query the corresponding fault cause list;

[0027] If not, determine whether there is fault cause information in the fault diagnosis question, if yes, continue, if not, prompt the user;

[0028] Based on the fault cause information obtained through analysis, find out whether there are fault causes for the entire station, entire line, or entire subsystem, and add them to the fault cause list;

[0029] According to the list of fault causes found, the corresponding troubleshooting steps and solutions are listed, and the results are output to the human-computer interaction interface.

[0030] It can be seen from the above technical solution that compared with the prior art, the present invention provides a subway communication professional equipment fault auxiliary diagnosis system based on knowledge graph, which realizes the intelligent and automated equipment fault auxiliary diagnosis of subway communication professional equipment by constructing a knowledge graph, improves diagnosis efficiency and accuracy, and reduces manpower consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0032] Figure 1 It is a schematic diagram of the principle of the fault auxiliary diagnosis system of the present invention;

[0033] Figure 2 It is a schematic diagram of the knowledge graph of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] The embodiment of the present invention discloses a knowledge graph-based auxiliary diagnosis system for subway communication professional equipment faults, such as Figure 1 As shown, it includes a knowledge extraction module, a knowledge storage module, a fault collection module, a fault diagnosis module and a human-computer interaction interface module;

[0036] The knowledge extraction module is used to extract the knowledge required for fault diagnosis from the database;

[0037] The knowledge storage module includes a graph database program and a database writing program, which is used to store the knowledge extracted by the knowledge extraction module into the graph database;

[0038] The fault alarm collection module collects alarms reported by the equipment side in real time and automatically through the network.

[0039] The fault diagnosis module is used to generate troubleshooting steps, fault causes, and solution information based on the fault organization;

[0040] The human-computer interaction interface module provides an interface for users to view fault information.

[0041] In an embodiment of the present invention, the human-computer interaction interface can be used to view the fault list, display the troubleshooting steps and solutions, and visualize the knowledge graph. It can update knowledge to the system and provide lifecycle management of historical faults, generate fault reports and other equipment operation and maintenance related information functions.

[0042] Furthermore, the database includes equipment information lists, fault work order data, and ledgers provided by subway operating units, equipment maintenance manuals provided by equipment manufacturers, and operation and maintenance experience documents provided by experts.

[0043] Furthermore, large model prompt words are written according to the characteristics of the document, and the large model is used to extract knowledge from the extracted text information, and manual review is added to ensure the accuracy of the data. After confirmation, the data is put into the warehouse. The system also supports manual configuration of some fault causes to achieve automatic detection by associating collectible indicators. For example, the cause of the downtime may be due to memory overflow.

[0044] In the embodiment of the present invention, the prompt words extracted from the existing materials are as follows:

[0045] You are a smart Chinese language assistant. You can judge the fault phenomenon, cause and solution in a sentence according to the logic of the sentence. For example, the video playback of the security platform failed because the Gbvideo program crashed and needed to be restarted. Here, the video playback failure of the security platform is the fault phenomenon, the Gbvideo program crash is the cause, and the solution is to restart Gbvideo. Please help me analyze the fault phenomenon, cause and solution in the following sentence.

[0046] If a sentence does not contain these elements, then the corresponding position in the result will be displayed as None. Do not include irrelevant explanations and descriptions in the result. For example, the failure of the security platform video playback is due to the Gbvideo program crash. Here, the failure of the security platform video playback is the fault phenomenon, the Gbvideo program crash is the cause, and the solution is None.

[0047] Furthermore, the knowledge storage module stores the knowledge in the graph database based on the knowledge graph schema. Specifically, for each piece of fault knowledge data, the knowledge storage module uses the SentenceTransformers library to calculate the embedding vector for each node in the knowledge graph schema, uses the embedding vector to calculate the cosine similarity of the fields in the document as the semantic text similarity, and merges the nodes whose semantic text similarity is greater than the preset threshold with the knowledge data. Figure 2 shown.

[0048] Furthermore, the fault collection module receives equipment fault information in real time through a standard protocol or a manufacturer's private protocol.

[0049] Furthermore, the user triggers the diagnosis by clicking a diagnosis operation in the question-and-answer mode or on the fault page on the human-computer interaction interface module.

[0050] Further, the user clicks the diagnosis operation on the fault page, and the fault diagnosis module performs fault diagnosis, specifically:

[0051] The fault collection module sends the fault alarm information and related basic information selected by the user to the fault diagnosis module, including line, site, equipment, alarm code, and subsystem data;

[0052] The fault diagnosis module searches for the corresponding device node and device type node in the graph database based on the basic information; matches the alarm type entity based on the device information and fault information;

[0053] Find the relevant fault cause entity according to the alarm type, locate the corresponding fault node according to the fault cause entity, add it to the fault node list, sort the fault nodes according to the occurrence rate attribute, and mark the node with the highest occurrence rate as the most common fault;

[0054] In an embodiment of the present invention, since multiple causes may be matched and the positioning operation is relatively cumbersome, a search is performed based on the fault scope and the parent device of the device to see whether there are any related faults, such as whether there is a fault that affects the entire site, or whether there is a fault that affects the entire subsystem. If there are related faults, the related faults are added to the front of the fault node list and marked as the most likely fault.

[0055] Arrange the list of fault nodes and find out the troubleshooting steps and solutions for the corresponding fault causes;

[0056] The above information is displayed in the order of fault nodes on the human-computer interaction interface, and the user performs positioning operations based on the information displayed on the page;

[0057] The system provides feedback on whether the actual fault has been resolved through the solution, and the statistical occurrence rate provides data support for subsequent diagnosis.

[0058] Furthermore, the user triggers diagnosis through the question-and-answer mode on the fault page, and the fault diagnosis module performs fault diagnosis, specifically:

[0059] The user inputs the fault diagnosis question he wants to ask on the human-computer interaction interface;

[0060] After receiving the request, the fault diagnosis module processes the input question text and identifies whether the question intent is a fault diagnosis question by combining keyword recognition and pattern matching. If so, it proceeds to the next step;

[0061] In the embodiment of the present invention, specifically, the fault diagnosis module defines semantic slots, breaks down the questions input by the user, and matches them with the semantic slots, mainly to verify keywords and entity objects.

[0062] Extract the subject in the fault diagnosis problem, calculate the similarity with the embedded vector calculated by the knowledge access module to determine whether it contains specific equipment information and alarm information. If so, perform fault diagnosis and query the corresponding fault cause list;

[0063] If not, determine whether there is fault cause information in the fault diagnosis question, if yes, continue, if not, prompt the user;

[0064] Based on the fault cause information obtained through analysis, find out whether there are fault causes for the entire station, entire line, or entire subsystem, and add them to the fault cause list;

[0065] According to the list of fault causes found, the corresponding troubleshooting steps and solutions are listed, and the results are output to the human-computer interaction interface.

[0066] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0067] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A knowledge graph-based auxiliary diagnosis system for subway communication professional equipment faults, characterized in that: It includes knowledge extraction module, knowledge storage module, fault collection module, fault diagnosis module and human-computer interaction interface module; The knowledge extraction module is used to extract the knowledge required for fault diagnosis from the database; The knowledge storage module includes a graph database program and a database writing program, which is used to store the knowledge extracted by the knowledge extraction module into the graph database; The fault alarm collection module collects alarms reported by the equipment side in real time and automatically through the network. The fault diagnosis module is used to generate troubleshooting steps, fault causes, and solution information based on the fault organization; The human-computer interaction interface module provides an interface for users to view fault information.

2. According to the knowledge graph-based subway communication equipment fault auxiliary diagnosis system of claim 1, it is characterized in that: The database includes equipment information lists, fault work order data, and ledgers provided by subway operating units, equipment maintenance manuals provided by equipment manufacturers, and operation and maintenance experience documents provided by experts.

3. According to the knowledge graph-based subway communication equipment fault auxiliary diagnosis system of claim 1, it is characterized in that: Write big model prompt words according to the characteristics of the document, use the big model to extract knowledge from the extracted text information, increase manual review to ensure the accuracy of the data, and put it into the warehouse after confirmation. The system also supports manual configuration of some fault causes to achieve automatic detection by associating collectible indicators.

4. According to claim 1, a knowledge graph-based subway communication equipment fault auxiliary diagnosis system is characterized in that: The knowledge storage module stores knowledge in the graph database based on the knowledge graph schema. Specifically: for each piece of fault knowledge data, the knowledge storage module uses the SentenceTransformers library to calculate the respective embedding vectors for each node in the knowledge graph schema, and uses the embedding vectors to calculate the cosine similarity of the fields in the document as the semantic text similarity, and merges the nodes whose semantic text similarity in the existing nodes is greater than the preset threshold with the knowledge data.

5. According to the knowledge graph-based subway communication equipment fault auxiliary diagnosis system of claim 1, it is characterized in that: The fault collection module receives equipment fault information in real time through standard protocols or manufacturer's private protocols.

6. According to the knowledge graph-based subway communication equipment fault auxiliary diagnosis system of claim 1, it is characterized in that: The user triggers the diagnosis by clicking the diagnosis operation in the question-and-answer mode or on the fault page on the human-computer interaction interface module.

7. According to claim 6, a knowledge graph-based subway communication equipment fault auxiliary diagnosis system is characterized in that: When the user clicks the diagnostic operation on the fault page, the fault diagnosis module performs fault diagnosis, specifically: The fault collection module sends the fault alarm information and related basic information selected by the user to the fault diagnosis module, including line, site, equipment, alarm code, and subsystem data; The fault diagnosis module searches for the corresponding device node and device type node in the graph database based on the basic information; matches the alarm type entity based on the device information and fault information; Find the relevant fault cause entity according to the alarm type, locate the corresponding fault node according to the fault cause entity, add it to the fault node list, sort the fault nodes according to the occurrence rate attribute, and mark the node with the highest occurrence rate as the most common fault; Arrange the list of fault nodes and find out the troubleshooting steps and solutions for the corresponding fault causes; The above information is displayed in the order of fault nodes on the human-computer interaction interface, and the user performs positioning operations based on the information displayed on the page; The system provides feedback on whether the actual fault has been resolved through the solution, and the statistical occurrence rate provides data support for subsequent diagnosis.

8. According to claim 7, a knowledge graph-based subway communication professional equipment fault auxiliary diagnosis system is characterized in that: The user triggers diagnosis through the question-and-answer mode on the fault page, and the fault diagnosis module performs fault diagnosis, specifically: The user inputs the fault diagnosis question he wants to ask on the human-computer interaction interface; After receiving the request, the fault diagnosis module processes the input question text and identifies whether the question intent is a fault diagnosis question by combining keyword recognition and pattern matching. If so, it proceeds to the next step; Extract the subject in the fault diagnosis problem, calculate the similarity with the embedded vector calculated by the knowledge access module to determine whether it contains specific equipment information and alarm information. If so, perform fault diagnosis and query the corresponding fault cause list; If not, determine whether there is fault cause information in the fault diagnosis question, if yes, continue, if not, prompt the user; Based on the fault cause information obtained through analysis, find out whether there are fault causes for the entire station, entire line, or entire subsystem, and add them to the fault cause list; According to the list of fault causes found, the corresponding troubleshooting steps and solutions are listed, and the results are output to the human-computer interaction interface.

Images