Accident knowledge graph construction method, device, storage medium and electronic device

Abstract

The present disclosure relates to the field of language processing technology, and specifically to a method, device, storage medium and electronic device for constructing an accident knowledge graph, the method comprising: acquiring historical emergency data, the historical emergency data comprising accident handling nodes for historical accident cases of a hydropower station and equipment states corresponding to the accident handling nodes, historical emergency strategies and corresponding induced accidents after adopting the historical emergency strategies; performing feature extraction on the historical emergency data to obtain semantic feature vectors and accident scenario feature vectors; based on a preset feature similarity algorithm, determining graph node feature information of the power equipment of the hydropower station and the similarity of the historical accident cases according to the semantic feature vectors and the accident scenario feature vectors; and mapping the graph node feature information to a preset graph structure according to the similarity of the historical accident cases to obtain the accident knowledge graph.

Description

Technical Field

[0001] The present disclosure relates to the field of language processing technology, and in particular, to a method, device, storage medium, and electronic device for constructing an accident knowledge graph. Background Art

[0002] In hydropower station management, the rapid and accurate handling of accidents is crucial. Currently, emergency response to hydropower station accidents relies heavily on the personal capabilities of emergency response personnel. However, equipment accidents are sudden, urgent, and unpredictable, placing significant demands on these personnel. Furthermore, the lack of targeted information collection leads to a lack of technical support for emergency decision-making when hydropower station accidents occur. This, in turn, results in inaccurate emergency strategies and prolonged development times, severely impacting the normal operation of hydropower stations. Summary of the Invention

[0003] According to a first aspect of an embodiment of the present disclosure, a method for constructing an accident knowledge graph is provided, comprising:

[0004] Acquire historical emergency data, the historical emergency data including accident handling nodes for historical accident cases of the hydropower station and equipment status corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies;

[0005] Extracting features from the historical emergency data to obtain semantic feature vectors and accident scene feature vectors;

[0006] Based on a preset feature similarity algorithm, determining the graph node feature information of the power equipment of the hydropower station and the similarity of the historical accident case according to the semantic feature vector and the accident scene feature vector;

[0007] According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain the accident knowledge graph.

[0008] Optionally, feature extraction is performed on the historical emergency data to obtain a semantic feature vector and an accident scene feature vector, including:

[0009] Performing knowledge extraction on the historical emergency data to obtain knowledge graph construction data, wherein the knowledge graph construction data includes power equipment entities in the historical emergency data, attribute information of the power equipment entities, and entity relationships between the power equipment entities;

[0010] Performing knowledge fusion based on the knowledge graph construction data and the graph data in the preset knowledge base to obtain fused knowledge graph construction data;

[0011] Feature extraction is performed on the fused knowledge graph construction data to obtain the semantic feature vector and the accident scene feature vector.

[0012] Optionally, the historical emergency data includes structured data, semi-structured data, and unstructured data. If the historical emergency data is structured data, the knowledge extraction of the historical emergency data to obtain knowledge graph construction data includes:

[0013] Extracting knowledge from the historical emergency data using preset mapping rules to obtain the knowledge graph construction data;

[0014] If the historical emergency data is semi-structured data or unstructured data, the knowledge extraction of the historical emergency data to obtain knowledge graph construction data includes:

[0015] Screening out historical accident cases targeting the hydropower station from the semi-structured data or unstructured data according to preset target information;

[0016] Through a preset association model algorithm based on semantic annotation, knowledge extraction is performed on the screened historical accident cases for the hydropower station to obtain the knowledge graph construction data.

[0017] Optionally, the performing knowledge fusion based on the knowledge graph construction data and graph data in a preset knowledge base to obtain fused knowledge graph construction data includes:

[0018] Merging the knowledge graph construction data with the graph data to obtain a plurality of merged power equipment entities, attribute information of each power equipment entity, and entity relationships between the plurality of power equipment entities;

[0019] comparing each of the merged plurality of electrical equipment entities with other electrical equipment entities according to the attribute information of each electrical equipment entity to obtain a comparison result;

[0020] screening the merged multiple power equipment entities according to the comparison result to obtain target power equipment entities, target attribute information of the target power equipment entities, and target entity relationships between the target power equipment entities;

[0021] According to the target power equipment entity, the target attribute information and the target entity relationship, the fused knowledge graph construction data is obtained.

[0022] Optionally, mapping the graph node feature information into a preset graph structure according to the similarity of the historical accident cases to obtain the accident knowledge graph includes:

[0023] According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an initial accident knowledge graph;

[0024] Performing knowledge reasoning through the initial accident knowledge graph to obtain a knowledge reasoning result;

[0025] Performing quality assessment on the knowledge reasoning result to obtain a quality assessment result;

[0026] Supplement the initial accident knowledge graph according to the quality assessment results to obtain a supplemented and complete accident knowledge graph;

[0027] The supplemented and complete accident knowledge graph is determined as the accident knowledge graph.

[0028] Optionally, it also includes:

[0029] Comparing the currently acquired historical emergency data with the historical emergency data in the accident knowledge graph to determine whether the currently acquired historical emergency data has changed;

[0030] In the case where the currently acquired historical emergency data changes, graph node feature information is obtained based on the currently acquired historical emergency data, and the accident knowledge graph is updated.

[0031] Optionally, obtaining historical emergency data includes:

[0032] The historical emergency data is obtained by using a web crawler tool to collect data from a preset data source.

[0033] According to a second aspect of an embodiment of the present disclosure, a device for constructing an accident knowledge graph is provided, comprising:

[0034] An acquisition module is used to acquire historical emergency data, wherein the historical emergency data includes accident handling nodes of historical accident cases of the hydropower station and the equipment status corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies;

[0035] A feature extraction module is used to extract features from the historical emergency data to obtain semantic feature vectors and accident scene feature vectors;

[0036] A determination module, configured to determine, based on a preset feature similarity algorithm and according to the semantic feature vector and the accident scene feature vector, the graph node feature information of the power equipment of the hydropower station and the similarity of the historical accident case;

[0037] A mapping module is used to map the graph node feature information to a preset graph structure according to the similarity of the historical accident cases to obtain the accident knowledge graph.

[0038] According to a third aspect of an embodiment of the present disclosure, a non-transitory computer-readable medium is provided, on which a computer program is stored, and when the program is executed by a processor, the steps of the method described in the first aspect are implemented.

[0039] According to a fourth aspect of an embodiment of the present disclosure, there is provided a device, including:

[0040] a memory having a computer program stored thereon;

[0041] A processor is used to execute the computer program in the memory to implement the steps of the method of the first aspect.

[0042] Through the above technical solution, historical accident cases at hydropower stations are collected and feature extracted to obtain semantic feature vectors and accident scenario feature vectors. The feature information of the graph nodes of the hydropower station's electrical equipment and the similarity of historical accident cases are determined. Then, based on the similarity of historical accident cases and a pre-set graph structure, an accident knowledge graph is constructed. This allows for visualization of each accident handling node, the key handling points for each handling node, the equipment operating mode, the fault location, and the nature of the accident, providing a guiding basis for the formulation of emergency response strategies, thereby improving the efficiency and accuracy of accident handling.

[0043] Other features and advantages of the present disclosure will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The accompanying drawings are used to provide a further understanding of the present disclosure and constitute a part of the specification. Together with the following detailed description, they are used to explain the present disclosure but do not constitute a limitation of the present disclosure. In the accompanying drawings:

[0045] Figure 1 It is a flowchart of a method for constructing an accident knowledge graph according to an exemplary embodiment.

[0046] Figure 2 It is a block diagram of an accident knowledge graph construction device according to an exemplary embodiment.

[0047] Figure 3 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0048] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

[0049] The inventors' research found that most of the existing knowledge graphs are general knowledge graphs. There is no complete knowledge graph construction method for highly targeted domain knowledge, especially knowledge graphs in the field of hydropower station safety and emergency response. When an accident occurs in a hydropower station, there is a lack of technical support for making emergency decisions, which leads to low accuracy of emergency strategies and long formulation time, seriously affecting the normal operation of the hydropower station.

[0050] In view of this, the present disclosure provides a method, device, medium and equipment for constructing an accident knowledge graph to solve the above technical problems.

[0051] Figure 1 A method for constructing an accident knowledge graph according to an exemplary embodiment includes the following steps:

[0052] In step S101, historical emergency data is obtained, which includes accident handling nodes for historical accident cases of the hydropower station and equipment states corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies.

[0053] In step S102, feature extraction is performed on the historical emergency data to obtain a semantic feature vector and an accident scene feature vector.

[0054] In step S103, based on a preset feature similarity algorithm, the graph node feature information of the power equipment of the hydropower station and the similarity of historical accident cases are determined according to the semantic feature vector and the accident scene feature vector.

[0055] In step S104, based on the similarity of historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an accident knowledge graph.

[0056] First of all, it should be understood that the present disclosure collects historical emergency data, extracts and processes the characteristic information in the emergency data according to a preset method, obtains multiple nodes and characteristic information of each node, and then maps the obtained characteristic information to a preset graph structure, thereby completing the construction of the accident knowledge graph.

[0057] For example, historical emergency data may include multiple historical accident cases for hydropower stations. Specifically, for each historical accident case, the accident handling node indicates at which nodes the power equipment of the hydropower station was handled in the historical accident. The equipment status corresponding to the accident handling node may include the status information of the power equipment corresponding to each accident handling node when the historical accident occurred, during the handling of the historical accident, and after the handling of the historical accident, wherein the status information of the power equipment may include the status information of the power equipment monitored by the monitoring system of the hydropower station, and the handling personnel information corresponding to each accident handling node for the historical accident. The historical emergency strategy may include the historical handling strategy corresponding to each accident handling node for the historical accident. The corresponding induced accident after the use of the historical emergency strategy may include the handling effect corresponding to each accident handling node for the historical accident after the adoption of the historical emergency strategy.

[0058] For example, the extraction of semantic feature vectors and accident scene feature vectors can be based on the structure and content of historical accident cases, analyzing the components of historical accident cases, annotating the historical accident cases with semantic information, and then performing feature extraction, and quantifying the extracted semantic features and accident scene features to obtain semantic feature vectors and accident scene feature vectors. The present embodiment does not specifically limit the method of feature extraction. In addition, the similarity of historical accident cases can be calculated by calculating cosine similarity, Euclidean distance, Manhattan distance, etc., and the present embodiment does not specifically limit this.

[0059] Using this technical solution, historical accident cases from hydropower stations are collected and feature extracted to obtain semantic feature vectors and accident scenario feature vectors. The feature information of the graph nodes of the hydropower station's electrical equipment and the similarity of historical accident cases are determined. Then, based on the similarity of historical accident cases and a pre-set graph structure, an accident knowledge graph is constructed. This allows for visualization of each accident handling node, the key handling points for each handling node, the equipment operating mode, the fault location, and the nature of the accident, providing a guiding basis for the formulation of emergency response strategies, thereby improving the efficiency and accuracy of accident handling.

[0060] In a possible manner, obtaining historical emergency data may be to use a web crawler tool to collect data from a preset data source to obtain the historical emergency data.

[0061] For example, the preset data source may include a local file storing historical accident cases for a hydropower station, an information recording manual for the power equipment of the hydropower station, and a network data source including data such as accident handling cases and accident handling methods for the hydropower station. The embodiments of the present disclosure are not limited to this.

[0062] In a possible way, feature extraction is performed on historical emergency data to obtain semantic feature vectors and accident scenario feature vectors. The method can be to first perform knowledge extraction on the historical emergency data to obtain knowledge graph construction data. The knowledge graph construction data includes power equipment entities in the historical emergency data, attribute information of power equipment entities, and entity relationships between power equipment entities. Then, knowledge fusion is performed based on the knowledge graph construction data and the graph data in the preset knowledge base to obtain fused knowledge graph construction data. Finally, feature extraction is performed on the fused knowledge graph construction data to obtain semantic feature vectors and accident scenario feature vectors.

[0063] It should be understood that knowledge extraction from historical emergency data mainly includes three elements: named entity extraction, entity relationship extraction, and attribute extraction. Among them, entity extraction can be to find multiple entities including time, place, power equipment, fault name, fault type, handling method, handling personnel, etc. in the text sentences included in the historical emergency data. The embodiment of the present disclosure does not specifically limit the method of entity extraction. Attribute extraction can be to use a python crawler to crawl web page information and monitoring data or other stored data of the hydropower station to determine the attribute information of each entity. Among them, the monitoring data or other stored data of the hydropower station can include real-time operation data of the power equipment of the hydropower station and monitoring data inside the hydropower station. The embodiment of the present disclosure does not specifically limit the method of entity attribute extraction. Relationship extraction can be to obtain a certain semantic relationship and relationship category between the extracted entities, and then form triple data consisting of entity pairs and the relationship connecting the pair of entities. The embodiment of the present disclosure does not specifically limit the method of relationship extraction.

[0064] For example, to ensure the integrity of the accident knowledge graph and resolve the problems of entity semantic duplication and semantic contradiction, the power equipment entities, attribute information of power equipment entities, and entity relationships between power equipment entities collected from historical emergency data can be fused with the graph data in a preset knowledge base. Feature extraction is then performed on the fused knowledge graph construction data to obtain semantic feature vectors and accident scenario feature vectors. The present disclosure does not specifically limit the feature extraction method.

[0065] In a possible manner, historical emergency data includes structured data, semi-structured data and unstructured data. If the historical emergency data is structured data, knowledge extraction is performed on the historical emergency data to obtain knowledge graph construction data. The knowledge extraction can be performed on the historical emergency data through a preset mapping rule to obtain knowledge graph construction data.

[0066] If the historical emergency data is semi-structured data or unstructured data, knowledge extraction is performed on the historical emergency data to obtain knowledge graph construction data. The method can be to first screen out historical accident cases for hydropower stations from the semi-structured data or unstructured data according to preset target information, and then use a preset semantic annotation-based association model algorithm to perform knowledge extraction on the screened historical accident cases for hydropower stations to obtain knowledge graph construction data.

[0067] It should be understood that due to different data sources, the degree of structure of the acquired historical emergency data varies. Specifically, historical emergency data includes structured data (such as linked data and databases), semi-structured data (such as tables and lists on web pages), or unstructured data (such as plain text data). The lower the degree of structure, the more complex the knowledge extraction. Different methods can be used to extract knowledge for different historical emergency data with different levels of structure.

[0068] For example, for structured historical emergency data, the meaning of its fields, the number of fields, and the data content are all clear. Knowledge extraction can be performed on the structured historical emergency data using preset mapping rules. The mapping rules can be set specifically based on the historical emergency data, and the embodiments of the present disclosure are not limited to this. Of course, knowledge extraction models for structured data can also be used to extract knowledge from structured historical emergency data, and the embodiments of the present disclosure are not limited to this.

[0069] For semi-structured data or unstructured data, they can be screened first to determine which ones are historical accident cases for hydropower stations, that is, based on the preset target information, the power equipment, status information of the power equipment, emergency strategies, operators, and the corresponding concept ontology of induced accidents after the adoption of historical emergency strategies, the meaning of each concept ontology, the relationship between the concept ontologies, and each node information included in the historical emergency data can be determined. Among them, the preset target information can be determined based on the element composition of the hydropower station accident case. Then, the preset association model algorithm based on semantic annotation is used to extract knowledge to obtain knowledge graph construction data. Of course, other entity extraction models or methods can also be used to extract knowledge from semi-structured or unstructured historical emergency data, and the embodiments of the present disclosure are not limited to this.

[0070] In a possible manner, knowledge fusion is performed based on the knowledge graph construction data and the graph data in the preset knowledge base to obtain the fused knowledge graph construction data. The knowledge graph construction data and the graph data can be first merged to obtain multiple merged power equipment entities, attribute information of each power equipment entity and entity relationships between multiple power equipment entities. Then, based on the attribute information of each power equipment entity, each power equipment entity in the merged multiple power equipment entities is compared with other power equipment entities to obtain comparison results. Then, based on the comparison results, the merged multiple power equipment entities are screened to obtain the target power equipment entity, the target attribute information of the target power equipment entity and the target entity relationship between the target power equipment entities. Finally, based on the target power equipment entity, the target attribute information and the target entity relationship, the fused knowledge graph construction data is obtained.

[0071] For example, due to the different data sources for obtaining historical emergency data, the multiple entities obtained may have repeated or contradictory semantic information or attribute information. For example, for a certain power equipment, the same parameter may be inconsistent, resulting in entity contradictions. The power equipment has multiple names or is identified by pronouns, resulting in the extraction of multiple duplicate entities.

[0072] Therefore, the knowledge graph construction data and the graph data can be first merged, and then each entity can be compared with other entities based on the entity's attribute information to obtain the target power equipment entity, the target attribute information of the target power equipment entity, and the target entity relationship between the target power equipment entities. The comparison method and the standard for determining the target power equipment entity based on the comparison results can be set according to the actual application scenario and are not limited in the embodiments of the present disclosure.

[0073] In a possible way, according to the similarity of historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an accident knowledge graph. It can be that the graph node feature information is first mapped to a preset graph structure according to the similarity of historical accident cases to obtain an initial accident knowledge graph, and then knowledge reasoning is performed through the initial accident knowledge graph to obtain a knowledge reasoning result, and then the knowledge reasoning result is quality evaluated to obtain a quality evaluation result, and the initial accident knowledge graph is supplemented according to the quality evaluation result to obtain a supplemented complete accident knowledge graph, and finally the supplemented complete accident knowledge graph is determined as the accident knowledge graph.

[0074] It should be understood that after mapping the graph node feature information to the preset graph structure, the initial accident knowledge graph obtained is incomplete. Therefore, knowledge reasoning can be performed around the existing concepts and relationship information in the initial accident knowledge graph to obtain new knowledge. Specifically, three types of methods, namely logical rules, embedded representations, and neural networks, can be used for knowledge reasoning. The disclosed embodiments do not specifically limit the methods of knowledge reasoning. Quality assessment of the results of knowledge reasoning can be to evaluate the confidence of the new knowledge obtained by reasoning, and retain knowledge with high confidence to ensure the accuracy and effectiveness of the accident knowledge graph. Of course, other methods can also be used for quality assessment, and the disclosed embodiments do not specifically limit this.

[0075] It should be understood that during the use of the hydropower station, the information collected by the monitoring system of the hydropower station is increasing, and the information in the preset data source is constantly updated. Therefore, in order to ensure the timeliness and accuracy of the knowledge graph, in another possible way, the accident knowledge graph construction method can also compare the currently acquired historical emergency data with the historical emergency data in the accident knowledge graph to determine whether the currently acquired historical emergency data has changed. In the case of changes in the currently acquired historical emergency data, the graph node feature information is obtained based on the currently acquired historical emergency data, and the accident knowledge graph is updated.

[0076] Figure 2 FIG. 2 is a block diagram of an accident knowledge graph construction device 200 according to an exemplary embodiment. Figure 2 The device includes an acquisition module 201 , a feature extraction module 202 , a determination module 203 and a mapping module 204 .

[0077] Acquisition module 201 is used to acquire historical emergency data, including accident handling nodes and equipment status corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies for historical accident cases of the hydropower station;

[0078] Feature extraction module 202, used to extract features from historical emergency data to obtain semantic feature vectors and accident scene feature vectors;

[0079] Determination module 203, for determining the similarity between graph node feature information of power equipment of the hydropower station and historical accident cases based on a preset feature similarity algorithm and the semantic feature vector and the accident scene feature vector;

[0080] The mapping module 204 is used to map the graph node feature information to a preset graph structure according to the similarity of historical accident cases to obtain an accident knowledge graph.

[0081] Optionally, the feature extraction module 202 is used to:

[0082] Performing knowledge extraction on the historical emergency data to obtain knowledge graph construction data, wherein the knowledge graph construction data includes power equipment entities in the historical emergency data, attribute information of the power equipment entities, and entity relationships between the power equipment entities;

[0083] Performing knowledge fusion based on the knowledge graph construction data and the graph data in the preset knowledge base to obtain fused knowledge graph construction data;

[0084] Feature extraction is performed on the fused knowledge graph construction data to obtain the semantic feature vector and the accident scene feature vector.

[0085] Optionally, the historical emergency data includes structured data, semi-structured data and unstructured data, and the feature extraction module 202 is used to:

[0086] In the case where the historical emergency data is structured data, knowledge extraction is performed on the historical emergency data using a preset mapping rule to obtain the knowledge graph construction data;

[0087] In a case where the historical emergency data is semi-structured data or unstructured data, historical accident cases targeting the hydropower station are screened from the semi-structured data or unstructured data according to preset target information;

[0088] Through a preset association model algorithm based on semantic annotation, knowledge extraction is performed on the screened historical accident cases for the hydropower station to obtain the knowledge graph construction data.

[0089] Optionally, the feature extraction module 202 is used to:

[0090] Merging the knowledge graph construction data with the graph data to obtain a plurality of merged power equipment entities, attribute information of each power equipment entity, and entity relationships between the plurality of power equipment entities;

[0091] comparing each of the merged plurality of electrical equipment entities with other electrical equipment entities according to the attribute information of each electrical equipment entity to obtain a comparison result;

[0092] screening the merged multiple power equipment entities according to the comparison result to obtain target power equipment entities, target attribute information of the target power equipment entities, and target entity relationships between the target power equipment entities;

[0093] According to the target power equipment entity, the target attribute information and the target entity relationship, the fused knowledge graph construction data is obtained.

[0094] The mapping module 204 is used to:

[0095] According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an initial accident knowledge graph;

[0096] Performing knowledge reasoning through the initial accident knowledge graph to obtain a knowledge reasoning result;

[0097] Performing quality assessment on the knowledge reasoning result to obtain a quality assessment result;

[0098] Supplement the initial accident knowledge graph according to the quality assessment results to obtain a supplemented and complete accident knowledge graph;

[0099] The supplemented and complete accident knowledge graph is determined as the accident knowledge graph.

[0100] Optionally, the knowledge graph construction device 200 further includes:

[0101] A comparison module is used to compare the currently acquired historical emergency data with the historical emergency data in the accident knowledge graph to determine whether the currently acquired historical emergency data has changed;

[0102] The knowledge updating module is used to obtain graph node feature information based on the currently acquired historical emergency data and update the accident knowledge graph when the currently acquired historical emergency data changes.

[0103] Optionally, the acquisition module 201 is used to:

[0104] The historical emergency data is obtained by using a web crawler tool to collect data from a preset data source.

[0105] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0106] Based on the same inventive concept, an embodiment of the present disclosure also provides a non-temporary computer-readable medium on which a computer program is stored. When the program is executed by a processor, the steps of the knowledge graph construction method provided by the present disclosure are implemented.

[0107] Regarding the apparatus in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0108] Based on the same inventive concept, an embodiment of the present disclosure also provides a non-temporary computer-readable medium on which a computer program is stored. When the program is executed by a processor, the steps of the knowledge graph construction method provided by the present disclosure are implemented.

[0109] Figure 3 FIG. 1 is a block diagram of an electronic device 300 according to an exemplary embodiment. Figure 3 As shown, the electronic device 300 may include: a processor 301 , a memory 302 , and may further include one or more of a multimedia component 303 , an input / output (I / O) interface 304 , and a communication component 305 .

[0110] The processor 301 is used to control the overall operation of the electronic device 300 to complete all or part of the steps in the above-mentioned knowledge graph construction method. The memory 302 is used to store various types of data to support the operation of the electronic device 300. These data may include, for example, instructions for any application or method operating on the electronic device 300, as well as application-related data, such as contact data, sent and received messages, pictures, audio, video, etc. The memory 302 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The multimedia component 303 may include a screen and an audio component. The screen may be, for example, a touch screen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 302 or transmitted via the communication component 305. The audio component also includes at least one speaker for outputting audio signals. The I / O interface 304 provides an interface between the processor 301 and other interface modules. The aforementioned other interface modules may be a keyboard, a mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 305 is used for wired or wireless communication between the electronic device 300 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IOT, eMTC, or other 5G, etc., or a combination of one or more thereof, is not limited here. Therefore, the corresponding communication component 305 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0111] In an exemplary embodiment, the electronic device 300 can be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to execute the above-mentioned knowledge graph construction method.

[0112] In another exemplary embodiment, a computer-readable storage medium including program instructions is further provided, and when the program instructions are executed by a processor, the steps of the above-mentioned knowledge graph construction method are implemented. For example, the computer-readable storage medium may be the above-mentioned memory 302 including the program instructions, and the above-mentioned program instructions may be executed by the processor 301 of the electronic device 300 to complete the above-mentioned knowledge graph construction method.

[0113] In another exemplary embodiment, a computer program product is also provided, which includes a computer program that can be executed by a programmable device, and the computer program has a code portion for executing the above-mentioned knowledge graph construction method when executed by the programmable device.

[0114] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings. However, the present disclosure is not limited to the specific details of the above embodiments. Within the technical concept of the present disclosure, various simple modifications can be made to the technical solutions of the present disclosure, and these simple modifications all fall within the scope of protection of the present disclosure.

[0115] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0116] In addition, the various embodiments of the present disclosure may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the contents disclosed by the present disclosure.

Claims

1. A method for constructing an accident knowledge graph, characterized in that: include: Acquire historical emergency data, the historical emergency data including accident handling nodes for historical accident cases of the hydropower station and equipment status corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies; Extracting features from the historical emergency data to obtain semantic feature vectors and accident scene feature vectors; Based on a preset feature similarity algorithm, determining the graph node feature information of the power equipment of the hydropower station and the similarity of the historical accident case according to the semantic feature vector and the accident scene feature vector; According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain the accident knowledge graph; The process of mapping the graph node feature information to a preset graph structure based on the similarity of the historical accident cases to obtain the accident knowledge graph includes: According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an initial accident knowledge graph; Performing knowledge reasoning through the initial accident knowledge graph to obtain a knowledge reasoning result; Performing quality assessment on the knowledge reasoning result to obtain a quality assessment result; Supplement the initial accident knowledge graph according to the quality assessment results to obtain a supplemented and complete accident knowledge graph; The supplemented and complete accident knowledge graph is determined as the accident knowledge graph.

2. The method according to claim 1, characterized in that Feature extraction is performed on the historical emergency data to obtain semantic feature vectors and accident scene feature vectors, including: Performing knowledge extraction on the historical emergency data to obtain knowledge graph construction data, wherein the knowledge graph construction data includes power equipment entities in the historical emergency data, attribute information of the power equipment entities, and entity relationships between the power equipment entities; Performing knowledge fusion based on the knowledge graph construction data and the graph data in the preset knowledge base to obtain fused knowledge graph construction data; Feature extraction is performed on the fused knowledge graph construction data to obtain the semantic feature vector and the accident scene feature vector.

3. The method according to claim 2, characterized in that The historical emergency data includes structured data, semi-structured data and unstructured data. If the historical emergency data is structured data, the knowledge extraction of the historical emergency data to obtain knowledge graph construction data includes: Extracting knowledge from the historical emergency data using preset mapping rules to obtain the knowledge graph construction data; If the historical emergency data is semi-structured data or unstructured data, the knowledge extraction of the historical emergency data to obtain knowledge graph construction data includes: Screening out historical accident cases targeting the hydropower station from the semi-structured data or unstructured data according to preset target information; Through a preset association model algorithm based on semantic annotation, knowledge extraction is performed on the screened historical accident cases for the hydropower station to obtain the knowledge graph construction data.

4. The method according to claim 2, characterized in that: The step of performing knowledge fusion based on the knowledge graph construction data and graph data in a preset knowledge base to obtain fused knowledge graph construction data includes: Merging the knowledge graph construction data with the graph data to obtain a plurality of merged power equipment entities, attribute information of each power equipment entity, and entity relationships between the plurality of power equipment entities; comparing each of the merged plurality of electrical equipment entities with other electrical equipment entities according to the attribute information of each electrical equipment entity to obtain a comparison result; screening the merged multiple power equipment entities according to the comparison result to obtain target power equipment entities, target attribute information of the target power equipment entities, and target entity relationships between the target power equipment entities; According to the target power equipment entity, the target attribute information and the target entity relationship, the fused knowledge graph construction data is obtained.

5. The method according to any one of claims 1 to 4, characterized in that Also includes: Comparing the currently acquired historical emergency data with the historical emergency data in the accident knowledge graph to determine whether the currently acquired historical emergency data has changed; In the case where the currently acquired historical emergency data changes, graph node feature information is obtained based on the currently acquired historical emergency data, and the accident knowledge graph is updated.

6. The method according to any one of claims 1 to 4, characterized in that The obtaining of historical emergency data includes: The historical emergency data is obtained by using a web crawler tool to collect data from a preset data source.

7. An accident knowledge graph construction device, characterized in that: include: An acquisition module is used to acquire historical emergency data, wherein the historical emergency data includes accident handling nodes of historical accident cases of the hydropower station and the equipment status corresponding to the accident handling nodes, historical emergency strategies, and induced accidents corresponding to the adoption of the historical emergency strategies; A feature extraction module is used to extract features from the historical emergency data to obtain semantic feature vectors and accident scene feature vectors; A determination module, configured to determine, based on a preset feature similarity algorithm and according to the semantic feature vector and the accident scene feature vector, the graph node feature information of the power equipment of the hydropower station and the similarity of the historical accident case; A mapping module, configured to map the graph node feature information to a preset graph structure based on the similarity of the historical accident cases to obtain the accident knowledge graph; The mapping module is used to: According to the similarity of the historical accident cases, the graph node feature information is mapped to a preset graph structure to obtain an initial accident knowledge graph; Performing knowledge reasoning through the initial accident knowledge graph to obtain a knowledge reasoning result; Performing quality assessment on the knowledge reasoning result to obtain a quality assessment result; Supplement the initial accident knowledge graph according to the quality assessment results to obtain a supplemented and complete accident knowledge graph; The supplemented and complete accident knowledge graph is determined as the accident knowledge graph.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

9. An electronic device, characterized in that: include: a memory having a computer program stored thereon; A processor, configured to execute the computer program in the memory to implement the steps of the method according to any one of claims 1 to 6.

Images