Hidden knowledge mining method for power generation full cycle and related equipment

Abstract

The invention relates to the technical field of power production intellectualization and artificial intelligence, in particular to a power generation full-cycle-oriented implicit knowledge mining method and related equipment, and the method comprises the steps: obtaining and analyzing power generation full-cycle multi-source data, and carrying out the classification modeling of the data as a knowledge graph core node according to an equipment identity label, a process bit number and a business label; extracting semantic elements of core nodes by using a large language model, and performing structured processing on a recessive causal logical relationship in a text to obtain a semantic triple consisting of faults, reasons and measures; and taking the extracted semantic elements as entities or attributes, associating the entities or attributes with the semantic triples, and fusing the entities or attributes into the knowledge graph. Based on the knowledge graph, causal reasoning of forward diagnosis or reverse traceability can be carried out according to real-time monitoring data or given fault phenomena, and effective support is provided for fault diagnosis and traceability of the power generation system.

Description

Technical Field

[0001] This invention relates to the fields of intelligent power production and artificial intelligence technology, specifically to a method and related equipment for mining implicit knowledge throughout the entire power generation cycle. Background Technology

[0002] In the field of intelligent operation and maintenance (O&M) of power generation enterprises, the currently commonly used technologies mainly fall into three categories: first, systems based on traditional rule bases, which rely on experts to manually write rules to achieve equipment O&M management; second, fault identification models based on deep learning, which train models to identify specific faults to assist in diagnosis; and third, experience-based documentation methods, which use unstructured documents such as maintenance reports, operation logs, and accident analysis reports to carry expert O&M experience. These three types of technologies correspond to the mainstream methods of rule-driven, data-driven, and document-driven approaches, respectively, forming the existing technological foundation for power generation equipment O&M management and providing important support for subsequent technological evolution.

[0003] However, the aforementioned existing technologies have significant inherent drawbacks: systems based on traditional rule bases struggle to cope with complex and ever-changing fault scenarios due to limited rule coverage, and the high cost of updating and maintaining rule bases makes them unsuitable for the rapid demands of equipment iteration and knowledge accumulation; while deep learning-based fault identification models can identify specific faults, their decision-making process lacks interpretability, making it difficult for maintenance personnel to judge the correctness of the results, and model training heavily relies on large amounts of high-quality labeled data, creating a data acquisition bottleneck in the low-frequency power industry; experience-based document-based methods lack effective automated extraction methods, failing to extract structured causal logic from unstructured documents, resulting in a large amount of knowledge that cannot be systematically reused and reasoned. More importantly, existing technologies as a whole cannot automatically and accurately extract implicit knowledge from multi-source heterogeneous data and transform it into computable and reasonable structured knowledge, nor do they offer comprehensive solutions serving the entire lifecycle of power generation equipment operation and maintenance. These shortcomings directly restrict the improvement of the intelligent operation and maintenance level of power generation enterprises. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method and related equipment for mining implicit knowledge throughout the entire power generation cycle, which addresses the shortcomings of the prior art. This method solves the technical problem that the existing technology as a whole cannot automatically and accurately extract implicit knowledge from multi-source heterogeneous data and transform it into computable and reasonable structured knowledge.

[0005] The objective of this invention is achieved through the following technical solutions: In a first aspect, the present invention provides a method for mining tacit knowledge throughout the entire power generation cycle, comprising: Acquire multi-source data throughout the entire power generation cycle, parse the multi-source data, and classify and model the multi-source data as core nodes in a knowledge graph based on equipment identification, process tag number, and business tag. By using a large language model to extract semantic elements from core nodes, the implicit causal logical relationships existing in the text of multi-source data are structured to obtain semantic triples; the semantic triples refer to fault, cause, and measure. The extracted semantic elements are treated as entities or attributes, associated with semantic triples, and then integrated into the knowledge graph. Based on the knowledge graph, causal reasoning for positive diagnosis or reverse tracing is performed according to real-time monitoring data or given fault phenomena.

[0006] As a further improvement of the present invention, the multi-source data includes structured data, unstructured data, and visual evidence data; The structured data includes monitoring data, power generation equipment, and operation logs used to describe the real-time operating conditions of the power generation equipment. The unstructured data includes technical documents, fault reports, expert experience texts, and operation and maintenance manuals; The visual evidence data includes photos of the power generation equipment, infrared thermal images, design drawings, and video data of the power generation equipment.

[0007] As a further improvement of the present invention, the core node includes parameter nodes, knowledge nodes, and multimodal external entity nodes; The parameter nodes are used to carry structured data; The knowledge nodes are used to carry unstructured data containing texts of expert experience; The multimodal external entity nodes are used to carry visual evidence data.

[0008] As a further improvement of the present invention, the extracted semantic elements are treated as entities or attributes and associated with semantic triples, and then integrated into the knowledge graph. This includes: establishing association relationships between the entities and attributes in the semantic triples and the parameter nodes and multimodal external entity nodes in the core nodes, respectively, so that real-time working condition data, expert tacit knowledge, and visual evidence data are interconnected to form a knowledge graph that integrates multi-dimensional information.

[0009] As a further improvement of the present invention, the positive diagnosis includes: When abnormalities occur in the real-time monitored parameter node data, the knowledge nodes associated with the abnormal parameter node are retrieved in the knowledge graph, the possible set of fault causes is inferred, and corresponding handling measures are recommended.

[0010] As a further improvement of the present invention, the reverse tracing includes: when a fault phenomenon is given, tracing back all possible cause chains that may lead to the fault phenomenon in the knowledge graph, and combining real-time parameters with historical cases to perform confidence assessment in order to locate the root cause of the fault.

[0011] As a further improvement of the present invention, the causal reasoning for forward diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena also includes: when a new maintenance report or handling result is generated, the new maintenance report or handling result is used as feedback data, and the causal association weights in the knowledge graph are strengthened or corrected based on the feedback data to realize the iterative update of the knowledge graph.

[0012] Secondly, the present invention provides a system for mining implicit knowledge throughout the entire power generation cycle, characterized in that it includes: The data classification and modeling module acquires multi-source data throughout the power generation cycle, parses the multi-source data, and classifies and models the multi-source data as core nodes in a knowledge graph based on equipment identification, process tag number, and business tag. The implicit knowledge extraction module uses a large language model to extract semantic elements from core nodes, and structures the implicit causal logical relationships existing in the text of multi-source data to obtain semantic triples; the semantic triples refer to fault, cause, and solution. The knowledge graph construction and reasoning module takes the extracted semantic elements as entities or attributes, associates them with semantic triples, and then integrates them into the knowledge graph. Based on the knowledge graph, it performs causal reasoning for positive diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena.

[0013] Thirdly, the present invention provides a computer-readable storage medium for storing one or more programs, the one or more programs including instructions that, when executed by a computing device, cause the computing device to perform the above-described method for mining implicit knowledge throughout the entire power generation cycle.

[0014] Fourthly, the present invention provides a computing device, comprising: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, and the one or more programs include steps for performing the above-described method for mining tacit knowledge throughout the power generation lifecycle.

[0015] The beneficial effects of this invention are as follows: This invention provides a method for mining implicit knowledge throughout the entire power generation cycle. By acquiring multi-source data from the entire power generation cycle and classifying and modeling the multi-source data into core nodes in a knowledge graph based on equipment identification, process tag number, and business tags, the efficiency of data structuring processing is improved. A large language model is used to extract semantic elements from the core nodes and structure the implicit causal logical relationships in the multi-source data text into semantic triples of fault, cause, and measure. This overcomes the technical defect in existing technologies where implicit logic is difficult to make explicit, and improves the accuracy and completeness of causal relationship extraction. The extracted semantic elements are used as entities or attributes. After associating sex and semantic triples, they are integrated into the knowledge graph to construct an associated knowledge system containing equipment status, fault causality, and countermeasures, enhancing the semantic richness and reasoning ability of the knowledge graph. Based on the knowledge graph, causal reasoning for forward diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena is performed, realizing intelligent and precise fault diagnosis and tracing, and improving the efficiency and accuracy of operation and maintenance decisions. The synergistic effect of the above technical features, through the deep integration of structured modeling and causal reasoning of the knowledge graph, realizes the optimization of the entire process from data parsing to intelligent diagnosis, significantly improving the intelligence level of power generation equipment operation and maintenance and the efficiency of fault handling. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the implicit knowledge mining method for the entire power generation cycle in an embodiment of the present invention; Figure 2 This is an internal structural diagram of a computer device in an embodiment of the present invention. Detailed Implementation

[0018] To make the objectives and technical solutions of this invention clearer and easier to understand, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. The specific embodiments described herein are for illustrative purposes only and are not intended to limit the invention.

[0019] The technical solution of the present invention will be clearly and completely described below with reference to the accompanying drawings and specific embodiments. The described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] Example 1 Existing technologies cannot automatically and accurately extract implicit knowledge from multi-source heterogeneous data, nor can they transform it into computable and reasonable structured knowledge. They also lack comprehensive solutions that serve the entire life cycle operation and maintenance of power generation equipment. Therefore, this embodiment provides a method for mining implicit knowledge throughout the entire power generation life cycle.

[0021] The implicit knowledge mining method for the entire power generation cycle includes the following steps: acquiring multi-source data for the entire power generation cycle; parsing the multi-source data; classifying and modeling the multi-source data as core nodes in a knowledge graph based on equipment identification, process tag number, and business tags; extracting semantic elements from the core nodes using a large language model; structuring the implicit causal logical relationships existing in the text of the multi-source data to obtain semantic triples; the semantic triples refer to faults, causes, and measures; associating the extracted semantic elements as entities or attributes with the semantic triples and integrating them into the knowledge graph; and based on the knowledge graph, performing causal reasoning for forward diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena.

[0022] The working principle of this embodiment is as follows: by acquiring multi-source data throughout the entire power generation cycle, and after parsing and classifying it, core nodes of a knowledge graph are formed. A large language model is used to extract semantic elements from the core nodes and structure them into semantic triples, thus making implicit causal logic explicit. Semantic elements and semantic triples are then associated and integrated into the knowledge graph, constructing a multi-dimensional information fusion knowledge network. Finally, based on this knowledge graph, causal reasoning is achieved through forward diagnosis or reverse tracing algorithms according to real-time monitoring data or given fault phenomena, providing support for fault handling. This achieves the integrated utilization of multi-source data throughout the entire power generation cycle, breaking down barriers between data sources; transforming implicit causal logic into structured semantic triples, enabling expert experience to be understood and reused by machines; and providing a solid foundation for reasoning through the constructed knowledge graph. Forward diagnosis and reverse tracing meet the fault handling needs in different scenarios, improving the efficiency and accuracy of fault diagnosis and reducing the workload of maintenance personnel.

[0023] This embodiment uses multi-source data to cover the entire lifecycle of power generation equipment, from planning and design, installation and commissioning, operation and maintenance to offline service. Data is acquired through various methods such as data interfaces, file import, and sensor acquisition.

[0024] This embodiment uses a data parsing engine to process the acquired multi-source data. Structured data is directly parsed into key-value pair format, and unstructured text is cleaned using text cleaning techniques in natural language processing to remove redundant characters and format symbols. Visual data is standardized and uniformly converted into JPG image format (video data is extracted and keyframes are converted into JPG format). During the parsing process, abnormal data (such as sensor readings that exceed reasonable ranges and unrecognizable garbled text) is removed using a data verification algorithm.

[0025] In this embodiment, the large language model adopted is the GPT-4 large model, which has powerful semantic understanding and entity recognition capabilities and can accurately extract the core semantic elements in the text. In practical applications, this model can also be selected from other models such as Wenxin Yiyan and Tongyi Qianwen, but this embodiment does not limit it.

[0026] Example 2 This embodiment further illustrates the implicit knowledge mining method for the entire power generation lifecycle described in Embodiment 1.

[0027] Multi-source data includes structured data, unstructured data, and visual evidence data. Structured data includes monitoring data describing the real-time operating conditions of power generation equipment, as well as power generation equipment and operation logs; unstructured data includes technical documents, fault reports, expert experience texts, and operation and maintenance manuals; visual evidence data includes photos of power generation equipment, infrared thermal images, design drawings, and video data of power generation equipment.

[0028] Furthermore, the core nodes include parameter nodes, knowledge nodes, and multimodal external entity nodes; parameter nodes are used to carry structured data; knowledge nodes are used to carry unstructured data containing expert experience text; and multimodal external entity nodes are used to carry visual evidence data.

[0029] The extracted semantic elements are treated as entities or attributes and associated with semantic triples, then integrated into the knowledge graph. This further includes establishing associations between the entities and attributes in the semantic triples and the parameter nodes and multimodal external entity nodes in the core nodes, respectively, so that real-time working condition data, expert tacit knowledge, and visual evidence data are interconnected to form a knowledge graph that integrates multi-dimensional information.

[0030] A graph fusion algorithm is used to integrate the aforementioned associated entities, attributes, nodes, and edges into a knowledge graph. During the fusion process, redundant data is removed (repeated attributes of the same entity), and conflicting data is processed (the confidence differences from different sources are weighted and averaged). Finally, a multi-dimensional fused knowledge graph is formed, which includes real-time operating data (parameter nodes), expert implicit knowledge (semantic triples, knowledge nodes), and visual evidence data (multimodal external entity nodes). The graph is stored in the Neo4j graph database, which supports efficient node query and association reasoning.

[0031] Positive diagnosis specifically includes: when abnormal data is detected in real-time monitored parameter nodes, retrieving knowledge nodes associated with the abnormal parameter nodes from the knowledge graph, inferring a set of possible fault causes, and recommending corresponding handling measures. A preset normal range threshold for parameter nodes (this threshold can be configured and adjusted based on equipment manuals, industry standards, and historical operating data) is used. By comparing the current value of a parameter node with the normal range threshold in real time, if the current value exceeds the threshold range, it is determined to be an abnormal parameter node. This embodiment uses a neighborhood query algorithm based on a graph database for retrieving associated knowledge nodes.

[0032] Reverse tracing specifically includes: when given a fault phenomenon, tracing back all possible causal chains that could lead to the fault phenomenon in the knowledge graph, and combining real-time parameters with historical cases to evaluate confidence and locate the root cause of the fault. In this embodiment, a depth-first search algorithm is used for reverse tracing of causal chains. The confidence of each causal chain is evaluated by combining real-time parameters and historical cases. All traced causal chains are sorted in descending order of overall confidence. The top-level cause of the causal chain with the highest overall confidence exceeding 0.8 is determined as the root cause of the fault. If multiple causal chains with an overall confidence exceeding 0.8 exist, multiple possible root causes are output simultaneously, with their respective confidence levels and supporting evidence (real-time parameters, historical cases, visual evidence) labeled. For example, "insufficient lubricating oil" is located as the root cause, supported by evidence including lubricating oil level parameters below the normal range, 26 out of 30 similar causal chains in history being verified as the true root cause, and infrared thermal imaging showing abnormal temperature at the bearing location.

[0033] Furthermore, the causal reasoning for positive diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena also includes: when a new maintenance report or handling result is generated, the new maintenance report or handling result is used as feedback data, and the causal association weights in the knowledge graph are strengthened or corrected based on the feedback data to achieve iterative updates of the knowledge graph.

[0034] Example 3 This embodiment provides a latent knowledge mining system for the entire power generation cycle, including: a data classification and modeling module, which acquires multi-source data for the entire power generation cycle, parses the multi-source data, and classifies and models the multi-source data into core nodes in a knowledge graph based on equipment identification, process tag number, and business tag; a latent knowledge extraction module, which uses a large language model to extract semantic elements from the core nodes, structures the implicit causal logical relationships existing in the text of the multi-source data, and obtains semantic triples; the semantic triples refer to faults, causes, and measures; a graph construction and reasoning module, which associates the extracted semantic elements as entities or attributes with the semantic triples and integrates them into the knowledge graph; and based on the knowledge graph, performs causal reasoning for forward diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena.

[0035] The three modules interact with each other through a message queue. The data classification and modeling module sends core node data to the message queue. The implicit knowledge extraction module retrieves data from the message queue, processes it, and then sends semantic triples to the message queue. The graph construction and reasoning module retrieves data from the message queue for graph fusion and reasoning. The running status and processing results of each module are recorded in real time through a log system, supporting system monitoring and fault diagnosis.

[0036] Example 4 In another embodiment of the present invention, a computer-readable storage medium is provided as a storage component within a terminal device, the function of which is to store programs and data. It should be noted that the computer-readable storage medium here encompasses not only the built-in storage components of the terminal device but also extended storage components supported by the device. Essentially, it is a tangible medium capable of containing or storing programs that can be invoked by or in conjunction with an instruction execution system, device, or apparatus. This storage medium provides storage areas for the terminal's operating system and stores one or more instructions suitable for processor loading and execution, which can constitute one or more computer programs containing program code.

[0037] Specifically, examples of computer-readable storage media (a non-exclusive list) include: electrical connections with one or more wires, portable disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable optical disc read-only memory, optical storage devices, magnetic storage devices, or any reasonable combination of the above types.

[0038] The storage medium may also include data signals propagated as part of a baseband portion or a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any reasonable combination of both. Furthermore, computer-readable storage medium may also refer to other readable media besides conventional readable storage media, capable of sending, propagating, or transmitting programs for use or operation by an instruction execution system, apparatus, or device. Program code on the storage medium can be transmitted via any suitable medium, including but not limited to wireless, wired, optical fiber, or any reasonable combination thereof.

[0039] The program code used to implement the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C. The execution modes of the program code include: running entirely on the user's computing device, running partially on the user's device as a standalone software package, running partially in a distributed manner on both the user's device and a remote computing device, or running entirely on a remote computing device or server. When a remote computing device is involved, the device can be connected to the user's computing device via any type of network such as a local area network (LAN) or a wide area network (WAN), or connected to an external computing device via the Internet through an Internet service provider.

[0040] The processor is capable of loading and executing one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the implicit knowledge mining method for the entire power generation cycle described in Example 1.

[0041] Example 5 Figure 2 This is a schematic diagram of a computer device provided according to an embodiment of the present invention.

[0042] Please see Figure 2 The terminal device is a computer device. In this embodiment, the computer device 60 includes a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61. When the processor 61 executes the computer program 63, it implements the implicit knowledge mining method for the entire power generation cycle described in this embodiment. To avoid repetition, these details are not elaborated here. Alternatively, when the processor 61 executes the computer program 63, it implements the functions of each model / unit in the computing system that constitutes the implicit knowledge mining method for the entire power generation cycle described in this embodiment. To avoid repetition, these details are not elaborated here.

[0043] Computer device 60 can be a desktop computer, laptop, handheld computer, cloud server, or other computing device. Computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art will understand that... Figure 2This is merely an example of computer device 60 and does not constitute a limitation on computer device 60. It may include more or fewer components than shown, or combine certain components, or different components. For example, computer device may also include input / output devices, network access devices, buses, etc.

[0044] The processor 61 may be a central processing unit (CPU), or other general-purpose processors, CPUs, graphics processing units (GPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, quantum computing-based data processing logic units, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0045] The memory 62 can be an internal storage unit of the computer device 60, such as a hard disk or RAM of the computer device 60. The memory 62 can also be an external storage device of the computer device 60, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc., equipped on the computer device 60.

[0046] Furthermore, the memory 62 may include both internal storage units and external storage devices of the computer device 60. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 can also be used to temporarily store data that has been output or will be output.

[0047] Any references to memory, databases, or other media used in the embodiments provided in this application may include at least one of non-volatile and volatile memory. Non-volatile memory may include read-only memory (Read-Only Memory). Memory includes ROM, magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0048] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

Claims

1. A method for mining tacit knowledge across the entire power generation lifecycle, characterized in that, include: Acquire multi-source data throughout the entire power generation cycle, parse the multi-source data, and classify and model the multi-source data as core nodes in a knowledge graph based on equipment identification, process tag number, and business tag. By using a large language model to extract semantic elements from core nodes, the implicit causal logical relationships existing in the text of multi-source data are structured to obtain semantic triples; the semantic triples include fault, cause, and measure. The extracted semantic elements are treated as entities or attributes, associated with semantic triples, and then integrated into the knowledge graph. Based on the knowledge graph, causal reasoning for positive diagnosis or reverse tracing is performed according to real-time monitoring data or given fault phenomena.

2. The method for mining implicit knowledge throughout the entire power generation lifecycle as described in claim 1, characterized in that, The multi-source data includes structured data, unstructured data, and visual evidence data; The structured data includes monitoring data, power generation equipment, and operation logs used to describe the real-time operating conditions of the power generation equipment. The unstructured data includes technical documents, fault reports, expert experience texts, and operation and maintenance manuals; The visual evidence data includes photos of the power generation equipment, infrared thermal images, design drawings, and video data of the power generation equipment.

3. The method for mining implicit knowledge throughout the entire power generation lifecycle according to claim 2, characterized in that, The core nodes include parameter nodes, knowledge nodes, and multimodal external entity nodes; The parameter nodes are used to carry structured data; The knowledge nodes are used to carry unstructured data containing texts of expert experience; The multimodal external entity nodes are used to carry visual evidence data.

4. The method for mining implicit knowledge throughout the entire power generation lifecycle according to claim 3, characterized in that, The extracted semantic elements are treated as entities or attributes and associated with semantic triples, then integrated into the knowledge graph. This includes establishing associations between the entities and attributes in the semantic triples and the parameter nodes and multimodal external entity nodes in the core nodes, respectively, so that real-time working condition data, expert tacit knowledge, and visual evidence data are interconnected to form a knowledge graph that integrates multi-dimensional information.

5. The method for mining implicit knowledge throughout the entire power generation lifecycle according to claim 1, characterized in that, The positive diagnosis includes: When abnormalities occur in the real-time monitored parameter node data, the knowledge nodes associated with the abnormal parameter node are retrieved in the knowledge graph, the possible set of fault causes is inferred, and corresponding handling measures are recommended.

6. The method for mining implicit knowledge throughout the entire power generation lifecycle according to claim 1, characterized in that, The reverse tracing includes: when a fault phenomenon is given, tracing back all possible cause chains in the knowledge graph, and combining real-time parameters with historical cases to conduct confidence assessment in order to locate the root cause of the fault.

7. The method for mining implicit knowledge throughout the entire power generation lifecycle according to any one of claims 1-6, characterized in that, The causal reasoning based on real-time monitoring data or given fault phenomena for positive diagnosis or reverse tracing also includes: when a new maintenance report or handling result is generated, the new maintenance report or handling result is used as feedback data, and the causal association weights in the knowledge graph are strengthened or corrected based on the feedback data to achieve iterative updates of the knowledge graph.

8. A system for mining tacit knowledge throughout the entire power generation lifecycle, characterized in that, include: The data classification and modeling module acquires multi-source data throughout the power generation cycle, parses the multi-source data, and classifies and models the multi-source data as core nodes in a knowledge graph based on equipment identification, process tag number, and business tag. The implicit knowledge extraction module uses a large language model to extract semantic elements from core nodes, and structures the implicit causal logical relationships existing in the text of multi-source data to obtain semantic triples; the semantic triples refer to fault, cause, and solution. The knowledge graph construction and reasoning module takes the extracted semantic elements as entities or attributes, associates them with semantic triples, and then integrates them into the knowledge graph. Based on the knowledge graph, it performs causal reasoning for positive diagnosis or reverse tracing based on real-time monitoring data or given fault phenomena.

9. A computer-readable storage medium for storing one or more programs, characterized in that, The one or more programs include instructions that, when executed by a computing device, cause the computing device to perform the implicit knowledge mining method for the entire power generation lifecycle as described in any one of claims 1 to 7.

10. A computing device, characterized in that, include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including steps for performing the implicit knowledge mining method for the entire power generation lifecycle as described in any one of claims 1 to 7.

Images