Thermal power plant fault analysis report generation method based on retrieval enhancement and related equipment

Abstract

The invention relates to the technical field of artificial intelligence and thermal power production, in particular to a thermal power plant fault analysis report generation method based on retrieval enhancement and related equipment, and the method adopts a distributed sensor to obtain time sequence operation data and equipment logs of thermal power plant equipment to construct multi-modal data. Multi-modal data are processed through a time sequence fault diagnosis model (including a retrieval enhancement generation module and a fault diagnosis module), the retrieval enhancement generation module retrieves in a thermal power plant equipment knowledge base and generates retrieval prompt data, and the fault diagnosis module performs diagnosis according to the retrieval prompt data. And finally, generating an equipment analysis report containing fault reasons, processing measures and prevention suggestions by adopting a predefined template according to a diagnosis result.

Description

Technical Field

[0001] The present invention relates to the technical fields of artificial intelligence and thermal power production, and particularly to a method for generating a fault analysis report for a thermal power plant based on retrieval enhancement and related equipment. Background Art

[0002] With the rapid development of the economy, the demand for energy is increasing day by day, and the inspection and maintenance work of power plants has become a key link to maintain the steady development of society. During the long-term operation of thermal power units, they will be affected by multiple factors such as equipment aging, environmental changes, and operation mode adjustments, and these factors may all cause faults.

[0003] In the production process of the power generation industry, it is particularly important to conduct professional analysis on faults and outage events, which is an indispensable part of avoiding risks and ensuring the stable operation of units in the future. However, it is not easy to write a high-quality expert report, which is time-consuming and complex. Professionals not only need to describe the event process in detail and put forward professional analysis suggestions, but also need to spend a lot of time consulting relevant official documents such as requirements or management measures to fill in the information.

[0004] In addition, the knowledge system of the power generation industry is extremely large, and experts are often only proficient in their own professional fields and it is difficult to meet the needs of all fault analyses in real time. Therefore, the existing manual methods have obvious deficiencies in making full use of the historical fault data and related knowledge of power plants. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method for generating a fault analysis report for a thermal power plant based on retrieval enhancement and related equipment in view of the above deficiencies in the prior art, so as to solve the technical problems of low efficiency in the process of generating a fault analysis report and deficiencies in the report content.

[0006] The object of the present invention is achieved by the following technical solutions: In a first aspect, the present invention provides a method for generating a fault analysis report for a thermal power plant based on retrieval enhancement, including: Obtaining the time-series operation data and equipment logs of the thermal power plant equipment according to distributed sensors, and constructing multimodal data according to the time-series operation data and equipment logs; Inputting the multimodal data into a time-series fault diagnosis model to obtain a fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the knowledge base of the thermal power plant equipment according to the multimodal data, and merge the retrieved files with the retrieval enhancement generation module to obtain retrieval hint data; the fault diagnosis module is used to diagnose according to the retrieval hint template generated by the retrieval enhancement generation module; According to the fault diagnosis results, a predefined report template is used to generate an equipment analysis report containing the cause of the fault, handling measures, and preventive suggestions.

[0007] As a further improvement of the present invention, according to the time-series operation data and equipment logs of the thermal power plant equipment obtained by distributed sensors, it specifically includes: Using the DCS system to obtain the time-series operation data of the thermal power plant; performing wavelet denoising processing on the time-series operation data; Using the Markov transfer field to convert the time-series signal into two-dimensional time-frequency image data; After fusing the two-dimensional time-frequency image data with the equipment logs of the thermal power plant, multi-modal data is obtained.

[0008] As a further improvement of the present invention, the knowledge base of thermal power plant equipment is a multi-source heterogeneous vector database; the vector database at least includes reports from thermal power plant domain experts, operation logs of thermal power equipment, thermal power equipment manuals, and historical fault data of thermal power equipment.

[0009] As a further improvement of the present invention, the retrieval enhancement generation module specifically includes: Identifying multi-modal data, converting the multi-modal data into vectors, querying the knowledge base of thermal power plant equipment according to the vector-form multi-modal data, and querying the most similar knowledge base segment according to the cosine similarity; Sorting the retrieved results according to the similarity score, and taking the knowledge base segment with the highest similarity as the first vector; Combining the first vector with the vector-form multi-modal data to obtain retrieval hint data.

[0010] As a further improvement of the present invention, the fault diagnosis module uses a large language model; the fault diagnosis module specifically includes: Performing semantic understanding according to the retrieval hint data, performing fault diagnosis according to the data after semantic understanding, and determining the fault phenomenon, occurrence event, and influence range of the thermal power equipment; Querying in the knowledge base of thermal power plant equipment according to the fault phenomenon, occurrence event, and influence range to obtain possible fault types, causes of faults, corresponding handling measures, and corresponding confidence levels; Sorting the diagnosis results according to the confidence level, and taking the fault type, cause of the fault, and corresponding handling measure with the highest confidence level as the fault diagnosis result according to the sorting result.

[0011] As a further improvement of the present invention, after obtaining the equipment analysis report, it also includes the steps of submitting the equipment analysis report for expert review and giving feedback according to the review results, specifically including: After the device analysis report is generated, an expert review mechanism is adopted to label the fault cases in the device analysis report, and the labeled fault cases are updated to the thermal power plant device knowledge base.

[0012] As a further improvement of the present invention, it further includes visual display of the device analysis report, and the visual interface includes: An expert review interface, which includes labeling error types for misdiagnosed cases in the expert review interface; A knowledge base setting interface, which is used to display in the form of a tree topology diagram, display the data annotation information in the thermal power plant device knowledge base, and perform information traceability according to the annotation information.

[0013] In a second aspect, the present invention provides a thermal power plant fault analysis report generation system based on retrieval enhancement, which is used to implement the above-mentioned thermal power plant fault analysis report generation method based on retrieval enhancement, including: A data acquisition module, which acquires the time-series operation data and device logs of thermal power plant devices according to distributed sensors, and constructs multimodal data according to the time-series operation data and device logs; A fault diagnosis module, which inputs the multimodal data into a time-series fault diagnosis model to obtain a fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the thermal power plant device knowledge base according to the multimodal data, and merge the retrieved files with the retrieval enhancement generation module to obtain a retrieval prompt template; the fault diagnosis module is used to diagnose according to the retrieval prompt template generated by the retrieval enhancement generation module; A report generation module, which generates a device analysis report including fault causes, handling measures and preventive suggestions according to the fault diagnosis result by using a predefined report template.

[0014] In a third aspect, the present invention provides a computing device, including: One or more processors, a memory and one or more programs, wherein 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 executing the above-mentioned thermal power plant fault analysis report generation method based on retrieval enhancement.

[0015] In a fourth aspect, the present invention provides a computer-readable storage medium storing one or more programs, and the one or more programs include instructions, which when executed by a computing device, cause the computing device to execute the above-mentioned thermal power plant fault analysis report generation method based on retrieval enhancement.

[0016] The beneficial effects of the present invention are as follows: By combining time-series operation data and device logs to construct multimodal data, the operating status of the device can be more comprehensively reflected. By retrieving in the knowledge base of thermal power plant equipment through the retrieval-enhanced generation module, historical cases, failure modes, and treatment methods related to the current device status can be obtained. After merging the retrieved information with the multimodal data, retrieval prompt data is generated, providing more comprehensive reference information for the fault diagnosis module and further improving the accuracy of diagnosis. According to the fault diagnosis results, an equipment analysis report containing the cause of the fault, treatment measures, and preventive suggestions is automatically generated. This not only reduces the time for manual report writing but also ensures the standardization and consistency of the report, improving the response speed of fault handling. Through automated diagnosis and report generation, the dependence on manual experience is reduced, and the labor cost is lowered. At the same time, accurate diagnosis and timely treatment measures can effectively reduce the device downtime, reducing the economic losses and risks brought by device failures.

[0017] Further, wavelet denoising is used to remove the noise in the signal while retaining the detailed information of the signal.

[0018] Further, by retrieving relevant information in the knowledge base of thermal power plant equipment, the retrieval-enhanced generation module can combine historical experience, expert knowledge, and the current device status to provide a more comprehensive basis for diagnosis. Through similarity calculation and sorting algorithms, the most relevant knowledge base fragments can be quickly retrieved, reducing the time for manual query and judgment. During each diagnosis process, the retrieved information and diagnosis results are recorded and updated in the knowledge base, forming a continuous learning and accumulation process.

[0019] Further, the expert review mechanism can effectively verify the accuracy and reliability of the diagnosis results by having experienced experts review the equipment analysis report. Experts can identify details that the model may overlook or misjudgment situations, thereby improving the quality of fault diagnosis. The annotation and feedback of experts on fault cases provide high-quality annotation data for the model, and these data can be used for further training and optimization of the fault diagnosis model. Through continuous learning and adjustment, the model can more accurately identify fault modes and improve the accuracy of diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0021] Figure 1It is a schematic flow diagram of a method for generating a fault analysis report of a thermal power plant based on retrieval enhancement provided by an embodiment of the present invention; Figure 2 It is a schematic flow diagram of a time-series fault diagnosis model provided by an embodiment of the present invention; Figure 3 It is a schematic structural diagram of an electronic device provided by an embodiment of the present invention. Specific embodiments

[0022] In order to make the purpose and technical solutions of the present invention clearer and easier to understand. The following further describes the present invention in detail with reference to the accompanying drawings and embodiments. The specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0023] The technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Among them, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments.

[0024] Embodiment 1 As Figure 1 、 Figure 2 shown, this embodiment provides a method for generating a fault analysis report of a thermal power plant based on retrieval enhancement. This method combines knowledge base retrieval and large language models to improve the automation and accuracy of fault analysis.

[0025] Obtain the time-series operation data and equipment logs of thermal power plant equipment according to distributed sensors, and construct multimodal data based on the time-series operation data and equipment logs.

[0026] Specifically, constructing multimodal data based on the time-series operation data and equipment logs specifically includes: Use the DCS system to obtain the time-series operation data of the thermal power plant; perform wavelet noise reduction processing on the time-series operation data; Use the Markov transfer field to convert the time-series signal into two-dimensional time-frequency image data; After fusing the two-dimensional time-frequency image data with the thermal power plant equipment logs, multimodal data is obtained.

[0027] Among them, the time-series operation data includes at least sensor data such as temperature, pressure, and flow. The equipment operation data includes data such as switch status and load rate, and the equipment logs include records of abnormal situations and events during equipment operation.

[0028] Fuse the two-dimensional time-frequency image data with the thermal power plant equipment logs to provide more comprehensive information. The specific steps are as follows: Data preprocessing: Preprocess the two-dimensional time-frequency image data and equipment logs, including data cleaning, format conversion, etc.

[0029] Align the two-dimensional time-frequency image data with the device log to ensure data consistency.

[0030] Extract key features from the device log, such as fault type, fault time, handling measures, etc.

[0031] Fuse the features of the two-dimensional time-frequency image data and the device log to form multimodal data. The following methods can be used for fusion: Feature-level fusion: Concatenate the features of different modalities to form a comprehensive feature vector; Decision-level fusion: Independently analyze the data of different modalities and then fuse the analysis results.

[0032] Input the multimodal data into the time-series fault diagnosis model to obtain the fault diagnosis result. Among them, the time-series fault diagnosis model includes a retrieval-enhanced generation module and a fault diagnosis module; the retrieval-enhanced generation module is used to retrieve in the thermal power plant equipment knowledge base according to the multimodal data, merge the retrieved files with the retrieval-enhanced generation module to obtain retrieval prompt data; the fault diagnosis module is used to diagnose according to the retrieval prompt template generated by the retrieval-enhanced generation module.

[0033] In this embodiment, the thermal power plant equipment knowledge base is a multi-source heterogeneous vector database. The vector database at least includes thermal power plant domain expert reports, thermal power equipment operation logs, thermal power equipment manuals, and thermal power equipment historical fault data. The data in the thermal power plant equipment knowledge base is in vector form. Specifically, a pre-trained text embedding model (such as BERT, RoBERTa, GPT, etc.) is used to convert text fragments into vectors, and these vectors are stored in a vector database (such as FAISS, Milvus, Pinecone, etc.).

[0034] The retrieval-enhanced generation module specifically includes: First, the module needs to identify and parse the input multimodal data. The multimodal data includes time-series operation data after wavelet denoising, two-dimensional time-frequency image data after Markov transfer field conversion, and thermal power plant equipment logs, etc. Preprocess the identified multimodal data, including data cleaning, format conversion, etc., to ensure data consistency and availability.

[0035] Use a deep learning model to convert the multimodal data into vectors. This vector representation can capture the complex features and relationships of the multimodal data. This embodiment also includes normalizing the generated vectors, such as normalization, zero-mean normalization, etc., to ensure comparability between vectors.

[0036] Query the knowledge base of thermal power plant equipment according to the multi-modal data in vector form, and query the most similar knowledge base fragments according to the cosine similarity; sort the retrieved knowledge base fragments according to the cosine similarity score, and the top N fragments with the highest scores are selected as the most similar knowledge base fragments. Set a similarity threshold to filter out the knowledge base fragments with similarity higher than the threshold to ensure the reliability of the retrieval results. Take the knowledge base fragment with the highest similarity as the first vector; merge the first vector with the multi-modal data in vector form to obtain retrieval prompt data. The merging method can adopt methods such as splicing and weighted summation to retain the original information of the multi-modal data and the relevant knowledge in the knowledge base.

[0037] In addition, the construction of the Retrieval-Augmented Generation (RAG) module: Combine a domestic open-source large language model with a professional knowledge corpus to achieve accurate information retrieval and natural language generation.

[0038] The fault diagnosis module uses a large language model; the fault diagnosis module specifically includes: Perform semantic understanding based on the retrieval prompt data, and perform fault diagnosis based on the data after semantic understanding to determine the fault phenomena, occurrence events, and influence ranges of thermal power equipment; Query in the knowledge base of thermal power plant equipment according to the fault phenomena, occurrence events, and influence ranges to obtain possible fault types, fault causes, corresponding treatment measures, and corresponding confidence levels; Sort the diagnosis results according to the confidence level, and take the fault type, fault cause, and corresponding treatment measures with the highest confidence level as the fault diagnosis results according to the sorting results.

[0039] Finally, according to the fault diagnosis results, use a predefined report template to generate an equipment analysis report including fault causes, treatment measures, and preventive suggestions.

[0040] After obtaining the equipment analysis report, it also includes the steps of submitting the equipment analysis report to expert review and providing feedback according to the review results, and optimizing the model through a feedback mechanism, specifically including: After the equipment analysis report is generated, use an expert review mechanism to annotate the fault cases in the equipment analysis report, and update the annotated fault cases to the knowledge base of thermal power plant equipment.

[0041] In addition, the method of this embodiment also includes visualizing the equipment analysis report, and the visualization interface includes: An expert review interface, which includes annotating the error types of misdiagnosed cases; A knowledge base setting interface, which is used to display in the form of a tree topology diagram, display the data annotation information in the knowledge base of thermal power plant equipment, and trace information according to the annotation information.

[0042] In addition, the method of this embodiment is integrated into the operation and maintenance platform of a thermal power plant to achieve intelligent fault analysis and automatic report generation.

[0043] Embodiment 2 This embodiment provides a system for generating a fault analysis report for a thermal power plant based on retrieval enhancement, which is used to implement the method for generating a fault analysis report for a thermal power plant based on retrieval enhancement in Embodiment 1, and includes: A data acquisition module, which acquires the time-series operation data and device logs of the thermal power plant equipment according to distributed sensors, and constructs multi-modal data according to the time-series operation data and device logs; A fault diagnosis module, which inputs the multi-modal data into a time-series fault diagnosis model to obtain a fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the knowledge base of thermal power plant equipment according to the multi-modal data, and merge the retrieved files with the retrieval enhancement generation module to obtain a retrieval prompt template; the fault diagnosis module is used to diagnose according to the retrieval prompt template generated by the retrieval enhancement generation module; A report generation module, which generates an equipment analysis report including the cause of the fault, treatment measures, and preventive suggestions according to the fault diagnosis result by using a predefined report template.

[0044] The specific implementation manners of each module have been described in detail in Embodiment 1 and will not be elaborated here.

[0045] Embodiment 3 In another embodiment of the present invention, the present invention also provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a terminal device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. It can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. And, in this storage space, there are also stored one or more instructions suitable for being loaded and executed by a processor, and these instructions can be one or more computer programs (including program codes). It should be noted that more specific examples (non-exhaustive list) of the computer-readable storage medium here include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0046] The computer-readable storage medium also includes a data signal propagated in a baseband or as part of a carrier wave, which carries the readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The readable storage medium can also be any readable medium other than the readable storage medium, and this readable medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination of the above.

[0047] The program code for performing the operations of the present invention can be written in any combination of one or more programming languages. The programming languages include object-oriented programming languages - such as Java, C++, etc., and also include conventional procedural programming languages - such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as an independent software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user's computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., by connecting through the Internet using an Internet service provider).

[0048] One or more instructions stored in the computer-readable storage medium can be loaded and executed by a processor to implement the corresponding steps of the method for generating a fault analysis report of a thermal power plant based on retrieval enhancement in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor as follows: Obtain the time-series operation data and device logs of the thermal power plant equipment according to the distributed sensors, and construct multimodal data according to the time-series operation data and device logs; Input the multimodal data into the time-series fault diagnosis model to obtain a fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the knowledge base of the thermal power plant equipment according to the multimodal data, and merge the retrieved files with the retrieval enhancement generation module to obtain retrieval hint data; the fault diagnosis module is used to diagnose according to the retrieval hint template generated by the retrieval enhancement generation module; According to the fault diagnosis result, use a predefined report template to generate an equipment analysis report including the fault cause, treatment measures, and preventive suggestions.

[0049] Embodiment 4 Please refer to Figure 3 , the terminal device is a computer device. The computer device 60 in this embodiment 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 computer program 63 is executed by the processor 61, it implements the method for generating a fault analysis report of a thermal power plant based on retrieval enhancement in the embodiment. To avoid repetition, it will not be elaborated here one by one. Alternatively, when the computer program 63 is executed by the processor 61, it implements the functions of each model / unit in the computing system of the sequential fault diagnosis model in Embodiment 1. To avoid repetition, it will not be elaborated here one by one.

[0050] The computer device 60 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The computer device 60 may include, but is not limited to, a processor 61 and a memory 62. Those skilled in the art can understand that Figure 3 merely examples of the computer device 60, which do not constitute a limitation on the computer device 60, may include more or fewer components than those shown in the figure, or combine some components, or different components. For example, the computer device may also include input / output devices, network access devices, a bus, etc.

[0051] The so-called processor 61 may be a central processing unit (CPU), or may also be other general-purpose processors, central processors, graphics processors, 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, data processing logics based on quantum computing, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0052] The memory 62 may be an internal storage unit of the computer device 60, such as the hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device 60.

[0053] Further, the memory 62 may also include both internal storage units of the computer device 60 and external storage devices. The memory 62 is used to store computer programs as well as other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.

[0054] Any reference to a memory, database, or other medium used in the embodiments provided in this application may include at least one of non-volatile and volatile memories. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, and the like. Volatile memory may include random access memory (RAM) or external cache memory, etc. 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), etc.

[0055] The databases involved in the embodiments provided in this application may include at least one of relational databases and non-relational databases. Non-relational databases may include distributed databases based on blockchain, etc., without limitation. 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, data processing logics based on quantum computing, etc., without limitation.

Claims

1. A method for generating a fault analysis report of a thermal power plant based on retrieval enhancement, characterized in that, Including: Obtain the time-series operation data and device logs of thermal power plant equipment according to distributed sensors, and construct multimodal data based on the time-series operation data and device logs; Input the multimodal data into the time-series fault diagnosis model to obtain the fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the thermal power plant equipment knowledge base according to the multimodal data, and merge the retrieved files with the retrieval enhancement generation module to obtain retrieval prompt data; The fault diagnosis module is used to diagnose according to the retrieval prompt template generated by the retrieval enhancement generation module; According to the fault diagnosis result, use a predefined report template to generate an equipment analysis report including fault causes, treatment measures, and preventive suggestions.

2. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 1, wherein, Obtaining the time-series operation data and device logs of thermal power plant equipment according to distributed sensors specifically includes: Use the DCS system to obtain the time-series operation data of the thermal power plant; perform wavelet denoising on the time-series operation data; Use the Markov transfer field to convert the time-series signal into two-dimensional time-frequency image data; After fusing the two-dimensional time-frequency image data with the thermal power plant equipment logs, obtain multimodal data.

3. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 2, wherein, The thermal power plant equipment knowledge base is a multi-source heterogeneous vector database; the vector database at least includes thermal power plant domain expert reports, thermal power equipment operation logs, thermal power equipment manuals, and thermal power equipment historical fault data.

4. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 3, wherein The retrieval enhancement generation module specifically includes: Identify the multimodal data, convert the multimodal data into vectors, query the thermal power plant equipment knowledge base according to the vector-form multimodal data, and query the most similar knowledge base fragments according to the cosine similarity; Sort the retrieved results according to the similarity score, and use the knowledge base fragment with the highest similarity as the first vector; Merge the first vector with the vector-form multimodal data to obtain retrieval prompt data.

5. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 4, wherein The fault diagnosis module uses a large language model; the fault diagnosis module specifically includes: Perform semantic understanding according to the retrieval prompt data, and perform fault diagnosis according to the data after semantic understanding to determine the fault phenomena, occurrence events, and influence scope of the thermal power equipment; Query in the thermal power plant equipment knowledge base according to the fault phenomena, occurrence events, and influence scope to obtain possible fault types, fault causes, corresponding treatment measures, and corresponding confidence levels; Sort the diagnosis results according to the confidence level, and use the fault type, fault cause, and corresponding treatment measure with the highest confidence level as the fault diagnosis result according to the sorting result.

6. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 1, wherein After obtaining the equipment analysis report, it also includes the steps of submitting the equipment analysis report for expert review and giving feedback according to the review results, specifically including: After the equipment analysis report is generated, use an expert review mechanism to annotate the fault cases in the equipment analysis report, and update the annotated fault cases to the thermal power plant equipment knowledge base.

7. The method for generating a fault analysis report of a thermal power plant based on retrieval enhancement according to claim 6, characterized in that, It also includes visualizing the equipment analysis report, and the visualization interface includes: An expert review interface, and the expert review interface includes marking the error types of misdiagnosed cases; The knowledge base setting interface is used to display in the form of a tree topology diagram, showing the data annotation information in the thermal power plant equipment knowledge base, and tracing information according to the annotation information.

8. A retrieval-enhanced fault analysis report generation system for thermal power plants, which is used to implement the retrieval-enhanced fault analysis report generation method for thermal power plants described in any one of claims 1 to 7, characterized in that, It includes: A data acquisition module that obtains the time-series operation data and equipment logs of thermal power plant equipment according to distributed sensors, and constructs multi-modal data based on the time-series operation data and equipment logs; A fault diagnosis module that inputs the multi-modal data into a time-series fault diagnosis model to obtain a fault diagnosis result; the time-series fault diagnosis model includes a retrieval enhancement generation module and a fault diagnosis module; the retrieval enhancement generation module is used to retrieve in the thermal power plant equipment knowledge base according to the multi-modal data, and merge the retrieved files with the retrieval enhancement generation module to obtain a retrieval prompt template; The fault diagnosis module is used to diagnose according to the retrieval prompt template generated by the retrieval enhancement generation module; A report generation module that generates an equipment analysis report including fault causes, handling measures, and preventive suggestions according to the fault diagnosis result and using a predefined report template.

9. A computing device, characterized in that, It includes: One or more processors, a memory, and one or more programs, where 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 executing any one of claims 1 to 7 of the method for generating a thermal power plant fault analysis report based on retrieval enhancement.

10. A computer-readable storage medium 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 execute any one of claims 1 to 7 of the method for generating a thermal power plant fault analysis report based on retrieval enhancement.

Images