Fault detection method, fault detection device and computer equipment

Abstract

This application relates to a fault detection method, a fault detection device, and a computer equipment. The method includes: acquiring operational status data of target power equipment in a target power grid area and a corresponding power grid operation and maintenance knowledge graph for the target power grid area; updating the power grid operation and maintenance knowledge graph based on the operational status data to obtain an updated power grid operation and maintenance knowledge graph; performing knowledge graph calculations based on the updated knowledge graph data to obtain calculation results; and determining the fault detection result for the target power grid area based on the calculation results. This method can improve fault detection efficiency.

Description

Technical Field

[0001] This application relates to the field of knowledge graph technology, and in particular to a fault detection method, a fault detection device, and a computer equipment. Background Technology

[0002] With the continuous advancement of power technology and the continuous improvement of users' living standards, the power grid has gradually become an indispensable infrastructure in modern society, economy, and daily life. The power grid includes various power equipment such as voltage transformation equipment, switching control equipment, transmission equipment, compensation and regulation equipment, and detection equipment. These various power equipment cooperate with each other to output power. For the power grid, operation and maintenance is the core work to ensure the stable operation of the power grid.

[0003] Currently, fault detection in the power grid mainly relies on manual methods based on power grid operation data, which results in low efficiency. Summary of the Invention

[0004] Therefore, it is necessary to provide a fault detection method, fault detection device, and computer equipment that can improve the fault detection efficiency of power grids, addressing the aforementioned technical problems.

[0005] Firstly, this application provides a fault detection method, comprising: acquiring operating status data of target power equipment in a target power grid area and a power grid operation and maintenance knowledge graph corresponding to the target power grid area, wherein the power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area, and each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment; updating the power grid operation and maintenance knowledge graph according to the operating status data to obtain an updated power grid operation and maintenance knowledge graph, wherein the updated power grid operation and maintenance knowledge graph includes updated knowledge graph data corresponding to the target power equipment; performing knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determining the fault detection result of the target power grid area based on the calculation results.

[0006] In one embodiment, the process of constructing a power grid operation and maintenance knowledge graph includes: acquiring historical operation and maintenance data of each power device in the target power grid area, the historical operation and maintenance data including historical fault records, electrical connection relationships between each power device and operation and maintenance association relationships between each power device; inputting the historical operation and maintenance data into a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0007] In one embodiment, historical operation and maintenance data is input into a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph output by the knowledge graph construction model. This includes: inputting historical operation and maintenance data into the knowledge graph construction model; identifying the historical operation and maintenance data through the knowledge graph construction model to obtain equipment entity information, equipment fault information, electrical connection information between the power equipment, and operation and maintenance relationship information between the power equipment; constructing triples corresponding to each power equipment based on the equipment entity information, equipment fault information, and operation and maintenance relationship information through the knowledge graph construction model; and connecting the triples based on the electrical connection information through the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph.

[0008] In one embodiment, the training process of the knowledge graph construction model includes: acquiring a training sample set, which includes multiple sample data and corresponding label data for each sample data. The sample data includes sample operation and maintenance data, sample fault records, and power technology documents of the sample power equipment. The label data includes label fault information of the sample power equipment, label electrical connection relationships between the sample power equipment, and label operation and maintenance relationships between the sample power equipment. The initial knowledge graph construction model is trained based on the training sample set to obtain the knowledge graph construction model.

[0009] In one embodiment, the updated knowledge graph data includes target device attribute information. Knowledge graph calculation is performed based on the target device attribute information, and the fault detection result of the target power device is determined based on the calculation result. This includes: comparing the target device attribute information with reference device attribute information, and determining whether the target power device includes a first abnormal power device based on the comparison result; if the target power device includes the first abnormal power device, determining a second abnormal power device that is associated with the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information; and generating a fault detection result based on the target device attribute information and the device attribute information of the second abnormal power device.

[0010] In one embodiment, determining a second abnormal power device that is associated with the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information includes: querying candidate fault devices that match both the fault type and the target device attribute information in the fault records included in the updated knowledge graph data; and determining, based on the relationship edges included in the updated knowledge graph data, power devices that have an electrical connection relationship and / or an operation and maintenance relationship with the first abnormal power device as the second abnormal power device from among the candidate fault devices.

[0011] In one embodiment, the method further includes: if the fault detection result indicates that the target power equipment has a target fault type, querying the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type; generating an operation and maintenance suggestion message based on the fault handling information, and outputting the operation and maintenance suggestion message.

[0012] In one embodiment, generating an operation and maintenance suggestion message based on fault handling information includes: combining the fault handling information to obtain multiple candidate operation and maintenance suggestion messages, each candidate operation and maintenance suggestion message including historical fault data; determining the similarity between each historical fault data and the target device attribute information, and determining the candidate operation and maintenance suggestion message with the highest similarity as the operation and maintenance suggestion message.

[0013] Secondly, this application also provides a fault detection device, comprising: a data acquisition module, used to acquire operating status data of target power equipment in a target power grid area and a power grid operation and maintenance knowledge graph corresponding to the target power grid area, wherein the power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area, and each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment; a knowledge graph updating module, used to update the power grid operation and maintenance knowledge graph according to the operating status data to obtain an updated power grid operation and maintenance knowledge graph, wherein the updated power grid operation and maintenance knowledge graph includes updated knowledge graph data corresponding to the target power equipment; and a fault detection module, used to perform knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determine the fault detection result of the target power grid area based on the calculation results.

[0014] Thirdly, this application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described in the first aspect.

[0015] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0016] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the method described in the first aspect.

[0017] The aforementioned fault detection method, fault detection device, and computer equipment first acquire the operating status data of the target power equipment in the target power grid area and the corresponding power grid operation and maintenance knowledge graph. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area, including entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment. The knowledge graph data enables the management of each power equipment in the target power grid area. The power grid operation and maintenance knowledge graph is updated based on the operating status data of the target power equipment. Based on the updated knowledge graph data of the target power equipment, knowledge graph calculations are performed to obtain the calculation results. The fault detection results of the target power grid area are determined based on the calculation results, realizing automated fault detection based on knowledge graphs and improving the fault detection efficiency of power equipment in the target power grid area of ​​the power system. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a diagram illustrating the application environment of a fault detection method in one embodiment;

[0020] Figure 2 This is a flowchart illustrating a fault detection method in one embodiment;

[0021] Figure 3 This is a flowchart illustrating the steps involved in constructing a power grid operation and maintenance knowledge graph in one embodiment.

[0022] Figure 4 This is a flowchart illustrating step 302 in one embodiment;

[0023] Figure 5 This is a flowchart illustrating the training steps of a knowledge graph construction model in one embodiment;

[0024] Figure 6 This is a flowchart illustrating step 203 in one embodiment;

[0025] Figure 7 This is a flowchart illustrating step 602 in one embodiment;

[0026] Figure 8 This is a flowchart illustrating the steps for generating operation and maintenance suggestion messages in one embodiment;

[0027] Figure 9 This is a flowchart illustrating step 802 in one embodiment;

[0028] Figure 10 This is a flowchart illustrating a fault detection method in another embodiment;

[0029] Figure 11 This is a structural block diagram of a fault detection device in one embodiment;

[0030] Figure 12 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0031] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0032] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0033] The fault detection method provided in this application embodiment can be applied to, for example, Figure 1 The application environment shown includes at least a server 101 and multiple power devices 102, wherein the server 101 includes a knowledge graph 101-1.

[0034] Server 101 is used to acquire the operating status data of target power equipment 102 in the target power grid area and the corresponding power grid operation and maintenance knowledge graph 101-1. Based on the operating status data, it updates the power grid operation and maintenance knowledge graph 101-1 to obtain the updated power grid operation and maintenance knowledge graph 101-1. Based on the updated knowledge graph data, it performs knowledge graph calculations to obtain calculation results and determines the fault detection results of the target power grid area based on the calculation results. The knowledge graph 101-1 may include the entity nodes corresponding to each power equipment, the equipment attribute information of each power equipment, and the relationship edges between each power equipment. Server 101 can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0035] Power equipment 102 is used for power regulation in the target power grid area. Server 101 can actively collect the operating status data of power equipment 102, and power equipment 102 can also send operating status data to server 101. Power equipment 102 includes, but is not limited to, generators, transformers, motors, circuit breakers, disconnect switches, load switches, fuses, power cables, busbars, reactors, capacitors, surge arresters, instrument transformers (current / voltage transformers), insulators, combined electrical appliances (such as GIS), grounding devices, measuring instruments (ammeters, voltmeters, energy meters, etc.), relay protection devices, automatic devices (such as automatic transfer switch), control and signaling devices, DC power supply equipment (such as battery banks), control cables, cooling equipment, etc.

[0036] In real-world scenarios, power equipment in a power system is intricately coupled with each other. A failure in one power device may lead to a cascading failure in multiple other power devices, or power dispatching / equipment adjustments made in response to a failure in one power device may cause cascading failures in other devices associated with that power device.

[0037] To address this, and to accurately detect faults in various power equipment, a power grid operation and maintenance knowledge graph can be constructed based on the electrical and operational connections between the power equipment. By leveraging the relationships between entity nodes in the knowledge graph, fault detection can be performed on multiple power equipment at once. Furthermore, by utilizing the equipment attribute information of each power equipment recorded in the knowledge graph, potential faults in other power equipment caused by abnormal operating status data of the power equipment can be detected, significantly improving the efficiency of power equipment fault detection.

[0038] In one exemplary embodiment, such as Figure 2 As shown, a fault detection method is provided, which can be applied to... Figure 1 The following steps, 201 to 203, are used as an example to illustrate the process of using a server in the example.

[0039] Step 201: Obtain the operating status data of the target power equipment in the target power grid area and the power grid operation and maintenance knowledge graph corresponding to the target power grid area.

[0040] In this application, the power grid operation and maintenance knowledge graph is constructed based on multi-dimensional data of various power equipment in the target power grid area. The power grid operation and maintenance knowledge graph includes entity nodes of each power equipment, equipment attribute information associated with the entity nodes, and relationship edges between the power equipment. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between the power equipment. Among these, the equipment attribute information includes the current operating status data of each power equipment, such as voltage data, current data, power data, temperature data, and current frequency. According to at least one of the data and runtime data, the equipment attribute information may also include historical fault information and / or historical fault handling information (the methods used to handle the faults) of each power equipment; the relationship edge may include the electrical connection relationship and / or operation and maintenance association relationship between each power equipment, wherein the electrical connection relationship may be a physical cable / electromagnetic connection, which can be connected by high voltage or low voltage, and the operation and maintenance association relationship may be a logical operation and maintenance relationship, such as: between cooling equipment and generator set, the cooling equipment is used to cool the generator set, if the generator set temperature is too high, it may be due to the failure of the cooling equipment, and the cooling equipment and generator set may not be electrically connected.

[0041] During implementation, the server acquires operational status data of target power equipment in the target power grid area at a preset frequency, and obtains the corresponding power grid operation and maintenance knowledge graph for the target power grid area. The target power equipment can be a critical power device prone to failure in the target power grid area, a critical electrical device with numerous electrical connections, or a critical electrical device with critical electrical functions. In this application, the operational status data may include at least one of the following: current voltage data, current data, current frequency data, temperature data, operating time, and power data of the power equipment.

[0042] Step 202: Update the power grid operation and maintenance knowledge graph based on the operating status data to obtain the updated power grid operation and maintenance knowledge graph.

[0043] During implementation, the power grid operation and maintenance knowledge graph includes equipment attribute information for each power device. This equipment attribute information includes the current operating status data of the power devices. After obtaining the operating status data, the server updates the equipment attribute information of the target power devices in the power grid operation and maintenance knowledge graph based on the operating status data of each target power device, resulting in an updated power grid operation and maintenance knowledge graph. The updated power grid operation and maintenance knowledge graph includes updated knowledge graph data corresponding to the target power devices, where the knowledge graph data includes the equipment attribute information and / or relationship edges of the target power devices.

[0044] During execution, the server can map the real-time collected operating status data to the corresponding equipment entity nodes in the power grid operation and maintenance knowledge graph, so as to update the equipment attribute information of the equipment nodes in the graph.

[0045] Step 203: Perform knowledge graph calculations based on the updated knowledge graph data to obtain the calculation results, and determine the fault detection results for the target power grid area based on the calculation results.

[0046] During implementation, the server performs knowledge graph calculations based on the updated knowledge graph data to obtain calculation results. Based on the calculation results, it determines whether the power equipment in the target power grid area has malfunctioned. If no malfunction has occurred, a normal fault detection result for the target power grid area is obtained. If there is a malfunctioning power equipment, the fault detection result for the target power grid area is determined based on the equipment attribute information of the malfunctioning power equipment.

[0047] During execution, the server can determine whether there are equipment faults in the target power grid area based on the calculation results of the graph, and locate the faulty equipment.

[0048] The aforementioned fault detection method first acquires the operating status data of the target power equipment in the target power grid area and the corresponding power grid operation and maintenance knowledge graph. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area, including entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment. Management of each power equipment in the target power grid area is realized through the knowledge graph data. The power grid operation and maintenance knowledge graph is updated based on the operating status data of the target power equipment. Based on the updated knowledge graph data of the target power equipment, knowledge graph calculation is performed to obtain the calculation result. The fault detection result of the target power grid area is determined based on the calculation result, realizing automated fault detection based on knowledge graph and improving the fault detection efficiency of power equipment in the target power grid area of ​​the power system.

[0049] Based on the above exemplary embodiment, the following provides a fault detection method in one or more exemplary embodiments, which is applied to... Figure 1 Taking the server in the example, the following content will be used for explanation.

[0050] In the process of constructing a power grid operation and maintenance knowledge graph, a power grid operation and maintenance knowledge graph can be generated based on the historical operation and maintenance data of each power device in the target power grid area and the relationships between the power devices, combined with a pre-trained knowledge graph construction model; one optional implementation method provided in this application is as follows: Figure 3As shown, the construction process of the power grid operation and maintenance knowledge graph includes steps 301 to 302:

[0051] Step 301: Obtain historical operation and maintenance data of each power device in the target power grid area.

[0052] In this application, historical operation and maintenance data refers to the data recorded when power equipment is operated and maintained in a historical period. Historical operation and maintenance data may include fault information when a fault occurs in a historical period, the methods used to resolve the fault, and the relationship between various power equipment. Historical operation and maintenance data includes historical fault records, electrical connection relationships between various power equipment, and operation and maintenance relationship between various power equipment.

[0053] During implementation, the server first obtains the operation and maintenance data of each power device in the target power grid area within a historical time period as historical operation and maintenance data.

[0054] Step 302: Input historical operation and maintenance data into the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0055] During implementation, the server inputs the historical operation and maintenance data of each power device into the knowledge graph construction model, and constructs the knowledge graph through the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0056] The knowledge graph construction model can be a pre-trained neural network model, which may include an entity recognition module and a relation extraction module. The entity recognition module can identify each power device in historical operation and maintenance data, and the relation extraction module can extract the relationships between the power devices. These relationships may include electrical connection relationships and / or operation and maintenance relationships. The entity recognition module and the relation extraction module can be trained separately or together.

[0057] It should be noted that the knowledge graph construction model can use any neural network architecture in the existing technology to implement its functions. The knowledge graph construction model can identify the entity name, fault record, fault handling record, operation and maintenance relationship, and electrical connection relationship of power equipment in text data / image data, and construct entity nodes and / or binary (triple) based on entity name, fault record and fault handling record, and connect each entity node and / or binary (triple) based on operation and maintenance relationship and / or electrical connection relationship to obtain the power grid operation and maintenance knowledge graph.

[0058] In one optional implementation provided by this application, a knowledge graph construction model is used to construct a power grid operation and maintenance knowledge graph, which improves the construction efficiency of the knowledge graph and thus improves the efficiency of fault detection.

[0059] In the process of constructing a power grid operation and maintenance knowledge graph, the knowledge graph construction model can first perform data identification, then construct triples based on the identified data, and connect the triples to obtain the power grid operation and maintenance knowledge graph; in one optional implementation method provided in this application, such as Figure 4 As shown, step 302 includes steps 401 to 403:

[0060] Step 401: Input historical operation and maintenance data into the knowledge graph construction model. The knowledge graph construction model is used to identify the historical operation and maintenance data to obtain the equipment entity information, equipment fault information, electrical connection information between the power equipment, and operation and maintenance relationship information between the power equipment.

[0061] During implementation, the server inputs historical operation and maintenance data into the knowledge graph construction model. The knowledge graph construction model first identifies the historical operation and maintenance data to obtain the equipment entity information corresponding to each power device, the equipment fault information corresponding to each power device, the electrical connection information between each power device, and the operation and maintenance relationship information between each power device. Among them, the equipment fault information may include the fault information of the power device when it fails, and / or the measures taken to resolve the fault. The operation and maintenance relationship information may include the association information of each power device when it fails. This association information can represent the related faults between faults. For example, if a power device A fails, it will cause a power device B to fail. Therefore, there is a related fault between power devices A and B. This association information can also represent the operation and maintenance association when resolving faults. For example, if a power device C fails, it needs to be resolved by adjusting power device D. Therefore, there is an operation and maintenance association between power devices C and D.

[0062] During execution, the knowledge graph construction model can identify the equipment entity information corresponding to each power device and the equipment fault information corresponding to each power device through the entity recognition module. The knowledge graph construction device can also extract the operation and maintenance relationship information between each power device through the relationship extraction module.

[0063] For example, for the entity recognition module, a sequence labeling structure can be used to predict the entity type of each word based on the encoding vector, thereby obtaining the entity information / entity node corresponding to each power device; for the relation extraction module, an entity pair classification structure can be used to classify the relationships between the identified entity combinations (such as determining whether there is a causal relationship between "transformer C" and "fuse failure").

[0064] Step 402: Construct a model using a knowledge graph to build triples for each power device based on the device entity information, device fault information, and maintenance relationship information.

[0065] During implementation, the server constructs a model using a knowledge graph. Based on the entity information of each device, it generates entity nodes corresponding to each power device. Based on the fault information and maintenance relationship information of each device, it constructs device attribute information for each power device. Finally, based on the generated entity nodes and device attribute information, it constructs triples for each power device. The device attribute information for each power device includes fault information and maintenance relationship information; therefore, the constructed data set can be called a triple. If the device attribute information is considered as a single data point, then a binary tuple corresponding to each power device can be constructed using the knowledge graph model based on the entity information, fault information, and maintenance relationship information.

[0066] Step 403: Construct a model using a knowledge graph, connecting each triplet based on electrical connection information and / or operation and maintenance related information to obtain a power grid operation and maintenance knowledge graph.

[0067] During implementation, the server knowledge graph construction model connects each triple or binary based on the electrical connection relationships between the identified power equipment to obtain the power grid operation and maintenance knowledge graph.

[0068] During the execution process, for the construction of the power grid operation and maintenance knowledge graph, historical operation and maintenance knowledge data and historical fault records can be used as text inputs to the knowledge graph construction model. The knowledge graph construction model processes the input text data to identify power grid equipment entities, fault entities, and their attribute information, and extracts the operation and maintenance related relationships between entities to form a structured set of entity and relationship triples. The topology connection relationship is parsed into structured triples representing the physical connection relationship between power grid equipment. The operation and maintenance related entity, relationship triple set, and topology connection relationship triples are integrated to construct the power grid operation and maintenance knowledge graph. The power grid operation and maintenance knowledge graph includes power grid equipment entity nodes, fault entity nodes, equipment attribute information, operation and maintenance relationship edges between equipment, and topology connection relationship edges between equipment.

[0069] For example, inputting the text: "On March 10, 2024, the cooling fan of the #1 main transformer in the 220kV substation stopped, causing the oil temperature to rise to 85℃. The SCADA system provides the connection relationship: <Main transformer #1, connection, 110kV bus>", firstly, entity recognition is performed to obtain: <Main transformer #1, equipment type, transformer>, <cooling fan stopped, fault type, mechanical defect>, <oil temperature 85℃, attribute, temperature over-limit>, <cooling fan stopped, caused, oil temperature 85℃>; the operation and maintenance relationship information is: <Main transformer #1, connection, 110kV bus>; the power grid operation and maintenance knowledge graph is: [Main transformer #1]-[connection]-[110kV bus], [cooling fan stopped]-[caused]-[oil temperature 85℃].

[0070] One optional implementation method provided in this application is to perform entity recognition, operation and maintenance relationship recognition, data group construction and data group connection in historical operation and maintenance data through a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph. The construction efficiency of the power grid operation and maintenance knowledge graph is improved by constructing the knowledge graph through the knowledge graph construction model.

[0071] During execution, the knowledge graph construction model can be pre-trained, using historical operation and maintenance data as sample data and entities and relation edges as labels for training the knowledge graph construction model; in one optional implementation provided by this application, such as Figure 5 As shown, the training process of the knowledge graph construction model includes steps 501 to 502:

[0072] Step 501: Obtain the training sample set.

[0073] In this application, the training sample set includes multiple sample data and corresponding label data for each sample data. The sample data includes sample operation and maintenance data of sample power equipment, sample fault records and power technology documents. The label data includes label fault information of sample power equipment, label electrical connection relationship between each sample power equipment and label operation and maintenance relationship between each sample power equipment.

[0074] For example, sample data can include historical power grid operation and maintenance reports, fault records, and technical documents. It can include annotations for power grid-specific entity types: equipment entities (such as transformers and circuit breakers), fault entities (such as overload and short circuit), and attribute entities (such as temperature and current). It can also include annotations for relationship types: topological connection relationships (such as physical connections between equipment), fault causal relationships (such as defects causing faults), and handling relationships (such as the association of measures with equipment). Label data can be entity boundaries (such as "circuit breaker" being labeled as an equipment entity) and annotations for entity relationship triples (such as <circuit breaker A, connection, bus B>). The training sample set can also be divided into a training set and a test set according to a preset ratio. The model is trained based on the training set, and the trained model is tested using the test set.

[0075] Step 502: Train the initial knowledge graph construction model based on the training sample set to obtain the knowledge graph construction model.

[0076] During implementation, for each set of training data (sample data and corresponding label data), the sample data is input into the initial knowledge graph construction model. The initial knowledge graph construction model performs entity recognition / data group recognition and knowledge graph construction on the sample data to obtain the predicted knowledge graph output by the initial knowledge graph construction model. Based on the predicted knowledge graph and label data, loss is calculated to obtain the training loss. The model parameters of the initial knowledge graph construction model are adjusted based on the training loss until the samples are exhausted or the training loss converges, thus obtaining the knowledge graph construction model.

[0077] During execution, the loss functions for entity recognition and relation extraction can be jointly optimized through the backpropagation algorithm. When the initial knowledge graph construction model has an entity F1 value greater than or equal to a preset threshold and a relation accuracy greater than or equal to a preset threshold in the test set, the performance is deemed to have met the target.

[0078] In one optional implementation provided by this application, historical operation and maintenance data of each power device are used as training data and label data. The initial knowledge graph construction model is trained using the historical operation and maintenance data to obtain the knowledge graph construction model, thereby improving the recognition accuracy of the knowledge graph construction model and thus improving the accuracy of fault detection in the target power grid area.

[0079] In the process of knowledge graph computation, reference equipment attribute information for each target power device can be pre-set. Based on the reference equipment attribute information, abnormal / faulty power devices among the target power devices are identified. Fault detection results are generated based on the equipment attribute information of the abnormal / faulty power devices. In one optional implementation provided by this application, such as… Figure 6 As shown, step 203 includes steps 601 to 603:

[0080] Step 601: Compare the target device attribute information and the reference device attribute information, and determine whether the target power device includes the first abnormal power device based on the comparison result.

[0081] During implementation, for each target power device, the server compares the target device attribute information with the reference device attribute information. It compares whether the target device attribute information exceeds the threshold range of the reference device attribute information. If the target device attribute information exceeds the threshold range of the reference device attribute information, the target power device is determined to be the first abnormal power device. If the target device attribute information does not exceed the threshold range of the reference device attribute information, the target power device is determined not to be the first abnormal power device. The above operation is performed for each target power device. If the target device attribute information exceeds the threshold range of the reference device attribute information, the target power device is determined to be the first abnormal power device.

[0082] During execution, the real-time status attributes of the device nodes can be compared with the preset normal operation threshold range to identify device nodes with abnormal status.

[0083] Step 602: If the target power equipment includes the first abnormal power equipment, determine the second abnormal power equipment that is associated with the first abnormal power equipment based on the fault type of the first abnormal power equipment and the attribute information of the target equipment.

[0084] During implementation, when the target power equipment includes a first abnormal power equipment, the server determines the fault type of the first abnormal power equipment based on the target equipment attribute information, and determines a second abnormal power equipment that is associated with the first abnormal power equipment based on the fault type and the target equipment attribute information. The target equipment attribute information may include operational status data exceeding the reference equipment attribute information, such as at least one of the following: abnormal voltage data, abnormal current data, abnormal current frequency data, abnormal temperature data, and abnormal power data.

[0085] During execution, if the target power equipment includes a first abnormal power equipment, the server can determine the fault type of the first abnormal power equipment based on the abnormal operating status data included in the target equipment attribute information of the first abnormal power equipment, and determine the second abnormal power equipment that is associated with the first abnormal power equipment based on the fault type and abnormal operating status data of the first abnormal power equipment.

[0086] For example, if the temperature of the first abnormal power device exceeds the reference temperature range, the fault type of the first abnormal power device can be determined as overheating. Then, based on the overheating fault type and the temperature of the first abnormal power device, a second abnormal power device that is associated with the first abnormal power device can be identified.

[0087] Step 603: Generate fault detection results based on the target device attribute information and the device attribute information of the second abnormal power device.

[0088] During implementation, the server generates a fault detection result corresponding to the fault type of the first abnormal power device based on the target device attribute information and the device attribute information of the second abnormal power device. During execution, the server generates the fault detection result based on the abnormal operating status data included in the target device attribute information of the first abnormal power device and the abnormal operating status data included in the device attribute information of the second abnormal power device.

[0089] For example, during the process of obtaining fault detection results through knowledge graph calculation, real-time data alarm: Main transformer #1 oil temperature: 82℃ → mapped to the "Main transformer #1" node in the graph; the graph reasoning process is as follows: threshold comparison: 82℃ > preset threshold (80℃) → marked as abnormal; topology analysis: locate related equipment along the connection edge, affecting the 110kV bus and downstream feeder; historical matching: search the graph to find similar fault chains (cooling fan stops → oil temperature exceeds limit); generate the fault detection result "Main transformer #1 temperature 82℃ too high".

[0090] It should be noted that the fault types of the first abnormal power equipment can include multiple types. For example, if there are simultaneously abnormal power equipment A with excessively high temperature, abnormal power equipment B with overcurrent, and abnormal power equipment C with overvoltage, then fault detection results are generated sequentially for abnormal power equipment A, B, and C. Furthermore, there may be correlations between the faults of abnormal power equipment A, B, and C. For example, the excessive temperature of abnormal power equipment A may be caused by the overcurrent of abnormal power equipment B, and the overcurrent of power equipment B may be caused by the overvoltage of abnormal power equipment C. In this regard, the fault detection results can include the fault correlations between the abnormal power equipment. Compared with the prior art, the beneficial effect of this application is that it can solve the fault detection in this situation, avoiding the drawback of single-point detection in traditional technology where repairing one power equipment may cause another to malfunction, thereby improving the fault detection efficiency of the entire power system.

[0091] In determining the second abnormal power device that is associated with the first abnormal power device, one can first query candidate fault devices associated with the first abnormal power device based on the fault records included in the power grid operation and maintenance knowledge graph, and then select the second abnormal power device that has a relationship edge with the first abnormal power device from the candidate fault devices; in one optional implementation provided by this application, such as Figure 7 As shown, step 602 includes steps 701 to 702:

[0092] Step 701: In the updated knowledge graph data, query the fault records to find candidate fault devices that match both the fault type and the target device attribute information.

[0093] During implementation, the server queries the updated knowledge graph data, including fault records, for candidate faulty devices that match both the fault type and the target device attribute information. During execution, the server can query fault records within the updated knowledge graph data, including the device attribute information of the first abnormal device, and then query these fault records for candidate faulty devices that match both the fault type and the abnormal operating status data included in the target device attribute information.

[0094] Furthermore, regarding the cause of the first abnormal device recorded in the fault record when the fault occurred, the first candidate fault device that matches the fault type is first determined, and then the data that matches the abnormal operating status data in the fault record is further determined from the first candidate fault device based on the abnormal operating status data.

[0095] For example, different equipment failures can cause power equipment to generate abnormal operating status data with different temperatures. First, identify the first candidate fault equipment that matches the fault type of excessive temperature, and then determine the candidate fault equipment from the first candidate fault equipment based on the specific temperature value.

[0096] Step 702: Based on the relationship edges included in the updated knowledge graph data, identify the power equipment that has an electrical connection relationship and / or operation and maintenance relationship with the first abnormal power equipment as the second abnormal power equipment among the candidate faulty equipment.

[0097] In real-world scenarios, fault records may be historical records that are no longer applicable in the updated knowledge graph. To address this, candidate fault devices connected by relational edges can be selected as abnormal power devices.

[0098] During implementation, the server, based on the relationship edges included in the updated knowledge graph data, identifies the power equipment among the candidate faulty equipment that has an electrical connection and / or operation and maintenance relationship with the first abnormal power equipment as the second abnormal power equipment. The relationship edges include both electrical connection and / or operation and maintenance relationships.

[0099] One optional implementation method provided in this application uses a knowledge graph combined with historical fault records to identify a second abnormal power device that has a fault association with the first abnormal power device, thereby improving the accuracy of abnormal power devices and thus improving the fault detection efficiency of the power system.

[0100] In real-world scenarios, after a fault is detected, a suggested message can be generated and output by combining a knowledge graph with historical fault handling information; in one optional implementation provided in this application, such as Figure 8 As shown, the method further includes steps 801 to 802:

[0101] Step 801: If the fault detection result indicates that the target power equipment has a target fault type, query the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type.

[0102] During implementation, if the fault detection result indicates that the target power equipment has a target fault type, the server queries the equipment attribute information of the target power equipment in the power grid operation and maintenance knowledge graph to find fault handling information that matches the target fault type.

[0103] During execution, multi-dimensional information (fault handling information) can be extracted from the power grid operation and maintenance knowledge graph based on the faulty equipment or faulty area. The multi-dimensional information package (fault handling information) includes the attribute information of the faulty equipment, the current real-time status data, the topology-related equipment information directly connected to the faulty equipment, similar cases matching the current fault in the historical operation and maintenance records and the operation and maintenance measures taken at that time, and the standard handling process for this type of fault predefined in the power grid operation and maintenance knowledge graph.

[0104] Step 802: Generate and output the operation and maintenance suggestion message based on the fault handling information.

[0105] During implementation, the server generates maintenance suggestion messages based on the retrieved fault handling information and outputs these messages to management users.

[0106] One optional implementation method provided in this application generates operation and maintenance suggestion messages by matching historical operation and maintenance information, and outputs suggestions based on the actual operation and maintenance process, thereby improving the reliability and effectiveness of the operation and maintenance suggestion messages.

[0107] During the generation of operation and maintenance suggestion messages, multiple fault handling methods may exist in historical faults. The fault handling method that best matches the fault detection results can be selected from these methods. In one optional implementation provided in this application, such as... Figure 9 As shown, step 802 includes steps 901 to 902:

[0108] Step 901: Combine the fault handling information to obtain multiple candidate operation and maintenance suggestion messages.

[0109] During implementation, the server can combine and process the acquired fault handling information to obtain multiple candidate maintenance suggestion messages; each candidate maintenance suggestion message includes historical fault data.

[0110] For example, for the fault type "overheating", the fault handling methods and corresponding data for multiple overheating conditions are combined to obtain multiple candidate maintenance suggestion messages that include fault handling methods and fault temperatures at historical fault times.

[0111] Step 902: Determine the similarity between each historical fault data and the target device attribute information, and determine the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0112] During implementation, the server can determine the similarity between each historical fault data and the target device attribute information, and identify the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0113] For example, based on the similarity calculation between abnormal temperature values ​​in abnormal operating status data and historical fault data, the candidate operation and maintenance suggestion message with the highest similarity is determined as the operation and maintenance suggestion message.

[0114] In addition, based on fault handling information, multiple operation and maintenance suggestions can be generated by matching historical best solutions or combining processes. The generated operation and maintenance suggestions are sorted by priority and output, and the source of each suggestion is marked to obtain operation and maintenance suggestion messages. That is, operation and maintenance suggestion messages can include multiple messages.

[0115] One optional implementation method provided in this application determines the operation and maintenance suggestion message that best matches the current fault scenario by matching historical optimal solutions, thereby improving the reliability and effectiveness of the operation and maintenance suggestion message.

[0116] In one embodiment, see Figure 10 The diagram illustrates a flowchart of a fault detection method provided in an embodiment of this application, which can be applied to... Figure 1 In the server shown. For example Figure 10 As shown, the fault detection method may include the following steps:

[0117] Step 1001: Obtain the operating status data of the target power equipment in the target power grid area and the power grid operation and maintenance knowledge graph corresponding to the target power grid area.

[0118] Step 1002: Update the power grid operation and maintenance knowledge graph based on the operating status data to obtain the updated power grid operation and maintenance knowledge graph.

[0119] Step 1003: Compare the target device attribute information and the reference device attribute information to obtain abnormal operating status data that exceeds the threshold range of the reference device attribute information, and determine whether the target power equipment includes the first abnormal power equipment based on the abnormal operating status data.

[0120] Step 1004: If the target power equipment includes the first abnormal power equipment, determine the second abnormal power equipment that is associated with the first abnormal power equipment based on the fault type and abnormal operating status data of the first abnormal power equipment.

[0121] Step 1005: Generate a fault detection result based on the abnormal operating status data of the first abnormal power equipment and the abnormal operating status data included in the equipment attribute information of the second abnormal power equipment.

[0122] Step 1006: Query the fault handling information that matches the fault type of the first abnormal power equipment in the power grid operation and maintenance knowledge graph.

[0123] Step 1007: Combine the fault handling information to obtain multiple candidate operation and maintenance suggestion messages, each of which includes historical fault data.

[0124] Step 1008: Determine the similarity between each historical fault data and the target device attribute information, and determine the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0125] It should be noted that any one or more of steps 1001 to 1008 can be combined to form a new implementation method according to the needs of implementation and deployment. Furthermore, any one or more technical features in the technical solution composed of steps 1001 to 1008 can also be combined to form a new implementation method according to the actual deployment needs, or technical features in one or more optional implementation methods provided by one or more of the above embodiments can be combined to form a new implementation method. These will not be elaborated on here.

[0126] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0127] Based on the same inventive concept, this application also provides a fault detection device for implementing the fault detection method described above. The solution provided by this device is similar to the solution described in the above method; therefore, the specific limitations in one or more fault detection device embodiments provided below can be found in the limitations of the fault detection method described above, and will not be repeated here.

[0128] In one exemplary embodiment, such as Figure 11As shown, a fault detection device is provided, including: a data acquisition module 1101, a knowledge graph update module 1102, and a fault detection module 1103. The data acquisition module 1101 is used to acquire the operating status data of target power equipment in a target power grid area and the corresponding power grid operation and maintenance knowledge graph of the target power grid area. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes the entity node corresponding to each power equipment, the equipment attribute information of each power equipment, and the relationship edges between each power equipment. The knowledge graph update module 1102 is used to update the power grid operation and maintenance knowledge graph according to the operating status data, obtaining an updated power grid operation and maintenance knowledge graph. The updated power grid operation and maintenance knowledge graph includes the updated knowledge graph data corresponding to the target power equipment. The fault detection module 1103 is used to perform knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determine the fault detection result of the target power grid area based on the calculation results.

[0129] In one embodiment, the device further includes a first knowledge graph construction unit and a second knowledge graph construction unit, wherein: the first knowledge graph construction unit is used to acquire historical operation and maintenance data of each power device in the target power grid area, the historical operation and maintenance data including historical fault records, electrical connection relationships between each power device and operation and maintenance association relationships between each power device; the second knowledge graph construction unit is used to input the historical operation and maintenance data into the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0130] In one embodiment, the second knowledge graph construction unit includes a third knowledge graph construction unit, a fourth knowledge graph construction unit, and a fifth knowledge graph construction unit, wherein: the third knowledge graph construction unit is used to input historical operation and maintenance data into the knowledge graph construction model, and to identify the historical operation and maintenance data through the knowledge graph construction model to obtain the equipment entity information, equipment fault information, electrical connection information between the power equipment, and operation and maintenance relationship information between the power equipment; the fourth knowledge graph construction unit is used to construct triples corresponding to each power equipment based on the equipment entity information, equipment fault information, and operation and maintenance relationship information through the knowledge graph construction model; the fifth knowledge graph construction unit is used to connect the triples based on the electrical connection information through the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph.

[0131] In one embodiment, the apparatus further includes a sample set acquisition module and a training module, wherein: the sample set acquisition module is used to acquire a training sample set, which includes multiple sample data and label data corresponding to each sample data. The sample data includes sample operation and maintenance data of sample power equipment, sample fault records and power technology documents. The label data includes label fault information of sample power equipment, label electrical connection relationships between sample power equipment, and label operation and maintenance relationships between sample power equipment; the training module is used to train the initial knowledge graph construction model based on the training sample set to obtain the knowledge graph construction model.

[0132] In one embodiment, the fault detection module 1103 includes a comparison unit, an abnormal device determination unit, and a fault detection unit, wherein: the comparison unit is used to compare the target device attribute information and the reference device attribute information, and determine whether the target power device includes a first abnormal power device based on the comparison result; the abnormal device determination unit is used to determine a second abnormal power device that is associated with a fault of the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information if the target power device includes the first abnormal power device; and the fault detection unit is used to generate a fault detection result based on the target device attribute information and the device attribute information of the second abnormal power device.

[0133] In one embodiment, the abnormal device determination unit includes a candidate abnormal device determination unit and an abnormal device selection unit, wherein: the candidate abnormal device determination unit is used to query candidate fault devices that match the fault type and target device attribute information in the fault records included in the updated knowledge graph data; the abnormal device selection unit is used to determine, based on the relation edges included in the updated knowledge graph data, the power device that has an electrical connection relationship and / or operation and maintenance association relationship with the first abnormal power device as the second abnormal power device.

[0134] In one embodiment, the device further includes a fault handling information query module and an operation and maintenance suggestion generation module, wherein: the fault handling information query module is used to query fault handling information matching the target fault type in the power grid operation and maintenance knowledge graph when the fault detection result indicates that the target power equipment has a target fault type; the operation and maintenance suggestion generation module is used to generate an operation and maintenance suggestion message based on the fault handling information and output the operation and maintenance suggestion message.

[0135] In one embodiment, the operation and maintenance suggestion generation module includes a candidate suggestion generation unit and an operation and maintenance suggestion selection unit, wherein: the candidate suggestion generation unit is used to combine fault handling information to obtain multiple candidate operation and maintenance suggestion messages, each candidate operation and maintenance suggestion message including historical fault data; the operation and maintenance suggestion selection unit is used to determine the similarity between each historical fault data and the target device attribute information, and to determine the candidate operation and maintenance suggestion message with the highest similarity as the operation and maintenance suggestion message.

[0136] Each module in the device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0137] In one exemplary embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 12 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores fault detection data for the target power grid area. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a fault detection method.

[0138] Those skilled in the art will understand that Figure 12 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0139] In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to perform the following steps: acquiring operating status data of target power equipment in a target power grid area and a power grid operation and maintenance knowledge graph corresponding to the target power grid area. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment; updating the power grid operation and maintenance knowledge graph according to the operating status data to obtain an updated power grid operation and maintenance knowledge graph, which includes updated knowledge graph data corresponding to the target power equipment; performing knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determining the fault detection results of the target power grid area based on the calculation results.

[0140] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring historical operation and maintenance data of each power device in the target power grid area, the historical operation and maintenance data including historical fault records, electrical connection relationships between each power device and operation and maintenance association relationships between each power device; inputting the historical operation and maintenance data into a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0141] In one embodiment, when the processor executes the computer program, it further performs the following steps: inputting historical operation and maintenance data into a knowledge graph construction model; identifying the historical operation and maintenance data through the knowledge graph construction model to obtain equipment entity information corresponding to each power equipment, equipment fault information corresponding to each power equipment, electrical connection information between each power equipment, and operation and maintenance relationship information between each power equipment; constructing triples corresponding to each power equipment through the knowledge graph construction model based on the equipment entity information, equipment fault information, and operation and maintenance relationship information; and connecting each triple according to the electrical connection information through the knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph.

[0142] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring a training sample set, which includes multiple sample data and label data corresponding to each sample data. The sample data includes sample operation and maintenance data of sample power equipment, sample fault records, and power technology documents. The label data includes label fault information of sample power equipment, label electrical connection relationships between sample power equipment, and label operation and maintenance relationships between sample power equipment. The initial knowledge graph construction model is trained based on the training sample set to obtain the knowledge graph construction model.

[0143] In one embodiment, when the processor executes the computer program, it further performs the following steps: comparing the target device attribute information and the reference device attribute information, and determining whether the target power device includes a first abnormal power device based on the comparison result; if the target power device includes the first abnormal power device, determining a second abnormal power device that is associated with a fault in the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information; and generating a fault detection result based on the target device attribute information and the device attribute information of the second abnormal power device.

[0144] In one embodiment, when the processor executes the computer program, it further performs the following steps: querying the fault records included in the updated knowledge graph data for candidate fault devices that match both the fault type and the target device attribute information; and determining, based on the relation edges included in the updated knowledge graph data, a power device that has an electrical connection relationship and / or an operation and maintenance association relationship with the first abnormal power device as the second abnormal power device.

[0145] In one embodiment, when the processor executes the computer program, it further performs the following steps: if the fault detection result indicates that the target power equipment has a target fault type, it queries the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type; it generates an operation and maintenance suggestion message based on the fault handling information and outputs the operation and maintenance suggestion message.

[0146] In one embodiment, when the processor executes the computer program, it further performs the following steps: combining fault handling information to obtain multiple candidate maintenance suggestion messages, each candidate maintenance suggestion message including historical fault data; determining the similarity between each historical fault data and the target device attribute information, and determining the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0147] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, it performs the following steps: acquiring operating status data of target power equipment in a target power grid area and a power grid operation and maintenance knowledge graph corresponding to the target power grid area. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment; updating the power grid operation and maintenance knowledge graph according to the operating status data to obtain an updated power grid operation and maintenance knowledge graph, which includes updated knowledge graph data corresponding to the target power equipment; performing knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determining the fault detection results of the target power grid area based on the calculation results.

[0148] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring historical operation and maintenance data of each power device in the target power grid area, the historical operation and maintenance data including historical fault records, electrical connection relationships between each power device and operation and maintenance association relationships between each power device; inputting the historical operation and maintenance data into a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0149] In one embodiment, when the processor executes the computer program, it further performs the following steps: inputting historical operation and maintenance data into a knowledge graph construction model; identifying the historical operation and maintenance data through the knowledge graph construction model to obtain equipment entity information corresponding to each power equipment, equipment fault information corresponding to each power equipment, electrical connection information between each power equipment, and operation and maintenance relationship information between each power equipment; constructing triples corresponding to each power equipment through the knowledge graph construction model based on the equipment entity information, equipment fault information, and operation and maintenance relationship information; and connecting each triple according to the electrical connection information through the knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph.

[0150] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring a training sample set, which includes multiple sample data and label data corresponding to each sample data. The sample data includes sample operation and maintenance data of sample power equipment, sample fault records, and power technology documents. The label data includes label fault information of sample power equipment, label electrical connection relationships between sample power equipment, and label operation and maintenance relationships between sample power equipment. The initial knowledge graph construction model is trained based on the training sample set to obtain the knowledge graph construction model.

[0151] In one embodiment, when the processor executes the computer program, it further performs the following steps: comparing the target device attribute information and the reference device attribute information, and determining whether the target power device includes a first abnormal power device based on the comparison result; if the target power device includes the first abnormal power device, determining a second abnormal power device that is associated with a fault in the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information; and generating a fault detection result based on the target device attribute information and the device attribute information of the second abnormal power device.

[0152] In one embodiment, when the processor executes the computer program, it further performs the following steps: querying the fault records included in the updated knowledge graph data for candidate fault devices that match both the fault type and the target device attribute information; and determining, based on the relation edges included in the updated knowledge graph data, a power device that has an electrical connection relationship and / or an operation and maintenance association relationship with the first abnormal power device as the second abnormal power device.

[0153] In one embodiment, when the processor executes the computer program, it further performs the following steps: if the fault detection result indicates that the target power equipment has a target fault type, it queries the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type; it generates an operation and maintenance suggestion message based on the fault handling information and outputs the operation and maintenance suggestion message.

[0154] In one embodiment, when the processor executes the computer program, it further performs the following steps: combining fault handling information to obtain multiple candidate maintenance suggestion messages, each candidate maintenance suggestion message including historical fault data; determining the similarity between each historical fault data and the target device attribute information, and determining the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0155] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps: acquiring operational status data of target power equipment in a target power grid area and a power grid operation and maintenance knowledge graph corresponding to the target power grid area, wherein the power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area, and each knowledge graph data includes entity nodes corresponding to each power equipment, equipment attribute information of each power equipment, and relationship edges between each power equipment; updating the power grid operation and maintenance knowledge graph according to the operational status data to obtain an updated power grid operation and maintenance knowledge graph, wherein the updated power grid operation and maintenance knowledge graph includes updated knowledge graph data corresponding to the target power equipment; performing knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and determining the fault detection results of the target power grid area based on the calculation results.

[0156] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring historical operation and maintenance data of each power device in the target power grid area, the historical operation and maintenance data including historical fault records, electrical connection relationships between each power device and operation and maintenance association relationships between each power device; inputting the historical operation and maintenance data into a knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

[0157] In one embodiment, when the processor executes the computer program, it further performs the following steps: inputting historical operation and maintenance data into a knowledge graph construction model; identifying the historical operation and maintenance data through the knowledge graph construction model to obtain equipment entity information corresponding to each power equipment, equipment fault information corresponding to each power equipment, electrical connection information between each power equipment, and operation and maintenance relationship information between each power equipment; constructing triples corresponding to each power equipment through the knowledge graph construction model based on the equipment entity information, equipment fault information, and operation and maintenance relationship information; and connecting each triple according to the electrical connection information through the knowledge graph construction model to obtain a power grid operation and maintenance knowledge graph.

[0158] In one embodiment, when the processor executes the computer program, it further performs the following steps: acquiring a training sample set, which includes multiple sample data and label data corresponding to each sample data. The sample data includes sample operation and maintenance data of sample power equipment, sample fault records, and power technology documents. The label data includes label fault information of sample power equipment, label electrical connection relationships between sample power equipment, and label operation and maintenance relationships between sample power equipment. The initial knowledge graph construction model is trained based on the training sample set to obtain the knowledge graph construction model.

[0159] In one embodiment, when the processor executes the computer program, it further performs the following steps: comparing the target device attribute information and the reference device attribute information, and determining whether the target power device includes a first abnormal power device based on the comparison result; if the target power device includes the first abnormal power device, determining a second abnormal power device that is associated with a fault in the first abnormal power device based on the fault type of the first abnormal power device and the target device attribute information; and generating a fault detection result based on the target device attribute information and the device attribute information of the second abnormal power device.

[0160] In one embodiment, when the processor executes the computer program, it further performs the following steps: querying the fault records included in the updated knowledge graph data for candidate fault devices that match both the fault type and the target device attribute information; and determining, based on the relation edges included in the updated knowledge graph data, a power device that has an electrical connection relationship and / or an operation and maintenance association relationship with the first abnormal power device as the second abnormal power device.

[0161] In one embodiment, when the processor executes the computer program, it further performs the following steps: if the fault detection result indicates that the target power equipment has a target fault type, it queries the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type; it generates an operation and maintenance suggestion message based on the fault handling information and outputs the operation and maintenance suggestion message.

[0162] In one embodiment, when the processor executes the computer program, it further performs the following steps: combining fault handling information to obtain multiple candidate maintenance suggestion messages, each candidate maintenance suggestion message including historical fault data; determining the similarity between each historical fault data and the target device attribute information, and determining the candidate maintenance suggestion message with the highest similarity as the maintenance suggestion message.

[0163] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0164] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0165] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0166] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A fault detection method, characterized in that, The method includes: Acquire the operating status data of target power equipment in the target power grid area and the power grid operation and maintenance knowledge graph corresponding to the target power grid area. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes the entity node corresponding to each power equipment, the equipment attribute information of each power equipment, and the relationship edges between each power equipment. Based on the operational status data, the power grid operation and maintenance knowledge graph is updated to obtain an updated power grid operation and maintenance knowledge graph, which includes the updated knowledge graph data corresponding to the target power equipment. The updated knowledge graph data is used to perform knowledge graph calculations to obtain calculation results, and the fault detection results of the target power grid area are determined based on the calculation results.

2. The method according to claim 1, characterized in that, The process of constructing the power grid operation and maintenance knowledge graph includes: Obtain historical operation and maintenance data of each power device in the target power grid area. The historical operation and maintenance data includes historical fault records, electrical connection relationships between each power device, and operation and maintenance association relationships between each power device. The historical operation and maintenance data is input into the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph output by the knowledge graph construction model.

3. The method according to claim 2, characterized in that, The step of inputting the historical operation and maintenance data into the knowledge graph construction model to obtain the power grid operation and maintenance knowledge graph output by the knowledge graph construction model includes: The historical operation and maintenance data is input into the knowledge graph construction model. The knowledge graph construction model is used to identify the historical operation and maintenance data to obtain the equipment entity information corresponding to each power equipment, the equipment fault information corresponding to each power equipment, the electrical connection information between each power equipment, and the operation and maintenance relationship information between each power equipment. The knowledge graph model is used to construct triplets corresponding to each power device based on the device entity information, device fault information, and maintenance relationship information. The knowledge graph is used to construct a model that connects the triples based on the electrical connection information to obtain the power grid operation and maintenance knowledge graph.

4. The method according to claim 2, characterized in that, The training process of the knowledge graph construction model includes: Obtain a training sample set, which includes multiple sample data and label data corresponding to each sample data. The sample data includes sample operation and maintenance data, sample fault records and power technology documents of sample power equipment. The label data includes label fault information of sample power equipment, label electrical connection relationship between each sample power equipment and label operation and maintenance relationship between each sample power equipment. The initial knowledge graph construction model is trained based on the training sample set to obtain the knowledge graph construction model.

5. The method according to any one of claims 1 to 4, characterized in that, The updated knowledge graph data includes target device attribute information. The step of performing knowledge graph calculations based on the target device attribute information and determining the fault detection result of the target power equipment based on the calculation results includes: The target device attribute information and the reference device attribute information are compared, and it is determined whether the target power device includes the first abnormal power device based on the comparison result. If the target power equipment includes a first abnormal power equipment, a second abnormal power equipment that is associated with the first abnormal power equipment is determined based on the fault type of the first abnormal power equipment and the attribute information of the target equipment. The fault detection result is generated based on the target device attribute information and the device attribute information of the second abnormal power device.

6. The method according to claim 5, characterized in that, The step of determining a second abnormal power device that is associated with the first abnormal power device based on the fault type of the first abnormal power device and the attribute information of the target device includes: In the updated knowledge graph data, query the fault records to find candidate fault devices that match both the fault type and the target device attribute information. Based on the relation edges included in the updated knowledge graph data, the power equipment that has an electrical connection relationship and / or operation and maintenance relationship with the first abnormal power equipment is identified as the second abnormal power equipment among the candidate faulty equipment.

7. The method according to claim 1, characterized in that, The method further includes: If the fault detection result indicates that the target power equipment has a target fault type, query the power grid operation and maintenance knowledge graph for fault handling information that matches the target fault type; Based on the fault handling information, an operation and maintenance suggestion message is generated and output.

8. The method according to claim 7, characterized in that, The generation of maintenance suggestion messages based on the fault handling information includes: The fault handling information is combined to obtain multiple candidate maintenance suggestion messages, each of which includes historical fault data. Determine the similarity between each of the historical fault data and the target equipment attribute information of the target power equipment, and determine the candidate operation and maintenance suggestion message with the highest similarity as the operation and maintenance suggestion message.

9. A fault detection device, characterized in that, The device includes: The data acquisition module is used to acquire the operating status data of target power equipment in the target power grid area and the power grid operation and maintenance knowledge graph corresponding to the target power grid area. The power grid operation and maintenance knowledge graph includes knowledge graph data of each power equipment in the target power grid area. Each knowledge graph data includes the entity node corresponding to each power equipment, the equipment attribute information of each power equipment, and the relationship edges between each power equipment. The knowledge graph update module is used to update the power grid operation and maintenance knowledge graph according to the operation status data to obtain the updated power grid operation and maintenance knowledge graph, wherein the updated power grid operation and maintenance knowledge graph includes the updated knowledge graph data corresponding to the target power equipment; The fault detection module is used to perform knowledge graph calculations based on the updated knowledge graph data to obtain calculation results, and to determine the fault detection results of the target power grid area based on the calculation results.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 8.

Images