Preventive maintenance method and system for daily field of nuclear power plant

Abstract

The invention discloses a preventive maintenance method and system in the daily field of a nuclear power plant, and the method comprises the steps: collecting first object data, and building a first identifier and a first database; and performing first information analysis and task planning according to the first database, and generating an analysis planning result. And integrating analysis and planning results, generating a first task work order, tracking the state of the first task work order, and updating the first database. According to the method, the interaction relationship of the multi-dimensional features is fully considered, and the accuracy of fault prediction is greatly improved, so that a reliable basis is provided for task planning. Through the comprehensive evaluation system, task planning is more accurate, and resource allocation is more efficient. A closed-loop management mechanism ensures the transparency and continuity of the maintenance process, and the maintenance efficiency is further improved through dynamic optimization. The method is not only suitable for preventive maintenance of nuclear power plant equipment, but also can be extensively applied to scenes of nuclear power plant environment monitoring, personnel behavior monitoring and the like, and has very high adaptability and popularization value.

Description

Technical Field

[0001] The present invention relates to the technical field of preventive maintenance of nuclear power plants, and specifically to a preventive maintenance method and system for the daily field of nuclear power plants. Background Art

[0002] With the increasing complexity of modern nuclear power plant equipment operation and the improvement of safety requirements, the daily maintenance and fault prevention of equipment have become key technical links to ensure the safe operation of nuclear power plants. Traditional nuclear power plant equipment maintenance methods mainly include regular manual inspections and regular planned maintenance. Although these methods can avoid the occurrence of major faults to a certain extent, they also have significant limitations. For example, it is difficult for manual inspections to capture the dynamic state information of equipment in real time, and due to relying on personnel experience judgment, potential hazards may not be discovered in time. And the planned maintenance based on a fixed cycle ignores the actual differences in equipment operation status, which may lead to over-maintenance or maintenance lag of some equipment, resulting in waste of resources or accumulation of risks.

[0003] In recent years, with the development of Internet of Things, big data analysis and artificial intelligence technologies, intelligent equipment condition monitoring and maintenance technologies have gradually emerged. For example, by deploying sensors on key equipment for real-time data collection and using data-driven models to evaluate the health status of equipment, the visualization and controllability of equipment operation status can be significantly improved. However, there are still some deficiencies in existing intelligent maintenance technologies: First, most existing fault prediction models are based on single-dimensional data analysis and do not fully consider multi-dimensional comprehensive factors such as equipment operation environment and maintenance resources, making it difficult to achieve precise task planning; Second, the task generation and execution process usually lacks closed-loop management and cannot dynamically optimize maintenance strategies and feedback them to the system, resulting in the inability to accumulate maintenance knowledge.

[0004] In summary, the current preventive maintenance technology for nuclear power plant equipment urgently needs further innovation to achieve more efficient and accurate fault prediction and task planning, and optimize the overall maintenance process through closed-loop work order management to improve the operation safety and economy of nuclear power plants. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is the problem of inaccurate fault prediction: through data normalization processing, principal component analysis and a classification model based on random forest, comprehensively analyzing the equipment operation status and historical fault data, the accuracy of fault prediction is improved.

[0007] Problem of lack of multi - dimensional comprehensive evaluation in task planning: By introducing comprehensive evaluation factors such as equipment importance, operating environment, and maintenance resources, and designing nested priority rules, the rationality and executability of task planning are improved.

[0008] Problem of lack of closed - loop management in the maintenance process: By generating maintenance work orders, tracking work order status, and dynamically updating the first database, closed - loop management from data collection to maintenance execution is achieved, optimizing the maintenance process and accumulating maintenance knowledge.

[0009] To solve the above - mentioned technical problems, the present invention provides the following technical solution: A preventive maintenance method in the daily field of nuclear power plants, including: collecting first object data, establishing a first identifier and a first database.

[0010] Conducting first - information analysis and task planning based on the first database to generate an analysis and planning result.

[0011] Integrating the analysis and planning result to generate a first task work order, tracking the status of the first task work order, and updating the first database.

[0012] As a preferred embodiment of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: the collecting of the first object data includes determining the data collection scope and deploying sensors to collect data.

[0013] As a preferred embodiment of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: the establishing of the first identifier and the first database includes generating corresponding first identifiers for all first objects, establishing a first database, and storing the first identifiers and status records of the first objects.

[0014] As a preferred embodiment of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: the conducting of first - information analysis and task planning based on the first database to generate an analysis and planning result includes extracting dynamic operation data and historical data of the first object from the first database, performing normalization processing, and extracting features. Inputting the features into a calculation model to output a first - information analysis result.

[0015] Conducting task planning based on the first - information analysis result to generate an analysis and planning result.

[0016] As a preferred embodiment of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: the first object includes but is not limited to key equipment of nuclear power plants, the first identifier is a preventive maintenance identifier PMID, and the first database is a preventive maintenance requirement database PMRD.

[0017] The first - information analysis is fault prediction, and the task planning is priority evaluation.

[0018] The first task work order is a maintenance work order.

[0019] As a preferred solution of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: the first information analysis includes extracting the dynamic operation data and historical fault data of the equipment from the first database PMRD, and performing normalization processing. Reducing the dimension of the high-dimensional operation parameter data, extracting key features, training a classification model based on random forest, and calculating the equipment failure probability.

[0020] Task planning includes evaluating based on the equipment failure probability, as well as equipment importance, operating environment, and maintenance resource availability, and outputting the priority result of the task.

[0021] If the equipment failure probability ≥ 0.8 and the equipment importance is high, set the priority to the highest. If the equipment failure probability ≥ 0.5 and the equipment importance is medium, set the equipment priority to relatively high.

[0022] If the equipment failure probability < 0.5 or the equipment importance is low, do not set it as a priority task for the time being and enter the observation stage.

[0023] If the operating environment is harsh and the maintenance resources are tense, raise the priority by one level. If the operating environment is good and the maintenance resources are sufficient, maintain the current priority. If the operating environment is ordinary, perform manual confirmation according to the health status and importance of the equipment.

[0024] According to the priority result, it is divided into emergency tasks, planned tasks, and observation tasks. Emergency tasks immediately generate work orders and allocate resources for maintenance. Planned tasks are added to the regular maintenance plan, and the system dynamically arranges time windows. Observation tasks continuously monitor the operating status and wait for further evaluation.

[0025] As a preferred solution of the preventive maintenance method in the daily field of nuclear power plants according to the present invention, wherein: integrating the analysis and planning results, generating the first task work order, tracking the status of the first task work order, and updating the first database includes storing each type of task in a structured manner, generating the first task work order for each type of task and distributing it for processing.

[0026] Monitor the processing status of the first task work order, and enter the maintenance result and equipment status into the first database.

[0027] A preventive maintenance system in the daily field of nuclear power plants, characterized in that it includes,

[0028] An acquisition module that acquires first object data and establishes a first identifier and a first database.

[0029] A calculation module that performs first information analysis and task planning based on the first database and generates an analysis and planning result.

[0030] An integration module that integrates the analysis and planning results, generates a first task work order, tracks the status of the first task work order, and updates the first database.

[0031] A computer device includes a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the steps of the method described above are implemented.

[0032] A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described above are implemented.

[0033] Advantages of the present invention: By normalizing device data, performing dimensionality reduction analysis, and training a random forest model, the device failure probability is generated. Compared with traditional single-dimensional prediction models, the present invention fully considers the interaction relationships of multi-dimensional features, greatly improving the accuracy of failure prediction, thus providing a reliable basis for task planning.

[0034] In task planning, in combination with four dimensions of device failure probability, device importance, operating environment, and maintenance resources, the task priority is refined and evaluated through a nested rule system. For example, when the operating environment is harsh and maintenance resources are scarce, the priority is appropriately increased. When the operating environment is good and resources are sufficient, the current priority is maintained. Through this comprehensive evaluation system, task planning is more accurate and resource allocation is more efficient.

[0035] Based on the analysis and planning results, three types of maintenance work orders, namely emergency tasks, planned tasks, and observation tasks, are generated, and the execution process of the work orders is tracked in real time. After the maintenance is completed, the maintenance results and device status are updated to the first database, and the newly added maintenance experience is entered into the knowledge base. This closed-loop management mechanism ensures the transparency and continuity of the maintenance process and further improves the maintenance efficiency through dynamic optimization.

[0036] The preventive maintenance system in the present invention adopts a modular design and includes a data acquisition module, a calculation module, and an integration module. This system is not only applicable to the preventive maintenance of nuclear power plant equipment, but also can be extended to applications such as nuclear power plant environmental monitoring and personnel behavior monitoring, with strong adaptability and promotion value. Description of the Drawings

[0037] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only 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. Among them:

[0038] Figure 1The overall flowchart of a preventive maintenance method and system in the daily field of nuclear power plants provided for the first embodiment of the present invention. Detailed implementation manners

[0039] To make the above objects, features, and advantages of the present invention more apparent and understandable, the following provides a detailed description of the specific implementation manners of the present invention with reference to the accompanying drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0040] Example 1, referring to Figure 1 , which is an embodiment of the present invention, provides a preventive maintenance method in the daily field of nuclear power plants, including:

[0041] S1: Collect first object data, establish a first identifier and a first database.

[0042] In the present invention, the first object is the key equipment of the nuclear power plant, the first identifier is the preventive maintenance identifier PMID, and the first database is the preventive maintenance requirement database PMRD.

[0043] Determine the scope of the key equipment of the nuclear power plant, including but not limited to generator sets, cooling systems, control systems, and sensor networks. Define key operating parameters for each type of equipment, such as temperature, pressure, vibration frequency, current, and voltage. Deploy multiple types of sensors to collect equipment operation data, including physical state data (such as temperature, pressure, etc.) and operating state data (such as switch state, operating duration, etc.).

[0044] Classify and organize the collected data, and extract the characteristic parameters of the equipment, for example:

[0045] Static characteristics: equipment model, manufacturer, installation date, etc.

[0046] Dynamic characteristics: real-time operating state parameters and historical operation records.

[0047] Build a feature library containing the characteristics of the entire life cycle of the equipment to support subsequent analysis.

[0048] Generate a unique planned maintenance identifier (PMID) for each piece of nuclear power plant equipment. The identifier should include the unique identification code of the equipment (such as the serial number), the type classification to which the equipment belongs, the installation location or area number, and the data collection timestamp.

[0049] Furthermore, the PMID format is: equipment type - area number - timestamp, such as Pump - 001 - 20250108.

[0050] Design a unified database PMRD structure to store device information. The database mainly includes the following tables:

[0051] Device information table: Store PMID, basic device information, and classification.

[0052] Operating status table: Store the collected real-time data and historical operating data.

[0053] Fault record table: Record the device fault history and its corresponding maintenance measures.

[0054] The data is stored in a standardized format (such as XML or JSON) to support subsequent cross-platform data interaction.

[0055] It should be noted that the first object includes, but is not limited to, the key equipment of nuclear power plants, and can also be the nuclear power plant environmental monitoring system and the nuclear power plant personnel behavior monitoring system.

[0056] In an alternative embodiment of the present invention, the first object is the key equipment of a nuclear power plant. First, determine the data collection scope of the environmental monitoring system, including key areas such as air quality monitoring, radiation monitoring, water quality monitoring, temperature and humidity monitoring inside and outside the plant building.

[0057] The key parameters include the following:

[0058] Radiation indicators: Such as gamma ray intensity, beta ray intensity, etc.

[0059] Air quality indicators: Such as PM2.5, CO2 concentration, etc.

[0060] Water quality indicators: Such as pH value, dissolved oxygen content, heavy metal content, etc.

[0061] Temperature and humidity: Such as ambient temperature, humidity.

[0062] Deploy different types of sensors in the environmental monitoring area, including:

[0063] Radiation sensor: Monitor radiation intensity.

[0064] Gas sensor: Monitor air composition.

[0065] Water quality sensor: Used to online monitor the physical and chemical parameters of water bodies.

[0066] Temperature and humidity sensor: Used to record the ambient temperature and humidity.

[0067] Use an edge device gateway for data aggregation.

[0068] Each sensor collects data in real time. The data includes the original collected value and the timestamp. Preprocess the collected data, including data cleaning, data smoothing, and data formatting.

[0069] Generate a unique identifier for each monitoring point, which includes the monitoring area number, monitoring type, and timestamp. The format is: monitoring type - area number - timestamp, such as Radiation - 001 - 20250108.

[0070] Design the storage structure of environmental monitoring data, mainly including:

[0071] Monitoring point information table: Record the unique identifier, geographical location, sensor type, etc. of the monitoring point.

[0072] Environmental status table: Record real - time monitoring data and historical monitoring records.

[0073] Abnormal record table: Record over - standard or abnormal monitoring results.

[0074] In an alternative embodiment of the present invention, the first object is the key equipment of a nuclear power plant. First, determine the data collection scope of personnel behavior monitoring, and clarify the areas and behaviors that need to be key - monitored within the nuclear power plant, including:

[0075] Sensitive areas: such as the reactor control room, radioactive material storage area.

[0076] Key behaviors: such as entering an authorized area, emergency operation records, etc.

[0077] Clarify the key parameters to be collected, such as:

[0078] Location trajectory data (such as personnel's real - time location, movement trajectory).

[0079] Behavior data (such as access card swiping records, operation records).

[0080] Biological information data (such as identity recognition, face information).

[0081] Deploy relevant hardware devices in the monitoring area:

[0082] RFID positioning system: Track the real - time location of staff.

[0083] Video monitoring equipment: Monitor operation behaviors.

[0084] Access control system: Record the access rights and time of entering the area.

[0085] Biometric devices: such as fingerprint, iris, and face recognition devices.

[0086] Establish a linkage interface with the security management system to collect behavior data.

[0087] The RFID device records the real - time location and movement trajectory of personnel, the access control system records the access rights and timestamp of personnel entering and leaving, and the video monitoring records the operation behaviors of personnel.

[0088] The location data is formatted into a coordinate trajectory, and the access control and operation records are associated with the personnel identity identification. Key events (such as unauthorized entry or emergency operations) are extracted from the behavior data.

[0089] A unique identifier is generated for each staff member, which includes an identity number, a permission level, and a timestamp.

[0090] The format is: identity number - permission level - timestamp, such as ID001 - Level3 - 20250108.

[0091] Design the storage structure of the personnel behavior data, mainly including:

[0092] Personnel information table: Records the identity identification, permission level, and associated device information of each staff member.

[0093] Behavior record table: Stores the real-time trajectory, operation behavior, and access control records.

[0094] Abnormal event table: Records abnormal behaviors (such as unauthorized entry into sensitive areas or illegal operations).

[0095] S2: Perform the first information analysis and task planning based on the first database to generate the analysis and planning results.

[0096] In the present invention, the first information analysis is fault prediction, and the task planning is priority evaluation.

[0097] Extract the dynamic operation data and historical fault data of the device from the first database (PMRD), including: x1, x2,..., xn, which are the device operation state parameters.

[0098] t is the time series information, recording the time of the device data.

[0099] Normalize the operation parameters to eliminate the influence of data magnitude differences.

[0100] Use principal component analysis (PCA) to reduce the dimension of the high-dimensional operation parameter data and extract key features.

[0101] Utilize the historical fault data and operation data to train a classification model based on random forest. According to the calculation results, compare the fault probability of the device with the preset threshold Pthreshold to judge the health state of the device.

[0102] If P(Failure∣X)≥Pthreshold, then mark the device as high risk.

[0103] If P(Failure∣X)<Pthreshold, then mark the device as low risk.

[0104] Conduct a priority assessment, and the assessment factors include the equipment failure probability, equipment importance, operating environment, and maintenance resource availability.

[0105] The first-level rule is the joint assessment of failure probability and equipment importance, including:

[0106] If the equipment failure probability ≥ 0.8 and the equipment importance is high, directly set the equipment priority to the highest, and enter the subsequent assessment process to determine specific maintenance measures.

[0107] If the equipment failure probability ≥ 0.5 and the equipment importance is medium, set the equipment priority to high, and conduct a refined assessment in combination with the operating environment factors.

[0108] If the equipment failure probability < 0.5 or the equipment importance is low, do not set it as a priority task for the time being and enter the observation stage.

[0109] The second-level rule is the joint assessment of operating environment and maintenance resources, including:

[0110] If the operating environment is harsh and the maintenance resources are scarce, raise the priority by one level to reduce potential risks.

[0111] If the operating environment is good and the maintenance resources are sufficient, maintain the current priority without adjustment.

[0112] If the operating environment is ordinary, it is handed over to the user for manual confirmation according to the health status and importance of the specific equipment.

[0113] The third-level rule is equipment classification and task integration, including:

[0114] According to the priority calculated by the first-level and second-level rules, the equipment is divided into three categories: emergency tasks, planned tasks, and observation tasks:

[0115] Emergency tasks: Immediately generate a work order and allocate resources for maintenance.

[0116] Planned tasks: Add them to the regular maintenance plan, and the system dynamically arranges the time window.

[0117] Observation tasks: Continuously monitor the operating status and wait for further evaluation.

[0118] S3: Integrate the analysis and planning results, generate the first work order, track the status of the first work order, and update the first database.

[0119] According to the priority assessment results, classify and sort the analysis and planning results by task category (emergency tasks, planned tasks, observation tasks):

[0120] Emergency tasks: Equipment that needs to be repaired immediately and its corresponding analysis results.

[0121] Scheduled tasks: Equipment that needs to be repaired within a certain time window and its analysis results.

[0122] Observation tasks: Equipment that only needs to continuously monitor the status and its analysis results without immediate repair.

[0123] Structurally store the detailed information of each type of task, including the following:

[0124] Unique equipment identifier (PMID).

[0125] Failure prediction probability and key operating parameters (such as temperature, vibration frequency, pressure, etc.).

[0126] Priority level and task category.

[0127] Recommended repair strategy (obtained from PMRD in the first database).

[0128] Current equipment status (such as running, stopped, awaiting maintenance, etc.).

[0129] It should be noted that task integration and conflict detection also need to be carried out, merging multiple task information belonging to the same equipment to avoid duplicate work order generation. Detect whether there are task conflicts caused by resource allocation or time arrangement, such as multiple tasks requiring the same time window or the same repair resource. Dynamically adjust the task execution order according to the task priority and time window to ensure that high-priority tasks are completed first.

[0130] Generate an independent repair work order for each classified task (emergency task, scheduled task), including the following information:

[0131] Work order number: Uniquely identify the work order, in the format such as work order type - equipment ID - timestamp.

[0132] Equipment information: Including equipment PMID, name, location, operating status, etc.

[0133] Task category: Indicate whether the task is emergency, scheduled, or observation.

[0134] Repair content: Retrieve the recommended repair measures from the first database (PMRD), including the required operation steps, spare parts list, tool requirements, etc.

[0135] Estimated execution time: Allocate a specific time window according to the analysis and planning results and resource availability.

[0136] Responsible personnel: Automatically assign a maintenance engineer from the maintenance personnel database and indicate the contact person and division of responsibilities.

[0137] After the work order is generated, store it in a standardized data format (such as XML or JSON) for easy subsequent cross-module invocation or data interaction in the system.

[0138] The system distributes work orders to the responsible personnel through the built-in messaging module or integrated enterprise communication tools.

[0139] The status of each work order from creation to completion needs to be recorded throughout the process, including the following statuses:

[0140] Not executed: The work order has been generated but has not yet started to be executed.

[0141] In progress: The maintenance personnel have started to execute the work order.

[0142] Paused: The maintenance is paused due to special reasons (such as insufficient resources or external condition restrictions).

[0143] Completed: The maintenance task is completed and the equipment resumes normal operation.

[0144] The status change is updated in real time by the maintenance personnel through the system interface or mobile terminal.

[0145] The system provides a real-time monitoring module for task progress, showing the execution status of each work order, including: the current status of the work order, the real-time feedback from the maintenance personnel (such as problem description, task completion percentage), and newly added abnormal situations during the maintenance process (such as discovering new potential faults). Visualize all progress data to generate a Gantt chart or other forms of dynamic progress reports, facilitating the administrator to comprehensively control the task execution situation.

[0146] If the work order is not completed beyond the expected execution time, the system automatically sends reminders to the responsible personnel and the administrator. If new faults are detected during the maintenance process or the work order has been paused for too long, the system triggers a warning and suggests further processing solutions.

[0147] According to the execution results of the work order, the following information is stored in the first database:

[0148] Maintenance result: Indicates whether the maintenance is successful, as well as the specific maintenance measures and effects.

[0149] Equipment status update: Update the equipment status (such as "normal operation", "need to observe", "shutdown") according to the operation data after maintenance.

[0150] New experience entry: If new fault modes or optimization measures are discovered during the maintenance process, enter them as new maintenance knowledge into PMRD.

[0151] By updating the first database, a task closed-loop is formed to ensure seamless connection of all links of analysis, planning, execution, and feedback.

[0152] The system inputs the operation data of the equipment after maintenance into the analysis module again for further optimization of the fault prediction model and priority evaluation rules.

[0153] The computer device can be a server. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the data cluster data of the power monitoring system. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a preventive maintenance method in the daily field of nuclear power plants.

[0154] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the various embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can 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, etc. Volatile memory can 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. The databases involved in the various embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the various embodiments provided in the present application can be general-purpose processors, central processors, graphics processors, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0155] Embodiment 2, which is an embodiment of the present invention, provides a preventive maintenance method and system in the daily field of nuclear power plants. In order to verify the beneficial effects of the present invention, scientific demonstrations are carried out through simulation experiments.

[0156] To verify the performance advantages of the method of the present invention, the following experiments are designed. By comparing and analyzing the fault prediction and maintenance planning of key equipment in nuclear power plants, such as cooling pumps, the differences between the present invention and the prior art in terms of efficiency and accuracy are evaluated.

[0157] Test object: Preventive maintenance test of the cooling pump in a nuclear power plant

[0158] Implementation process of the prior art:

[0159] Data collection: Traditional methods mainly rely on manual inspections. The physical operating status (temperature, vibration, pressure) of the cooling pump is checked weekly, and abnormal phenomena are recorded. Due to the fixed inspection cycle, the accuracy of data collection is limited.

[0160] Fault prediction: Manually analyze based on historical data and use the empirical threshold method to judge whether there are potential faults. For example, when the vibration amplitude exceeds 2 mm / s or the temperature exceeds 70 °C, it is considered that there is a risk of failure. Due to the simple analysis model, it is vulnerable to interference from abnormal data.

[0161] Task planning: Manually generate maintenance tasks according to the risk assessment results, and give priority to dealing with equipment with obvious faults. The task allocation lacks the ability of dynamic adjustment, resulting in low resource utilization.

[0162] Execution process: Track the task status through maintenance records. There is no automated feedback process, and it is difficult to accumulate maintenance knowledge.

[0163] Implementation process of the present invention:

[0164] Data collection: Deploy multiple types of sensors to collect the operating status data of the cooling pump in real time, including key parameters such as temperature, vibration frequency, pressure, and current. The data is updated and uploaded to the database every minute.

[0165] Establish a database: Generate a unique preventive maintenance identification (PMID) for each cooling pump and store it in the preventive maintenance requirements database (PMRD). The database design includes equipment information tables, operating status tables, and fault record tables.

[0166] Fault prediction: Extract the dynamic operating data and historical fault data of the cooling pump from the PMRD. After normalization, use the random forest classification model to predict the fault probability, and extract characteristic variables (such as the non-linear relationship between vibration and temperature) by combining principal component analysis (PCA).

[0167] Task planning: Based on four dimensions of fault probability, equipment importance, operating environment, and maintenance resources, generate task priorities through a nested rule system. For example, when the fault probability ≥ 0.8 and the operating environment is harsh, the task priority is set to the highest.

[0168] Task execution and feedback: Generate maintenance work orders according to the plan, track the status of the work orders in real time, and dynamically update the maintenance results to the PMRD to provide support for subsequent maintenance tasks.

[0169] The experimental results are shown in Table 1.

[0170] Table 1 Experimental results

[0171]

[0172] By comparing the data in the table, it can be seen that the present invention is superior to the prior art in multiple parameters and performance indicators:

[0173] The present invention uses a random forest model for prediction, combined with multi-dimensional feature extraction, and the accuracy of the fault probability is significantly higher than the empirical threshold method of the prior art. For example, the failure probability of cooling pump 7 is 95%, while the prior art only judges it as "high priority" based on a single vibration parameter (3.5mm / s), which cannot effectively capture the linkage characteristics of temperature and vibration, and may cause delayed processing.

[0174] The present invention introduces factors such as equipment importance, operating environment and maintenance resources for priority assessment. For cooling pump 5 (failure probability 70%, temperature 70°C), the prior art simply relies on failure probability assessment as "medium priority", while the present invention combines environmental factors (harsh environment) to adjust it to "high priority", which is more targeted.

[0175] The present invention dynamically adjusts the task order through priority rules, so that high-risk equipment (such as cooling pump 1, cooling pump 3, cooling pump 7) is given priority, and the average maintenance completion time is 5.33 hours. However, due to the lack of dynamic task adjustment capability in the prior art, the average maintenance time is extended to 8.14 hours.

[0176] The present invention supports real-time update and subsequent analysis through PMRD dynamic storage of device status and fault data. Compared with the prior art which relies on manual inspection and results in a low data update frequency (0.14 times / day), the data update frequency of the present invention reaches 1440 times / day (updated once every minute), ensuring the real-time and reliability of the data.

[0177] In the fault prediction model, the present invention effectively reduces the data error rate (average error rate 3.4%) through normalization processing and multi-dimensional data noise reduction, while the error rate of the single parameter analysis method in the prior art is as high as 6.7%, further demonstrating the advantage of the present invention in data accuracy.

[0178] In summary, the present invention significantly improves the efficiency and accuracy of preventive maintenance of cooling pumps in nuclear power plants through intelligent fault prediction models and comprehensive task planning methods, while optimizing resource allocation and maintenance processes. Compared with the prior art, the present invention effectively solves problems such as delayed data updates, single analysis, and inaccurate task planning, and has obvious innovation and practical value.

[0179] Embodiment 3 is an embodiment of the present invention, comprising a preventive maintenance system for daily use in a nuclear power plant, specifically:

[0180] The acquisition module acquires the first object data and establishes a first identification and a first database.

[0181] A calculation module analyzes the first information and plans tasks according to the first database, and generates an analysis and planning result.

[0182] An integration module integrates the analysis and planning result, generates a first task work order, tracks the status of the first task work order, and updates the first database.

[0183] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A preventive maintenance method in the daily field of nuclear power plants, characterized in that, Including: Collect the first object data, establish the first identifier and the first database; Perform the first information analysis and task planning based on the first database, and generate the analysis and planning results; Integrate the analysis and planning results, generate the first task work order, track the status of the first task work order, and update the first database.

2. The preventive maintenance method for the daily operation field of a nuclear power plant according to claim 1, characterized in that: The collecting the first object data includes determining the data collection scope and deploying sensors to collect data.

3. The preventive maintenance method for the daily operation field of a nuclear power plant according to claim 2, characterized in that: The establishing the first identifier and the first database includes generating corresponding first identifiers for all first objects, establishing the first database, and storing the first identifiers and the status records of the first objects.

4. The preventive maintenance method for the daily operation field of a nuclear power plant according to claim 3, characterized in that: The performing the first information analysis and task planning based on the first database to generate the analysis and planning results includes extracting the dynamic operation data and historical data of the first object from the first database, performing normalization processing, and extracting features; Input the features into the calculation model and output the first information analysis results; Perform task planning based on the first information analysis results to generate the analysis and planning results.

5. The preventive maintenance method in the daily operation field of a nuclear power plant according to claim 4, characterized in that: The first object includes but is not limited to the key equipment of a nuclear power plant, the first identifier is the preventive maintenance identifier PMID, and the first database is the preventive maintenance requirement database PMRD; The first information analysis is fault prediction, and the task planning is priority evaluation; The first task work order is a maintenance work order.

6. The preventive maintenance method for the daily operation field of a nuclear power plant according to claim 5, characterized in that: The first information analysis includes extracting the dynamic operation data and historical fault data of the equipment from the first database PMRD and performing normalization processing; Reduce the dimension of the high-dimensional operation parameter data, extract key features, train a classification model based on a random forest, and calculate the equipment failure probability; The task planning includes evaluating based on the equipment failure probability, as well as the equipment importance, operating environment, and maintenance resource availability, and outputting the priority result of the task; If the equipment failure probability ≥ 0.8 and the equipment importance is high, set the priority to the highest; If the equipment failure probability ≥ 0.5 and the equipment importance is medium, set the equipment priority to be higher; If the equipment failure probability < 0.5 or the equipment importance is low, do not set it as a priority task for the time being and enter the observation stage; If the operating environment is harsh and the maintenance resources are scarce, increase the priority by one level; If the operating environment is good and the maintenance resources are sufficient, maintain the current priority; If the operating environment is ordinary, perform manual confirmation according to the health status and importance of the equipment; Classify the tasks into emergency tasks, planned tasks, and observation tasks according to the priority results. The emergency tasks immediately generate work orders and allocate resources for maintenance. The planned tasks are added to the regular maintenance plan, and the system dynamically arranges the time window. The observation tasks continuously monitor the operating status and wait for further evaluation.

7. The preventive maintenance method for the daily operation field of a nuclear power plant according to claim 6, characterized in that: The integrating the analysis and planning results, generating the first task work order, tracking the status of the first task work order, and updating the first database includes storing each type of task in a structured manner, generating the first task work order for each type of task and distributing it for processing; Monitor the processing status of the first task work order, and enter the maintenance results and the equipment status into the first database.

8. A preventive maintenance system in the daily field of a nuclear power plant adopting the method according to any one of claims 1-7, characterized in that: A collection module that collects the first object data, establishes the first identifier and the first database; A calculation module that performs first information analysis and task planning based on a first database to generate an analysis and planning result; An integration module that integrates the analysis and planning result to generate a first task work order, tracks the status of the first task work order, and updates the first database.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7.

Images