A method and system for intelligent power plant operation and maintenance decision support based on knowledge graph

Abstract

The present invention discloses a method and system for intelligent power plant operation and maintenance decision support based on knowledge graph, which relates to the technical field of power plant operation and maintenance decision-making, including data collection and information extraction; editing knowledge graphs and performing intelligent diagnosis; performing adaptability analysis of models in power plant operation and maintenance, and updating and optimizing knowledge graphs. The intelligent power plant operation and maintenance decision support method based on knowledge graph provided by the present invention can realize dynamic fusion of new information and existing graphs by calculating the similarity between the entity relationship to be integrated and the existing knowledge graph, ensuring that only relationships with a similarity exceeding a threshold are integrated into the graph, maintaining the accuracy and relevance of the knowledge graph, and being able to automatically adapt to and update new information, so that the graph always maintains a high degree of accuracy and relevance. The present invention achieves better results in terms of accuracy and reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of power plant operation and maintenance decision-making, and specifically to an intelligent power plant operation and maintenance decision support method and system based on knowledge graph. Background Art

[0002] With the development of automation and intelligence of power systems, the operation and maintenance of power plant equipment has gradually relied on efficient fault detection, status monitoring and intelligent decision support technologies. In recent years, data-driven intelligent operation and maintenance technology for power plants has made significant progress, gradually shifting from traditional periodic maintenance and passive fault handling to active operation and maintenance mode with real-time data monitoring and predictive analysis as the core. As a tool for expressing complex data structures, knowledge graphs can represent the multidimensional information and historical fault records of power plant equipment in the form of entities and relationships, which is suitable for equipment operation and maintenance scenarios with rich data and complex structures. By combining knowledge graphs with machine learning models, dynamic monitoring and multidimensional analysis of the operating status of power plant equipment can be achieved, improving the accuracy and timeliness of equipment operation and maintenance. However, the application of knowledge graphs in intelligent operation and maintenance of power plants is still in the exploratory stage, especially in data integration, diagnostic reasoning and model adaptability optimization. There is still room for improvement.

[0003] Most of the existing power plant operation and maintenance systems are based on traditional numerical monitoring and rule reasoning. Although they can achieve a certain degree of fault detection and early warning, they have significant limitations in data diversification and information linkage. First, the existing technology is difficult to process structured data (such as sensor data) and unstructured data (such as maintenance reports, expert opinions) at the same time, resulting in a lack of multi-source data integration in the operation and maintenance decision-making process. In addition, in the fault diagnosis link, traditional operation and maintenance systems mostly rely on fixed rules or single thresholds, and these methods are often difficult to cope with the complex power plant equipment operating environment and various uncertain factors, prone to false alarms or missed alarms, and lack of dynamic adaptability. In terms of intelligent diagnosis of equipment health status, the existing technology usually lacks a fault correlation analysis mechanism based on multi-dimensional data linkage and historical knowledge feedback, resulting in the isolation of the diagnosis process and the inability to conduct a comprehensive correlation analysis of abnormal events. As for the adaptability of the model, the existing technology has insufficient adaptability in the migration application between different devices, lacks adaptive capabilities, and is difficult to ensure the accuracy of predictions in different equipment environments. In addition, the knowledge update in the existing technology is mostly static manual addition, lacks an automated learning feedback mechanism, and cannot expand and optimize the knowledge graph in real time. Therefore, in order to address the issues of data integration, intelligent diagnosis, model adaptability and knowledge updating in power plant operation and maintenance, there is an urgent need for an intelligent operation and maintenance decision support method based on knowledge graphs to provide more accurate and dynamic equipment health monitoring and fault warning support. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: the existing intelligent operation and maintenance methods of power plants have difficulties in integrating data diversity, insufficient correlation of intelligent diagnosis, limited model adaptability, and problems in how to achieve real-time knowledge updating and feedback.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for intelligent power plant operation and maintenance decision support based on knowledge graph, comprising data collection and information extraction; editing the knowledge graph and performing intelligent diagnosis; analyzing the adaptability of the model in power plant operation and maintenance, and updating and optimizing the knowledge graph.

[0007] As a preferred solution of the knowledge graph-based intelligent power plant operation and maintenance decision support method described in the present invention, the data collection and integration into the knowledge graph includes the simultaneous collection of two types of data including structured data and unstructured data; structured data includes tables and charts of sensor data and equipment parameter database; unstructured data includes text records, maintenance reports and expert opinions, etc. After acquisition, information extraction is performed through natural language processing and semantic analysis; the data is integrated into the knowledge graph.

[0008] As a preferred solution of the intelligent power plant operation and maintenance decision support method based on knowledge graph described in the present invention, the information extraction includes identifying the equipment and event types in the data to construct entity relationships, constructing feature f t (d i ), expressed as:

[0009] f t (d i )={(e p ,r q ,e r )|e p ,e r ∈Entity in the text, r q ∈Relation}

[0010] Among them, d i represents the i-th data source, e p and e r are the start and target entities extracted from the text, r q Represents the relationship between an entity and an event type.

[0011] As a preferred solution of the intelligent power plant operation and maintenance decision support method based on knowledge graph described in the present invention, the editing of knowledge graph includes calculating the similarity between the entity relationship to be integrated and the existing entities and relationships, which is expressed as:

[0012]

[0013] Among them, R(eold ) is the original knowledge graph φ is the cosine similarity function r k is the entity relationship to be integrated, w(r) represents the adaptive dynamic weight of the relationship, where γ is the adjustment factor, r represents the relationship between the entity and the event type, and θ is the balance coefficient;

[0014] After the similarity calculation is completed, if Sim′(r k ,R(e old )) exceeds the preset similarity threshold δ, the relationship is considered to be close to the existing relationship in the knowledge graph, and the new relationship r k Integrate with related entities and relationships in the graph to k The corresponding entity node e integrated into the knowledge graph old , the new relationship r k Add to the knowledge graph in the form of new edges;

[0015] If the calculated similarity Sim′(r k ,R(e old )) is less than the set threshold δ, it indicates that the new relationship r k The semantic correlation between r and the existing knowledge graph is low. k is a new relationship different from any existing relationship, then r k The relevant entity information is added to the knowledge graph as a new node and connected with other appropriate nodes with low weight.

[0016] As a preferred solution of the knowledge graph-based intelligent power plant operation and maintenance decision support method described in the present invention, the intelligent diagnosis includes applying historical fault data for fault diagnosis, using the knowledge graph as a knowledge base for fault detection, matching the fault data with the data of the knowledge graph, and performing time correlation analysis based on abnormal events and operating condition correlation detection respectively.

[0017] As a preferred solution of the knowledge graph-based intelligent power plant operation and maintenance decision support method described in the present invention, the adaptability analysis of the model in power plant operation and maintenance includes testing the actual performance of the model on the target device, collecting the prediction accuracy as the performance indicator on the device, and after obtaining the performance indicator of the target device, comparing it with the benchmark performance on the original training device, calculating the performance difference of the model between different devices. If the adaptability index is higher than 0.85, it means that the model adaptability meets the requirements. If it is lower than the standard value, it means that the knowledge graph needs to be updated.

[0018] When the adaptability index shows that the knowledge graph adaptability meets the requirements, it is officially released to the target device for real-time monitoring and health status prediction.

[0019] As a preferred solution of the knowledge graph-based intelligent power plant operation and maintenance decision support method described in the present invention, wherein: the updating and optimization of the knowledge graph includes that after each fault warning or health diagnosis, the system automatically adds new fault causes, maintenance plans and health status information items to the knowledge graph, and evaluates the knowledge density of the equipment entities in the graph, that is, the ratio of the knowledge items contained in each equipment class to the total number of equipment. When the knowledge density does not meet the preset value, historical data is input for matching and new knowledge items are added.

[0020] Another object of the present invention is to provide an intelligent power plant operation and maintenance decision support system based on knowledge graph, which can solve the problems of lack of multi-source data integration capability and insufficient diagnostic correlation in current power plant intelligent operation and maintenance technology by integrating structured and unstructured data into the knowledge graph and applying it to intelligent diagnosis.

[0021] As a preferred solution of the knowledge graph-based intelligent power plant operation and maintenance decision support system described in the present invention, it includes: a data acquisition module, an editing module, and an update optimization module; the data acquisition module is used to collect data and extract information; the editing module is used to edit the knowledge graph and perform intelligent diagnosis; the update optimization module is used to analyze the adaptability of the model in power plant operation and maintenance, and update and optimize the knowledge graph.

[0022] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of a knowledge graph-based intelligent power plant operation and maintenance decision support method.

[0023] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a knowledge graph-based intelligent power plant operation and maintenance decision support method.

[0024] Beneficial effects of the present invention: The knowledge graph-based intelligent power plant operation and maintenance decision support method provided by the present invention can realize the dynamic fusion of new information and existing graphs by calculating the similarity between the entity relationship to be integrated and the existing knowledge graph, ensuring that only the relationship with a similarity exceeding a threshold is integrated into the graph, maintaining the accuracy and relevance of the knowledge graph, and being able to automatically adapt and update new information, so that the graph always maintains a high degree of accuracy and relevance. By testing the performance of the model on the target device and the prediction accuracy, the system can evaluate the cross-device adaptability of the model, avoiding performance degradation due to environmental differences, and improving the success rate of knowledge graph deployment and operation and maintenance effects. The present invention achieves better results in terms of accuracy and reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0026] Figure 1 An overall flow chart of a knowledge graph-based intelligent power plant operation and maintenance decision support method provided for the first embodiment of the present invention.

[0027] Figure 2 A fault library interface diagram of a knowledge graph-based intelligent power plant operation and maintenance decision support method provided in the first embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0029] Example 1, reference Figure 1-Figure 2 , which is an embodiment of the present invention, provides a method for intelligent power plant operation and maintenance decision support based on knowledge graph, including:

[0030] S1: perform data collection and information extraction;

[0031] Furthermore, the data collection and integration into the knowledge graph includes simultaneously collecting two types of data including structured data and unstructured data;

[0032] Structured data includes tables and charts of sensor data and equipment parameter databases;

[0033] Unstructured data includes text records, maintenance reports, and expert opinions. After acquisition, information is extracted through natural language processing and semantic analysis.

[0034] Integrate data into knowledge graphs.

[0035] It should be noted that the information extraction includes identifying the device and event type in the data to construct entity relationships and construct feature f t (d i ), expressed as:

[0036] f t (d i )={(e p,r q ,e r )|e p ,e r ∈Entity in the text, r q ∈Relation}

[0037] Among them, d i represents the i-th data source, e p and e r are the start and target entities extracted from the text, r q Represents the relationship between an entity and an event type.

[0038] S2: Edit the knowledge graph and perform intelligent diagnosis.

[0039] Furthermore, the editing of the knowledge graph includes calculating the similarity between the entity relationship to be integrated and the existing entities and relationships, which is expressed as:

[0040]

[0041] Among them, R(e old ) is the original knowledge graph φ is the cosine similarity function r k is the entity relationship to be integrated, w(r) represents the adaptive dynamic weight of the relationship, where γ is the adjustment factor, r represents the relationship between the entity and the event type, and θ is the balance coefficient;

[0042] After the similarity calculation is completed, if Sim′(r k ,R(e old )) exceeds the preset similarity threshold δ, the relationship is considered to be close to the existing relationship in the knowledge graph, and the new relationship r k Integrate with related entities and relationships in the graph to k The corresponding entity node e integrated into the knowledge graph old , the new relationship r k Add to the knowledge graph in the form of new edges;

[0043] If the calculated similarity Sim′(r k ,R(e old )) is less than the set threshold δ, it indicates that the new relationship r k The semantic correlation between r and the existing knowledge graph is low. k is a new relationship different from any existing relationship, then r k The relevant entity information is added to the knowledge graph as a new node and connected with other appropriate nodes with low weight.

[0044] It should be noted that intelligent diagnosis includes applying historical fault data for fault diagnosis, using the knowledge graph as the knowledge base for fault detection, matching the fault data with the data of the knowledge graph, and performing time correlation analysis based on abnormal events and operating condition correlation detection respectively.

[0045] Based on the time correlation analysis of abnormal events, when temperature or vibration abnormalities occur, the knowledge graph system can query the historical data of the abnormal device, analyze the time distribution and processing methods of similar events, and determine whether the current abnormality may be related to other abnormalities. The analysis steps are as follows:

[0046] 1. Abnormal type identification: When the system detects that the temperature or vibration exceeds the standard, it searches the knowledge graph for similar event records that have occurred in the past, for example, historical events marked as "temperature is too high" or "abnormal vibration".

[0047] 2. Time series query: The system queries the event time series recorded in the knowledge graph based on the anomaly type and calculates the frequency and time interval of such anomalies. For example, if similar events are found to occur frequently during each season transition, it may point to anomalies caused by specific environmental or operating conditions.

[0048] 3. Processing method matching: For the current anomaly, the system searches for similar processing methods through the graph and analyzes their effectiveness. For example, if the previous vibration anomaly was solved by replacing the bearing, the system can recommend this solution first.

[0049] Operational condition association detection including knowledge graphs can help the detection system identify potential fault event associations under specific operating conditions (such as high temperature and high pressure), thereby providing early warning under similar conditions.

[0050] The system performs association analysis through the following steps:

[0051] 1. Associated query of operating conditions: When the system detects a specific operating condition (such as a temperature above 85°C), it automatically queries the knowledge graph for historical failure events under the same conditions. For example, if a cooling system failure occurs multiple times at high temperatures, the event will be marked as a "condition-sensitive failure."

[0052] 2. Priority assessment of potential faults: The system queries the most common fault types under the operating conditions and establishes a fault priority list. For example, if the seal failure rate of the pump station is high under high pressure conditions, the seal failure will be listed as a key detection item.

[0053] 3. Linkage triggering conditions: Once the system detects specific conditions that meet the historical fault association, such as the equipment is in a high temperature and high pressure environment, the corresponding linkage detection operation is triggered.

[0054] It should also be noted that the adaptability analysis of the model in power plant operation and maintenance includes testing the actual performance of the model on the target equipment, collecting the prediction accuracy as the performance indicator on the equipment, and after obtaining the performance indicator of the target equipment, comparing it with the benchmark performance on the original training equipment, calculating the performance difference of the model between different equipment. If the adaptability index is higher than 0.85, it means that the model adaptability meets the requirements. If it is lower than the standard value, it means that the knowledge graph needs to be updated.

[0055] When the adaptability index shows that the knowledge graph adaptability meets the requirements, it is officially released to the target device for real-time monitoring and health status prediction.

[0056] S3: Conduct adaptability analysis of the model in power plant operation and maintenance, and update and optimize the knowledge graph.

[0057] Furthermore, the updating and optimization of the knowledge graph includes that after each fault warning or health diagnosis, the system automatically adds new fault causes, maintenance plans and health status information items to the knowledge graph, evaluates the knowledge density of equipment entities in the graph, that is, the ratio of knowledge items contained in each equipment class to the total number of equipment, and when the knowledge density does not meet the preset value, inputs historical data for matching and adds new knowledge items.

[0058] Example 2, an embodiment of the present invention, provides an intelligent power plant operation and maintenance decision support method based on knowledge graph. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0059] First, in order to verify the effectiveness and innovation of the intelligent power plant operation and maintenance decision support method based on knowledge graph, the operation and maintenance system first collected data and constructed knowledge graphs from the operation data of a large power plant. The collected data is divided into structured and unstructured data: structured data mainly includes real-time operating parameters of equipment (such as temperature, pressure, vibration, etc.) and equipment parameter database information, and unstructured data includes equipment maintenance records, expert opinions and operation logs. To process unstructured data, the operation and maintenance system uses natural language processing (NLP) and semantic analysis technology to extract information to ensure the comprehensiveness and efficiency of data.

[0060] After data collection is completed, the operation and maintenance system integrates and processes structured data and unstructured data, mapping the equipment, fault events and other information in the data into entities and relationships in the knowledge graph. Through data extraction, the operation and maintenance system identifies and constructs multiple key entities, such as "boiler", "pump station", "cooling operation and maintenance system" and other equipment entities, and analyzes the frequency and impact of fault events, and then establishes the relationship between the entities. The construction of the knowledge graph enables the operation and maintenance system to obtain comprehensive references from various events, equipment status and historical data when analyzing equipment failures.

[0061] Based on the construction of the graph, the operation and maintenance system accesses historical fault data as a fault detection module in the knowledge graph for intelligent diagnosis. When an equipment abnormality occurs, the operation and maintenance system automatically matches similar fault information by querying historical fault events in the knowledge graph, thereby providing fault causes and diagnostic solutions based on historical events. In addition, the operation and maintenance system conducts time correlation analysis and operating condition correlation detection on abnormal events. For example, when the temperature of a pump station exceeds the normal range, it combines historical fault records under the same operating conditions to achieve early warning.

[0062] The operation and maintenance system then conducted an adaptability test of the knowledge graph, deployed the knowledge graph on the target device, and monitored its performance under actual working conditions. The operation and maintenance system collected the accuracy of the knowledge graph prediction and the response time of the fault diagnosis, calculated the adaptability index and compared it with the performance of the original training device. In the adaptability test, if the performance of the knowledge graph on the target device is better than the set threshold (adaptability index is 0.85), the knowledge graph is successfully adapted and can be officially used for operation and maintenance decision support of the target device. If the requirements are not met, the operation and maintenance system will supplement and update the data through the feedback mechanism of the knowledge graph to further optimize the knowledge graph.

[0063] After fault warning and health diagnosis, the operation and maintenance system automatically updates the knowledge graph and adds new information items about equipment failure causes, solutions, and health status. By introducing a knowledge density evaluation mechanism, the operation and maintenance system introduces historical data and maintenance records of other equipment to supplement when the proportion of equipment knowledge items is low, continuously improving the integrity and practicality of the graph.

[0064] Table 1 Experimental data comparison table

[0065]

[0066] As can be seen from Table 1, the system maintains a high consistency in the accuracy of fault prediction and the speed of diagnostic response under different operating parameters on various devices, including temperature, pressure, vibration, and number of fault events. For example, the historical warning accuracy of Boiler A and Pump Station 1 reached 90% and 85% respectively, showing the high efficiency of knowledge graphs in multi-source data integration and fault prediction. In addition, the diagnostic response time of the equipment was controlled within 2.5 seconds, indicating that the knowledge graph can quickly match historical events and provide reference diagnosis, effectively improving the efficiency of operation and maintenance response.

[0067] The knowledge density data further demonstrates the innovative advantages of the content of the present invention in data integration and knowledge accumulation. In the experimental data, the knowledge density of cooling systems A and B is relatively high (0.88 and 0.87), reflecting that the system can dynamically update and expand the content of the knowledge graph in the process of continuous historical event accumulation and fault diagnosis. This automated knowledge update mechanism enables the knowledge graph to not only have a comprehensive device information recording function, but also to introduce supplementary data in devices with low knowledge density, further enhancing the adaptability of the knowledge graph and the practicality of knowledge.

[0068] The experiment also showed that during the adaptability test, the feedback mechanism of the knowledge graph effectively supported the dynamic optimization and cross-device adaptability of the model. For example, on a device with an adaptability index lower than 0.85, by supplementing historical data and updating knowledge, the performance of Pump Station 2 in the adaptability test met the standard requirements, providing a reliable guarantee for the successful deployment of the model on the device.

[0069] In summary, it can be clearly seen from the experimental data that the knowledge graph-based intelligent power plant operation and maintenance decision support method of the present invention can effectively integrate different types of operation and maintenance data, and realize rapid fault diagnosis, response and dynamic adaptive optimization in actual operation. This method is not only far superior to traditional systems in terms of early warning accuracy and response speed, but also further improves the system's knowledge accumulation and decision support capabilities through the automatic update of the knowledge graph and the knowledge density evaluation mechanism, thereby achieving innovative breakthroughs and significant technological progress in the field of power plant operation and maintenance.

[0070] Example 3, an embodiment of the present invention, provides an intelligent power plant operation and maintenance decision support system based on knowledge graph, including a data acquisition module, an editing module, and an update optimization module.

[0071] Among them, the data acquisition module is used to collect data and extract information; the editing module is used to edit the knowledge graph and perform intelligent diagnosis; the update and optimization module is used to analyze the adaptability of the model in power plant operation and maintenance, and update and optimize the knowledge graph.

[0072] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0073] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0074] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0075] 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-mentioned 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, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0076] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for intelligent power plant operation and maintenance decision support based on knowledge graph, characterized in that: include: Conduct data collection and extract information; Edit knowledge graphs and conduct intelligent diagnosis; Conduct adaptability analysis of the model in power plant operation and maintenance, and update and optimize the knowledge graph.

2. The method for intelligent power plant operation and maintenance decision support based on knowledge graph according to claim 1, characterized in that: The data collection and integration into the knowledge graph includes the simultaneous collection of two types of data including structured data and unstructured data; Structured data includes tables and charts of sensor data and equipment parameter databases; Unstructured data includes text records, maintenance reports, and expert opinions. After acquisition, information is extracted through natural language processing and semantic analysis. Integrate data into knowledge graphs.

3. The method for supporting intelligent power plant operation and maintenance decision based on knowledge graph according to claim 2, characterized in that: The information extraction includes identifying the equipment and event types in the data, constructing entity relationships, and constructing feature f t (d i ), expressed as: f t (d i )={(e p ,r q ,e r )|e p ,e r ∈Entity in the text, r q ∈Relation} Among them, d i represents the i-th data source, e p and e r are the start and target entities extracted from the text, r q Represents the relationship between an entity and an event type.

4. The method for supporting intelligent power plant operation and maintenance decision based on knowledge graph according to claim 3, characterized in that: The editing of the knowledge graph includes calculating the similarity between the entity relationship to be integrated and the existing entities and relationships, which is expressed as: Among them, R(e old ) is the original knowledge graph φ is the cosine similarity function r k is the entity relationship to be integrated, w(r) represents the adaptive dynamic weight of the relationship, where γ is the adjustment factor, r represents the relationship between the entity and the event type, and θ is the balance coefficient; After the similarity calculation is completed, if Sim′(r k ,R(eold)) exceeds the preset similarity threshold δ, the relationship is considered to be close to the existing relationship in the knowledge graph, and the new relationship r k Integrate with related entities and relationships in the graph to k The corresponding entity node e integrated into the knowledge graph old , the new relationship r k Add to the knowledge graph in the form of new edges; If the calculated similarity Sim′(r k ,R(e old )) is less than the set threshold δ, it indicates that the new relationship r k The semantic correlation between r and the existing knowledge graph is low. k is a new relationship different from any existing relationship, then r k The relevant entity information is added to the knowledge graph as a new node and connected with other appropriate nodes with low weight.

5. The method for supporting intelligent power plant operation and maintenance decision based on knowledge graph according to claim 4, characterized in that: The intelligent diagnosis includes applying historical fault data to perform fault diagnosis, using the knowledge graph as a knowledge base for fault detection, matching the fault data with the data of the knowledge graph, and performing time correlation analysis and operating condition correlation detection based on abnormal events respectively.

6. The method for supporting intelligent power plant operation and maintenance decision based on knowledge graph according to claim 5, characterized in that: The adaptability analysis of the model in power plant operation and maintenance includes testing the actual performance of the model on the target equipment, collecting the prediction accuracy as the performance indicator on the equipment, and after obtaining the performance indicator of the target equipment, comparing it with the benchmark performance on the original training equipment, calculating the performance difference of the model between different equipment. If the adaptability index is higher than 0.85, it means that the model adaptability meets the requirements. If it is lower than the standard value, it means that the knowledge graph needs to be updated. When the adaptability index shows that the knowledge graph adaptability meets the requirements, it is officially released to the target device for real-time monitoring and health status prediction.

7. The method for supporting intelligent power plant operation and maintenance decision based on knowledge graph according to claim 6, characterized in that: The updating and optimization of the knowledge graph includes that after each fault warning or health diagnosis, the system automatically adds new fault causes, maintenance plans and health status information items to the knowledge graph, evaluates the knowledge density of equipment entities in the graph, that is, the ratio of knowledge items contained in each equipment class to the total number of equipment, and when the knowledge density does not meet the preset value, inputs historical data for matching and adds new knowledge items.

8. A system using the knowledge graph-based intelligent power plant operation and maintenance decision support method as claimed in any one of claims 1 to 7, characterized in that: Including data collection module, editing module, update and optimization module; The data acquisition module is used to collect data and extract information; The editing module is used to edit the knowledge graph and perform intelligent diagnosis; The update optimization module is used to analyze the adaptability of the model in power plant operation and maintenance, and to update and optimize the knowledge graph.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the knowledge graph-based intelligent power plant operation and maintenance decision support method described in 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 knowledge graph-based intelligent power plant operation and maintenance decision support method described in any one of claims 1 to 7 are implemented.

Images