Intelligent maintenance system and method for power plant equipment

Abstract

The invention relates to an intelligent maintenance system and method for power plant equipment, belongs to the technical field of intelligent maintenance, and solves the problems of scattered maintenance process, low document generation efficiency, difficulty in field and background collaboration, difficulty in data security guarantee and the like in a traditional power plant equipment maintenance system. The invention discloses an intelligent maintenance system for power plant equipment, and the system comprises a PC terminal management platform which is connected to a third-party system to obtain reference data of to-be-detected equipment, and calls an intelligent agent platform based on the obtained reference data to generate a maintenance planning document; analyzing the maintenance planning document, generating a corresponding execution instruction according to the maintenance process and the maintenance content, issuing the execution instruction to a mobile terminal operation platform, receiving corresponding data uploaded by the mobile terminal operation platform according to the execution instruction, and generating a maintenance process record based on the corresponding data of all node execution instructions; and forming full-process data based on the maintenance planning document and the maintenance process record, and calling the intelligent agent platform to generate a maintenance summary document based on the full-process data.

Description

Technical Field

[0001] This invention relates to the field of intelligent maintenance technology, and in particular to an intelligent maintenance system and method for power plant equipment. Background Technology

[0002] In traditional power plant equipment maintenance, the planning, execution, and summarization of maintenance tasks typically rely on manual methods. This approach is inefficient and prone to errors. Maintenance personnel need to manually write numerous planning documents, record the maintenance process, and write summary reports. The creation of these documents is not only time-consuming and labor-intensive but also of inconsistent quality. Furthermore, the collection and transmission of on-site maintenance data lacks effective digital means, resulting in untimely information exchange between the field and the back-end systems, impacting the collaborative efficiency of maintenance work. Regarding data security, traditional maintenance systems lack encryption and access control for maintenance data, making them susceptible to data leaks and unauthorized access.

[0003] In implementing the embodiments of the present invention, the prior art has at least the following problems or defects: the maintenance process is scattered and lacks intelligent management, the document generation efficiency is low, the collaboration between the field and the back-end is difficult, and data security is difficult to guarantee. Summary of the Invention

[0004] Based on the above analysis, the embodiments of the present invention aim to provide an intelligent maintenance system and method for power plant equipment, in order to solve the problems of scattered maintenance processes, low document generation efficiency, difficulty in on-site and back-end collaboration, and difficulty in ensuring data security in existing traditional power plant equipment maintenance systems.

[0005] On one hand, embodiments of the present invention provide an intelligent maintenance system for power plant equipment, comprising:

[0006] The PC-based management platform is used to: During the maintenance planning phase, connect to a third-party system to obtain reference data for the equipment to be tested, and use the obtained reference data to call the intelligent agent platform to generate a maintenance planning document; During the maintenance process management phase, parse the maintenance planning document, generate corresponding node execution instructions according to the maintenance content, and send them to the mobile operation platform; Receive the corresponding data uploaded by the mobile operation platform according to the execution instructions, and generate maintenance process records based on the corresponding data of all node execution instructions; During the maintenance summary phase, form full-process data based on the maintenance planning document and the maintenance process records, and call the intelligent agent platform to generate a maintenance summary document based on the full-process data.

[0007] The mobile terminal operation platform is used to receive and upload corresponding data to the PC terminal management platform according to the node execution instructions sent by the PC terminal management platform;

[0008] The intelligent agent platform is used to respond to the call request from the PC-side management platform, generate the maintenance planning document during the maintenance planning phase, and generate the maintenance summary document during the maintenance summary phase.

[0009] Furthermore, the PC-based management platform includes:

[0010] The maintenance planning module is used to obtain reference data of the equipment to be tested from the third-party system, construct maintenance tasks based on the obtained reference data, and call the intelligent agent platform to generate maintenance planning documents.

[0011] The maintenance process management module is used to parse the maintenance planning document, create multiple process nodes based on the maintenance content described in the maintenance plan and the predefined process template, and issue node execution instructions corresponding to each process node to the mobile terminal operation platform in sequence according to the maintenance progress; receive the corresponding data uploaded by the mobile terminal operation platform according to the instructions of each node, and perform structured processing on all node data to generate a complete maintenance process record.

[0012] The maintenance summary module is used to gather the complete maintenance planning document and maintenance process records corresponding to this maintenance task, form the full-process data, and call the intelligent agent platform to generate a maintenance summary document based on the full-process data.

[0013] Furthermore, the maintenance planning module includes:

[0014] The task creation unit is used to determine the maintenance type and scope based on reference data of the device to be tested obtained from a third-party system, and to create a maintenance task containing task parameters based on a predefined process template.

[0015] The resource allocation unit is used to allocate personnel, equipment, and time resources to the maintenance task based on its requirements.

[0016] The planning document request unit is used to extract the task parameters and allocated resource information in the maintenance task, call the intelligent agent platform to generate a maintenance planning document, and store the maintenance planning document in the object storage server.

[0017] Furthermore, the intelligent agent platform includes:

[0018] The service gateway unit is used to receive and route call requests from the PC-side management platform;

[0019] Vector databases are used to store vectorized data of device manuals, historical maintenance documents, and industry standards.

[0020] The retrieval enhancement generation unit is used to receive task parameters and resource information sent by the maintenance planning module or full-process data sent by the maintenance summary module. Based on the task parameters and resource information or full-process data, it retrieves relevant reference document fragments from the vector database, combines the reference document fragments with preset document generation prompts, and inputs them into the large language model to generate the corresponding maintenance planning document or maintenance summary document.

[0021] Furthermore, the maintenance process management module includes:

[0022] The process execution unit is used to determine the current process node to be executed based on the maintenance content and the predefined process template, and to generate the corresponding node execution instruction.

[0023] The data acquisition unit is used to receive corresponding data uploaded by the mobile terminal operation platform. The corresponding data is associated with the current process node determined by the process execution unit and includes images, videos and text records.

[0024] The record generation unit is used to generate a maintenance process record corresponding to a specific process node based on the corresponding data received by the data acquisition unit and associated with that process node, and to store the maintenance process record to an object storage server.

[0025] Furthermore, the maintenance summary module includes:

[0026] The data aggregation unit is used to retrieve the maintenance planning document from the object storage server and the maintenance process record from the object storage server or the relational database after the maintenance task is completed.

[0027] The summary document request unit is used to take the aggregated maintenance planning document and maintenance process record as full-process data, call the intelligent agent platform to generate a maintenance summary document, and receive the returned maintenance summary document.

[0028] The archiving unit is used to store the maintenance summary document to an object storage server.

[0029] Furthermore, the management domain also includes a system management module, which includes:

[0030] The access control unit is used to determine the user's access permissions to the maintenance planning module, maintenance process management module, and maintenance summary module based on the user and role correspondence defined in the system, and to cache the access permission data in the Redis database.

[0031] The log management unit is used to store the user's operation logs for the maintenance planning module, maintenance process management module, and maintenance summary module into a relational database.

[0032] Furthermore, the PC-based management platform also includes a service monitoring module, which includes:

[0033] The indicator acquisition unit is used to periodically collect the performance indicators of the maintenance planning module, maintenance process management module and maintenance summary module, and store the collected performance indicators in a relational database.

[0034] The health detection unit is used to call the health check interfaces of the maintenance planning module, maintenance process management module and maintenance summary module through a scheduled task to determine the service status of each module.

[0035] The alarm triggering unit is used to generate and send alarm information when performance indicators exceed preset thresholds or service status is abnormal.

[0036] Furthermore, the PC-based management platform also includes a timed scheduling module, which includes:

[0037] The data synchronization unit is used to synchronize temporary maintenance data cached in Redis to the relational database;

[0038] The cache cleanup unit is used to periodically clean up expired cached data in the Redis database;

[0039] The document backup unit is used to periodically back up maintenance documents from the object storage server to a separate backup server.

[0040] Secondly, embodiments of the present invention provide an intelligent maintenance method for power plant equipment, comprising:

[0041] The PC-based management platform connects to a third-party system to obtain reference data for the equipment under test. Based on this reference data, it constructs maintenance tasks and calls the intelligent agent platform to generate a maintenance planning document. The PC-based management platform parses the maintenance planning document to obtain the maintenance content, generates execution instructions according to the maintenance memory nodes, and sends them to the mobile operation platform. The mobile operation platform receives the instructions and uploads the corresponding data. The PC-based management platform receives the corresponding data for all process nodes, generates and updates the maintenance process record, and monitors the progress in real time. When the maintenance task is completed, the PC-based management platform aggregates the maintenance planning document and the complete maintenance process record to form full-process data, and calls the intelligent agent platform to generate a maintenance summary document based on the full-process data.

[0042] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0043] The PC-based management platform can create ordered execution instructions based on predefined process templates and maintenance content, and accurately push these instructions to the mobile work platform. On-site personnel strictly follow these instructions via mobile devices and upload standardized corresponding data (images, videos, text, etc.). This model ensures consistency between on-site operations and the planned scheme, achieves real-time and structured collection of progress, quality, and safety data, significantly improves the efficiency of cross-departmental and cross-level collaborative work, and reduces errors and delays in information transmission.

[0044] Based on predefined process templates and large-scale models, maintenance planning documents, process records, and summary documents are generated, improving the quality and efficiency of document generation and avoiding errors and non-standard issues that may occur with manual document writing. Meanwhile, the automatic aggregation and analysis of full-process data provides a scientific basis for maintenance decision-making, further optimizing the management process of power plant equipment maintenance, reducing operation and maintenance costs, and improving equipment reliability and availability.

[0045] In this invention, the above-described technical solutions can be combined with each other to achieve more preferred combinations. Other features and advantages of this invention will be set forth in the following description, and some advantages may become apparent from the description or be learned by practicing the invention. The objects and other advantages of this invention can be realized and obtained from what is particularly pointed out in the description and drawings. Attached Figure Description

[0046] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0047] Figure 1 This is a schematic diagram of an intelligent maintenance system for power plant equipment provided in an embodiment of the present invention. Detailed Implementation

[0048] Preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, which form part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

[0049] Example 1

[0050] A specific embodiment of the present invention discloses an intelligent maintenance system for power plant equipment, such as... Figure 1 As shown.

[0051] The system comprises a PC-based management platform, a mobile operation platform, and an intelligent agent platform. The PC-based management platform is deployed within the power plant's high-security production management network area, executing all business logic and data services, and providing a web-based interface for management personnel. The mobile operation platform refers to a dedicated WeChat mini-program deployed on the smartphones of field workers for receiving instructions and uploading data. The intelligent agent platform, deployed independently as a unified AI capability center, is compatible with various mainstream language model architectures and interfaces, providing intelligent document generation services for the PC-based management platform. The system is built on the Java language and the Spring Boot microservice development framework.

[0052] The mobile work platform receives and executes node commands sent by the PC-based management platform, and uploads corresponding data to the PC-based management platform. The mobile work platform communicates directly with the PC-based management platform via the network's API interface, receiving commands such as initiating A-level maintenance planning for a specific unit or executing specific procedures. Field personnel use this platform to collect corresponding data such as images, videos, text records, and electronic signatures, and then upload them after associating them with the currently executed process nodes. Various documents generated by the PC-based management platform after processing (such as maintenance planning documents) can also be pushed to the mobile device for viewing via the interface.

[0053] The PC-based management platform receives business requests from its web frontend or via API, and parses these requests to determine their type. Business access requests include maintenance planning requests, process execution requests, and maintenance summary requests. For example, a project manager clicking "Create Turbine Overhaul Task" on the PC web interface initiates a maintenance planning request. Process execution requests include the PC issuing work instructions to the mobile device and receiving data submitted by the mobile device after completing a work procedure. Maintenance summary requests are instructions to generate a final maintenance summary report and close the task after all predetermined work procedures and process nodes for a maintenance task have been completed. This can be done manually by the project manager clicking the "Apply to Close Task" button on the system interface, or automatically triggered by the system when it detects that all task nodes are complete. The maintenance summary report is generated based on all documents generated throughout the task's lifecycle, including planning documents and all process records, and is then archived.

[0054] In some embodiments, the backend services of the PC-based management platform and the mobile operating platform, as well as the intelligent agent platform, adopt a microservice architecture and can be containerized and deployed in a Kubernetes cluster. The relational database, vector database, object storage server, and Redis database used by the system can be deployed in a cluster to meet high availability requirements.

[0055] The PC-based management platform employs a microservices design for its various business modules, allowing each module to run and scale independently. Relational databases, vector databases, object storage servers, and Redis databases are deployed in a cluster, better adapting to the complex needs of power plant equipment maintenance while reducing operational costs.

[0056] Containerized deployment architecture is a software deployment method. Docker containers are lightweight, portable software containers that package applications and their dependencies together, ensuring consistent operation across different environments. Kubernetes clusters are container orchestration platforms used for automated deployment, scaling, and management of containerized applications. In this system, the various modules of the PC-side management platform, such as the maintenance planning module and the maintenance process management module, can be encapsulated as independent Docker containers, each managed and scheduled uniformly through Kubernetes. Clustered deployment refers to deploying multiple database instances on different servers to form a cluster. For example, the relational database PostgreSQL uses a master-slave replication cluster deployment. The vector database uses a HubbleVector distributed sharding cluster. The object storage server MinIO uses a multi-node distributed cluster based on erasure coding. The Redis database uses a distributed caching cluster deployment in Cluster mode, specifically using Redis version 7.x, with a three-master, three-slave architecture to provide distributed caching capabilities, supporting automatic failover, and meeting the caching needs of high-concurrency scenarios.

[0057] During implementation, each service module on each platform, such as the maintenance planning module and the maintenance process management module, is written as an independent Dockerfile, defining the module's runtime environment and dependencies. Docker images are built and pushed to private or public image repositories. In the Kubernetes cluster, these containers are deployed and managed by defining Deployment and Service resources. For example, a Deployment is defined for the maintenance planning module, specifying parameters such as the number of replicas, resource requests, and limits to ensure the module can automatically scale according to load.

[0058] Relational databases employ a master-slave replication architecture, with the master node handling write operations and the slave nodes handling read operations. Vector databases and object storage servers utilize distributed storage technologies such as MinIO. The system is also configured with a load balancer, such as Nginx, to evenly distribute user requests across the container instances.

[0059] In some embodiments, the PC-based management platform includes:

[0060] The maintenance planning module is used to acquire reference data from third-party systems, construct maintenance tasks based on the acquired reference data, and call the intelligent agent platform to generate a maintenance planning document. Reference data includes status data (such as measured parameters), defect data (reported defects), hazard data (potential risk points), and technical supervision report data (periodic professional assessment reports) obtained from other power plant business systems. Third-party systems typically refer to the power plant's defect management system (such as ERP), hazard management system, technical supervision service platform, and real-time monitoring system (such as SIS). The maintenance planning document is a comprehensive technical document prepared to guide the implementation of the entire maintenance project, clearly defining the maintenance objectives, scope, list of standard and non-standard items, required resource plans, technical specifications, safety measures, etc. The maintenance summary document is a comprehensive report that reviews, analyzes, and evaluates the entire process after all process nodes of the maintenance task have been completed, including work completion status, major problems found, handling results, and equipment status.

[0061] In some embodiments, the maintenance planning module includes:

[0062] The task creation unit is used to determine the maintenance type and scope based on reference data of the equipment to be tested obtained from a third-party system, and to create a maintenance task containing task parameters based on a predefined process template. Equipment information refers to the specific parameters of the equipment to be maintained, such as equipment model, system, and location. The process template is matched with the equipment information. Maintenance types are categorized according to maintenance purpose and needs. For example, preventative maintenance is a periodic inspection and maintenance performed to prevent equipment failure, while fault maintenance is an emergency repair after equipment failure.

[0063] The resource allocation unit is used to allocate personnel, equipment, and time resources based on the needs of the maintenance task. The resource allocation unit allocates resources according to the needs of the maintenance task. Personnel resources can include the number of maintenance personnel and their professional skill requirements; equipment resources include the specific tools and equipment required for maintenance; and time resources are time windows allocated based on the urgency of the maintenance task and the expected workload. During implementation, after the task is created, based on the complexity and duration of the maintenance task, the unit allocates specific executors (teams, supervisors), required equipment (specialized testing instruments, cranes, etc.), and time resources (planned start / completion dates, etc.) to the task from the power plant's human resources database, tool inventory, and calendar system, through a rule engine or manual specification.

[0064] The planning document request unit is used to extract task parameters from the maintenance task, and based on the task parameters and the allocated personnel, equipment, and time resources, it calls the intelligent agent platform to generate a maintenance planning document and stores the maintenance planning document on an object storage server. The request is sent to the intelligent agent platform via an internal API call to its unified service interface, automatically generating the maintenance planning document. After generation, the document is stored on an object storage server, such as a MinIO cluster, and the storage path is recorded back into the task data.

[0065] The intelligent agent platform includes a vector database and a retrieval-enhanced generation unit;

[0066] A vector database is used to store vectorized data from equipment manuals, historical maintenance documents, and industry standards. This database represents knowledge within the storage domain in a vectorized manner. Dedicated databases such as HubbleVector are used to convert unstructured text knowledge into a machine-readable and efficiently searchable format. Text extraction and cleaning are performed on technical manuals, annual maintenance reports, and industry standards such as DL / T standards and safety regulations provided by equipment manufacturers. Embedding models, such as text2vec-large-chinese and BGE, are used to segment the text into paragraphs and convert them into high-dimensional vectors, such as 768-dimensional vectors. The vectors, their corresponding original text fragments, and metadata are stored together in the vector database and indexed.

[0067] The retrieval enhancement generation unit is used to receive task parameters and resource information sent by the maintenance planning module or full-process data sent by the maintenance summary module. Based on the task parameters and resource information or full-process data, it retrieves relevant reference document fragments from the vector database, combines the reference document fragments with preset document generation prompts, and inputs them into the large language model to generate the corresponding maintenance planning document or maintenance summary document.

[0068] When receiving task parameters and resource information from the maintenance planning module or full-process data from the maintenance summary module, the query text is converted into a vector. Semantic retrieval is then performed in the vector database to find the top K most relevant text fragments. A hybrid retrieval strategy combining semantic retrieval and keyword retrieval (such as equipment model or component name) is employed. The retrieved relevant fragments are used as context, combined with targeted, pre-defined document generation prompts, and input into a large language model. This intelligent agent platform is built on a self-developed MaaS (Model-as-a-Service) platform, can utilize locally deployed large language models, and is compatible with various types of large language models, using historical power plant data for training. This platform encapsulates knowledge base retrieval and large language model invocation capabilities into a unified service, automatically generating planning and summary documents tailored to the actual power plant scenario.

[0069] The maintenance process management module parses the maintenance planning document, creates multiple process nodes based on predefined process templates, and sequentially issues node execution instructions corresponding to each process node to the mobile work platform according to the maintenance progress. It receives the corresponding data uploaded by the mobile work platform based on the instructions for each node, performs structured processing on all node data, and generates a complete maintenance process record. Maintenance tasks are created based on reference data and matching process templates; task parameters include equipment information, maintenance type, and required resources. The process templates are designed based on the BPMN 2.0 standard workflow engine, supporting visual process design, instance execution, and task scheduling, and are easily integrated with the Spring Boot framework.

[0070] Furthermore, the maintenance process management module includes:

[0071] The process execution unit is used to determine the current process node to be executed based on the maintenance content and the predefined process template, and generate the corresponding node execution instruction. According to the process logic, when the process progresses to a certain node, such as on-site disassembly and inspection, the process execution unit determines the current process node to be executed, updates the task status in the database, and prepares the data acquisition template required for that node.

[0072] The data acquisition unit receives corresponding data uploaded from the mobile work platform. This data is associated with the current process node determined by the process execution unit and includes images, videos, and text records. The data acquisition unit receives the corresponding data uploaded from the mobile work platform via an API interface. The data is associated with a specific process node ID during transmission. For example, when uploading a photo of a turbine blade, the maintenance task ID and process node ID to which the photo belongs are simultaneously submitted. Data formats include images (for equipment status and construction process documentation), videos (for recording key operational dynamics), and text records (for filling in measurement data, observation descriptions, etc.).

[0073] The record generation unit is used to generate a maintenance process record corresponding to a specific process node based on the data received by the data acquisition unit and associated with that node, and then stores the maintenance process record on the object storage server. The record generation unit can extract key items from the text record using simple natural language processing. It integrates all data associated with the same node, populates it into a predefined report template, automatically generates a maintenance process record for that node, and stores it on the object storage server, completing the archiving of field data.

[0074] The maintenance summary module is used to gather the complete maintenance planning document and maintenance process records corresponding to this maintenance task, form the full-process data, and call the intelligent agent platform to generate a maintenance summary document based on the full-process data.

[0075] In some embodiments, the maintenance summary module includes:

[0076] The data aggregation unit is used to retrieve maintenance planning documents from the object storage server and maintenance process records from the object storage server or relational database after the maintenance task is completed.

[0077] The summary document request unit is used to take the aggregated maintenance planning document and maintenance process records as full-process data, call the intelligent agent platform to generate a maintenance summary document, and receive the returned maintenance summary document. The summary document request unit takes the full-process data as input, calls the intelligent agent platform to generate a maintenance summary document, and the maintenance summary document includes task objectives, execution process, problems found and solutions, etc.

[0078] The archiving unit is used to store maintenance summary documents to the object storage server. The archiving unit stores the maintenance summary documents in the long-term archive directory of the object storage server and establishes an association index between all digital records of this maintenance and physical equipment assets. The object storage server is a distributed storage system suitable for storing and managing large amounts of unstructured data, such as documents, images, and videos.

[0079] In some embodiments, the PC-based management platform further includes a system management module, which includes:

[0080] The access control unit determines user access permissions for the maintenance planning module, maintenance process management module, and maintenance summary module based on the user-role mapping defined in the system, and caches the access permission data in a Redis database. Users refer to system operators, and roles are categories based on user responsibilities and permission scope, such as administrators and maintenance engineers. Each role is assigned different permissions, defining the functional modules that users can access and operate. For example, administrators have full access to all modules, while maintenance engineers can only access modules related to maintenance tasks. Access permission data is cached in a Redis database, a high-performance distributed caching system capable of quickly reading and updating cached data, thereby improving system response speed.

[0081] The log management unit stores user operation logs for the maintenance planning module, maintenance process management module, and maintenance summary module in a relational database, and synchronizes the operation log content to a log analysis database that supports multi-dimensional retrieval. A relational database is a structured data storage system suitable for storing and querying complex structured data. Operation logs include user identity information, operation time, operation type, etc. Brief log entries are stored in relational databases such as PostgreSQL. More detailed operation content requiring complex queries and analysis can be synchronized to a dedicated log analysis database such as Elasticsearch to support multi-dimensional auditing and traceability.

[0082] In some embodiments, the PC-based management platform further includes a service monitoring module, which includes:

[0083] The performance metrics collection unit is used to periodically collect performance metrics from the maintenance planning module, maintenance process management module, and maintenance summary module, and store the collected performance metrics in a relational database. The performance metrics collection unit integrates Spring Boot Actuator or Prometheus client to periodically and automatically collect performance metrics data of each business microservice, including CPU utilization, memory usage, interface response time, etc., and stores these metrics data in a time-series database or a specific monitoring storage.

[0084] The health monitoring unit is used to call the health check interfaces of the maintenance planning module, maintenance process management module, and maintenance summary module via scheduled tasks to determine the service status of each module. The health monitoring unit sequentially calls the health check interfaces provided by each microservice and determines whether the real-time operating status of each service is normal based on the returned response status code and content. The health check interfaces return status information such as normal, warning, and fault, used to evaluate the operating status of each module.

[0085] The alarm triggering unit is used to generate and send alarm messages when performance indicators exceed preset thresholds or service status becomes abnormal. The preset thresholds are upper limits for performance indicators set based on system design and operational experience; for example, CPU utilization exceeding 80% or response time exceeding 5 seconds may trigger an alarm. Alarm messages can be sent in various ways, such as email, SMS, or in-system notifications. In implementation, when any indicator exceeds the preset threshold or service status becomes abnormal, an alarm event is immediately generated, and the alarm information is sent to system maintenance personnel in real time via email, SMS, or instant messaging tools.

[0086] In some embodiments, the PC-based management platform further includes a timed scheduling module, which includes:

[0087] The data synchronization unit is used to synchronize temporary maintenance data cached in Redis to the relational database. Redis is a distributed caching system used for fast storage and access to temporary data, but this data needs to be periodically synchronized to the relational database to ensure consistency.

[0088] The cache cleanup unit is used to periodically clean up expired cached data in the Redis database. Regularly cleaning up expired cached data in the Redis database frees up memory space and avoids performance degradation caused by excessive cached data.

[0089] The document backup unit is used to periodically back up maintenance documents from the object storage server to a dedicated backup server. This backup server stores copies of the documents, enabling rapid data recovery in the event of a failure in the primary storage system. The operation cycle can be configured according to the system's actual needs, such as by configuring scheduled tasks. For example, data synchronization can be set to occur hourly, cache clearing can be set to run daily at midnight, and document backups can be set to occur weekly.

[0090] Example 2

[0091] A method for intelligent maintenance of power plant equipment, comprising:

[0092] The PC-based management platform connects to a third-party system to obtain reference data for the equipment under test. Based on this reference data, it constructs maintenance tasks and calls the intelligent agent platform to generate a maintenance planning document. The PC-based management platform parses the execution instructions of the maintenance planning document and sends them to the mobile operation platform. It receives the corresponding data uploaded by the mobile operation platform according to the execution instructions and generates maintenance process records based on the corresponding data of all node execution instructions. When the maintenance task is completed, the PC-based management platform aggregates the maintenance planning document and the complete maintenance process records to form full-process data, and calls the intelligent agent platform to generate a maintenance summary document based on the full-process data.

[0093] It is understandable that the steps and references recorded in the intelligent maintenance method for power plant equipment are... Figure 1 The various modules and units described correspond to the components of the intelligent maintenance system for power plant equipment. Therefore, the operations, characteristics, and beneficial effects described above for the intelligent maintenance method for power plant equipment are also applicable to the intelligent maintenance system for power plant equipment and the methods and steps it contains, and will not be repeated here.

[0094] Those skilled in the art will understand that all or part of the processes of the methods described in the above embodiments can be implemented by a computer program instructing related hardware, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a disk, optical disk, read-only memory, or random access memory, etc.

[0095] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. An intelligent maintenance system for power plant equipment, characterized in that, include: The PC-based management platform is used to connect to a third-party system during the maintenance planning phase to obtain reference data of the equipment to be tested, and to call the intelligent agent platform to generate maintenance planning documents based on the obtained reference data. During the maintenance process management phase, the maintenance planning document is parsed to generate node execution instructions and sent to the mobile operation platform. The corresponding data uploaded by the mobile operation platform according to the execution instructions is received, and maintenance process records are generated based on the corresponding data of all node execution instructions. During the maintenance summary phase, the maintenance planning document and the maintenance process records are used to form full-process data, and the intelligent agent platform is called to generate a maintenance summary document based on the full-process data. The mobile terminal operation platform is used to receive and upload corresponding data to the PC terminal management platform according to the node execution instructions sent by the PC terminal management platform; The intelligent agent platform is used to respond to the call request from the PC-side management platform, generate the maintenance planning document during the maintenance planning phase, and generate the maintenance summary document during the maintenance summary phase.

2. The system according to claim 1, characterized in that, The PC-based management platform includes a maintenance planning module, which is used to obtain reference data of the equipment to be tested from the third-party system, construct maintenance tasks based on the obtained reference data, and call the intelligent agent platform to generate maintenance planning documents. The maintenance process management module is used to parse the maintenance planning document to obtain the maintenance content, create multiple process nodes according to the predefined process template, and issue node execution instructions corresponding to each process node to the mobile terminal operation platform in sequence according to the maintenance progress and maintenance content; receive the corresponding data uploaded by the mobile terminal operation platform according to the instructions of each node, and perform structured processing on all node data to generate a complete maintenance process record. The maintenance summary module is used to gather the complete maintenance planning document and maintenance process records corresponding to this maintenance task, form the full-process data, and call the intelligent agent platform to generate a maintenance summary document based on the full-process data.

3. The system according to claim 2, characterized in that, The maintenance planning module includes a task creation unit, which is used to determine the maintenance type and scope based on reference data of the equipment to be tested obtained from a third-party system, and to create maintenance tasks containing task parameters based on a predefined process template. The resource allocation unit is used to allocate personnel, equipment, and time resources to the maintenance task based on its requirements. The planning document request unit is used to extract the task parameters and allocated resource information in the maintenance task, call the intelligent agent platform to generate a maintenance planning document, and store the maintenance planning document in the object storage server.

4. The system according to claim 3, characterized in that, The intelligent agent platform includes: The service gateway unit is used to receive and route call requests from the PC-side management platform; Vector databases are used to store vectorized data of device manuals, historical maintenance documents, and industry standards. The retrieval enhancement generation unit is used to receive task parameters and resource information sent by the maintenance planning module or full-process data sent by the maintenance summary module. Based on the task parameters and resource information or full-process data, it retrieves relevant reference document fragments from the vector database, combines the reference document fragments with preset document generation prompts, and inputs them into the large language model to generate the corresponding maintenance planning document or maintenance summary document.

5. The system according to claim 4, characterized in that, The maintenance process management module includes: The process execution unit is used to determine the current process node to be executed based on the maintenance content and the predefined process template, and generate the corresponding node execution instructions; The data acquisition unit is used to receive corresponding data uploaded by the mobile terminal operation platform. The corresponding data is associated with the current process node determined by the process execution unit and includes images, videos and text records. The record generation unit is used to generate a maintenance process record corresponding to a specific process node based on the corresponding data received by the data acquisition unit and associated with that process node, and to store the maintenance process record to an object storage server.

6. The system according to claim 5, characterized in that, The maintenance summary module includes: The data aggregation unit is used to retrieve the maintenance planning document from the object storage server and the maintenance process record from the object storage server or the relational database after the maintenance task is completed. The summary document request unit is used to take the aggregated maintenance planning document and maintenance process record as full-process data, call the intelligent agent platform to generate a maintenance summary document, and receive the returned maintenance summary document. The archiving unit is used to store the maintenance summary document to an object storage server.

7. The system according to claim 6, characterized in that, The management domain also includes a system management module, which includes: The access control unit is used to determine the user's access permissions to the maintenance planning module, maintenance process management module, and maintenance summary module based on the user and role correspondence defined in the system, and to cache the access permission data in the Redis database. The log management unit is used to store the user's operation logs for the maintenance planning module, maintenance process management module, and maintenance summary module into a relational database.

8. The system according to claim 7, characterized in that, The PC-based management platform also includes a service monitoring module, which comprises: The indicator acquisition unit is used to periodically collect the performance indicators of the maintenance planning module, maintenance process management module and maintenance summary module, and store the collected performance indicators in a relational database. The health detection unit is used to call the health check interfaces of the maintenance planning module, maintenance process management module and maintenance summary module through a scheduled task to determine the service status of each module. The alarm triggering unit is used to generate and send alarm information when performance indicators exceed preset thresholds or service status is abnormal.

9. The system according to claim 8, characterized in that, The PC-based management platform also includes a timed scheduling module, which includes: The data synchronization unit is used to synchronize temporary maintenance data cached in Redis to the relational database; The cache cleanup unit is used to periodically clean up expired cached data in the Redis database; The document backup unit is used to periodically back up maintenance documents from the object storage server to a separate backup server.

10. A method for intelligent maintenance of power plant equipment, characterized in that, The method includes: The PC-based management platform connects to a third-party system to obtain reference data for the equipment under test. Based on this reference data, it constructs maintenance tasks and calls the intelligent agent platform to generate a maintenance planning document. The PC-based management platform parses the execution instructions of the maintenance planning document and sends them to the mobile operation platform. It receives the corresponding data uploaded by the mobile operation platform according to the execution instructions and generates maintenance process records based on the corresponding data of all node execution instructions. When the maintenance task is completed, the PC-based management platform aggregates the maintenance planning document and the complete maintenance process records to form full-process data, and calls the intelligent agent platform to generate a maintenance summary document based on the full-process data.

Images