Power equipment fault intelligent diagnosis method, system and equipment based on knowledge graph

Abstract

The invention provides a power equipment fault intelligent diagnosis method, system and equipment based on a knowledge graph, and the method comprises the steps: collecting historical work order data, equipment real-time operation data and expert maintenance record data of power equipment, and carrying out the preprocessing; power equipment, fault types, fault reasons and maintenance measure entities are identified from the preprocessed data, and the relation between the entities is extracted; constructing a fault tree and displaying a fault causal relationship by taking a fault type as a root node, a fault reason as an intermediate node and a fault phenomenon as a leaf node based on a relationship between entities; associating fault types with maintenance measures by using a fault causal relationship, and constructing a solution map; according to the method, feature extraction is carried out on the preprocessed equipment real-time operation data, the corresponding fault type is determined, the fault type is input to the solution graph, the solution is generated, automatic fault recognition is carried out in a knowledge graph mode, and the fault recognition efficiency is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power equipment fault detection, and in particular to a knowledge graph-based intelligent diagnosis method, system and equipment for power equipment faults. Background Art

[0002] The stable operation of power equipment is crucial for ensuring a secure and reliable power supply. However, various faults are inevitable during long-term operation. Traditional methods for power equipment fault diagnosis rely primarily on manual experience and simple data analysis, resulting in low diagnostic efficiency, poor accuracy, and difficulty handling complex faults. On the one hand, with the continuous expansion of power systems and the increasing complexity of equipment, the volume of fault data has exploded, making manual processing of this data increasingly difficult. On the other hand, faults manifest themselves in diverse ways across different equipment, and complex relationships may exist between faults. Traditional methods struggle to fully and accurately capture these relationships, compromising the accuracy and timeliness of fault diagnosis. Furthermore, power equipment fault diagnosis requires extensive expertise and experience, which is often dispersed across different experts and historical data, making it difficult to effectively integrate and utilize.

[0003] Therefore, how to improve the fault diagnosis efficiency of power equipment has become a technical problem that technical personnel in this field urgently need to solve. Summary of the Invention

[0004] The present invention provides a method, system and device for intelligent diagnosis of power equipment faults based on knowledge graphs, which are used to solve the defect of low efficiency in power equipment fault diagnosis in the prior art.

[0005] In a first aspect, the present invention provides a method for intelligent diagnosis of power equipment faults based on a knowledge graph, comprising: Collect historical work order data of power equipment, real-time equipment operation data, and expert maintenance record data, and perform pre-processing; Identifying the entities of power equipment, fault type, fault cause and repair measure in the preprocessed data, and extracting the relationship between the entities; Based on the relationship between the entities, a fault tree is constructed with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node to show the fault causal relationship; Using the fault causal relationship, the fault type and maintenance measures are associated to build a solution map; Feature extraction is performed on the pre-processed real-time operation data of the equipment, and the corresponding fault type is determined, and the fault type is input into the solution map to generate a solution.

[0006] According to a knowledge graph-based intelligent diagnosis method for power equipment faults provided by the present invention, the inputting of the fault type into the solution graph to generate a solution includes: Based on the fault type and cause, find the repair measures and solutions corresponding to the fault type in the solution map and sort them according to effectiveness and feasibility; Based on the sorting results, the maintenance measures and solutions are pushed to the operation and maintenance personnel, and relevant knowledge explanations and operation guides are provided.

[0007] According to the present invention, a method for intelligent diagnosis of power equipment faults based on a knowledge graph further includes: Monitor the operating status and fault data of power equipment in real time, and update new fault data and maintenance data to the fault tree and solution map; The fault tree and the solution map are optimized regularly to adjust the weights of entities and relationships.

[0008] According to a knowledge graph-based intelligent diagnosis method for power equipment faults provided by the present invention, the preprocessing includes: The historical work order data, real-time equipment operation data and expert maintenance record data are cleaned to remove noise, duplication and erroneous data, and are standardized to unify the data format and unit, and then the text data is segmented and tagged with parts of speech.

[0009] According to a knowledge graph-based intelligent diagnosis method for power equipment faults provided by the present invention, the power equipment, fault type, fault cause, and maintenance measure entities are identified in the preprocessed data, including: Extracting nouns and noun phrases from the text of the preprocessed data, matching the nouns and noun phrases with a power equipment dictionary, identifying power equipment model and specification patterns, and determining the power equipment entity based on the context; Identifying the fault description portion in the preprocessed data, extracting fault keywords and phrases, and classifying them into a predefined fault type system to determine the fault type entity; Locating cause indicators in the pre-processed data, extracting cause description phrases, and associating them with specific devices and faults to determine the fault cause entity; Identify maintenance action verbs in the pre-processed data, extract maintenance objects and contents, associate them with specific equipment and faults, and determine maintenance measure entities.

[0010] According to a knowledge graph-based intelligent fault diagnosis method for power equipment provided by the present invention, the relationship between extracted entities includes: Extract the power equipment-fault type relationship based on co-occurrence analysis, verb pattern matching and dependency analysis; Extract the fault type-fault cause relationship based on causal connective identification, result-cause pattern analysis, and semantic role labeling; Extract the relationship between fault type and maintenance measures based on action object analysis, purpose adverbial recognition and measure effectiveness judgment; Based on targeted analysis of measures and application of professional domain knowledge, the relationship between fault cause and maintenance measures is extracted.

[0011] According to the present invention, a method for intelligent diagnosis of power equipment faults based on a knowledge graph is provided. The method constructs a fault tree based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, including: Select the top-level fault type entity as the root node; Query all fault cause entities pointing to the root node and construct intermediate nodes; Query the fault cause entity starting from the root node and add leaf nodes; Based on the leaf nodes, each of the intermediate nodes is recursively executed to detect deeper causes, expand the self-description structure, and construct a fault tree.

[0012] According to the present invention, a method for intelligent diagnosis of power equipment faults based on a knowledge graph is provided, wherein the method utilizes the fault causal relationship, associates the fault type with the maintenance measures, and constructs a solution graph, including: Mapping causal relationships to solutions is achieved through forward tracing, reverse correlation, and effectiveness verification; Conduct multi-dimensional relationship modeling and dynamic weight matching; Based on the mapping, the multidimensional relationship modeling and the dynamic weight matching, a basic framework is established, fault type nodes and maintenance measure nodes are added to the basic framework, and causal relationship edges and maintenance measure edges are constructed to build a solution map.

[0013] In a second aspect, the present invention further provides an intelligent fault diagnosis system for power equipment based on a knowledge graph, comprising: The collection module is used to collect historical work order data of power equipment, real-time equipment operation data and expert maintenance record data, and perform pre-processing; an extraction module, configured to identify the entities of power equipment, fault type, fault cause and repair measure in the preprocessed data, and extract the relationships between the entities; A construction module is used to construct a fault tree based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, to show the fault causal relationship; using the fault causal relationship, the fault type is associated with the maintenance measures to construct a solution map; The determination module is used to extract features from the pre-processed real-time operation data of the equipment, determine the corresponding fault type, input the fault type into the solution map, and generate a solution.

[0014] In a third aspect, the present invention also provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, an intelligent diagnosis method for power equipment faults based on a knowledge graph as described above is implemented.

[0015] In a fourth aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described knowledge graph-based intelligent diagnosis methods for power equipment faults.

[0016] In a fifth aspect, the present invention also provides a computer program product, comprising a computer program, which, when executed by a processor, implements any of the above-mentioned knowledge graph-based intelligent diagnosis methods for power equipment faults.

[0017] The present invention provides a method, system and equipment for intelligent diagnosis of power equipment faults based on knowledge graphs, including collecting historical work order data of power equipment, real-time operation data of equipment and expert maintenance record data, and performing preprocessing; identifying power equipment, fault type, fault cause and maintenance measure entities in the preprocessed data, and extracting the relationship between the entities; based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, a fault tree is constructed to display the fault causal relationship; using the fault causal relationship, the fault type and maintenance measures are associated to construct a solution map; feature extraction is performed on the preprocessed real-time operation data of the equipment, and the corresponding fault type is determined, the fault type is input into the solution map, a solution is generated, and faults are automatically identified through the knowledge graph, effectively improving the efficiency of fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, a brief introduction is given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0019] Figure 1 This is a flow chart of the knowledge graph-based intelligent diagnosis method for power equipment faults provided in this embodiment; Figure 2 This is a schematic diagram of the structure of the knowledge graph-based intelligent diagnosis system for power equipment faults provided in this embodiment; Figure 3 Schematic diagram of the structure of the electronic device provided in this embodiment. DETAILED DESCRIPTION

[0020] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0021] Figure 1 This is a flow chart of the knowledge graph-based intelligent diagnosis method for power equipment faults provided in this embodiment.

[0022] like Figure 1 As shown, the flowchart of the method for intelligent diagnosis of power equipment faults based on knowledge graph provided by an embodiment of the present invention mainly includes the following steps: 101. Collect historical work order data of power equipment, real-time equipment operation data and expert maintenance record data, and perform pre-processing.

[0023] In a specific implementation, historical work order data includes fault occurrence time, fault symptoms, and repair measures. Real-time operating parameters include voltage, current, and temperature. Expert maintenance records include fault causes and repair measures. For example, historical work order data for power equipment from the past year was collected, covering fault repair records for various equipment such as transformers and circuit breakers. Furthermore, sensors were used to collect equipment operating parameters in real time, and senior experts were invited to share their long-term experience in fault diagnosis.

[0024] After collecting relevant data, historical work order data, real-time equipment operation data, and expert maintenance record data are cleaned to remove noise, duplication, and errors, and standardized to unify data formats and units. The text data is then segmented and tagged with parts of speech to complete data preprocessing for subsequent knowledge extraction.

[0025] 102. Identify the entities of power equipment, fault type, fault cause and maintenance measures in the preprocessed data, and extract the relationship between the entities.

[0026] Identification can be rule-based or feature-based. Rule-based identification uses a dictionary of equipment names (transformers, circuit breakers, disconnectors, etc.) to match model specification patterns (such as "SF6-220kV / 3150A") to identify power equipment. Define fault keywords (short circuit, overload, oil leakage, discharge, etc.) and construct regular expressions (such as "*phase [short circuit|grounding] fault") to identify fault types. Use cause indicator word recognition ("due to," "due to," "the reason is," etc.) and cause phrase pattern matching to identify the cause of the fault. Identify maintenance measures by identifying action verbs ("replace," "repair," "adjust," etc.) and matching them with maintenance content patterns.

[0027] The specific process involves extracting nouns and noun phrases from the preprocessed data, matching them to a power equipment dictionary, identifying power equipment model and specification patterns, and then identifying the power equipment entity based on the context. For example, "main transformer B phase bushing" is identified as "power equipment."

[0028] Identify the fault description portion of the preprocessed data, extract fault keywords and phrases, and classify them into a predefined fault type system to determine the fault type entity. For example, "C phase ground short circuit occurred" is identified as "Fault type: ground short circuit."

[0029] Locate the cause indicator words in the preprocessed data, extract the cause description phrases, and associate them with specific equipment and faults to determine the fault cause entity. For example, "insulation aging due to long-term operation" is identified as "Fault cause: insulation aging."

[0030] Identify maintenance action verbs in preprocessed data, extract the maintenance object and content, associate them with specific equipment and faults, and identify the maintenance measure entity. For example, "Replaced a damaged casing" is identified as "Maintenance measure: Replace casing."

[0031] The relationship between power equipment and fault types is extracted based on co-occurrence analysis, verb pattern matching, and dependency analysis. Co-occurrence analysis identifies equipment and fault types that appear in the same sentence or paragraph. For example, the sentence "#equipment#main transformer bushing occurrence#fault#flashover discharge" is used. Verb pattern matching identifies conjunctions such as "occurrence," "appearance," and "produce." An example rule is: equipment + [occurrence-related verb] + fault → establishes a relationship. Dependency analysis uses equipment as the subject and fault as the object. An example dependency path is: nsubj(occurrence) - dobj(occurrence).

[0032] The fault type-fault cause relationship is extracted based on causal connective identification, result-cause pattern analysis, and semantic role labeling. Causal connective identification includes identifying phrases such as "owing to," "because," and "the reason is." For example, "#fault# short circuit is due to #reason# insulation aging." Result-cause pattern analysis includes phrases such as "fault" followed by "cause" with a demonstrative in between, or "cause" followed by "fault," using phrases such as "result" and "cause." Semantic role labeling identifies "cause" as an adverbial or complement component.

[0033] Based on action object analysis, purpose adverbial recognition, and measure effectiveness judgment, the relationship between fault type and maintenance measure is extracted. Action object analysis is based on maintenance verbs, such as "replace," "repair," and "adjust." For example, "For the #fault#grounding fault, #measure#replaced the cable." Purpose adverbial recognition uses patterns such as "in order to address..." and "taken measures to address..." For example, "For the #fault#overheating, #measure#heat dissipation modification was implemented." Measure effectiveness judgment identifies whether the measures are truly targeted at the specific fault, eliminating general maintenance measures.

[0034] Based on targeted action analysis and application of specialized domain knowledge, the relationship between fault causes and maintenance measures is extracted. Targeted action analysis involves determining whether the measures directly address the specific cause. For example, "Due to #reason# seal aging, #measure# was taken to replace the seal." Application of specialized domain knowledge involves verifying the rationale based on the power maintenance knowledge base. For example, "Insulation aging → additional insulation testing; mechanical wear → component replacement."

[0035] After relationship extraction, relationship verification and integration are performed, including consistency checks and relationship graph construction. Consistency checks include logical verification and conflict resolution. Logic verification involves checking whether the action matches the fault type and verifying whether the cause can cause the fault. Conflict resolution involves using confidence scores or contextual evidence strength when the same fault has multiple possible causes.

[0036] Relationship graph construction involves triple storage and knowledge graph construction. Triple storage includes (equipment, occurrence, fault)(fault, cause, cause description)(fault, treatment, repair measures)(cause, corresponding measures, measure description). Knowledge graph construction includes equipment --> |occurrence|fault, fault --> |cause| cause, fault --> |treatment| measures, cause --> |corresponding measures| measures.

[0037] This enables the systematic extraction of the complex relationship network between equipment, faults, causes and measures from power equipment fault data, providing a solid foundation for fault knowledge graph construction and intelligent diagnosis.

[0038] 103. Based on the relationship between entities, a fault tree is constructed with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node to show the cause-effect relationship of the fault.

[0039] Specifically, constructing a fault tree based on extracted entity relationships is a systematic process that can intuitively display the causal chain from fault symptoms to root causes. Fault trees are composed of the following elements: Root: The top-level fault event (fault type entity); Intermediate nodes: The direct causes of the upper-level events (fault cause entities); Leaf nodes: The bottom-level observable phenomena or basic events; Logic gates: Representing the logical relationships between events (AND / OR, etc.).

[0040] Select the top-level fault type entity as the root node. For example, select "Winding Short Circuit" as the root node. Query all fault cause entities pointing to the root node and construct intermediate nodes. For example, "Insulation Aging" → "Winding Short Circuit" and "Overvoltage Surge" → "Winding Short Circuit" are examples.

[0041] Query the fault cause entity starting from the root node and add leaf nodes. For example, "Winding Short Circuit" → "Temperature Abnormal" and "Winding Short Circuit" → "Protection Action".

[0042] Based on the leaf nodes, recursively execute each intermediate node to detect deeper causes. For example, "oil paper deterioration" → "insulation aging", "lightning strike" → "overvoltage shock", expand the self-description structure, and build a fault tree. The example is as follows: [Winding short circuit] ├─[Insulation aging]───[Oil paper deterioration] ├─[Overvoltage shock]───[Lightning strike] ├─[Temperature abnormality] └─[Protection Action] For example, for a transformer fault, a corresponding fault tree is constructed using "winding short circuit" as the fault type, "insulation aging" and "overvoltage shock" as the fault causes, and "abnormal temperature rise" and "abnormal oil gas content" as the fault phenomena. Furthermore, maintenance measures such as "winding replacement" and "enhanced insulation testing" are associated with the "winding short circuit" fault type to form a solution map.

[0043] Finally, a completeness check is performed, including top-down verification: each node must have at least one child node or be marked as a basic event, and the root node must cover all major failure modes; bottom-up verification: each phenomenon must be traceable to at least one failure type, and no isolated nodes exist. Logical consistency checks include: gate type verification: AND gate subnodes should not have mutually exclusive relationships, and OR gate subnodes should have independent occurrence possibilities; causal strength assessment: domain expert knowledge is used to verify the legitimacy of causal relationships, and confidence scores are added to weak associations.

[0044] Example of the final transformer fault tree: [Transformer failure] ├─[Insulation fault]─┬─[Winding short circuit]─┬─[Abnormal temperature] ││└─[Gas Action] │└─[Casing discharge]───[Local overheating] ├─[Cooling system failure]─┬─[Oil pump failure] │└─[Fan stopped] └─[Tap changer fault]───[Gear position abnormality] Circuit breaker failure example: [Circuit breaker refuses to operate] ├─[Mechanical failure]─┬─[Mechanical jam] │└─[Connecting rod broken] ├─[Electrical failure]─┬─[Coil burnt] │└─[Auxiliary contact failure] └─[Control circuit failure]─┬─[Power failure] └─[Relay failure] Through a systematic construction process, discrete entity relationships can be transformed into intuitive fault tree models, providing a powerful support tool for fault diagnosis, preventive maintenance, and reliability analysis of power equipment. The quality of the fault tree construction directly affects its application effectiveness, so it requires repeated verification and optimization combined with domain expert knowledge.

[0045] 104. Use the cause-effect relationship of faults to associate fault types with maintenance measures and build a solution map.

[0046] Through forward tracing, backward correlation, and validation, we map causal relationships to solutions. Forward tracing involves following the chain from fault type to root cause, for example, "winding short circuit → insulation aging → oil-paper deterioration." Reverse correlation involves following the chain from root cause to repair measures, for example, "oil-paper deterioration → [replace oil-paper, vacuum drying]." Validation involves checking the frequency of occurrence of this combination in historical work orders and verifying the recommended ratings in the expert knowledge base.

[0047] Multidimensional relationship modeling and dynamic weight matching were performed. Weight influencing factors included historical implementation times, repair success rate, cost and time indicators, and expert ratings. Therefore, the calculation formula was: Weight = 0.4 * Success Rate + 0.3 * Timeliness + 0.2 * Cost Coefficient + 0.1 * Expert Rating.

[0048] Based on mapping, multidimensional relationship modeling and dynamic weight matching, a basic framework is established, fault type nodes and maintenance measure nodes are added to the basic framework, and causal relationship edges and maintenance measure edges are constructed to build a solution map.

[0049] Through this systematic construction process, the solution map transforms discrete maintenance knowledge into a computable, optimizable intelligent decision-making system. This results in: improved troubleshooting efficiency (reducing diagnosis time by over 50%), more scientific maintenance solutions (reducing recurring failure rates by 30%), and standardized knowledge transfer (enabling new employees to quickly access expert experience). The resulting solution map is not just a static knowledge base, but a continuously evolving intelligent operation and maintenance support system, providing strong support for reliability management of power equipment.

[0050] 105. Extract features from the pre-processed real-time equipment operation data, determine the corresponding fault type, input the fault type into the solution map, and generate a solution.

[0051] Extract fault-related features from real-time collected equipment operating parameters, such as parameter change trends and abnormal fluctuations. These features are then matched with fault patterns in the knowledge graph. Leveraging the reasoning capabilities of the knowledge graph, the possible fault types and causes are inferred. For example, if a device's temperature suddenly rises, the knowledge graph can quickly identify possible causes, such as a cooling system failure or excessive load.

[0052] Based on the fault type and cause, the system searches the solution map for corresponding repair measures and solutions, and sorts them by effectiveness and feasibility. Based on the sorting results, the repair measures and solutions are pushed to the operation and maintenance personnel, and relevant knowledge explanations and operation guides are provided. This helps operation and maintenance personnel quickly understand and implement solutions. For example, based on the diagnostic results, the system searches the solution map for repair measures corresponding to "winding short circuit" and "cooling system failure," such as "check winding insulation" and "clean radiator." These solutions are sorted by feasibility and effectiveness and recommended to the operation and maintenance personnel.

[0053] Furthermore, this embodiment also includes real-time monitoring of the operating status and fault data of power equipment, updating newly occurring fault data and repair data into the fault tree and solution graph, and continuously enriching and improving the knowledge graph. The fault tree and solution graph are regularly optimized to remove outdated knowledge and adjust the weights of entities and relationships to improve the accuracy and practicality of the knowledge graph. For example, if a new fault cause or effective repair measure is discovered, the entities and relationships in the knowledge graph are promptly updated. The knowledge graph is regularly evaluated and optimized to ensure the accuracy and practicality of the knowledge.

[0054] The method of the present invention can achieve the following effects: 1. Improve the accuracy and efficiency of fault diagnosis: By integrating multi-source data to build a knowledge graph, the causal relationship and correlation information of power equipment faults can be comprehensively and accurately captured. The reasoning ability of the knowledge graph can be used to achieve fast and accurate fault diagnosis, greatly improving diagnostic efficiency.

[0055] 2. Provide scientific solutions: The solution map provides operation and maintenance personnel with targeted maintenance measures and operating guidelines, avoiding blind maintenance, improving the success rate and efficiency of maintenance, and reducing equipment downtime and maintenance costs.

[0056] 3. Effective integration and inheritance of knowledge: Integrating scattered historical work orders, equipment parameters, and expert experience into the knowledge graph enables structured storage and sharing of knowledge, making it easier for new employees to learn and draw lessons, and promoting the inheritance and accumulation of knowledge.

[0057] 4. Support continuous optimization of the system: The knowledge graph update and maintenance mechanism enables the system to continuously adapt to changes in power equipment and new failure modes, ensuring that the system's diagnostic capabilities and the effectiveness of solutions continue to improve over time.

[0058] Based on the same general inventive concept, the present invention also protects an intelligent diagnosis system for power equipment faults based on a knowledge graph. The intelligent diagnosis system for power equipment faults based on a knowledge graph described below and the intelligent diagnosis method for power equipment faults based on a knowledge graph described above can refer to each other.

[0059] Figure 2 This is a structural diagram of the knowledge graph-based intelligent diagnosis system for power equipment faults provided in this embodiment.

[0060] like Figure 2 As shown, this embodiment provides an intelligent diagnosis system for power equipment faults based on a knowledge graph, including: The collection module 201 is used to collect historical work order data of power equipment, real-time equipment operation data and expert maintenance record data, and perform pre-processing; Extraction module 202, for identifying power equipment, fault type, fault cause, and maintenance measure entities in the pre-processed data, and extracting the relationships between the entities; Construction module 203 is used to construct a fault tree based on the relationship between entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, to show the fault causal relationship; using the fault causal relationship, the fault type is associated with the maintenance measures to construct a solution map; The determination module 204 is used to extract features from the pre-processed real-time operation data of the equipment, determine the corresponding fault type, input the fault type into the solution map, and generate a solution.

[0061] Figure 3 Schematic diagram of the structure of the electronic device provided in this embodiment.

[0062] like Figure 3 As shown, the electronic device may include: a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 communicate with each other via the communications bus 340. The processor 310 may call logic instructions in the memory 330 to execute a knowledge graph-based intelligent diagnosis method for power equipment faults. The method includes: collecting historical work order data, real-time equipment operation data, and expert maintenance record data of power equipment and performing preprocessing; identifying power equipment, fault type, fault cause, and maintenance measure entities in the preprocessed data, and extracting the relationships between the entities; based on the relationships between the entities, constructing a fault tree with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node to display the fault causal relationship; using the fault causal relationship, associating the fault type with the maintenance measure to construct a solution map; performing feature extraction on the preprocessed real-time equipment operation data, determining the corresponding fault type, and inputting the fault type into the solution map to generate a solution.

[0063] Furthermore, the logic instructions in the aforementioned memory 330 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product, stored in a storage medium, includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0064] On the other hand, the present invention also provides a computer program product, which includes a computer program, which can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the knowledge graph-based intelligent diagnosis method for power equipment faults provided by the above methods, which includes: collecting historical work order data, real-time equipment operation data and expert maintenance record data of power equipment, and preprocessing them; identifying power equipment, fault type, fault cause and maintenance measure entities in the preprocessed data, and extracting the relationship between the entities; based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, a fault tree is constructed to display the fault causal relationship; using the fault causal relationship, the fault type and maintenance measures are associated to construct a solution map; feature extraction is performed on the preprocessed real-time equipment operation data, and the corresponding fault type is determined, and the fault type is input into the solution map to generate a solution.

[0065] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the knowledge graph-based intelligent diagnosis method for power equipment faults provided by the above-mentioned methods, the method comprising: collecting historical work order data, real-time equipment operation data and expert maintenance record data of power equipment, and performing preprocessing; identifying power equipment, fault type, fault cause and maintenance measure entities in the preprocessed data, and extracting the relationship between the entities; based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, constructing a fault tree to display the fault causal relationship; utilizing the fault causal relationship, associating the fault type with the maintenance measures to construct a solution map; performing feature extraction on the preprocessed real-time equipment operation data, and determining the corresponding fault type, inputting the fault type into the solution map to generate a solution.

[0066] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0067] Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented using software plus a necessary general-purpose hardware platform, or of course, hardware. Based on this understanding, the essence of the above technical solution, or the portion that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, or an optical disk, and includes a number of instructions for causing a computer device (such as a personal computer, server, or network device) to execute the methods described in each embodiment or certain portions of the embodiments.

[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for intelligent diagnosis of power equipment faults based on knowledge graph, characterized in that: include: Collect historical work order data of power equipment, real-time equipment operation data, and expert maintenance record data, and perform pre-processing; Identifying the entities of power equipment, fault type, fault cause and repair measure in the preprocessed data, and extracting the relationship between the entities; Based on the relationship between the entities, a fault tree is constructed with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node to show the fault causal relationship; Using the fault causal relationship, the fault type and maintenance measures are associated to build a solution map; Feature extraction is performed on the pre-processed real-time operation data of the equipment, and the corresponding fault type is determined, and the fault type is input into the solution map to generate a solution.

2. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: The inputting the fault type into the solution map to generate a solution includes: Based on the fault type and cause, find the repair measures and solutions corresponding to the fault type in the solution map and sort them according to effectiveness and feasibility; Based on the sorting results, the maintenance measures and solutions are pushed to the operation and maintenance personnel, and relevant knowledge explanations and operation guides are provided.

3. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: Also includes: Monitor the operating status and fault data of power equipment in real time, and update new fault data and maintenance data to the fault tree and solution map; The fault tree and the solution map are optimized regularly, and the weights of entities and relationships are adjusted.

4. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: The pre-processing comprises: The historical work order data, real-time equipment operation data and expert maintenance record data are cleaned to remove noise, duplication and erroneous data, and are standardized to unify the data format and unit, and then the text data is segmented and tagged with parts of speech.

5. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: The identifying of the power equipment, fault type, fault cause and repair measure entity in the pre-processed data includes: Extracting nouns and noun phrases from the text of the preprocessed data, matching the nouns and noun phrases with a power equipment dictionary, identifying power equipment model and specification patterns, and determining the power equipment entity based on the context; Identifying the fault description portion in the preprocessed data, extracting fault keywords and phrases, and classifying them into a predefined fault type system to determine the fault type entity; Locating cause indicators in the pre-processed data, extracting cause description phrases, and associating them with specific devices and faults to determine the fault cause entity; Identify maintenance action verbs in the pre-processed data, extract maintenance objects and contents, associate them with specific equipment and faults, and determine maintenance measure entities.

6. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 5 is characterized in that: The relationship between the extracted entities includes: Extract the power equipment-fault type relationship based on co-occurrence analysis, verb pattern matching and dependency analysis; Extract the fault type-fault cause relationship based on causal connective identification, result-cause pattern analysis, and semantic role labeling; Extract the relationship between fault type and maintenance measures based on action object analysis, purpose adverbial recognition and measure effectiveness judgment; Based on targeted analysis of measures and application of professional domain knowledge, the relationship between fault cause and maintenance measures is extracted.

7. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: The fault tree is constructed based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, including: Select the top-level fault type entity as the root node; Query all fault cause entities pointing to the root node and construct intermediate nodes; Query the fault cause entity starting from the root node and add leaf nodes; Based on the leaf nodes, each of the intermediate nodes is recursively executed to detect deeper causes, expand the self-description structure, and construct a fault tree.

8. The method for intelligent diagnosis of power equipment faults based on knowledge graph according to claim 1 is characterized in that: The method of utilizing the fault causal relationship, associating fault types with maintenance measures, and constructing a solution map includes: Mapping causal relationships to solutions is achieved through forward tracing, reverse correlation, and effectiveness verification; Conduct multi-dimensional relationship modeling and dynamic weight matching; Based on the mapping, the multidimensional relationship modeling and the dynamic weight matching, a basic framework is established, fault type nodes and maintenance measure nodes are added to the basic framework, and causal relationship edges and maintenance measure edges are constructed to build a solution map.

9. An intelligent diagnosis system for power equipment faults based on knowledge graph, characterized in that: include: The collection module is used to collect historical work order data of power equipment, real-time equipment operation data and expert maintenance record data, and perform pre-processing; an extraction module, configured to identify the entities of power equipment, fault type, fault cause and repair measure in the preprocessed data, and extract the relationships between the entities; A construction module is used to construct a fault tree based on the relationship between the entities, with the fault type as the root node, the fault cause as the intermediate node, and the fault phenomenon as the leaf node, to show the fault cause-effect relationship; Using the fault causal relationship, the fault type and maintenance measures are associated to build a solution map; The determination module is used to extract features from the pre-processed real-time operation data of the equipment, determine the corresponding fault type, input the fault type into the solution map, and generate a solution.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the intelligent diagnosis method for power equipment faults based on knowledge graph as described in any one of claims 1 to 8.

Images