Method and system for constructing power operation personal risk dynamic evolution through key factor analysis

Abstract

The invention discloses a method and system for constructing power operation personal risk dynamic evolution through key factor analysis, and relates to the field of artificial intelligence, and the method comprises the steps: constructing an operation risk analysis knowledge graph, and carrying out the entity recognition; extracting knowledge graph entities; performing accident chain replay of the knowledge graph and the expert model; and constructing a risk dynamic evolution model based on the space-time knowledge graph. According to the method, previous electric power operation accident cases are analyzed, influences of different time and different environments on operation of electric power personnel are researched, risk factors in historical accidents are analyzed, an analysis method based on a space-time knowledge graph is provided, historical accident chain deduction is achieved, and the risk factors in the historical accidents are analyzed. And providing a power operation personal risk dynamic evolution model based on key risk factors.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence, and particularly to a method and system for dynamically evolving the personal risk of electric power operation by analyzing key factors and constructing the same. Background Art

[0002] Behind the rapid development of the electric power industry, the safety of electric power production is the fundamental basis and important prerequisite for all work of electric power enterprises. Improving the safety production level of electric power enterprises is an urgent task for relevant researchers and electric power enterprise managers. According to Heinrich's law, due to the sharp expansion of the total amount of electric power generation and consumption, the potential safety hazards existing in electric power production will also increase accordingly, and the number of safety incidents and accidents brought about thereby may also increase significantly.

[0003] The electric power industry has the industry characteristics of complex systems and large chain effects of accident consequences. The modern electric power industry is an industry that operates by integrating various modern scientific and technological means. It is a socialized joint system with highly centralized and unified production, conversion, transmission, and distribution of electric power energy, and is a dynamically complex system involving various factors. The electric power production process is a complex process including power generation, power transmission, voltage transformation, power distribution, and power consumption. In a power system, power generation, power supply, and power consumption equipment are interconnected and coupled with each other on the entire large power grid. Therefore, any failure or malfunction of any equipment will affect and spread to the entire system in a short time. In severe cases, it will cause a chain reaction and lead to the expansion of the accident. In serious cases, a large-scale power outage accident will occur, which is incomparable to accidents in other industries.

[0004] With the development of information technology, data collection means have been continuously upgraded, and more and more text data generated during the operation of various industries can be efficiently collected, greatly promoting the intelligent process of industries based on big data technology. There are various types of text data, which can be divided into three types: structured text, semi-structured text, and unstructured text according to the organizational form of the text content. When performing data analysis, the more structured the data is, the easier it is to extract valuable information. However, in most scenarios, the proportion of unstructured data in the total data volume is the highest, followed by semi-structured text, and structured text is often difficult to directly obtain. In order to better manage the knowledge in a large amount of unstructured text, data is usually structured based on big data technology, and then the information described in natural language text can be converted into a unified format. On this basis, the knowledge contained in the text is extracted and stored in the form of triples, and finally a knowledge graph is constructed to provide knowledge analysis and reasoning capabilities. The knowledge graph stores a large amount of structured knowledge in the form of graph data. It can display the relationships between knowledge in the form of triples. Based on the intelligent model of the knowledge graph, strong logical reasoning capabilities can be obtained. Therefore, the knowledge graph is a key support technology for applications such as user portraits, intelligent question answering, associated text analysis, and text summary generation.

[0005] With the electronicization of text data related to power grid equipment status assessment, operation and maintenance, testing and overhaul, and the rapid increase in the amount of text data in the field of power grid equipment, such as power grid equipment defect specifications and power grid equipment design specifications, these texts lay the data foundation for the research and development of intelligent technology in the field of power grid equipment. However, these power grid equipment text data have the characteristics of rich types, fast iteration speed, and low information density. They come from different sources and have different structures. Most of them are unstructured and semi-structured data. Although they contain a lot of valuable professional knowledge, they cannot be directly used for knowledge analysis and reasoning due to their low level of structuring. How to extract domain expertise from these multi-source heterogeneous power grid equipment text data and build a vertical domain knowledge graph for power grid equipment is the key focus and difficulty in achieving efficient management of power grid equipment knowledge. Summary of the invention

[0006] In view of the above-mentioned existing problems, the present invention analyzes previous power operation accident cases, studies the impact of different times and different environments on the operations of power personnel, analyzes the risk factors in historical accidents, and proposes an analysis method based on spatiotemporal knowledge graphs to realize historical accident chain deduction, and then proposes a dynamic evolution model of personal risks in power operations based on key risk factors.

[0007] In order to solve the above technical problems, a key factor analysis method for constructing the dynamic evolution of personal risk in power operations is proposed, including:

[0008] Construct a knowledge graph for job hazard analysis and perform entity recognition; extract knowledge graph entities; replay accident chains using knowledge graphs and expert models; and construct a risk dynamic evolution model based on spatiotemporal knowledge graphs.

[0009] As a preferred solution of the key factor analysis method for constructing the dynamic evolution of personal risk in power operations described in the present invention, the construction of the knowledge graph of operation hazard analysis includes a process stage, an output stage and an implementation stage;

[0010] Said operations include historical incidents;

[0011] The process stages include document analysis of the problems and coverage that the operation needs to solve, including in-depth reading of historical accident documents and operating procedures, extracting key concepts and terms corresponding to the operation hazard analysis process, referring to existing meta-ontologies, defining entities, attributes and relationships in the operation hazard analysis process, and compiling a glossary through knowledge acquisition through expert interviews and literature reviews;

[0012] The output stage includes classifying the output concepts, building rules to determine the relationships and constraints between the concepts, and outputting the overall framework;

[0013] The implementation stage includes defining a classification system and a role model for job hazard analysis. After selecting or graph query language, ontology implementation is carried out, creating an ontology for job hazard analysis, and importing the ontology and data into the knowledge graph platform for verification;

[0014] The ontology implementation includes creating a schema file and a data file for the ontology. The data file contains hazard and control measures and information on personal risks in electric power operations.

[0015] As a preferred solution for constructing a dynamic evolution method for personal risks in electric power operations through key factor analysis according to the present invention, wherein: the extraction of knowledge graph entities includes using the LDA model. The first layer is the document layer, and the training model deeply explores the associations with key entities in the documents;

[0016] The second layer is the theme layer, which presents the theme content in the documents in the form of keywords;

[0017] The third layer is the distribution layer of theme keywords. The weights of theme keywords are solved through a probability distribution model to extract keywords.

[0018] As a preferred solution for constructing a dynamic evolution method for personal risks in electric power operations through key factor analysis according to the present invention, wherein: the extraction of keywords includes using the TF-IDF method to extract keywords from documents, i.e., historical accidents. The extracted keywords represent historical accidents:

[0019]

[0020] wherein, tf ij represents the term frequency of word i in document j, n i,j is the number of the word in document j, and ∑ k n k,j represents the sum of all words in document j. idf is the inverse document frequency:

[0021]

[0022] wherein, |D| is the total number of all documents in the corpus, |{j:t∈d j}| represents the number of documents containing word t, and d j represents the jth document in the corpus, calculating the information on the frequency of word occurrences in the document;

[0023] Normalize the elements of the judgment matrix column by column to calculate the weight b ij :

[0024]

[0025] wherein, a ijis an element of the pairwise comparison matrix, and the weight vector w is estimated using the arithmetic mean of all column vectors i :

[0026]

[0027] When AW = λW, the maximum eigenvalue λ max is:

[0028]

[0029] The consistency index CI is:

[0030]

[0031] Among them, n represents the order of the judgment matrix, that is, the number of rows or columns of the matrix, AW represents the result of multiplying the judgment matrix by the weight vector W, and W represents the weight vector or eigenvector, specifically the weight distribution of the judgment matrix.

[0032] As a preferred solution for constructing the dynamic evolution method of personal risks in electric power operations through key factor analysis according to the present invention, wherein: the accident chain replay includes extracting the relationships of all risk factors based on the constructed knowledge graph of job hazard analysis to form an accident replay simulation time chain;

[0033] Adopt the scenario-based continuous operation risk assessment method for scenario re-assessment, analyze the root cause and other reasons for the accident, and update and store the knowledge graph.

[0034] As a preferred solution for constructing the dynamic evolution method of personal risks in electric power operations through key factor analysis according to the present invention, wherein: the re-assessment includes calculating the single-attribute similarity and the comprehensive layer similarity;

[0035] The single-attribute similarity uses the KNN algorithm. By calculating the similarity between the target case and the source case, the top k cases with the highest similarity are output. The single-attribute similarity Sim(s,t) between the target situation and the source situation is:

[0036] Sim(s,t) = 1 - D(s,t)

[0037]

[0038] Among them, D(s,t) is the normalized Euclidean distance between the target situation and the source situation, m is the retrieval attribute number, M is the total number of retrieval attributes, and w m is the weight of attribute m in D m (s,t), and D m (s,t) is the Euclidean distance between the normalized target situation and the source situation in the m attribute dimensions;

[0039] The data types of the retrieved attributes include symbolic types and numerical types. The symbolic types are determined as follows:

[0040]

[0041] The numerical types are determined as follows:

[0042]

[0043] D m (s, t) = ∣P sm -P tm ∣

[0044] where P sk is the k-attribute value of the source case, P tm is the m-attribute value of the target case, d m (s, t) is the distance between the source case and the target case in m attribute dimensions, Max m is the maximum value of the m-attribute values in the case base, Min m is the minimum value of the m-attribute values in the case base.

[0045] The similarity at the comprehensive layer is as follows:

[0046]

[0047] where r i is the attribute weight, n is the number of all attributes compared between the target case and the source case, o is the variable index, Sim(s, t) is the similarity of a single attribute, and the global attribute similarity SIM(G) is obtained as follows:

[0048] SIM(G) = Sim(C)c + Sim(C)g

[0049] where c represents the similarity of the target case and g represents the similarity of the global context.

[0050] As a preferred solution for constructing the dynamic evolution method of the personal risk in electric power operations by analyzing the key factors of the present invention, wherein: constructing the risk dynamic evolution model includes constructing a spatio-temporal knowledge graph based on the constructed operation hazard analysis knowledge graph, which is dynamic, stores spatio-temporal dynamic data, and queries the entity relationship information at any moment or time period;

[0051] Storing the graph information in a dynamic visualization form and updating the data, performing accident replay and reasoning. During the deduction process, when any missing information is found in the original data, the spatio-temporal knowledge graph completes the information to form a dynamic evolution model;

[0052] The information completion of the spatio-temporal knowledge graph detects the missing entities, attributes, and relationships in the knowledge graph, combines historical data, real-time data, and similar case patterns, uses symbolic and numerical data processing, KNN algorithm, and graph embedding reasoning technology to complete the missing information, and verifies the accuracy of the information through expert review, and dynamically updates the completion results to the spatio-temporal knowledge graph.

[0053] Another object of the present invention is to provide a key factor analysis for constructing a dynamic evolution system of personal risks in electric power operations. The present invention aims to solve the spatio-temporal matching problem of supply and demand in the distribution network. The static cluster division module initially completes the pairing of resources and loads, and then the dynamic optimization cluster division module uses the particle swarm algorithm for real-time adjustment, and the voltage control module calculates the reactive power compensation demand and guides the controllable resources to participate in voltage regulation, so as to improve the voltage stability of the distribution network, optimize the resource utilization efficiency, achieve spatio-temporal matching of supply and demand, ensure the orderly participation of controllable resources in voltage regulation, and comprehensively improve the economy, reliability, and coordination of the power grid operation.

[0054] As a preferred embodiment of the key factor analysis for constructing a dynamic evolution system of personal risks in electric power operations according to the present invention, it is characterized by including a knowledge graph construction module for job hazard analysis, a knowledge graph entity extraction module, an accident chain replay module, and a risk dynamic evolution module;

[0055] The knowledge graph construction module for job hazard analysis analyzes the job environment and historical accident documents, extracts key concepts and terms, defines entities, attributes, and relationships, creates a glossary, constructs a classification system and a role model, realizes an ontology, and imports the ontology and data into the knowledge graph platform for verification;

[0056] The knowledge graph entity extraction module uses the LDA model to perform hierarchical processing on the documents, including the document layer, the topic layer, and the topic keyword distribution layer, extracts keywords using the TF-IDF method, and determines the key entities and their associations;

[0057] The accident chain replay module uses the constructed knowledge graph to extract the relationships of risk factors, forms an accident replay simulation time chain, uses a scenario-based evaluation method for re-evaluation, analyzes the accident causes, and updates the knowledge graph;

[0058] The risk dynamic evolution module constructs a spatio-temporal knowledge graph to store and query entity relationship information, dynamically visualizes the graph information, updates the data, performs accident replay and reasoning, completes the missing information, and forms a dynamic evolution model.

[0059] A computer device includes a memory and a processor, wherein the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of analyzing key factors to construct a dynamic evolution method for personal risks in power operations are implemented.

[0060] 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 analyzing key factors to construct a dynamic evolution method for personal risks in power operations are implemented.

[0061] Beneficial effects of the present invention: The spatiotemporal knowledge graph method proposed in the present invention can reproduce the accident chain of historical accidents and analyze the main causes of accidents. In addition, accidents can also be deduced and analyzed. Compared with the static model constructed by ordinary knowledge graphs, our method has achieved significant improvements and higher flexibility compared with existing methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0063] Figure 1 An overall flow chart of a method for constructing a dynamic evolution of personal risk in power operations by analyzing key factors provided in one embodiment of the present invention.

[0064] Figure 2 A diagram of the construction process of a work hazard analysis knowledge graph for constructing a dynamic evolution method of personal risk in power operations through key factor analysis provided by an embodiment of the present invention.

[0065] Figure 3 An LDA probability graph model diagram of a method for constructing a dynamic evolution of personal risk in power operations through key factor analysis provided by an embodiment of the present invention.

[0066] Figure 4 An embodiment of the present invention provides a knowledge graph and an expert model for constructing a dynamic evolution method of personal risk in power operations through key factor analysis to perform an accident chain replay flow chart.

[0067] Figure 5 An expression diagram of a spatiotemporal knowledge graph of a method for constructing a dynamic evolution of personal risks in power operations through key factor analysis provided in an embodiment of the present invention.

[0068] Figure 6 A structural block diagram of an electronic device for constructing a dynamic evolution system of personal risks in power operations through key factor analysis provided by an embodiment of the present invention. Detailed implementation manners

[0069] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of them. Based on the embodiments in the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0070] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0071] Secondly, the so-called "one embodiment" or "embodiment" herein refers to specific features, structures, or characteristics that may be included in at least one implementation manner of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it an embodiment that is mutually exclusive of other embodiments alone or selectively.

[0072] The present invention is described in detail in conjunction with schematic diagrams. When detailing the embodiments of the present invention, for ease of explanation, the cross-sectional views showing the device structure will be enlarged locally not in accordance with the general scale, and the schematic diagrams are only examples and should not limit the protection scope of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0073] At the same time, in the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0074] Unless otherwise clearly defined and limited in the present invention, the terms "mounted, connected, and coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0075] Example 1, referring to Figures 1-5, which is the first embodiment of the present invention. This embodiment provides a method for constructing a dynamic evolution of personal risks in electric power operations through key factor analysis, including:

[0076] S1: Construct a knowledge graph for job hazard analysis and perform entity recognition.

[0077] As Figure 2 shown, furthermore, the process stage, the output stage, and the implementation stage;

[0078] The job includes historical accidents;

[0079] The process stage includes performing document analysis on the problems to be solved and the coverage of the job, including in-depth study of historical accident documents and operating procedures, and extracting key concepts and terms corresponding to the job hazard analysis process, ensuring that the basis for ontology construction is solid and accurate.

[0080] Refer to existing meta-ontologies, which provide the basis and framework for constructing ontologies in specific domains.

[0081] Define entities, attributes, and relationships in the job hazard analysis process, and further enrich and verify these basic elements through the knowledge acquisition link of expert interviews and literature reviews.

[0082] Compile a glossary to ensure the consistency and accuracy of all terms in the ontology.

[0083] The output stage includes classifying the output concepts, constructing rules to determine the relationships and constraints between concepts, and outputting the overall framework;

[0084] The implementation stage includes defining the classification system and role model for job hazard analysis, selecting or a graphical query language, then performing ontology implementation, creating a job hazard analysis ontology, and importing the ontology and data into the knowledge graph platform to verify the scientificity, accuracy of the model, and whether its output results conform to the actual scenario;

[0085] Through quantitative and qualitative evaluations of the verification result model effect (such as whether the themes and keywords are accurate), provide a basis for model optimization, verify that the model can support subsequent knowledge graph construction and risk analysis, and improve the efficiency and reliability of job safety analysis.

[0086] The ontology implementation includes creating a schema file and a data file for the ontology, and the data file contains hazard and control measures and personal risk information for electric power operations.

[0087] S2: Extract knowledge graph entities.

[0088] It should be noted that as Figure 3As shown, the LDA model adopted extracts key entities automatically from the case-related information through entity recognition technology since the keywords of each entity are not clearly defined in the accident cases. Based on this, it is preset that each case focuses on several core areas or topics of the power grid for in-depth elaboration, and this feature exactly fits the application scenario of the LDA (Latent Dirichlet Allocation) topic model, facilitating the efficient automatic extraction of keywords.

[0089] The first layer is the document layer, which is the basis of the entire model. Through relevant training models, the association between each document and its potential topic words is deeply explored.

[0090] The second layer is the topic layer, which presents the topic content in the document in the form of keywords; it can explore and reveal the proportion of the topic distribution hidden in the document to help understand the core issues of the document.

[0091] The third layer is the distribution layer of topic keywords. The weights of topic keywords are solved through a probability distribution model to extract keywords; it is based on the assumption of the linear topological structure of the potential topics in the article by the LDA model.

[0092] Furthermore, the TF-IDF method is used to extract keywords from the documents, that is, historical accidents. The extracted keywords represent historical accidents:

[0093]

[0094] Among them, tf ij represents the word frequency of word i in document j, n i,j is the number of the word in document j, and ∑ k n k,j represents the sum of all words in document j. idf is the inverse document frequency:

[0095]

[0096] Among them, |D| is the total number of all documents in the corpus, and |{j:t∈d j}| represents the number of documents containing word t, and d j represents the j-th document in the corpus, calculating the information of the word occurrence frequency in the document;

[0097] Normalize the elements of the judgment matrix column by column and calculate the weight b ij :

[0098]

[0099] Among them, a ij is the element of the pairwise comparison matrix, and the weight vector w i is estimated using the arithmetic mean of all column vectors.:

[0100]

[0101] When AW=λW, the maximum characteristic root λ max for:

[0102]

[0103] The consistency index CI is:

[0104]

[0105] Wherein, n represents the order of the judgment matrix, that is, the number of rows or columns of the matrix, AW represents the result of multiplying the judgment matrix by the weight vector W, and W represents the weight vector or eigenvector, specifically, the weight distribution of the judgment matrix.

[0106] S3: Perform accident chain replay using knowledge graph and expert model.

[0107] Furthermore, Figure 4 As shown, based on the constructed knowledge graph of job hazard analysis, all risk factors are extracted to form a time chain for accident replay simulation;

[0108] A scenario-based continuous operation risk assessment method is used to conduct scenario re-evaluation, analyze the root causes and other causes of accidents, and update and store the knowledge graph.

[0109] It should be noted that the single attribute similarity and the comprehensive layer similarity are calculated; among them, the single attribute similarity is determined by comparing the similarity of the target scenario and the source scenario in each attribute dimension, and is calculated using the normalized Euclidean distance formula, and the calculation rules are set for symbolic and numerical attributes respectively; the comprehensive layer similarity is achieved by weighted summation of the single attribute similarities, and the weights are allocated according to the importance of each attribute in the scenario assessment, and finally the overall similarity between the target scenario and the source scenario is obtained, which is used to update the risk assessment results and support the dynamic evolution process;

[0110] The single attribute similarity adopts the KNN algorithm, by calculating the similarity between the target case and the source case, the output is the top k cases with the highest similarity. The single attribute similarity Sim(s,t) between the target case and the source case is:

[0111] Sim(s,t)=1-D(s,t)

[0112]

[0113] Where D(s,t) is the normalized Euclidean distance between the target case and the source case, m is the retrieval attribute number, M is the total number of retrieval attributes, and w mIt is the attribute m in D m (s,t)'s weight, D m (s,t) is the Euclidean distance between the normalized target situation and the source situation in the m attribute dimensions;

[0114] The data types of the retrieved attributes include symbolic types and numerical types. The symbolic type is determined as:

[0115]

[0116] The numerical type is determined as:

[0117]

[0118] D m (s,t) = ∣P sm - P tm ∣

[0119] where P sk is the k attribute value of the source situation, P tm is the m attribute value of the target situation, d m (s,t) is the distance between the source situation and the target situation in the m attribute dimensions, Max m is the maximum value of the m attribute value in the case base, Min m is the minimum value of the m attribute value in the case base.

[0120] The similarity of the comprehensive layer is:

[0121]

[0122] where r i is the attribute weight, n is the number of all attributes compared in the target situation and the source situation, o is the variable index, Sim(s,t) is the similarity of a single attribute, and the global attribute similarity SIM(G) is obtained:

[0123] SIM(G) = Sim(C)c + Sim(C)g

[0124] where c represents the similarity of the target situation and g represents the similarity of the global context.

[0125] S4: Construct a risk dynamic evolution model based on the spatio-temporal knowledge graph.

[0126] It should be noted that, as Figure 5 shown, based on the constructed job hazard analysis knowledge graph, construct a spatio-temporal knowledge graph that is dynamic, stores spatio-temporal dynamic data, and queries entity relationship information at any moment or time period;

[0127] The atlas information is stored in a dynamic visualization form, the data is updated, and the accident is replayed and inferred. During the deduction process, when any missing information is found in the original data, the spatio-temporal knowledge graph completes the information to form a dynamic evolution model.

[0128] The information completion of the spatio-temporal knowledge graph is achieved by detecting the missing entities, attributes, and relationships in the knowledge graph, combining historical data, real-time data, and similar case patterns, using symbolic and numerical data processing, KNN algorithm, and graph embedding reasoning technology to complete the missing information, and verifying the accuracy of the information through expert review. The completion results are dynamically updated to the spatio-temporal knowledge graph. This process improves the integrity of the knowledge graph, supports accurate accident replay and inference, and provides comprehensive data support and optimized risk prediction capabilities for the risk dynamic evolution model.

[0129] Example 2, the second embodiment of the present invention, which is different from the previous embodiment in that:

[0130] If the described functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0131] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device.

[0132] More specific examples (non-exhaustive list) of computer-readable media include the following: electrical connections (electronic devices) having one or more wirings, portable computer disk cartridges (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0133] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0134] Example 3, referring to Figure 6 , is the third embodiment of the present invention. This embodiment provides a key factor analysis for constructing a dynamic evolution system of personal risks in electric power operations, including a job hazard analysis knowledge graph construction module 10, a knowledge graph entity extraction module 20, an accident chain replay module 30, and a risk dynamic evolution module 40;

[0135] The job hazard analysis knowledge graph construction module 10 analyzes the job environment and historical accident documents, extracts key concepts and terms, defines entities, attributes, and relationships, creates a glossary, constructs a classification system and a role model, realizes an ontology, and imports the ontology and data into a knowledge graph platform for verification;

[0136] The knowledge graph entity extraction module 20 uses the LDA model to perform hierarchical processing on the documents, including a document layer, a theme layer, and a theme keyword distribution layer, extracts keywords using the TF-IDF method, and determines key entities and associations;

[0137] The accident chain replay module 30 uses the constructed knowledge graph to extract relationships of risk factors, forms an accident replay simulation time chain, performs re-evaluation using a scenario-based evaluation method, analyzes the accident causes, and updates the knowledge graph;

[0138] The risk dynamic evolution module 40 constructs a spatio-temporal knowledge graph to store and query entity relationship information, dynamically visualizes the graph information, updates the data, performs accident replay and reasoning, complements the missing information, and forms a dynamic evolution model.

[0139] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not restrictive. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. Analysis of key factors to construct a dynamic evolution method for personal risks in power operations, characterized in that: include, Construct a knowledge graph for job hazard analysis and perform entity recognition; Extract knowledge graph entities; Conduct accident chain replay using knowledge graphs and expert models; Construct a risk dynamic evolution model based on spatiotemporal knowledge graph.

2. The dynamic evolution method for personal risks in power operations by analyzing key factors as described in claim 1, characterized in that: The construction of the operation hazard analysis knowledge graph includes a process stage, an output stage, and an implementation stage; Said operations include historical incidents; The process stages include document analysis of the problems and coverage that the operation needs to solve, including in-depth reading of historical accident documents and operating procedures, extracting key concepts and terms corresponding to the operation hazard analysis process, referring to existing meta-ontologies, defining entities, attributes and relationships in the operation hazard analysis process, and compiling a glossary through knowledge acquisition through expert interviews and literature reviews; The output stage includes classifying the output concepts, building rules to determine the relationships and constraints between the concepts, and outputting the overall framework; The implementation phase includes defining the classification system and role model of job hazard analysis, selecting or graphical query language, implementing the ontology, creating the job hazard analysis ontology, and importing the ontology and data into the knowledge graph platform for verification; The ontology implementation includes creating a schema file and a data file of the ontology, wherein the data file contains information on hazards and control measures and personal risks of power operations.

3. The dynamic evolution method for personal risks in power operations by analyzing key factors as claimed in claim 2, characterized in that: The extraction of knowledge graph entities includes adopting an LDA model, the first layer of which is a document layer, and training the model to deeply mine the document to determine the association with the key entity; The second layer is the subject layer, which displays the subject content in the document in the form of keywords; The third layer is the distribution layer of topic keywords, which uses the probability distribution model to solve the weight of topic keywords and extract keywords.

4. The method for constructing a dynamic evolution of personal risks in power operations by analyzing key factors as described in claim 3, characterized in that: The keyword extraction includes extracting keywords from the document, i.e., the historical accident, using the TF-IDF method, and the extracted keywords represent the historical accident: Among them, tf ij represents the term frequency of word i in document j, n i,j is the number of the word in document j, ∑ k n k,j represents the sum of all words in document j, and idf is the inverse document frequency: Among them, |D| is the total number of all documents in the corpus, and |{j: t ∈ d j}| represents the number of documents containing the word t, where d j is the j-th document in the corpus, calculating the information of the word occurrence frequency in the document; Normalize the elements of the judgment matrix column by column and calculate the weight b ij : where a ij is an element of the pairwise comparison matrix, and the weight vector w is estimated using the arithmetic mean of all column vectors i : When AW = λW, the maximum eigenvalue λ max is: The consistency index CI is: Wherein, n represents the order of the judgment matrix, that is, the number of rows or columns of the matrix, AW represents the result of multiplying the judgment matrix by the weight vector W, and W represents the weight vector or eigenvector, specifically, the weight distribution of the judgment matrix.

5. The method for constructing a dynamic evolution of personal risks in power operations by analyzing key factors according to claim 4, characterized in that: The accident chain replay includes extracting relationships among all risk factors based on the constructed operation hazard analysis knowledge graph to form an accident replay simulation time chain; A scenario-based continuous operation risk assessment method is used to conduct scenario re-evaluation, analyze the root causes and other causes of accidents, and update and store the knowledge graph.

6. The method for dynamically evolving the personal risk in electric power operation by analyzing key factors as claimed in claim 5, wherein: The re-evaluation includes calculating single attribute similarity and comprehensive layer similarity; The single attribute similarity adopts the KNN algorithm, by calculating the similarity between the target case and the source case, the output is the top k cases with the highest similarity. The single attribute similarity Sim(s,t) between the target case and the source case is: Sim(s,t)=1-D(s,t) Among them, D(s,t) is the normalized Euclidean distance between the target situation and the source situation, m is the retrieval attribute number, M is the total number of retrieval attributes, and w m is the weight of attribute m in D m (s,t), and D m (s,t) is the Euclidean distance between the normalized target situation and the source situation in the m attribute dimensions; The data types of the retrieved attributes include symbolic type and numerical type, and the symbolic type is determined as: The value type is determined as: D m (s,t) = |P sm - P tm | where P sk is the k-attribute value of the source case, and P tm is the m-attribute value of the target case, and d m (s, t) is the distance between the source case and the target case in m attribute dimensions, and Max m is the maximum value of the m-attribute values in the case base, and Min m is the minimum value of the m-attribute values in the case base. The comprehensive layer similarity is: where r i is the attribute weight, n is the number of all attributes being compared in the target and source cases, o is the variable index, Sim(s,t) is the similarity of a single attribute, and the global attribute similarity SIM(G) is obtained as follows: SIM(G)=Sim(C)c+Sim(C)g Among them, c represents the similarity of the target situation, and g represents the similarity of the global context.

7. The method for constructing a dynamic evolution of personal risks in power operations through critical factor analysis according to claim 6, characterized in that: The construction of the risk dynamic evolution model includes constructing a spatio-temporal knowledge graph based on the constructed job hazard analysis knowledge graph, which is dynamic, stores spatio-temporal dynamic data, and queries entity relationship information at any moment or time period. Store the graph information in a dynamic visualization form, update the data, and conduct accident replay and reasoning. During the deduction process, when any missing information is found in the original data, the spatio-temporal knowledge graph completes the information to form a dynamic evolution model. The information completion of the spatio-temporal knowledge graph is achieved by detecting the missing entities, attributes, and relationships in the knowledge graph, combining historical data, real-time data, and similar case patterns, using symbolic and numerical data processing, KNN algorithm, and graph embedding reasoning technology to complete the missing information, and verifying the accuracy of the information through expert review, and dynamically updating the completion result to the spatio-temporal knowledge graph.

8. A system for constructing a dynamic evolution method of personal risks in power operations by using the key factor analysis as described in any one of claims 1 to 7, characterized in that: It includes a job hazard analysis knowledge graph construction module, a knowledge graph entity extraction module, an accident chain replay module, and a risk dynamic evolution module. The job hazard analysis knowledge graph construction module analyzes the job environment and historical accident documents, extracts key concepts and terms, defines entities, attributes, and relationships, creates a glossary, constructs a classification system and a role model, realizes the ontology, and imports the ontology and data into the knowledge graph platform for verification. The knowledge graph entity extraction module uses the LDA model to perform hierarchical processing on the documents, including the document layer, the topic layer, and the topic keyword distribution layer, extracts keywords using the TF-IDF method, and determines the key entities and associations. The accident chain replay module uses the constructed knowledge graph to extract the relationships of risk factors, forms an accident replay simulation time chain, uses a scenario-based evaluation method for re-evaluation, analyzes the accident causes, and updates the knowledge graph. The risk dynamic evolution module constructs a spatio-temporal knowledge graph to store and query entity relationship information, visualizes the graph information dynamically, updates the data, conducts accident replay and reasoning, completes the missing information, and forms a dynamic evolution model.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the key factor analysis for constructing the dynamic evolution method of the personal risk in electric power operation according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the key factor analysis for constructing the dynamic evolution method of the personal risk in electric power operation according to any one of claims 1 to 7.

Images