Association method, system and equipment of electric power safety knowledge and storage medium

Abstract

The invention relates to an association method, system and device for electric power safety knowledge and a storage medium, and the method comprises the following steps: obtaining multi-dimensional data of the electric power safety knowledge, and carrying out the fusion of the multi-dimensional data, and obtaining the fusion data of the electric power safety knowledge; performing graph modeling based on the electric power safety knowledge fusion data to obtain an electric power safety knowledge graph; acquiring power fault information, and performing fault correlation analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; performing causal relationship mining on the power failure propagation path to obtain power failure equipment and an induction mechanism corresponding to the power failure equipment; performing influence range analysis on the power grid equipment based on the induction mechanism to obtain an influence power grid region range, and formulating a management strategy corresponding to the equipment in the influence power grid region range; the technical problem that comprehensive analysis and prediction cannot be carried out timely and accurately in the face of complex power faults is solved.

Description

Technical Field

[0001] The present invention relates to the field of power safety technology, and in particular to a method, system, device and storage medium for associating power safety knowledge. Background Art

[0002] As an important part of national infrastructure, the safety of the power system is directly related to the stability and development of society. However, with the continuous expansion and complexity of the power system, the frequency and impact of power failures are gradually increasing, and traditional power safety management methods can no longer meet current needs. Traditional methods rely on experience and manual judgment, and have limited ability to identify and handle complex faults, especially in the correlation analysis and propagation path tracking of power failures.

[0003] At present, one of the main problems in power safety management is the lack of effective data fusion and intelligent analysis methods. The multi-dimensional data of power safety knowledge is often scattered in different systems and platforms, and lacks a unified analysis framework, making it impossible to conduct comprehensive analysis and prediction in a timely and accurate manner when facing complex power failures. In addition, the existing power failure analysis methods mostly rely on post-analysis, which makes it difficult to provide real-time warnings and decision support, which is particularly untimely when dealing with sudden failures.

[0004] To solve these problems, this study proposed a method for associating power safety knowledge based on graph theory algorithms. By integrating multi-dimensional data of power safety knowledge and performing graph modeling, it aims to improve the accuracy and efficiency of power fault analysis. This method uses the power safety knowledge graph to perform association analysis on fault information, which can more effectively track the propagation path of power faults and conduct causal mining, thereby formulating more accurate management strategies. However, in practical applications, there may be challenges in data quality, complexity of graph modeling, and real-time processing capabilities. These problems need to be further resolved in research and practice. Summary of the invention

[0005] The main purpose of the present invention is to provide a method, system, device and storage medium for associating power safety knowledge, which solves the technical problem that comprehensive analysis and prediction cannot be carried out in a timely and accurate manner when facing complex power failures.

[0006] To achieve the above object, the present invention provides a method for associating power safety knowledge, comprising the following steps: Acquire multi-dimensional data of power safety knowledge, and fuse the multi-dimensional data to obtain power safety knowledge fusion data; By using a preset graph theory algorithm, graph modeling is performed based on the power safety knowledge fusion data to obtain a power safety knowledge graph; Acquire power fault information of power grid equipment, and perform fault correlation analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; Perform causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment; Based on the inducing mechanism, the influence range of the power grid equipment is analyzed to obtain the area range of the affected power grid, and a management strategy corresponding to the equipment within the area range of the affected power grid is formulated.

[0007] Furthermore, the multi-dimensional data includes power grid operation data, equipment maintenance records, power safety standard documents, accident case database, and power grid environment monitoring data; the multi-dimensional data is fused to obtain power safety knowledge fusion data, including: Performing a multi-scale time series analysis on the multi-dimensional data of the power grid to obtain power grid operation status characteristics; wherein the power grid operation status characteristics include power grid voltage fluctuations, power grid frequency changes, and power grid load fluctuations; By using a conditional random field algorithm, the power grid equipment is sequence-labeled based on the equipment maintenance record to obtain the health status of the power grid equipment; wherein the health status of the power grid equipment includes a normal operating state, a minor fault state, and a serious fault state; The electric power safety standard document is semantically parsed by natural language processing technology to obtain electric power safety operation procedures; wherein the electric power safety operation procedures include operating step specifications, emergency plans and safety inspection points; Performing data mining on the accident case database to obtain historical accident patterns; wherein the historical accident patterns include historical failure types, historical failure inducing conditions, and historical failure consequences; Extracting environmental features from the power grid environment monitoring data to obtain environmental extraction features; The power grid operation status characteristics, the health status of the power grid equipment, the power safety operation procedures, the historical accident patterns and the environmental extraction characteristics are subjected to data fusion processing to obtain power safety knowledge fusion data.

[0008] Furthermore, a graph modeling is performed based on the power safety knowledge fusion data through a preset graph theory algorithm to obtain a power safety knowledge graph, including: The power safety knowledge fusion data is quantitatively represented by a preset graph theory algorithm to obtain a multi-dimensional dynamic feature matrix; Performing space-time correlation analysis on power grid equipment based on the multi-dimensional dynamic feature matrix to obtain a space-time correlation matrix between power grid equipment; Performing attribute graph modeling on the power grid devices based on the spatiotemporal association matrix between the power grid devices to obtain a power grid device relationship graph with additional attribute information; Performing deep feature extraction on the power grid equipment relationship graph of the additional attribute information to obtain a multi-level embedded vector representation of the power grid equipment; The multi-level embedding vector representation is subjected to data fusion and graph modeling through a preset heterogeneous information network to obtain a preliminary draft of the electric power safety knowledge graph; Based on the graph attention mechanism, the preliminary draft of the power safety knowledge graph is optimized and adjusted to obtain the power safety knowledge graph.

[0009] Furthermore, based on the power safety knowledge graph, a fault correlation analysis is performed on the power fault information to obtain a power fault propagation path, including: Perform fault correlation analysis on the power grid equipment based on the power fault information to obtain potential power fault equipment; wherein the potential power fault equipment includes potential power fault equipment with short circuit, open circuit and overload; Topologically sorting the potential power failure devices based on the power safety knowledge graph to obtain potential power failure sorting devices; constructing a cause-and-effect chain of faulty devices based on the potential power fault sequencing device; Using a preset Markov chain Monte Carlo method to perform path sampling on the causal chain, a plurality of potential fault propagation paths are obtained; In the causal chain, probability calculation is performed on each of the potential power fault sequencing devices to obtain the failure probability of the potential power fault sequencing device; Based on the failure probability of the potential power fault sorting device, probability calculation is performed on the multiple potential fault propagation paths to obtain the probability of each potential fault propagation path; The potential fault propagation path corresponding to the highest probability is determined as the power fault propagation path.

[0010] Furthermore, the causal relationship of the power fault propagation path is mined to obtain the power fault device and the inducing mechanism corresponding to the power fault device, including: Conduct preliminary causal relationship mining on the power fault propagation path to form a preliminary causal relationship map; wherein the preliminary causal relationship map includes the relationship between potential power fault equipment and the inducing mechanism corresponding to the potential power fault equipment; The Granger causality test algorithm is used to verify the relationship between the potential power failure device and the inducing mechanism corresponding to the potential power failure device, so as to obtain a verification causal relationship network; wherein the verification causal relationship network includes the potential power failure device and the significant inducing factors of the potential power failure device; Performing dimensionality reduction processing on the verification causal relationship network to obtain the main inducing mechanism corresponding to each potential power fault device; Performing time series analysis on the main inducing mechanism characteristics to obtain the time inducing mechanism; combining the time induction mechanism and the main induction mechanism to obtain a combined induction mechanism; The combined inducing mechanisms are scored to obtain an inducing mechanism score, and the combined inducing mechanism with the highest score is determined as the inducing mechanism, and the potential power fault device corresponding to the combined inducing mechanism with the highest score is determined as the power fault device.

[0011] Further, based on the induction mechanism, the influence range of the power grid equipment is analyzed to obtain the influence range of the power grid area, and a management strategy corresponding to the power grid equipment within the influence range of the power grid area is formulated, including: Using a preset geographic information system, the spatial distribution of power grid equipment is mapped based on the induction mechanism to obtain a spatial distribution map of power fault equipment; wherein the spatial distribution map includes the geographical location of the power fault equipment in the power grid and the impact range of the power fault equipment; Performing network topology analysis on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices; wherein the correlation matrix includes connectivity and mutual influence between the power fault devices; Using the shortest path algorithm in graph theory, the path impact analysis of the power fault device is performed based on the association matrix to obtain the affected power grid area range; wherein the affected power grid area range is a set of impact paths from the power fault device to each power grid device, and the affected power grid area range includes the power grid fault impact path length and the power grid fault impact intensity; Comprehensively evaluate the length of the grid fault impact path and the grid fault impact intensity to obtain a risk assessment matrix for grid equipment; Perform risk analysis on the power grid equipment based on the risk assessment matrix to obtain a risk division scheme affecting the power grid area; wherein the risk division scheme is the distribution of high-risk equipment and low-risk equipment in each area; Based on the risk classification scheme, a management strategy is formulated for the power grid equipment in each of the areas to obtain management strategies for power grid equipment of different risk levels; wherein the management strategy includes a preventive maintenance plan and an emergency response plan.

[0012] Furthermore, a network topology analysis is performed on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices, including: Extracting the interaction relationship between the devices corresponding to the power fault devices in the spatial distribution diagram to obtain the interaction network relationship between the power fault devices; Modeling the dependency relationship between the power failure devices based on the interactive network relationship between the power failure devices to obtain the dependency relationship between the devices; Using different colors to depict the dependency relationship, obtaining the depicted dependency relationship and the strength of the depicted dependency relationship; Dynamically predicting the depicted dependency relationship and the strength of the depicted dependency relationship to obtain a dynamic interaction prediction network between power fault devices; A correlation matrix between power failure devices is constructed based on the dynamic interaction prediction network between power failure devices.

[0013] The present invention also provides a power safety knowledge association system, comprising: An acquisition module, used to acquire multi-dimensional data of power safety knowledge, and fuse the multi-dimensional data to obtain power safety knowledge fusion data; A modeling module, used to perform graph modeling based on the power safety knowledge fusion data through a preset graph theory algorithm to obtain a power safety knowledge graph; A first analysis module is used to obtain power fault information and perform fault association analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; A mining module, used to mine the cause-effect relationship of the power fault propagation path to obtain the power fault device and the inducing mechanism corresponding to the power fault device; The second analysis module is used to analyze the impact range of the power grid equipment based on the inducing mechanism, obtain the impact area of ​​the power grid, and formulate management strategies corresponding to the equipment within the impact area of ​​the power grid.

[0014] The present invention also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the above methods when executing the computer program.

[0015] The present invention also provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the steps of any of the above-mentioned methods are implemented.

[0016] The method for associating power safety knowledge provided by the present invention comprises the following steps: obtaining multi-dimensional data of power safety knowledge, and fusing the multi-dimensional data to obtain power safety knowledge fusion data; performing graph modeling based on the power safety knowledge fusion data through a preset graph theory algorithm to obtain a power safety knowledge graph; obtaining power fault information, and performing fault association analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; performing causal relationship mining on the power fault propagation path to obtain power fault equipment and the inducing mechanism corresponding to the power fault equipment; performing an impact range analysis on the power grid equipment based on the inducing mechanism to obtain the impact grid area range, and formulating a management strategy corresponding to the equipment within the impact grid area range; solving the technical problem of being unable to conduct comprehensive analysis and prediction in a timely and accurate manner when facing complex power faults, and realizing that by mining the causal relationship of the power fault propagation path, the fault equipment and its inducing mechanism can be clarified, and the cause of the fault can be fundamentally understood. This capability is helpful for the maintenance and optimization of power system equipment, and effectively prevents the recurrence of similar faults. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a schematic diagram of steps of a method for associating power safety knowledge in one embodiment of the present invention; Figure 2 is a structural block diagram of an association system of power safety knowledge in one embodiment of the present invention; Figure 3 It is a schematic block diagram of the structure of a computer device according to an embodiment of the present invention.

[0018] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0020] like Figure 1 As shown, Figure 1 It is a schematic diagram of the steps of a method for associating power safety knowledge in one embodiment of the present invention; In one embodiment of the present invention, a method for associating power safety knowledge is provided, comprising the following steps: Step S1, acquiring multi-dimensional data of power safety knowledge, and fusing the multi-dimensional data to obtain power safety knowledge fused data.

[0021] Specifically, in the step of "obtaining multi-dimensional data of power safety knowledge and fusing the multi-dimensional data to obtain power safety knowledge fusion data", it is first necessary to collect power safety knowledge data from different sources. These data may include the operating status of power equipment, historical fault records, maintenance logs, environmental monitoring data, and user reports. In order to achieve this goal, a data acquisition system must be built that can automatically obtain data from sensors, monitoring platforms, and databases of various power systems. These data usually exist in different formats and structures, and may include structured data (such as database tables) and unstructured data (such as text reports). Once the data is acquired, these multi-dimensional data need to be fused. The process of data fusion involves integrating data from these different sources into a unified data set, namely "power safety knowledge fusion data". This process usually includes data cleaning, data matching, and data conversion. Data cleaning is used to remove redundant and erroneous data and ensure the accuracy of the data. Data matching involves aligning data from different sources according to their common attributes, for example, matching the equipment number with the equipment number in the maintenance log. Data conversion converts the data into a consistent format for subsequent analysis. For example, fault records in different formats are unified into a standardized table format. In practical applications, for example, a power company may collect a large amount of data from substations, transmission lines, and distribution networks. These data include voltage and current readings at substations, load information on transmission lines, and equipment status of distribution networks. These data are integrated into a central database through a data acquisition system. During the data fusion process, the system converts data from different sources into the same format, and uses algorithmic processing to fill in missing values ​​and correct errors, ultimately forming a comprehensive power safety knowledge fusion data. This data set not only includes the real-time status of the equipment, but also integrates historical fault records and maintenance information, providing a comprehensive foundation for subsequent power safety analysis.

[0022] Step S2, using a preset graph theory algorithm, graph modeling is performed based on the power safety knowledge fusion data to obtain a power safety knowledge graph.

[0023] Specifically, in the step of "using a preset graph theory algorithm to perform graph modeling based on the power safety knowledge fusion data to obtain a power safety knowledge graph", the "power safety knowledge fusion data" needs to be input into the graph theory algorithm first. Graph theory algorithm is a mathematical tool for processing graph data structure, which can present complex data relationships in the form of graphs. The core task of graph modeling is to map the data elements of power safety knowledge and the relationships between them into a graph through a graph theory algorithm, thereby obtaining a "power safety knowledge graph". Specifically, the graph modeling process starts with the definition of nodes and edges of the data. In this process, each important information point (such as equipment, fault records, maintenance information, etc.) in the power safety knowledge fusion data is defined as a node in the graph. The relationship between nodes (such as the power transmission relationship between devices, the association between equipment and fault records, etc.) is defined as an edge in the graph. The preset graph theory algorithm, such as the connectivity algorithm, the shortest path algorithm or the community detection algorithm of the graph, will be applied to these nodes and edges to form a structured knowledge graph. For example, it is assumed that the "power safety knowledge fusion data" includes substations, transmission lines, distribution equipment and the connection relationship between them. Through the graph modeling process, substations and transmission lines will be represented as nodes in the graph, and the power transmission relationship between them will be connected by edges. In addition, the fault records of the substations will also be added to the graph as nodes and connected to the relevant substation nodes by edges to represent the relationship between them. Graph theory algorithms will play a vital role in this modeling process. For example, the shortest path algorithm can identify the path of fault propagation, while the community detection algorithm can help discover the potential group relationship between devices, thereby revealing the implicit structure in the power system. The power safety knowledge graph processed by these graph theory algorithms can clearly show the relationship between the various components in the power system and their complex interactions, providing an intuitive and effective tool for subsequent power fault analysis and management.

[0024] Step S3, obtaining power fault information of power grid equipment, and performing fault association analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path.

[0025] Specifically, in the step of "obtaining power fault information of power grid equipment, and performing fault correlation analysis on the power fault information based on the power safety knowledge graph to obtain the power fault propagation path", it is first necessary to obtain the latest "power fault information" from the power system. This information may include key data such as the time, location, type of fault, and equipment involved in the fault. This data usually comes from a real-time monitoring system, fault report, or maintenance record. Once the power fault information is obtained, it can be input into the constructed "power safety knowledge graph". The power safety knowledge graph contains the connection relationship and interaction between the various devices in the power system. The nodes and edges in the graph can fully understand the structure of the system and the potential propagation path of the fault. Based on these graph data, the fault correlation analysis technology can be used to identify the relationship between the fault information and other nodes in the graph. Specifically, the fault correlation analysis will track the propagation path of the power fault through the connection relationship in the graph. For example, if a substation fails, the system can use the power safety knowledge graph to find the transmission lines, distribution equipment, and related fault records directly connected to the substation. Through the edge and node relationships in the graph, it is possible to trace how the fault propagates from the substation to other equipment or areas, thereby determining the propagation path of the fault. Graph theory algorithms in the graph (such as the shortest path algorithm, graph traversal algorithm, etc.) can help analyze the diffusion path of the fault and identify the equipment and areas that may be affected. For example, suppose a short circuit fault occurs in a substation. By obtaining the fault information and combining it with the power safety knowledge graph, the system can analyze that the short circuit fault may spread to the downstream distribution equipment through the connected transmission lines. The graph can show the specific connection relationship between these transmission lines and distribution equipment, thereby helping engineers to clarify the path of fault propagation and the possible affected areas. This association analysis can not only help quickly locate the source of the fault, but also predict the impact of the fault on the overall stability of the power grid, so as to formulate targeted emergency measures and maintenance plans.

[0026] Step S4, performing causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment.

[0027] Specifically, in the step of "carrying out causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment", it is first necessary to use the previously obtained power fault propagation path to conduct an in-depth analysis of the cause and propagation process of the fault. The power fault propagation path reveals the way and scope of the fault propagation from one device to other devices in the power system, and the core of causal relationship mining is to identify the key equipment that causes the fault and its inducing mechanism through the analysis of these propagation paths. The process of causal relationship mining mainly depends on the association between nodes and edges in the "power safety knowledge graph". Based on the power safety knowledge graph, the power fault equipment that plays a key role in the propagation path can be identified, and its interaction relationship with other equipment can be traced. For example, using the historical fault records, maintenance information and connection relationships between devices in the graph, it is possible to analyze which equipment plays a role in initiating or aggravating the fault propagation process. At this time, the preset causal relationship mining algorithm will be used to find the causal chain between these devices and the fault, so as to determine which equipment is the root cause of the fault and what their inducing mechanism is. For example, suppose that in a certain power grid area, a fault occurs in substation A. After the propagation path analysis of the power safety knowledge graph, it is found that the fault propagates to transmission line B and distribution equipment C. In order to further explore the cause and effect relationship of the fault, the system will use the equipment historical data and related event information stored in the graph to analyze the connection characteristics between substation A and transmission line B. It may be found that the short circuit fault of substation A is caused by equipment aging or overload, and is directly related to the failure of the protection device of transmission line B. This mining process can identify substation A as a power fault device and clarify that its corresponding inducing mechanism is a short circuit caused by equipment aging. Through this causal mining, a comprehensive understanding of the propagation of power faults can be obtained, providing a basis for formulating effective maintenance and prevention strategies.

[0028] Step S5, analyzing the impact range of the power grid equipment based on the inducing mechanism, obtaining the impact area of ​​the power grid, and formulating management strategies corresponding to the equipment within the impact area of ​​the power grid.

[0029] Specifically, in the step of "analyzing the impact range of the power grid equipment based on the inducing mechanism, obtaining the impact area of ​​the power grid, and formulating the corresponding management strategy for the equipment within the impact area", it is first necessary to use the "power fault equipment and its inducing mechanism" obtained by causal relationship mining as the basis for analysis, and further evaluate its impact on the entire power grid. Equipment in the power grid is often interdependent, and the failure of a certain equipment may have a chain reaction on the equipment connected to it and even a larger area of ​​the power grid. Therefore, the core of the impact range analysis is to determine the specific impact of the faulty equipment on other equipment and areas in the power grid based on the inducing mechanism. By analyzing the key position of the faulty equipment in the power grid and the connection with other equipment, the power safety knowledge graph can be used to simulate the spread of the fault. Specifically, the impact range analysis will simulate the interaction between the power grid topology and the equipment to determine whether the impact of the faulty equipment will spread to downstream equipment or other important nodes. For example, if the inducing mechanism indicates that the failure of a substation may cause voltage fluctuations or power outages, the impact range analysis will consider the transmission lines and distribution networks connected to the substation to determine which areas may be affected. Based on these analyses, the system can generate a range map of the affected grid area, showing the areas and equipment that may be affected by the fault. Once the scope of the affected grid area is obtained, the next step is to develop corresponding management strategies for these equipment. The development of management strategies will be based on the importance of the equipment, the characteristics of the fault propagation path, and the maintenance history of the equipment. For example, for important substations or critical transmission lines, more stringent monitoring measures can be taken to increase the frequency of real-time monitoring and preventive maintenance. If a distribution device is in the affected area, it may be necessary to arrange preventive maintenance immediately to reduce the further propagation of the fault. In addition, for minor equipment within the affected range, gradual load switching or temporary isolation can be adopted to reduce the impact of the fault on the entire grid. For example, suppose that in a certain area of ​​the power grid, the aging of equipment in substation A caused a fault, and the impact range analysis showed that its fault may spread to the connected transmission line B and distribution equipment C. Based on this impact range, the power company may develop a corresponding management strategy, first conducting emergency inspections and preventive maintenance on transmission line B to ensure that it will not be further affected by the fault in subsequent operations. At the same time, for distribution equipment C, isolation measures may be taken to ensure stable power supply to users. This management strategy based on impact range analysis can effectively reduce the impact of faults on the overall operation of the power grid and ensure the continuous and stable operation of the power system.

[0030] In a specific embodiment, the multi-dimensional data includes power grid operation data, equipment maintenance records, power safety standard documents, accident case database, and power grid environment monitoring data; the multi-dimensional data is fused to obtain power safety knowledge fusion data, including: Performing a multi-scale time series analysis on the multi-dimensional data of the power grid to obtain power grid operation status characteristics; wherein the power grid operation status characteristics include power grid voltage fluctuations, power grid frequency changes, and power grid load fluctuations; By using a conditional random field algorithm, the power grid equipment is sequence-labeled based on the equipment maintenance record to obtain the health status of the power grid equipment; wherein the health status of the power grid equipment includes a normal operating state, a minor fault state, and a serious fault state; The electric power safety standard document is semantically parsed by natural language processing technology to obtain electric power safety operation procedures; wherein the electric power safety operation procedures include operating step specifications, emergency plans and safety inspection points; Performing data mining on the accident case database to obtain historical accident patterns; wherein the historical accident patterns include historical failure types, historical failure inducing conditions, and historical failure consequences; Extracting environmental features from the power grid environment monitoring data to obtain environmental extraction features; The power grid operation status characteristics, the health status of the power grid equipment, the power safety operation procedures, the historical accident patterns and the environmental extraction characteristics are subjected to data fusion processing to obtain power safety knowledge fusion data.

[0031] Specifically, in the step of "fusing the multi-dimensional data to obtain the fusion data of power safety knowledge", it is first necessary to process and fuse the power grid operation data, equipment maintenance records, power safety standard documents, accident case database and power grid environment monitoring data. First, multi-scale time series analysis is performed on the multi-dimensional data of the power grid to extract the power grid operation status characteristics, which include power grid voltage fluctuations, power grid frequency changes and power grid load fluctuations. This analysis can help identify the operation status of the power grid at different time scales and reveal the dynamic behavior and potential problems of the power grid. Next, the equipment maintenance records are sequence labeled through the conditional random field algorithm to obtain the health status of the power grid equipment. These health statuses include normal operation status, minor fault status and serious fault status, which help to understand the actual operation status and maintenance needs of the equipment. At the same time, natural language processing technology will be used to perform semantic analysis on the power safety standard documents to extract power safety operation procedures from them. These procedures include operating procedure specifications, emergency plans and safety inspection points, providing systematic guidance for equipment operation and maintenance. In addition, data mining is performed on the accident case database to identify historical accident patterns, which include historical fault types, historical fault-inducing conditions, and historical fault consequences, providing the background and regularity of fault occurrence. Finally, environmental feature extraction is performed on the power grid environment monitoring data to obtain the extracted features of the power grid environment, which may affect the external factors of power grid operation, such as climate change or environmental pollution. After the fusion of this information, comprehensive power safety knowledge fusion data can be obtained. These data integrate the power grid operation status, power grid equipment health status, power safety operation procedures, historical accident patterns, and environmental characteristics to form a comprehensive power safety knowledge base. For example, assuming that the power grid operation status shows abnormal voltage fluctuations, by integrating the equipment health status information and historical accident patterns, it can be determined whether there is a potential risk of failure, and corresponding emergency measures and prevention strategies can be formulated in combination with the power safety operation procedures. This fusion processing can improve the safety and reliability of the power system and ensure the stability of power operation.

[0032] In a specific embodiment, a graph modeling is performed based on the power safety knowledge fusion data through a preset graph theory algorithm to obtain a power safety knowledge graph, including: The power safety knowledge fusion data is quantitatively represented by a preset graph theory algorithm to obtain a multi-dimensional dynamic feature matrix; Performing space-time correlation analysis on power grid equipment based on the multi-dimensional dynamic feature matrix to obtain a space-time correlation matrix between power grid equipment; Performing attribute graph modeling on the power grid devices based on the spatiotemporal association matrix between the power grid devices to obtain a power grid device relationship graph with additional attribute information; Performing deep feature extraction on the power grid equipment relationship graph of the additional attribute information to obtain a multi-level embedded vector representation of the power grid equipment; The multi-level embedding vector representation is subjected to data fusion and graph modeling through a preset heterogeneous information network to obtain a preliminary draft of the electric power safety knowledge graph; Based on the graph attention mechanism, the preliminary draft of the power safety knowledge graph is optimized and adjusted to obtain the power safety knowledge graph.

[0033] Specifically, in the step of "using a preset graph theory algorithm to perform graph modeling based on the power safety knowledge fusion data to obtain a power safety knowledge graph", it is first necessary to quantitatively characterize the power safety knowledge fusion data to construct a multi-dimensional dynamic feature matrix. In this step, by applying a preset graph theory algorithm, different features in the power safety knowledge fusion data are digitally processed to obtain a feature matrix representing the power grid operation status, equipment health status, operating procedures, accident modes and environmental characteristics. This feature matrix can comprehensively and systematically capture the dynamic characteristics of the power grid and form a feature representation of the power system in different dimensions. Next, based on the multi-dimensional dynamic feature matrix, a space-time association analysis is performed on the power grid equipment to obtain a space-time association matrix between the power grid equipment. The space-time association matrix reveals the timeliness connection and spatial distribution characteristics between the equipment by analyzing the mutual relationship between the power grid equipment in space and time. This matrix helps to understand the mutual influence of equipment at different times and places, and how they work together on the overall safety and stability of the power grid. After obtaining the space-time association matrix, attribute graph modeling is further performed on the power grid equipment based on the matrix to generate a power grid equipment relationship graph with additional attribute information. In the process of attribute graph modeling, various attribute information of power grid equipment (such as equipment type, operating status, maintenance history, etc.) is incorporated into the graph, and the relationship network between equipment is constructed through graph theory methods. This relationship graph not only shows the interconnection between equipment, but also incorporates the additional attribute information of each equipment, providing detailed equipment background for subsequent analysis. Deep feature extraction of the power grid equipment relationship graph with additional attribute information can obtain a multi-level embedded vector representation of the power grid equipment. This step converts the nodes in the graph (i.e., power grid equipment) into multi-level vector representations through complex graph neural networks or embedding algorithms. These vectors can capture the deep features and relationships of the equipment in the graph. These embedded vectors not only reflect the basic attributes of the equipment, but also integrate the role of the equipment in the power grid and its complex relationship with other equipment. Through the preset heterogeneous information network, the multi-level embedded vector representation is data fused and graph modeled to obtain the first draft of the power safety knowledge graph. The use of heterogeneous information networks allows different types of data (such as equipment attributes, environmental factors, historical accidents, etc.) to be effectively integrated to form a comprehensive first draft of the power safety knowledge graph. This first draft graph integrates various information of the power grid and shows the complex relationships and interactions between equipment. Finally, based on the graph attention mechanism, the draft of the power safety knowledge graph was optimized and adjusted to obtain the final power safety knowledge graph. The graph attention mechanism optimizes the structure and content of the graph by assigning different weights to different nodes and edges in the graph, thereby improving the graph's ability to predict and analyze power safety.The final power safety knowledge graph can accurately reflect the status of power grid equipment, the relationship between equipment, and potential safety risks, providing strong support for the monitoring and management of power grid operations. For example, assuming that in a power grid, the power safety knowledge graph constructed through the above steps shows that the failure modes of certain equipment under specific environmental conditions frequently occur. This graph can help operation and maintenance personnel identify key faulty equipment and take targeted maintenance measures based on the multi-level embedded vectors provided in the graph, thereby effectively preventing and responding to potential problems in the power system.

[0034] In a specific embodiment, a fault association analysis is performed on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path, including: Perform fault correlation analysis on the power grid equipment based on the power fault information to obtain potential power fault equipment; wherein the potential power fault equipment includes potential power fault equipment with short circuit, open circuit and overload; Topologically sorting the potential power failure devices based on the power safety knowledge graph to obtain potential power failure sorting devices; constructing a cause-and-effect chain of faulty devices based on the potential power fault sequencing device; Using a preset Markov chain Monte Carlo method to perform path sampling on the causal chain, a plurality of potential fault propagation paths are obtained; In the causal chain, probability calculation is performed on each of the potential power fault sequencing devices to obtain the failure probability of the potential power fault sequencing device; Based on the failure probability of the potential power fault sorting device, probability calculation is performed on the multiple potential fault propagation paths to obtain the probability of each potential fault propagation path; The potential fault propagation path corresponding to the highest probability is determined as the power fault propagation path.

[0035] Specifically, in this step, it is first necessary to perform fault association analysis on the power fault information based on the power safety knowledge graph to identify potential power fault devices. This means that by analyzing the existing fault information, it is possible to determine the devices in the power grid that may have short circuits, circuit breaks or overloads. For example, if an abnormal signal appears in a generator, according to the association analysis of the power safety knowledge graph, the generator and the substation and other related equipment connected to it will be identified as potential power fault devices, and these devices may become nodes for fault propagation. After identifying these potential fault devices, these devices are topologically sorted based on the power safety knowledge graph. The purpose of topological sorting is to arrange these devices in the order in which the faults occur. For example, if a short circuit of a generator causes an overload of a substation, the generator will be ranked before the substation. In this way, the potential fault devices in the power grid can be clearly arranged according to their association and order of occurrence, forming the so-called potential power fault sorting devices. After the sorting is completed, the next step is to build a causal chain of fault devices based on these potential power fault sorting devices. The causal chain reflects the transmission process of the fault, that is, how the failure of a certain device causes the failure of other devices. For example, a short circuit of a generator may cause other equipment in the power grid to overload, which in turn may cause some substations to disconnect. The causal chain maps these complex equipment failure associations into a clear fault propagation path. To further analyze the path, the preset Markov chain Monte Carlo method is used to sample the paths of the causal chain. This method generates multiple potential fault propagation paths through multiple simulation samplings, thereby obtaining different possible combinations. For example, in a power grid, possible fault paths include a short circuit of a generator causing an overload of a substation, which then affects the downstream power supply equipment. Through path sampling, the fault propagation of these different paths can be evaluated. Next, on the causal chain, a probability calculation is performed for each potential power fault sorting device to obtain the probability of failure of each device. By combining these probabilities, the probability of failure of each device can be evaluated. For example, if a generator is improperly maintained or overloaded, its probability of failure may be higher than that of other devices. Finally, based on these failure probabilities, all potential fault propagation paths are probabilistically calculated to obtain the fault propagation probability of each path. Finally, the potential fault propagation path corresponding to the highest probability is determined as the power fault propagation path. This process helps grid managers identify the most likely fault transmission mode in advance, so that they can make effective prevention and response measures in actual operations. For example, when generators and related equipment in a grid are assessed as having a high risk of failure, managers can carry out preventive maintenance in advance to reduce the risk of power outages.

[0036] In a specific embodiment, causal relationship mining is performed on the power fault propagation path to obtain the power fault device and the inducing mechanism corresponding to the power fault device, including: Conduct preliminary causal relationship mining on the power fault propagation path to form a preliminary causal relationship map; wherein the preliminary causal relationship map includes the relationship between potential power fault equipment and the inducing mechanism corresponding to the potential power fault equipment; The Granger causality test algorithm is used to verify the relationship between the potential power failure device and the inducing mechanism corresponding to the potential power failure device, so as to obtain a verification causal relationship network; wherein the verification causal relationship network includes the potential power failure device and the significant inducing factors of the potential power failure device; Performing dimensionality reduction processing on the verification causal relationship network to obtain the main inducing mechanism corresponding to each potential power fault device; Performing time series analysis on the main inducing mechanism characteristics to obtain the time inducing mechanism; combining the time induction mechanism and the main induction mechanism to obtain a combined induction mechanism; The combined inducing mechanisms are scored to obtain an inducing mechanism score, and the combined inducing mechanism with the highest score is determined as the inducing mechanism, and the potential power fault device corresponding to the combined inducing mechanism with the highest score is determined as the power fault device.

[0037] Specifically, in the step of "carrying out causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment", it is first necessary to carry out preliminary causal relationship mining on the power fault propagation path to form a preliminary causal relationship map. In this process, by analyzing the equipment and its fault information in the power fault propagation path, the relationship between the potential power fault equipment and the inducing mechanisms corresponding to these equipment can be identified, thereby constructing a preliminary causal relationship map. The map includes the preliminary association between the potential power fault equipment and their inducing mechanisms, providing a basis for the subsequent causal relationship verification. Next, the Granger causality test algorithm is used to verify the relationship between the potential power fault equipment and its corresponding inducing mechanism in the preliminary causal relationship map to obtain a verification causal relationship network. The Granger causality test algorithm uses a statistical test method to determine whether a variable can be used to predict another variable, thereby verifying the validity of the causal relationship in the preliminary causal relationship map. After this verification process, the obtained verification causal relationship network includes the potential power fault equipment and its significant inducing factors, providing a more accurate and reliable causal relationship network. Dimensionality reduction processing of the verification causal relationship network is an important next step. Dimensionality reduction can simplify the causal network and remove redundant information to obtain the main inducing mechanism corresponding to each potential power fault device. In this step, several main inducing mechanisms that are most closely related to potential power fault devices are extracted through data dimensionality reduction technology, thereby reducing data complexity and improving analysis efficiency. Subsequently, time series analysis is performed on the characteristics of the main inducing mechanisms to obtain the time inducing mechanism. Time series analysis identifies the characteristics of the inducing mechanism at a specific time point or time period by analyzing the law of data changes over time. These time inducing mechanisms can reveal the relationship between the failure and time of power fault equipment and provide important information for further comprehensive analysis. After obtaining the time inducing mechanism, it is combined with the main inducing mechanism to obtain the combined inducing mechanism. The combined inducing mechanism integrates the time inducing mechanism with the main inducing mechanism to form a more comprehensive and complex inducing mechanism model. This combination can more accurately reflect the fault inducement of power fault equipment, thereby improving the accuracy of fault prediction. Scoring the combined inducing mechanism is an important next step. By scoring the combined inducing mechanism, the inducing mechanism score can be obtained. The scoring process calculates the score value of each combined inducing mechanism by comprehensively considering the significance and influence of each inducing mechanism. Finally, the combined inducing mechanism with the highest score is determined as the inducing mechanism, and the corresponding potential power failure equipment is determined as the power failure equipment. For example, in a power system, through the above steps analysis, some transformers and transmission lines are identified as potential power failure equipment. The preliminary causal relationship map shows that the failure of these devices may be related to factors such as overload and ambient temperature changes.Through the verification of the Granger causality test algorithm, it was determined that the change in ambient temperature was a significant inducing factor for these equipment failures. After dimensionality reduction, overload and ambient temperature change were extracted as the main inducing mechanisms. Time series analysis further showed that ambient temperature had a significant impact on equipment failures during the high temperature period in summer. After combining these time inducing mechanisms with the main inducing mechanisms, the combined inducing mechanism obtained showed that high temperature in summer and long-term overload were the main causes of transformer failures. Through scoring, it was determined that this combined inducing mechanism had the highest score, so it was determined as the inducing mechanism, and finally the transformer was confirmed as a power failure device. This analysis process provides a scientific basis for the preventive maintenance of the power system and can effectively reduce the risk of failures.

[0038] In a specific embodiment, the influence range of the power grid equipment is analyzed based on the induction mechanism to obtain the affected power grid area, and a management strategy corresponding to the power grid equipment within the affected power grid area is formulated, including: Using a preset geographic information system, the spatial distribution of power grid equipment is mapped based on the induction mechanism to obtain a spatial distribution map of power fault equipment; wherein the spatial distribution map includes the geographical location of the power fault equipment in the power grid and the impact range of the power fault equipment; Performing network topology analysis on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices; wherein the correlation matrix includes connectivity and mutual influence between the power fault devices; Using the shortest path algorithm in graph theory, the path impact analysis of the power fault device is performed based on the association matrix to obtain the affected power grid area range; wherein the affected power grid area range is a set of impact paths from the power fault device to each power grid device, and the affected power grid area range includes the power grid fault impact path length and the power grid fault impact intensity; Comprehensively evaluate the length of the grid fault impact path and the grid fault impact intensity to obtain a risk assessment matrix for grid equipment; Perform risk analysis on the power grid equipment based on the risk assessment matrix to obtain a risk division scheme affecting the power grid area; wherein the risk division scheme is the distribution of high-risk equipment and low-risk equipment in each area; Based on the risk classification scheme, a management strategy is formulated for the power grid equipment in each of the areas to obtain management strategies for power grid equipment of different risk levels; wherein the management strategy includes a preventive maintenance plan and an emergency response plan.

[0039] Specifically, in this step, the influence range of the power grid equipment is analyzed based on the inducing mechanism. First, a preset geographic information system (GIS) is used to perform spatial distribution mapping to generate a spatial distribution map of the power fault equipment. This map not only shows the geographical location of the power fault equipment in the power grid, but also identifies the influence range of these equipment. For example, after a generator fails, the surrounding substations and distribution lines may be affected. Next, the network topology analysis of the power fault equipment is performed based on the spatial distribution map to obtain the correlation matrix between the power fault equipment, reflecting the connectivity and mutual influence between the equipment. After using the shortest path algorithm in graph theory, the path impact analysis of these devices can be performed to form the scope of the area affecting the power grid. This impact area not only includes the set of impact paths from the power fault equipment to each power grid equipment, but also covers the length and intensity of the grid fault impact path. For example, assuming that the failure of a substation will cause the power outage of multiple nearby distribution lines, the path analysis can quantify the severity of these impacts. After a comprehensive evaluation of the length and intensity of the impact path of the power grid fault, a risk assessment matrix of the power grid equipment can be constructed. The risk level of each device in the matrix is ​​determined by considering its position and impact degree in the fault propagation path. Based on this risk assessment matrix, risk analysis can be performed and a risk division plan that affects the scope of the power grid area can be formed. The plan describes in detail the distribution of high-risk and low-risk equipment in each area. Finally, according to the risk division plan, management strategies are formulated for power grid equipment in each area. These management strategies are formulated for equipment with different risk levels. For example, for high-risk equipment, more frequent maintenance and monitoring measures may be taken, while for low-risk equipment, the inspection frequency can be appropriately reduced, so as to achieve optimal allocation of resources and ensure the safety and stability of the power grid. Through this series of analysis and strategy formulation, the ability of the power grid to respond to failures can be effectively improved.

[0040] In a specific embodiment, a network topology analysis is performed on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices, including: Extracting the interaction relationship between the devices corresponding to the power fault devices in the spatial distribution diagram to obtain the interaction network relationship between the power fault devices; Modeling the dependency relationship between the power failure devices based on the interactive network relationship between the power failure devices to obtain the dependency relationship between the devices; Using different colors to depict the dependency relationship, obtaining the depicted dependency relationship and the strength of the depicted dependency relationship; Dynamically predicting the depicted dependency relationship and the strength of the depicted dependency relationship to obtain a dynamic interaction prediction network between power fault devices; A correlation matrix between power failure devices is constructed based on the dynamic interaction prediction network between power failure devices.

[0041] Specifically, in this step, the network topology analysis of the power fault devices in the spatial distribution diagram is first performed to obtain the correlation matrix between the power fault devices. Specifically, the analysis process begins with the extraction of the interactive relationship between the power fault devices and the corresponding devices in the spatial distribution diagram. This process involves identifying the interactive network relationship between the power fault devices. For example, the failure of a generator may directly affect the substation and other distribution equipment connected to it. Therefore, through the extraction of the interactive relationship, the mutual influence between these devices can be clarified. Next, based on the interactive network relationship between the power fault devices, the dependency relationship between the power fault devices is modeled. This modeling process reflects how the devices are interdependent. For example, if the failure of a generator causes an overload of a substation, the relationship between the two will be significantly identified in the dependency model. By depicting these dependencies in different colors, their strength can be intuitively displayed. Dark colors may indicate strong dependencies, while light colors indicate weak dependencies. Then, the depicted dependencies and their strengths are dynamically predicted to generate a dynamic interactive prediction network between power fault devices. This dynamic prediction not only considers the current equipment status, but also integrates various factors that may affect equipment interaction in the future, such as load changes, equipment maintenance status, etc. Through these dynamic predictions, potential fault propagation paths and their possible impact ranges can be identified, so as to better understand how faults propagate in the power grid. Based on the dynamic interaction prediction network between the power fault devices, the correlation matrix between the power fault devices is finally constructed. This matrix brings together all the information of the above analysis and prediction to form a comprehensive equipment relationship diagram, which clarifies the degree of mutual influence and correlation of each power fault device. For example, in a practical scenario, if a generator failure causes a chain reaction of multiple transmission lines, the correlation matrix can show which devices play a key role in the propagation of the fault, thereby providing important data support for power grid management and fault response. Through this systematic network topology analysis, power grid managers can more effectively identify potential fault propagation paths and take preventive and emergency response measures in advance to ensure the safety and stability of the power system. This method not only improves the accuracy of fault identification, but also enhances the risk resistance of the power grid, thereby providing a guarantee for maintaining the continuity of power supply.

[0042] The above describes the method for associating power safety knowledge in the embodiment of the present invention. The following describes the system for associating power safety knowledge in the embodiment of the present invention. Figure 2 , an embodiment of the power safety knowledge association system in the embodiment of the present invention includes: An acquisition module 21 is used to acquire multi-dimensional data of power safety knowledge and fuse the multi-dimensional data to obtain power safety knowledge fusion data; A modeling module 22, configured to perform graph modeling based on the power safety knowledge fusion data by using a preset graph theory algorithm to obtain a power safety knowledge graph; A first analysis module 23 is used to obtain power fault information of power grid equipment, and perform fault association analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; A mining module 24 is used to mine the cause-effect relationship of the power fault propagation path to obtain the power fault device and the inducing mechanism corresponding to the power fault device; The second analysis module 25 is used to analyze the impact range of the power grid equipment based on the induction mechanism, obtain the impact area of ​​the power grid, and formulate management strategies corresponding to the equipment within the impact area of ​​the power grid.

[0043] In this embodiment, for the specific implementation of each unit in the above system embodiment, please refer to the above method embodiment, which will not be repeated here.

[0044] Reference Figure 3 The present invention also provides a computer device in an embodiment, wherein the internal structure of the computer device can be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a display screen, an input device, a network interface and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the corresponding data in this embodiment. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0045] Those skilled in the art will understand that Figure 3 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied.

[0046] An embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, the above method is implemented. It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0047] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided by the present invention and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double-speed data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM.

[0048] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the presence of other identical elements in the process, device, article or method including the element.

[0049] The above description is only a preferred embodiment of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for associating power safety knowledge, characterized in that: The following steps are involved: Acquire multi-dimensional data of power safety knowledge, and fuse the multi-dimensional data to obtain power safety knowledge fusion data; By using a preset graph theory algorithm, graph modeling is performed based on the power safety knowledge fusion data to obtain a power safety knowledge graph; Acquire power fault information of power grid equipment, and perform fault correlation analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; Perform causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment; Based on the inducing mechanism, the influence range of the power grid equipment is analyzed to obtain the area range of the affected power grid, and a management strategy corresponding to the equipment within the area range of the affected power grid is formulated.

2. The method for associating power safety knowledge according to claim 1, characterized in that: The multi-dimensional data includes power grid operation data, equipment maintenance records, power safety standard documents, accident case database, and power grid environment monitoring data; the multi-dimensional data is fused to obtain power safety knowledge fusion data, including: Performing a multi-scale time series analysis on the multi-dimensional data of the power grid to obtain power grid operation status characteristics; wherein the power grid operation status characteristics include power grid voltage fluctuations, power grid frequency changes, and power grid load fluctuations; By using a conditional random field algorithm, the power grid equipment is sequence-labeled based on the equipment maintenance record to obtain the health status of the power grid equipment; wherein the health status of the power grid equipment includes a normal operating state, a minor fault state, and a serious fault state; The electric power safety standard document is semantically parsed by natural language processing technology to obtain electric power safety operation procedures; wherein the electric power safety operation procedures include operating step specifications, emergency plans and safety inspection points; Performing data mining on the accident case database to obtain historical accident patterns; wherein the historical accident patterns include historical failure types, historical failure inducing conditions, and historical failure consequences; Extracting environmental features from the power grid environment monitoring data to obtain environmental extraction features; The power grid operation status characteristics, the health status of the power grid equipment, the power safety operation procedures, the historical accident patterns and the environmental extraction characteristics are subjected to data fusion processing to obtain power safety knowledge fusion data.

3. The method for associating power safety knowledge according to claim 2, characterized in that: Through a preset graph theory algorithm, graph modeling is performed based on the power safety knowledge fusion data to obtain a power safety knowledge graph, including: The power safety knowledge fusion data is quantitatively represented by a preset graph theory algorithm to obtain a multi-dimensional dynamic feature matrix; Performing space-time correlation analysis on power grid equipment based on the multi-dimensional dynamic feature matrix to obtain a space-time correlation matrix between power grid equipment; Performing attribute graph modeling on the power grid devices based on the spatiotemporal association matrix between the power grid devices to obtain a power grid device relationship graph with additional attribute information; Performing deep feature extraction on the power grid equipment relationship graph of the additional attribute information to obtain a multi-level embedded vector representation of the power grid equipment; The multi-level embedding vector representation is subjected to data fusion and graph modeling through a preset heterogeneous information network to obtain a preliminary draft of the electric power safety knowledge graph; Based on the graph attention mechanism, the preliminary draft of the power safety knowledge graph is optimized and adjusted to obtain the power safety knowledge graph.

4. The method for associating power safety knowledge according to claim 1, characterized in that: Based on the power safety knowledge graph, a fault correlation analysis is performed on the power fault information to obtain a power fault propagation path, including: Perform fault correlation analysis on the power grid equipment based on the power fault information to obtain potential power fault equipment; wherein the potential power fault equipment includes potential power fault equipment with short circuit, open circuit and overload; Topologically sorting the potential power failure devices based on the power safety knowledge graph to obtain potential power failure sorting devices; constructing a cause-and-effect chain of faulty devices based on the potential power fault sequencing device; Using a preset Markov chain Monte Carlo method to perform path sampling on the causal chain, a plurality of potential fault propagation paths are obtained; In the causal chain, probability calculation is performed on each of the potential power fault sequencing devices to obtain the failure probability of the potential power fault sequencing device; Based on the failure probability of the potential power fault sorting device, probability calculation is performed on the multiple potential fault propagation paths to obtain the probability of each potential fault propagation path; The potential fault propagation path corresponding to the highest probability is determined as the power fault propagation path.

5. The method for associating power safety knowledge according to claim 1, characterized in that: Performing causal relationship mining on the power fault propagation path to obtain the power fault equipment and the inducing mechanism corresponding to the power fault equipment includes: Conduct preliminary causal relationship mining on the power fault propagation path to form a preliminary causal relationship map; wherein the preliminary causal relationship map includes the relationship between potential power fault equipment and the inducing mechanism corresponding to the potential power fault equipment; The Granger causality test algorithm is used to verify the relationship between the potential power failure device and the inducing mechanism corresponding to the potential power failure device, so as to obtain a verification causal relationship network; wherein the verification causal relationship network includes the potential power failure device and the significant inducing factors of the potential power failure device; Performing dimensionality reduction processing on the verification causal relationship network to obtain the main inducing mechanism corresponding to each potential power fault device; Performing time series analysis on the main inducing mechanism characteristics to obtain the time inducing mechanism; combining the time induction mechanism and the main induction mechanism to obtain a combined induction mechanism; The combined inducing mechanisms are scored to obtain an inducing mechanism score, and the combined inducing mechanism with the highest score is determined as the inducing mechanism, and the potential power fault device corresponding to the combined inducing mechanism with the highest score is determined as the power fault device.

6. The method for associating power safety knowledge according to claim 1, characterized in that: Based on the induction mechanism, the influence range of the power grid equipment is analyzed to obtain the influence area of ​​the power grid, and a management strategy corresponding to the power grid equipment within the influence area is formulated, including: Using a preset geographic information system, the spatial distribution of power grid equipment is mapped based on the induction mechanism to obtain a spatial distribution map of power fault equipment; wherein the spatial distribution map includes the geographical location of the power fault equipment in the power grid and the impact range of the power fault equipment; Performing network topology analysis on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices; wherein the correlation matrix includes connectivity and mutual influence between the power fault devices; Using the shortest path algorithm in graph theory, the path impact analysis of the power fault device is performed based on the association matrix to obtain the affected power grid area range; wherein the affected power grid area range is a set of impact paths from the power fault device to each power grid device, and the affected power grid area range includes the power grid fault impact path length and the power grid fault impact intensity; Comprehensively evaluate the length of the grid fault impact path and the grid fault impact intensity to obtain a risk assessment matrix for grid equipment; Perform risk analysis on the power grid equipment based on the risk assessment matrix to obtain a risk division scheme affecting the power grid area; wherein the risk division scheme is the distribution of high-risk equipment and low-risk equipment in each area; Based on the risk classification scheme, a management strategy is formulated for the power grid equipment in each of the areas to obtain management strategies for power grid equipment of different risk levels; wherein the management strategy includes a preventive maintenance plan and an emergency response plan.

7. The method for associating power safety knowledge according to claim 6, characterized in that: Performing network topology analysis on the power fault devices in the spatial distribution diagram to obtain a correlation matrix between the power fault devices, including: Extracting the interactive relationship between the devices corresponding to the power fault devices in the spatial distribution diagram to obtain the interactive network relationship between the power fault devices; Modeling the dependency relationship between the power failure devices based on the interactive network relationship between the power failure devices to obtain the dependency relationship between the devices; Using different colors to depict the dependency relationship, obtaining the depicted dependency relationship and the strength of the depicted dependency relationship; Dynamically predicting the depicted dependency relationship and the strength of the depicted dependency relationship to obtain a dynamic interaction prediction network between power fault devices; A correlation matrix between power failure devices is constructed based on the dynamic interaction prediction network between power failure devices.

8. A power safety knowledge association system, characterized in that: include: An acquisition module, used to acquire multi-dimensional data of power safety knowledge, and fuse the multi-dimensional data to obtain power safety knowledge fusion data; A modeling module, used to perform graph modeling based on the power safety knowledge fusion data through a preset graph theory algorithm to obtain a power safety knowledge graph; A first analysis module is used to obtain power fault information of power grid equipment, and perform fault association analysis on the power fault information based on the power safety knowledge graph to obtain a power fault propagation path; A mining module, used to mine the cause-effect relationship of the power fault propagation path to obtain the power fault device and the inducing mechanism corresponding to the power fault device; The second analysis module is used to analyze the impact range of the power grid equipment based on the inducing mechanism, obtain the impact area of ​​the power grid, and formulate management strategies corresponding to the equipment within the impact area of ​​the power grid.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images