Fault judgment system, method and equipment based on knowledge graph and medium

Abstract

The invention relates to a fault judgment system, method and device based on a knowledge graph and a medium, and belongs to the technical field of reliability analysis, and the system comprises a data collection module which is used for obtaining to-be-diagnosed information, the to-be-diagnosed information comprises a device type, a working condition and real-time monitoring data of a target device, the real-time monitoring data comprises operation parameters and state information; the fault diagnosis reasoning module is used for inputting the to-be-diagnosed information into a pre-constructed knowledge graph and performing fault diagnosis on the to-be-diagnosed information by adopting a graph reasoning algorithm to obtain a reasoning result of the target equipment, and the reasoning result comprises a fault type, a fault reason and an influence factor of the target equipment; and the diagnosis result analysis module is used for evaluating the reasoning result to obtain an evaluation result, displaying the reasoning result and the evaluation result based on a visual tool, and generating a fault diagnosis report. According to the method, the accuracy of fault judgment is effectively improved, and misjudgment caused by insufficient experience in a traditional method is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of reliability analysis, and particularly to a fault judgment system, method, device and medium based on a knowledge graph. Background Art

[0002] With the rapid development of information technology, Fault Tree Analysis (FTA), as an effective reliability analysis tool, is widely used in fields such as engineering, manufacturing, aerospace, etc. A fault tree is a graphical logical model used to analyze the causes and effects of system failures.

[0003] Traditional fault tree generation methods mostly rely on expert experience and manual construction. Experts need to determine the most undesirable fault event as the top event according to the characteristics and requirements of the system, and then gradually analyze the direct and indirect causes leading to its occurrence to construct a fault tree.

[0004] Although this method can reflect the fault modes of the system to a certain extent, it has disadvantages such as time-consuming, low efficiency, and being easily affected by subjective factors. In addition, for complex large-scale systems, the fault tree may be very large and complex, making it difficult to construct and analyze, and there may be errors due to the overly complex model. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a fault judgment system, method, device and medium based on a knowledge graph, aiming to solve at least one of the above technical problems.

[0006] The technical solution of the present invention to solve the above technical problems is as follows:

[0007] In a first aspect, the present application provides a fault judgment system based on a knowledge graph, adopting the following technical solution:

[0008] A fault judgment system based on a knowledge graph, comprising.

[0009] A data acquisition module, configured to obtain information to be diagnosed, where the information to be diagnosed includes the device type, working conditions, and real-time monitoring data of the target device, and the real-time monitoring data includes operation parameters and status information;

[0010] A fault diagnosis reasoning module, which is used to input the information to be diagnosed into a pre-constructed knowledge graph, and use a graph reasoning algorithm to perform fault diagnosis on the information to be diagnosed to obtain the reasoning result of the target device. The pre-constructed knowledge graph includes multiple nodes corresponding to the target device, and each of the nodes is connected by an edge. The multiple nodes include a fault mode node, a fault cause node, and a fault impact node. Each edge represents the corresponding relationship between the corresponding nodes. The reasoning result includes the fault type, fault cause, and influencing factors of the target device;

[0011] A diagnosis result analysis module, which is used to evaluate the reasoning result to obtain an evaluation result, and display the reasoning result and the evaluation result based on a visualization tool to generate a fault diagnosis report.

[0012] The beneficial effects of the present invention are as follows: The system can comprehensively collect and integrate information such as the device type, working conditions, and real-time monitoring data of the target device, and input it into a pre-constructed knowledge graph. With the help of the graph reasoning algorithm, the system can efficiently perform fault diagnosis on this information, quickly obtain the fault type, fault cause, and influencing factors of the target device. Further, the diagnosis result analysis module evaluates the reasoning result, generates a detailed fault diagnosis report and displays it through a visualization tool, greatly improving the accuracy and efficiency of fault diagnosis. This process not only reduces the need for manual intervention but also significantly improves the flexibility and adaptability of fault diagnosis, and is particularly suitable for fault management and decision support of complex systems.

[0013] On the basis of the above technical solutions, the present invention can be further improved as follows.

[0014] Further, the data acquisition module is also used to obtain the device basic information and historical fault records of the target device. The historical fault records include multiple historical faults and the corresponding maintenance records and fault causes for each historical fault;

[0015] The system further includes:

[0016] A data preprocessing module, which is used to perform preprocessing operations on the device basic information and the historical fault records respectively to obtain the target device basic information and the target historical fault records;

[0017] A knowledge graph construction module, which is used to construct a knowledge graph based on the target device basic information and the target historical fault records by using data mining technology and natural language processing technology.

[0018] The beneficial effects of adopting the above further solution are as follows: The data acquisition module can not only obtain the real-time monitoring data of the target device, but also obtain the basic information of the device and historical fault records. This enables the system to comprehensively understand the operating status of the device and past maintenance situations, providing detailed data support for subsequent fault diagnosis. The data preprocessing module preprocesses the basic information of the device and historical fault records to ensure the quality and consistency of the data, further improving the accuracy of fault diagnosis. The knowledge graph construction module uses data mining technology and natural language processing technology to structure the basic information of the device and historical fault records, forming a knowledge graph containing nodes such as fault modes, fault causes, and fault impacts, providing rich background knowledge and reasoning basis for fault diagnosis.

[0019] Further, the knowledge graph construction module is specifically used for:

[0020] Based on the basic information of the target device, the target historical fault records, and natural language processing technology, determine multiple entities, each of which is a fault mode, a fault cause, or a fault impact;

[0021] Based on data mining technology, determine the corresponding relationships between each of the entities, and the corresponding relationships are causal relationships or dependency relationships;

[0022] Based on the corresponding relationships between each of the entities, construct a knowledge graph, with each entity as a node and each of the corresponding relationships as an edge connecting the nodes, and store the knowledge graph in a graph database.

[0023] The beneficial effects of adopting the above further solution are: When constructing the knowledge graph, the present invention can make full use of the basic information of the target device and historical fault records, combine natural language processing technology, accurately identify multiple entities, and use data mining technology to determine the causal relationships or dependency relationships between each entity. The knowledge graph constructed in this way not only covers the static information of the device, but also contains dynamic fault modes and their causes, forming a structured and semantically rich knowledge system. Further, storing the constructed knowledge graph in a graph database ensures the efficient access and rapid retrieval of the knowledge graph, providing a solid data foundation and technical support for subsequent fault diagnosis.

[0024] Further, the system further includes:

[0025] A feedback optimization module, used to obtain user feedback information and fault records corresponding to the fault diagnosis report;

[0026] Based on the embedded knowledge graph completion algorithm TransR and the feedback information, update the knowledge graph.

[0027] The beneficial effects of adopting the above further solution are as follows: The feedback optimization module can obtain the feedback information of users on the fault diagnosis report and the latest fault records in real time. By using the embedded knowledge graph completion algorithm TransR and combining this feedback information and fault records, it dynamically updates the node and edge relationships in the knowledge graph. This not only improves the accuracy and integrity of the knowledge graph, but also enhances the self-learning ability and adaptability of the system, and further improves the accuracy and reliability of fault diagnosis.

[0028] Further, the fault diagnosis and reasoning module is specifically used for:

[0029] Performing entity recognition on the text in the information to be diagnosed based on the LSTM algorithm to obtain multiple target entities of the target device;

[0030] Based on the search algorithm, searching for the nodes and corresponding relationships corresponding to each of the target entities in the pre-constructed knowledge graph;

[0031] Constructing a fault tree based on the nodes and corresponding relationships corresponding to each of the target entities;

[0032] Based on the fault tree and the logical reasoning algorithm, determining the fault type and fault cause of the target device;

[0033] Based on the corresponding relationship of the fault cause of the target device in the knowledge graph, performing influence factor reasoning to obtain the influence factors of the target device;

[0034] Based on the fault type, fault cause and influence factors of the target device, determining the reasoning result of the target device.

[0035] The beneficial effects of adopting the above further solution are as follows: Performing entity recognition on the text in the information to be diagnosed based on the LSTM algorithm can accurately extract the key fault-related information of the target device; By searching for the nodes and corresponding relationships corresponding to each target entity in the pre-constructed knowledge graph, the fault mode and cause can be quickly located, improving the speed and accuracy of fault diagnosis; Constructing a fault tree based on the nodes and corresponding relationships corresponding to each target entity can intuitively display the logical relationship of the fault, facilitating understanding and analysis; By determining the fault type and cause of the target device based on the fault tree and the logical reasoning algorithm, the accuracy of fault diagnosis is further enhanced; Performing influence factor reasoning based on the corresponding relationship of the fault cause of the target device in the knowledge graph can comprehensively evaluate the influence degree of the fault, providing a scientific basis for subsequent repair and maintenance.

[0036] Further, the process of the fault diagnosis and reasoning module constructing a fault tree based on the nodes and corresponding relationships of the target entity includes:

[0037] Construct a fault tree based on the nodes, corresponding relationships, and graph theory algorithms of the target entity.

[0038] The beneficial effect of adopting the above further solution is that constructing a fault tree based on the nodes, corresponding relationships, and graph theory algorithms of the target entity can accurately reflect the logical relationships between various fault modes inside the device, thereby improving the accuracy and reliability of fault diagnosis.

[0039] Further, the data acquisition module includes:

[0040] Device sensors for obtaining the operating parameters and status information of the target device in real time;

[0041] An operation log module for obtaining the maintenance records and fault causes corresponding to each historical fault of the target device;

[0042] An alarm module for obtaining the historical fault records during the operation of the target device;

[0043] A data port module for obtaining the basic device information of the target device.

[0044] The beneficial effect of adopting the above further solution is that device sensors can obtain the operating parameters and status information of the target device in real time, ensuring that the basic data required for fault diagnosis is comprehensive and timely. The operation log module records the maintenance records and fault causes of each historical fault, providing an important basis for the identification and reasoning of fault modes. The alarm module captures the historical faults during the operation of the device, helping to quickly respond to and handle emergencies. The data port module obtains the basic device information, ensuring the accuracy and integrity of the knowledge graph construction. In summary, this solution significantly improves the accuracy and real-time performance of fault diagnosis, enhancing the reliability and maintenance efficiency of the system.

[0045] In a second aspect, the present application provides a fault judgment method based on a knowledge graph, adopting the following technical solution:

[0046] A fault judgment method based on a knowledge graph includes:

[0047] Obtain the information to be diagnosed, where the information to be diagnosed includes the device type, working conditions, and real-time monitoring data of the target device, and the real-time monitoring data includes operating parameters and status information;

[0048] Input the to-be-diagnosed information into a pre-constructed knowledge graph, and use a graph reasoning algorithm to perform fault diagnosis on the to-be-diagnosed information to obtain the reasoning result of the target device. The pre-constructed knowledge graph includes multiple nodes corresponding to the target device, and each of the nodes is connected by edges. The multiple nodes include a fault mode node, a fault cause node, and a fault impact node. Each of the edges represents the corresponding relationship between the corresponding nodes. The reasoning result includes the fault type, fault cause, and influencing factors of the target device;

[0049] Evaluate the reasoning result to obtain an evaluation result, and display the reasoning result and the evaluation result based on a visualization tool to generate a fault diagnosis report.

[0050] In a third aspect, the present application provides an electronic device, adopting the following technical solution:

[0051] An electronic device includes a memory and a processor. A computer program capable of being loaded and executed by the processor for the method of a fault judgment system based on a knowledge graph described in the second aspect is stored on the memory.

[0052] In a fourth aspect, the present application provides a computer-readable storage medium, adopting the following technical solution:

[0053] A computer-readable storage medium stores a computer program capable of being loaded and executed by the processor for any one of the methods of a fault judgment based on a knowledge graph described in the second aspect.

[0054] Additional aspects and advantages of the present application will be given in part in the following description, which will become apparent from the following description, or will be understood through the practice of the present application. Description of the Drawings

[0055] Figure 1 It is a schematic structural diagram of a fault judgment system based on a knowledge graph provided by an embodiment of the present invention;

[0056] Figure 2 It is a schematic composition diagram of a knowledge graph provided by an embodiment of the present invention;

[0057] Figure 3 It is a schematic flowchart of a fault judgment method based on a knowledge graph provided by an embodiment of the present invention;

[0058] Figure 4 It is an industrial pump failure fault tree instance diagram provided by an embodiment of the present invention.

[0059] Figure 5 It is a wind turbine fault tree instance diagram provided by an embodiment of the present invention.

[0060] Figure 6 Schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. Specific implementation manners

[0061] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0062] In addition, the term "and / or" herein is only an associative relationship describing associated objects, indicating that three relationships may exist. For example, A and / or B may represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after unless otherwise specified.

[0063] As Figure 1 shown, an embodiment of the present application provides a fault judgment system based on a knowledge graph, including:

[0064] A data acquisition module, configured to obtain information to be diagnosed, where the information to be diagnosed includes the device type, working conditions, and real-time monitoring data of the target device, and the real-time monitoring data includes operating parameters and status information;

[0065] A fault diagnosis and reasoning module, configured to input the information to be diagnosed into a pre-constructed knowledge graph, and perform fault diagnosis on the information to be diagnosed by using a graph reasoning algorithm to obtain an inference result of the target device. The pre-constructed knowledge graph includes multiple nodes corresponding to the target device, and the nodes are connected by edges. The multiple nodes include a fault mode node, a fault cause node, and a fault impact node. Each edge represents the corresponding relationship between the corresponding nodes. The inference result includes the fault type, fault cause, and influencing factors of the target device;

[0066] A diagnosis result analysis module, configured to evaluate the inference result to obtain an evaluation result, and display the inference result and the evaluation result based on a visualization tool to generate a fault diagnosis report.

[0067] The system can comprehensively collect and integrate information such as the device type, working conditions, and real-time monitoring data of the target device, and input it into a pre-constructed knowledge graph. With the graph reasoning algorithm, the system can efficiently perform fault diagnosis on this information, quickly obtain the fault type, fault cause, and its influencing factors of the target device. Further, the diagnosis result analysis module evaluates the reasoning result, generates a detailed fault diagnosis report, and displays it through a visualization tool, greatly improving the accuracy and efficiency of fault diagnosis. This process not only reduces the need for manual intervention but also significantly enhances the flexibility and adaptability of fault diagnosis, especially suitable for fault management and decision support of complex systems.

[0068] In the embodiments of the present application, the user can input the device type, working conditions, operating parameters, etc. from the graphical interface of the system, and the data acquisition module obtains the information to be diagnosed. It can also collect the information to be diagnosed from sensors, alarm modules, etc., and is not limited thereto.

[0069] In the embodiments of the present application, the fault diagnosis reasoning module is specifically used for:

[0070] Performing entity recognition on the text in the information to be diagnosed based on the LSTM algorithm to obtain multiple target entities of the target device, where each target entity is a fault mode, a fault cause, or a fault impact;

[0071] Searching for the nodes and corresponding relationships of each target entity in a pre-constructed knowledge graph based on a search algorithm;

[0072] Constructing a fault tree based on the nodes and corresponding relationships of each target entity;

[0073] Determining the fault type and fault cause of the target device based on the fault tree and a logical reasoning algorithm;

[0074] Performing influence factor reasoning based on the corresponding relationship of the fault cause of the target device in the knowledge graph to obtain the influence factors of the target device;

[0075] Determining the reasoning result of the target device based on the fault type, fault cause, and influence factors of the target device.

[0076] In the above embodiments, a large number of labeled data sets are used to train the LSTM model so that it can recognize named entities in the text, such as device names, component names, fault types, etc.

[0077] First, preprocess the information to be diagnosed, and input the preprocessed information to be diagnosed into the LSTM model for entity recognition, so as to obtain multiple target entities of the target device.

[0078] After that, according to the characteristics of the knowledge graph and the search requirements, select a search algorithm, such as depth-first search (DFS), breadth-first search (BFS), or a graph database query language (such as Cypher), etc., to search for the nodes corresponding to each target entity in the knowledge graph and find the corresponding relationships between these nodes.

[0079] After that, use the nodes corresponding to the target entities as failure modes, and match their corresponding relationships with the logic gates of the fault tree. Based on the matching mode, find the "AND gate" and "OR gate" relationships, and map the nodes and corresponding relationships corresponding to the searched target entities into the fault tree to form a complete fault tree. For example, if the failure of a device may be caused by multiple factors, an AND gate can be constructed; if any failure factor will result in a top-level failure, an OR gate is used. Finally, generate a fault tree that matches the information to be diagnosed.

[0080] After that, according to the fault tree and the inference requirements, select a logical inference algorithm, such as forward inference, backward inference, or hybrid inference, etc., to determine the fault type and fault cause of the target device, and search for other entities and relationships related to the fault cause in the knowledge graph. Based on these entities and relationships, perform influence factor inference to obtain the influence factors of the target device.

[0081] In the embodiment of the present application, the process of constructing a fault tree by the fault diagnosis inference module based on the nodes and corresponding relationships of the target entities includes:

[0082] Construct a fault tree based on the nodes, corresponding relationships of the target entities, and graph theory algorithms. Analyze the initial fault tree through graph theory algorithms, identify and delete redundant nodes, and optimize the structure of the tree to make it more concise and clear, so as to obtain the final fault tree.

[0083] Optionally, the data acquisition module is further configured to obtain the device basic information and historical fault records of the target device, and the historical fault records include multiple historical faults and the repair records and fault causes corresponding to each historical fault.

[0084] Specifically, the data acquisition module includes:

[0085] Device sensors, which are used to obtain the operation parameters and status information of the target device in real time;

[0086] The operation log module, which is used to obtain the repair records and fault causes corresponding to each historical fault of the target device;

[0087] The alarm module, which is used to obtain the historical fault records during the operation of the target device;

[0088] The data port module, which is used to obtain the device basic information of the target device.

[0089] In the embodiments of the present application, the device basic information includes extracting the device functions, performance indicators, and common fault information from the device manuals provided by the manufacturer, and forming a knowledge base by collecting the experiences and insights of industry experts through forms such as interview questionnaires.

[0090] The operating parameters and status information of the target device can be obtained in real time through device sensors, ensuring that the basic data required for fault diagnosis is comprehensive and timely. The operation log module records the repair records and fault causes of each historical fault, providing an important basis for the identification and reasoning of fault modes. The alarm module captures the historical faults during the operation of the device, helping to quickly respond to and handle emergencies. The data port module obtains the device basic information, ensuring the accuracy and integrity of the knowledge graph construction. Ensure that the original data required for fault diagnosis is comprehensive and real-time.

[0091] In this embodiment, the confidence information provided by the graph reasoning algorithm (such as rule-based reasoning, statistics-based reasoning, etc.) can be used to evaluate the reasoning results to obtain the evaluation results.

[0092] The reasoning results can also be compared with historical fault cases, calculate the similarity between the reasoning results and each historical fault case, and use the highest similarity as the evaluation result to verify the rationality of the reasoning results.

[0093] The system further includes:

[0094] A data preprocessing module, configured to perform preprocessing operations on the device basic information and the historical fault records respectively to obtain the target device basic information and the target historical fault records;

[0095] A knowledge graph construction module, configured to construct a knowledge graph based on the target device basic information and the target historical fault records by using data mining technology and natural language processing technology.

[0096] In the embodiments of the present application, the preprocessing operations include outlier detection operations, missing value filling operations, data standardization operations, such as denoising, cleaning, formatting, etc., to ensure data quality.

[0097] In the embodiments of the present application, as Figure 2 shown, the knowledge graph construction module is specifically configured to:

[0098] Based on the target device basic information, the target historical fault records, and natural language processing technology, determine multiple entities, each of which is a fault mode, a fault cause, or a fault impact;

[0099] Based on data mining technology, determine the corresponding relationships between each of the entities, and the corresponding relationships are causal relationships or dependency relationships;

[0100] Based on the corresponding relationships between each of the said entities, a knowledge graph is constructed, with each entity as a node and each of the said corresponding relationships as an edge connecting the nodes, and the knowledge graph is stored in a graph database. This ensures the efficient access and rapid retrieval of the knowledge graph, providing a solid data foundation and technical support for subsequent fault diagnosis.

[0101] Optionally, the system further includes:

[0102] A feedback optimization module, configured to obtain user feedback information and fault records corresponding to the fault diagnosis report;

[0103] Based on the embedded knowledge graph completion algorithm TransR, the feedback information, and the fault records, update the knowledge graph.

[0104] Through the feedback optimization module, it is possible to obtain in real time the user's feedback information on the fault diagnosis report and the latest fault records. Using the embedded knowledge graph completion algorithm TransR, combining this feedback information and fault records, dynamically update the node and edge relationships in the knowledge graph. This not only improves the accuracy and integrity of the knowledge graph, but also enhances the self-learning ability and adaptability of the system, further improving the accuracy and reliability of fault diagnosis.

[0105] Based on the same principle as the system shown in Figure 1 , an embodiment of the present invention also provides a fault judgment method based on a knowledge graph, as shown in Figure 3 The fault judgment method based on a knowledge graph includes:

[0106] Step S1, obtain the information to be diagnosed, where the information to be diagnosed includes the device type, working conditions, and real-time monitoring data of the target device, and the real-time monitoring data includes operating parameters and status information;

[0107] Step S2, input the information to be diagnosed into the pre-constructed knowledge graph, and use a graph reasoning algorithm to perform fault diagnosis on the information to be diagnosed to obtain the reasoning result of the target device. The pre-constructed knowledge graph includes multiple nodes corresponding to the target device, and each of the nodes is connected by an edge. The multiple nodes include a fault mode node, a fault cause node, and a fault impact node. Each of the edges represents the corresponding relationship between the corresponding nodes. The reasoning result includes the fault type, fault cause, and influencing factors of the target device;

[0108] Step S3, evaluate the reasoning result to obtain an evaluation result, and display the reasoning result and the evaluation result based on a visualization tool to generate a fault diagnosis report.

[0109] It should be noted that the execution subject of this method can be an electronic device, and its implementation principle is the same as that of a fault judgment system based on a knowledge graph described above, so it will not be elaborated here.

[0110] Taking the target device as an industrial pump as an example, the method for generating a fault tree of the industrial pump will be illustrated below.

[0111] As Figure 4 shown, this embodiment provides a method for generating a fault tree for an industrial pump, including:

[0112] In the first step, obtain the device structure, key components, and normal operating parameters of the industrial pump from the device manual, obtain the fault cases of the industrial pump from the historical fault record database, where the fault cases include fault causes and occurrence frequencies, etc., and obtain the expert experience and common fault modes of the industrial pump from the industry expert knowledge base.

[0113] In the second step, use natural language processing technology and data mining technology to perform entity recognition on the collected industrial pump data and construct relationships. In entity recognition, entities include device components (such as motors, pump casings), fault modes (such as leakage, overheating), and potential causes (such as wear, blockage), etc. Entity relationships such as "pump casing leakage" may be caused by "seal wear".

[0114] In the third step, the user selects a specific industrial pump (for example, model "ABC-200") on the interface and inputs the following working conditions: working temperature 50°C, working pressure 5 bar, working environment chemical plant, and operating parameters, etc., to obtain the information to be diagnosed.

[0115] In the fourth step, input the information to be diagnosed entered by the user into the knowledge graph to automatically generate a preliminary fault tree of the industrial pump, as Figure 4 shown.

[0116] First is the fault path reasoning. According to the logical relationship of "AND gate" or "OR gate" in the fault tree, infer the specific causes and conditions for the occurrence of the fault, that is, the following 4 types: Industrial pump failure ← Motor failure ← Overload; Industrial pump failure ← Motor failure ← Cable damage; Industrial pump failure ← Pump casing leakage ← Seal wear; Industrial pump failure ← Pump casing leakage ← Corrosion.

[0117] Then, perform influence factor reasoning on the fault causes, extract the relationship nodes corresponding to the fault causes in the knowledge graph. For example, extract the relationship nodes for "corrosion" to obtain "Corrosion ← Temperature change", "Corrosion ← Moisture and air oxidation", "Corrosion ← Local turbulence and cavitation", "Corrosion ← Frequent start and stop", etc., and then the specific causes of the industrial pump failure due to pump casing corrosion can be obtained.

[0118] In the fifth step, a graph theory algorithm is adopted to identify redundant nodes in the initial fault tree and merge fault modes with a similarity greater than a set similarity threshold between each pair of fault modes, ensuring that nodes corresponding to each fault mode with a similarity greater than the set similarity threshold appear only once. The optimized industrial pump fault tree is more concise and easier for users to understand.

[0119] In the sixth step, the fault tree is displayed based on a visualization tool, and users can edit, add, or modify fault modes through a graphical interface. Additionally, feedback information (such as adding a new fault mode "excessive vibration") will be recorded, and the knowledge graph will be updated based on the embedded knowledge graph completion algorithm TransR, the feedback information, and the fault records to improve the accuracy of fault tree generation.

[0120] Next, taking a wind turbine as the target device as an example, the method for generating a fault tree of the wind turbine will be illustrated.

[0121] As Figure 5 shown, this embodiment demonstrates a method for generating a fault tree for a wind turbine, including:

[0122] In the first step, extract the device type, model, power, key components, and normal operating parameters of the wind turbine from the device manual; search for historical fault cases of each model of wind turbine in the historical fault record database, where the historical fault cases include fault causes and occurrence frequencies; obtain expert experience and common fault modes of wind turbines from the industry expert knowledge base.

[0123] In the second step, use natural language processing technology and data mining technology to perform entity recognition on the data collected for the wind turbine. In the entity recognition, entities include the fault types of the wind turbine (blade damage, generator failure, control system failure), fault causes (strong wind, overload, material fatigue), fault consequences (decrease in output power, equipment shutdown, safety accident), etc. Entity relationships such as blade damage is caused by strong wind, blade damage leads to a decrease in output power, and vibration monitoring is used to detect generator failure.

[0124] In the third step, the user selects a wind turbine on the interface and inputs fault diagnosis information such as operating conditions. For example, the model is "WT-3000", the power is 3000 kW, the operating time is 8000 hours, and the recent maintenance record is an electrical system inspection and blade cleaning.

[0125] In the fourth step, input the diagnostic information entered by the user into the knowledge graph to automatically generate a preliminary fault tree of the wind turbine, as Figure 6 shown.

[0126] In the fifth step, a graph theory algorithm is adopted to identify redundant nodes in the initial fault tree and merge fault modes with a similarity greater than a set similarity threshold between each pair of fault modes, ensuring that nodes corresponding to each fault mode with a similarity greater than the set similarity threshold appear only once. The optimized fault tree is more concise.

[0127] First, fault path reasoning is carried out. According to the logical relationship of "AND gate" or "OR gate" in the fault tree, the specific causes and conditions for the occurrence of the fault are deduced, namely the following 4 types: Wind turbine failure ← Electrical failure ← Short circuit; Wind turbine failure ← Electrical failure ← Loose connection; Wind turbine failure ← Blade damage ← Excessive wind force; Wind turbine failure ← Blade damage ← Material fatigue;

[0128] Then, influence factor reasoning is carried out on the fault causes, and the relationship nodes corresponding to the fault causes in the knowledge graph are extracted. For example, for "material fatigue", the relationship nodes "Material fatigue ← Unstable wind speed", "Material fatigue ← Unbalanced blades", "Material fatigue ← Unqualified raw material quality", "Material fatigue ← Start-stop load impact", "Material fatigue ← Overload operation", etc. are extracted, and the specific reasons for the failure of the industrial pump due to pump shell corrosion can be obtained.

[0129] In the sixth step, the fault tree is displayed based on a visualization tool, and users can edit, add, or modify fault modes through the graphical interface, and feedback information (such as adding new fault modes "gearbox failure", "bearing wear") will be recorded. Based on the embedded knowledge graph completion algorithm TransR, the feedback information, and the fault records, the knowledge graph is updated to improve the accuracy of fault tree generation.

[0130] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working process of the method described above can refer to the corresponding process in the foregoing system embodiment and will not be elaborated here.

[0131] Figure 6 This is a structural block diagram of an electronic device 300 according to an embodiment of the present application.

[0132] As Figure 6 shown, the electronic device 300 includes a processor 301 and a memory 302, and may further include one or more of an information input / output (I / O) interface 303, a communication component 304, and a communication bus 305.

[0133] Among them, the processor 301 is used to control the overall operation of the electronic device 300 to complete all or part of the steps in the above-mentioned fault judgment method based on the knowledge graph; the memory 302 is used to store various types of data to support the operation of the electronic device 300. These data may include, for example, instructions for any application or method operating on the electronic device 300, as well as application-related data. The memory 302 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, a magnetic disk, or an optical disk, or one or more of them.

[0134] The I / O interface 303 provides an interface between the processor 301 and other interface modules. The above-mentioned other interface modules can be a keyboard, a mouse, buttons, etc. These buttons can be virtual buttons or physical buttons. The communication component 304 is used to test the wired or wireless communication between the electronic device 300 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, near field communication (NFC), 2G, 3G, or 4G, or a combination of one or more of them. Therefore, the corresponding communication component 304 can include: a Wi-Fi component, a Bluetooth component, and an NFC component.

[0135] The communication bus 305 may include a path for transmitting information between the above-mentioned components. The communication bus 305 can be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The communication bus 305 can be divided into an address bus, a data bus, a control bus, etc.

[0136] The electronic device 300 can be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components, and is used to execute the fault judgment method based on the knowledge graph given in the above embodiments.

[0137] The computer-readable storage medium provided by the embodiments of the present application will be introduced below. The computer-readable storage medium described below can be correspondingly referred to the fault judgment method based on the knowledge graph described above.

[0138] The present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned fault judgment method based on the knowledge graph are implemented.

[0139] The computer-readable storage medium may include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical discs.

[0140] The term "including", "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements but also other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0141] The above description is only the preferred embodiments of the present application and the description of the applied technical principles. Those skilled in the art should understand that the scope of the application involved in the present application is not limited to the technical solutions formed by the specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the foregoing application concept. For example, the technical solutions formed by mutually replacing the above features with other technical features having similar functions (but not limited to) applied in the present application.

Claims

1. A fault diagnosis system based on knowledge graph, characterized in that: include: A data acquisition module, used to obtain information to be diagnosed, wherein the information to be diagnosed includes the device type, working conditions and real-time monitoring data of the target device, wherein the real-time monitoring data includes operating parameters and status information; A fault diagnosis reasoning module is used to input the information to be diagnosed into a pre-constructed knowledge graph, and use a graph reasoning algorithm to perform fault diagnosis on the information to be diagnosed to obtain a reasoning result of the target device, wherein the pre-constructed knowledge graph includes a plurality of nodes corresponding to the target device, each of the nodes is connected by an edge, and the plurality of nodes include a fault mode node, a fault cause node, and a fault impact node, each of the edges represents a corresponding relationship between the corresponding nodes, and the reasoning result includes a fault type, a fault cause, and an influencing factor of the target device; The diagnostic result analysis module is used to evaluate the reasoning result, obtain the evaluation result, and display the reasoning result and the evaluation result based on a visualization tool to generate a fault diagnosis report.

2. According to the knowledge graph-based fault diagnosis system of claim 1, it is characterized in that: The data acquisition module is further used to obtain basic device information and historical fault records of the target device, wherein the historical fault records include multiple historical faults and maintenance records and fault causes corresponding to each of the historical faults; The system further comprises: A data preprocessing module, used to perform preprocessing operations on the basic equipment information and the historical fault records respectively, to obtain the target equipment basic information and the target historical fault records; The knowledge graph construction module is used to construct a knowledge graph based on the target device basic information and the target historical fault records using data mining technology and natural language processing technology.

3. According to the knowledge graph-based fault diagnosis system of claim 2, it is characterized in that: The knowledge graph construction module is specifically used for: Based on the target device basic information, the target historical fault records and natural language processing technology, multiple entities are determined, each of which is a fault mode, a fault cause or a fault impact; Based on data mining technology, determine the corresponding relationship between each of the entities, each of the corresponding relationships is a causal relationship or a dependency relationship; Based on the corresponding relationship between each of the entities, a knowledge graph is constructed, with each entity as a node and each of the corresponding relationship as an edge connecting the nodes, and the knowledge graph is stored in a graph database.

4. According to claim 1, a fault diagnosis system based on knowledge graph is characterized in that: The system further comprises: Feedback optimization module, used to obtain user feedback information and fault records corresponding to the fault diagnosis report; The knowledge graph is updated based on the embedded knowledge graph completion algorithm TransR, the feedback information and the fault record.

5. According to the knowledge graph-based fault diagnosis system of claim 1, it is characterized in that: The fault diagnosis reasoning module is specifically used for: Perform entity recognition on the text in the information to be diagnosed based on the LSTM algorithm to obtain multiple target entities of the target device; Based on the search algorithm, search for nodes and corresponding relationships corresponding to each of the target entities in the pre-built knowledge graph; Constructing a fault tree based on the nodes and corresponding relationships corresponding to each of the target entities; Based on the fault tree and logical reasoning algorithm, determine the fault type and fault cause of the target device; Based on the corresponding relationship between the fault cause of the target device in the knowledge graph, inference of influencing factors is performed to obtain the influencing factors of the target device; Based on the fault type, fault cause and influencing factors of the target device, an inference result of the target device is determined.

6. A fault diagnosis system based on knowledge graph according to claim 5, characterized in that: The fault diagnosis reasoning module constructs a processing process of a fault tree based on the nodes and corresponding relationships of the target entity, including: A fault tree is constructed based on the nodes, corresponding relationships and graph theory algorithms of the target entity.

7. A fault diagnosis system based on knowledge graph according to claim 2, characterized in that: The data acquisition module comprises: Equipment sensors are used to obtain the operating parameters and status information of the target equipment in real time; Operation log module, used to obtain the maintenance record and fault cause corresponding to each historical fault of the target device; Alarm module, used to obtain historical fault records during the operation of the target equipment; The data port module is used to obtain basic device information of the target device.

8. A fault judgment method based on knowledge graph, characterized in that: include: Acquiring information to be diagnosed, wherein the information to be diagnosed includes a device type, working conditions, and real-time monitoring data of the target device, wherein the real-time monitoring data includes operating parameters and status information; The information to be diagnosed is input into a pre-constructed knowledge graph, and a graph reasoning algorithm is used to perform fault diagnosis on the information to be diagnosed to obtain a reasoning result of the target device, wherein the pre-constructed knowledge graph includes a plurality of nodes corresponding to the target device, each of the nodes is connected by an edge, and the plurality of nodes include a fault mode node, a fault cause node, and a fault impact node, and each of the edges represents a corresponding relationship between the corresponding nodes, and the reasoning result includes a fault type, a fault cause, and an influencing factor of the target device; The reasoning result is evaluated to obtain an evaluation result, and the reasoning result and the evaluation result are displayed based on a visualization tool to generate a fault diagnosis report.

9. An electronic device, characterized in that: comprising a processor coupled to a memory; The processor is configured to execute the computer program stored in the memory, so that the electronic device performs the method according to claim 1.

10. A computer-readable storage medium, characterized in that: The method comprises a computer program or an instruction, which, when being executed on a computer, causes the computer to execute the method as claimed in claim 1.

Images