Unit refined intelligent operation and maintenance management system based on digital twinning and RCM

Abstract

The invention discloses a unit refined intelligent operation and maintenance management system based on digital twinning and RCM, and relates to the technical field of unit operation and maintenance management, the system comprises a data sensing layer, a digital twinning layer, an RCM analysis layer and a maintenance strategy generation layer; the data sensing layer is used for acquiring data of a unit by deploying a sensor and docking an internet of things; the digital twinborn layer establishes a digital twinborn model for the unit based on artificial intelligence and a digital twinborn technology; the RCM analysis layer analyzes the risk level of the fault and the maintenance priority of the faulted part based on the part state data of the digital twin model; the maintenance strategy generation layer is used for generating differentiated maintenance strategies according to components generating faults of different risk levels and performing fine adjustment on the maintenance strategies according to the actual operation state of a unit; the problem of maintenance hysteresis of a traditional maintenance mode is solved; the fault occurrence probability is corrected close to the actual situation, and the accuracy of risk level assessment is improved.

Description

Technical Field

[0001] This invention relates to the field of unit operation and maintenance management technology, specifically a refined intelligent operation and maintenance management system for units based on digital twins and RCM. Background Technology

[0002] As core equipment in production operations, the quality of operation and maintenance management of industrial units directly impacts production efficiency and safety. Traditional unit operation and maintenance relies heavily on manual inspections and periodic maintenance, which has significant limitations: Firstly, manual inspections are inefficient and subjective, making it difficult to capture subtle abnormalities in the unit's condition in real time, easily leading to missed or misdiagnosed faults. Secondly, periodic maintenance does not consider the actual health status and fault risks of the unit, potentially resulting in over-maintenance increasing costs or under-maintenance causing equipment downtime. Furthermore, it suffers from the following drawbacks: incomplete data collection; high false alarm and missed alarm rates in fault warnings; maintenance strategies based on experience or fixed cycles, leading to over-maintenance or under-maintenance; and the prevalence of periodic maintenance or post-fault maintenance models without considering the unit's reliability priorities to develop differentiated maintenance strategies, resulting in under-maintenance of high-risk components, such as downtime caused by untimely replacement of wind turbine main shaft bearings; and over-maintenance of low-risk components, such as frequent disassembly of conventional valves leading to shortened lifespan; ultimately resulting in high maintenance costs and low efficiency. Summary of the Invention

[0003] The purpose of this invention is to provide a refined and intelligent operation and maintenance management system for generator units based on digital twins and RCM, so as to solve the problems raised in the prior art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a refined intelligent operation and maintenance management system for generating units based on digital twin and RCM, the system comprising a data perception layer, a digital twin layer, an RCM analysis layer, and a maintenance strategy generation layer; The data perception layer preprocesses and stores the collected data in a database by deploying sensors and connecting to the IoT data collection unit to collect real-time operating data, historical maintenance data, and environmental data. The digital twin layer relies on artificial intelligence and digital twin technology to deeply mine data and establish a digital twin model for the unit. The RCM analysis layer uses component status data from a digital twin model to identify all potential faults in the unit and analyzes the risk level of the faults and the maintenance priority of the faulty components. The maintenance strategy generation layer is used to generate differentiated maintenance strategies for components that experience faults of different risk levels, and to fine-tune the maintenance strategies according to the actual operating status of the unit.

[0005] Furthermore, the data perception layer includes a data acquisition module and a data preprocessing module. The data acquisition module includes a real-time data acquisition unit, a historical maintenance data acquisition unit, and an environmental data acquisition unit. The real-time data acquisition unit collects operating parameters of unit components by deploying an IoT sensor network. The historical maintenance data acquisition unit collects historical fault records and maintenance records of different components of the unit by connecting to the enterprise's MES and CMMS systems. The historical fault records include fault type, occurrence time, and impact range. The maintenance records include consumable usage and maintenance duration. The environmental data acquisition unit collects temperature, humidity, and dust concentration in the unit's operating environment by deploying sensors, and obtains the unit's load demand and operating time through the unit's operating records. The data preprocessing module is used to clean and spatiotemporally align the data. Outliers are removed from the collected data using the 3σ rule, and the collected multi-source data is associated by timestamp and component location.

[0006] Furthermore, the digital twin layer includes a digital twin model construction module and a real-time state mapping module. The digital twin model construction module is used to construct geometric twin models, physical twin models, and behavioral twin models. Based on CAD drawings and laser scanning data, the digital twin model construction module constructs a 1:1 scale three-dimensional geometric model of the unit, refined to the component level, such as bearings, gears, and valves. It supports model scaling, disassembly, and perspective switching to intuitively display the unit structure. Based on physical field simulation algorithms, a physical property model of the unit is constructed to simulate the mechanical, thermal, and fluid characteristics of the unit components. By integrating historical operation and maintenance data with real-time status data, the unit behavior prediction model is trained using LSMT. The real-time state mapping module realizes real-time data synchronization between the physical unit and the digital twin model through edge computing nodes. When an anomaly occurs in a physical unit component, such as when the temperature of a component exceeds the normal range, the corresponding component in the digital twin model automatically highlights the warning, and at the same time, the behavioral twin model is triggered to analyze the abnormal diffusion path, such as whether the excessive temperature has caused wear on adjacent gears.

[0007] Furthermore, the RCM analysis layer includes a Failure Mode and Effects Analysis (FMEA) module, a maintenance priority assessment module, and a maintenance strategy generation module. The FMEA module, based on component status data from a digital twin model, identifies all potential failure modes of the unit, such as bearing wear, motor overload, and valve jamming, and analyzes the severity of each failure. The maintenance priority assessment module is used to adjust the risk priority assessment values ​​of different components in the unit in conjunction with production requirements.

[0008] Furthermore, the Failure Mode and Effects Analysis (FMEA) unit first identifies all failure types of unit components using a digital twin model, denoting all identified failure types as {A1, A2, ..., A...}. n, where n represents the total number of fault types; analyze the severity, occurrence frequency, and detection difficulty of each fault respectively, and denote a certain fault type as A i , where i represents the number of different types of faults, i = 1, 2,... n; then the severity of the i-th fault is denoted as W Ai , the fault occurrence frequency is denoted as R Ai , the fault detection difficulty is denoted as S Ai ; Evaluate the risk priority assessment value of the i-th fault based on the obtained data analysis ; Calculate the risk priority assessment value according to the following formula: ; Among them, represents the severity of the i-th fault, represents the actual fault occurrence frequency of the i-th fault; represents the fault detection difficulty of the i-th fault.

[0009] Furthermore, during the actual operation of the unit, the fault occurrence frequency in the actual situation is affected by the unit operation environment and unit load, and there will be differences from the values obtained through historical operation and maintenance data during real-time data collection; calculate the actual fault occurrence frequency according to the following formula: ; Among them, represents the basic fault occurrence frequency obtained from historical operation and maintenance data; T represents the environmental temperature of the unit operation, T0 represents the optimal environmental temperature of the unit operation; G represents the actual dust concentration in the unit operation environment; G0 represents the moderate dust concentration value in the unit operation environment; P represents the actual load value of the unit operation; P0 represents the optimal load value during the unit operation.

[0010] Furthermore, set the grading thresholds Q1 and Q2 for the risk priority assessment value; compare and analyze the obtained risk priority assessment value with the set thresholds, and the analysis results are as follows: If ≥ Q1, judge that the risk level of the i-th fault is high risk, and the maintenance priority of the component where the i-th fault occurs is high priority; If Q2 ≤ < Q1, judge that the risk level of the i-th fault is medium risk, and the maintenance priority of the component where the i-th fault occurs is medium priority; If < Q2, judge that the risk level of the i-th fault is low risk, and the maintenance priority of the component where the i-th fault occurs is low priority.

[0011] Furthermore, the maintenance strategy generation layer sets differentiated maintenance strategies for unit components that experience faults of different risk levels: For components prone to high-risk failures, a predictive maintenance and real-time monitoring strategy is implemented: Based on a constructed behavioral twin model, the system predicts potential failures of actual unit components according to parameter changes in different components within the behavioral twin model, with a failure prediction lead time of [time period missing]. ; For components prone to medium-risk failures, implement a maintenance strategy of preventative maintenance and periodic inspections: set the interval to [duration to be specified]. The maintenance cycle is to regularly detect and maintain medium-risk faults; For components prone to low-risk failures, a maintenance strategy of post-failure repair and periodic spot checks is implemented: after a faulty component is detected, it is repaired or replaced, with a set time interval. The duration of regular spot checks.

[0012] Furthermore, the fault prediction allowance, maintenance cycle, and periodic inspection duration for unit components are dynamic values. The changes in the maintenance cycle and periodic inspection duration are affected by the vibration frequency and real-time temperature of the unit components. The correction index is calculated according to the following formula: ; Where H represents the fault prediction allowance, maintenance cycle, and periodic inspection duration correction index for unit components; X represents the collected component vibration frequency; X0 represents the vibration frequency value of the unit component during normal operation; Y represents the actual temperature of the unit component during operation; and Y0 represents the temperature of the component during normal operation. The time allotted for fault prediction after adjustments to the core components of the unit, based on the operating status of the unit components, is: ; The maintenance cycle for key components of the unit has been adjusted based on the operating status of the unit's components. ; Based on the operating status of unit components, the periodic sampling inspection time for ordinary components of the unit is adjusted as follows: ; After adjusting the fault prediction allowance, maintenance cycle, and periodic inspection duration of the unit components based on their actual operating status, the adjusted duration is fed back to the maintenance strategy generation layer to update the maintenance strategy in real time.

[0013] Compared with the prior art, the beneficial effects of the present invention are: This application constructs digital twin models of unit components based on digital twin technology. Through real-time mapping between the digital twin model and the physical unit components, it can predict potential faults in core components. Furthermore, after predicting high-risk faults in core components, it generates corresponding maintenance strategies in advance, achieving predictive maintenance for core components. This avoids the problems of unit downtime or further damage caused by untimely maintenance of core components in traditional maintenance methods, solving the problem of maintenance lag in traditional methods. Moreover, this application combines historical operation and maintenance data with real-time unit operation data when performing fault risk level analysis. This approach modifies the probability of failure to closely approximate actual conditions, improving the accuracy of risk level assessment. The risk level assessment can adapt to changes in the environment and operating status during actual unit operation, resulting in more accurate assessments. Furthermore, it improves the accuracy of maintenance level judgment, addressing the problem that current technologies often rely on fixed rules or single thresholds for fault diagnosis, making it difficult to cope with complex operating environments and uncertainties. Moreover, the generated maintenance strategy updates and adjusts the sampling cycle or sampling duration of unit components in real time based on their operating conditions, achieving timely fault detection by adjusting the fault detection frequency according to the unit component status. This improves the timeliness of fault detection. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the unit refined intelligent operation and maintenance management system based on digital twin and RCM of the present invention. Detailed Implementation

[0015] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0016] like Figure 1 As shown, the present invention provides a technical solution, a refined intelligent operation and maintenance management system for generator units based on digital twin and RCM, the system including a data perception layer, a digital twin layer, an RCM analysis layer and a maintenance strategy generation layer; The data perception layer preprocesses and stores the collected data in a database by deploying sensors and connecting to IoT data collection units to collect real-time operating data, historical maintenance data, and environmental data. The digital twin layer relies on artificial intelligence and digital twin technology to deeply mine data and establish digital twin models for the units; The RCM analysis layer uses component status data from a digital twin model to identify all potential faults in the unit and analyzes the risk level of the faults and the maintenance priority of the faulty components. The maintenance strategy generation layer is used to generate differentiated maintenance strategies for components that experience failures of different risk levels, and to fine-tune the maintenance strategies based on the actual operating status of the unit.

[0017] The data perception layer includes a data acquisition module and a data preprocessing module. The data acquisition module includes a real-time data acquisition unit, a historical maintenance data acquisition unit, and an environmental data acquisition unit. The real-time data acquisition unit collects operating parameters of unit components by deploying an IoT sensor network. The historical maintenance data acquisition unit collects historical fault records and maintenance records of different components of the unit by connecting to the enterprise's MES and CMMS systems. The historical fault records include the fault type, occurrence time, and scope of impact. The maintenance records include consumable usage and maintenance duration. The environmental data acquisition unit collects temperature, humidity, and dust concentration in the unit's operating environment by deploying sensors, and obtains the unit's load demand and operating time through the unit's operating records. The data preprocessing module is used to clean and spatiotemporally align the data. Outliers are removed from the collected data using the 3σ rule, and the collected multi-source data is correlated by timestamp and component location.

[0018] The digital twin layer includes a digital twin model construction module and a real-time state mapping module. The digital twin model construction module is used to build geometric twin models, physical twin models, and behavioral twin models. Based on CAD drawings and laser scanning data, the digital twin model construction module builds a 1:1 scale 3D geometric model of the unit, detailed to the component level, such as bearings, gears, and valves. It supports model scaling, disassembly, and perspective switching to intuitively display the unit structure. Based on physical field simulation algorithms, it builds a physical property model of the unit to simulate the mechanical, thermal, and fluid characteristics of the unit components. By integrating historical operation and maintenance data with real-time status data, it uses LSMT to train the unit behavior prediction model. The real-time state mapping module realizes real-time data synchronization between the physical unit and the digital twin model through edge computing nodes. When an anomaly occurs in the physical unit component, such as the temperature of a component exceeding the normal range, the corresponding component in the digital twin model automatically highlights the warning, and at the same time, it triggers the behavioral twin model to analyze the abnormal diffusion path, such as whether the excessive temperature has caused wear on adjacent gears.

[0019] The RCM analysis layer includes a Failure Mode and Effects Analysis (FMEA) module, a maintenance priority assessment module, and a maintenance strategy generation module. The FMEA module uses component status data from a digital twin model to identify all potential failure modes of the unit, such as bearing wear, motor overload, and valve jamming, and analyzes the severity of each failure. The maintenance priority assessment module is used to adjust the risk priority assessment values ​​of different components in the unit in accordance with production requirements.

[0020] The Failure Mode and Effects Analysis (FMDA) unit first identifies all failure types of unit components using a digital twin model, denoting all identified failure types as {A1, A2, ..., A...}. n}, where n represents the total number of fault types; the severity, frequency of occurrence, and detection difficulty of each fault are analyzed separately, and a certain fault type is denoted as A. i Where i represents the number of different types of faults, i = 1, 2, ..., n; then the severity of the i-th type of fault is represented by W. Ai The frequency of failures is expressed as R. Ai The difficulty of fault detection is represented by S. Ai ;Analyze the risk priority assessment value of the i-th type of failure based on the acquired data. Calculate the risk priority assessment value according to the following formula: ; in, This indicates the severity of the i-th type of fault. This represents the actual frequency of occurrence of the i-th type of fault; This represents the difficulty of detecting the i-th type of fault.

[0021] During actual operation, the frequency of faults is affected by the unit's operating environment and load, resulting in differences between real-time data acquisition and historical maintenance data. The actual fault frequency can be calculated using the following formula: ; in, This indicates the frequency of basic faults obtained from historical operation and maintenance data; T represents the ambient temperature of the unit operation, and T0 represents the optimal ambient temperature for unit operation; G represents the actual dust concentration in the unit operation environment; G0 represents the moderate dust concentration in the unit operation environment; P represents the actual load value of the unit operation; and P0 represents the optimal load value for unit operation.

[0022] Set threshold values ​​Q1 and Q2 for risk priority assessment; compare the obtained risk priority assessment values ​​with the set threshold values, and the analysis results are as follows: like If Q1 ≥, determine that the risk level of the i-th type of fault is high risk, and the maintenance priority of the component where the i-th type of fault occurs is high priority; If Q2 ≤ < Q1, determine that the risk level of the i-th type of fault is medium risk, and the maintenance priority of the component where the i-th type of fault occurs is medium priority; If < Q2, determine that the risk level of the i-th type of fault is low risk, and the maintenance priority of the component where the i-th type of fault occurs is low priority.

[0023] The maintenance strategy generation layer sets different maintenance strategies for the components of the unit that generate faults of different risk levels: For the components that generate high-risk faults, implement a maintenance strategy of predictive maintenance and real-time monitoring: Based on the constructed behavior twin model, according to the parameter changes of different components of the unit in the behavior twin model, predict the faults that will occur in the actual unit components, and the fault prediction reserve time is ; For the components that generate medium-risk faults, implement a maintenance strategy of preventive maintenance and regular inspection: Set a maintenance cycle of and regularly detect and maintain medium-risk faults; For the components that generate low-risk faults, implement a maintenance strategy of post-fault repair and regular spot checks: After detecting the faulty component, repair and replace the faulty component, and set a regular spot check time of ;

[0024] The fault prediction reserve time, maintenance cycle, and regular spot check time for the unit components are dynamic change values. The changes in the maintenance cycle and regular spot check time are affected by the vibration frequency and real-time temperature of the unit components; Calculate the correction index according to the following formula: ; Among them, H represents the correction index of the fault prediction reserve time, maintenance cycle, and regular spot check time for the unit components; X represents the collected vibration frequency of the component; X0 represents the vibration frequency value of the unit component during normal operation; Y represents the actual temperature of the unit component during operation; Y0 represents the temperature of the component during normal operation; The adjusted fault prediction reserve time for the core components of the unit according to the operating state of the unit components is: ; The adjusted maintenance cycle of the important components of the unit according to the operating state of the unit components ; The adjusted regular spot check time for the general components of the unit according to the operating state of the unit components is adjusted to: ; After adjusting the fault prediction allowance, maintenance cycle, and periodic inspection duration of the unit components based on their actual operating status, the adjusted duration is fed back to the maintenance strategy generation layer to update the maintenance strategy in real time.

[0025] Example 1: The Failure Mode and Effects Analysis (FMDA) unit first identifies all failure types of unit components using a digital twin model, denoting all identified failure types as {A1, A2, ..., A...}. n}, where n represents the total number of fault types; the severity, frequency of occurrence, and detection difficulty of each fault are analyzed separately, and a certain fault type is denoted as A. i Where i represents the number of different types of faults, i = 1, 2, ..., n; then the severity of the i-th type of fault is represented by W. Ai The frequency of failures is expressed as R. Ai The difficulty of fault detection is represented by S. Ai ;Analyze the risk priority assessment value of the i-th type of failure based on the acquired data. Calculate the risk priority assessment value according to the following formula: ; Where, =0.8 represents the severity of the i-th type of fault. This represents the actual frequency of occurrence of the i-th type of fault; =0.8 indicates the difficulty of fault detection for the i-th type of fault.

[0026] During actual operation, the frequency of faults is affected by the unit's operating environment and load, resulting in differences between real-time data acquisition and historical maintenance data. The actual fault frequency can be calculated using the following formula: ; in, =0.5 represents the frequency of basic faults obtained from historical operation and maintenance data; T=30℃ represents the ambient temperature of unit operation, and T0=25℃ represents the optimal ambient temperature for unit operation; G=3 mg / m³ represents the actual dust concentration in the unit's operating environment; G0=4 mg / m³ represents the moderate dust concentration in the unit's operating environment; P=60kW represents the actual load value of the unit's operation; P0=80kW represents the optimal load value for unit operation; analysis yields... =0.436; then =0.279.

[0027] Set the risk priority assessment value's tiered thresholds Q1=0.4 and Q2=0.15; compare the obtained risk priority assessment values ​​with the set thresholds, and the analysis results are as follows: Analysis yields Q2≤ <If Q1 determines that the risk level of the i-th type of fault is medium risk, the maintenance priority of the component where the i-th type of fault occurs is medium priority; Then, for the components with medium-risk faults determined, implement preventive maintenance and regular inspection maintenance strategies: set a maintenance cycle with a duration of = 120 days, and regularly detect and maintain medium-risk faults; The maintenance cycle and the duration of regular spot checks for the components of the unit are dynamically changing values. The changes in the maintenance cycle and the duration of regular spot checks are affected by the vibration frequency and real-time temperature of the unit components; calculate the correction index according to the following formula: ; where H represents the correction index of the maintenance cycle and the duration of regular spot checks for the unit components; X = 1000 Hz represents the vibration frequency of the collected component; X0 = 800 Hz represents the vibration frequency value when the unit components are operating normally; Y = 60 °C represents the actual temperature when the unit components are working; Y0 = 80 °C represents the temperature when the components are operating normally; and the calculated H = 0.94; The adjusted maintenance cycle for the important components of the unit according to the operating status of the unit components ; = 113 days.

[0028] After adjusting the fault prediction reserved duration, maintenance cycle, and duration of regular spot checks for the unit components according to the actual operating status of the unit components, feedback the adjusted duration to the maintenance strategy generation layer to update the maintenance strategy in real time.

[0029] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be encompassed within the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

Claims

1. A refined intelligent operation and maintenance management system for generating units based on digital twins and RCM, characterized by: The system includes a data perception layer, a digital twin layer, an RCM analysis layer, and a maintenance strategy generation layer; The data perception layer preprocesses and stores the collected data in a database by deploying sensors and connecting to the IoT data collection unit to collect real-time operating data, historical maintenance data, and environmental data. The digital twin layer relies on artificial intelligence and digital twin technology to deeply mine data and establish a digital twin model for the unit. The RCM analysis layer uses component status data from a digital twin model to identify all potential faults in the unit and analyzes the risk level of the faults and the maintenance priority of the faulty components. The maintenance strategy generation layer is used to generate differentiated maintenance strategies for components that experience faults of different risk levels, and to fine-tune the maintenance strategies according to the actual operating status of the unit.

2. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 1, characterized in that: The data perception layer includes a data acquisition module and a data preprocessing module. The data acquisition module includes a real-time data acquisition unit, a historical maintenance data acquisition unit, and an environmental data acquisition unit. The real-time data acquisition unit collects operating parameters of unit components by deploying an IoT sensor network. The historical maintenance data acquisition unit collects historical fault records and maintenance records of different components of the unit by connecting to the enterprise's MES and CMMS systems. The historical fault records include fault type, occurrence time, and scope of impact. The maintenance records include consumable usage and maintenance duration. The environmental data acquisition unit collects temperature, humidity, and dust concentration in the unit's operating environment by deploying sensors, and obtains the unit's load demand and operating time through the unit's operation records; the data preprocessing module is used to clean and align the data in time and space; outliers are removed from the collected data using the 3σ rule, and the collected multi-source data is associated by timestamp and component location.

3. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 1, characterized in that: The digital twin layer includes a digital twin model construction module and a real-time state mapping module. The digital twin model construction module is used to construct geometric twin models, physical twin models, and behavioral twin models. Based on CAD drawings and laser scanning data, the digital twin model construction module constructs a 1:1 scale three-dimensional geometric model of the unit, refined to the component level, supporting model scaling, disassembly, and perspective switching, for intuitive display of the unit structure. Based on physics field simulation algorithms, a physical property model of the unit is constructed to simulate the mechanical, thermal, and fluid characteristics of the unit components. By fusing historical operation and maintenance data with real-time status data, LSMT is used to train the unit behavior prediction model. The real-time state mapping module realizes real-time data synchronization between the physical unit and the digital twin model through edge computing nodes.

4. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 1, characterized in that: The RCM analysis layer includes a failure mode and impact analysis module, a maintenance priority assessment module, and a maintenance strategy generation module. The failure mode and impact analysis module, based on component status data from a digital twin model, sorts out all potential failure modes of the unit and analyzes the severity of each failure. The maintenance priority assessment module is used to adjust the risk priority assessment values ​​of different components in the unit in combination with production needs.

5. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 4, characterized in that: The Failure Mode and Effects Analysis (FMEA) unit first identifies all failure types of unit components using a digital twin model, denoting all identified failure types as {A1, A2, ..., A...}. n }, where n represents the total number of fault types; the severity, frequency of occurrence, and detection difficulty of each fault are analyzed separately, and a certain fault type is denoted as A. i Where i represents the number of different types of faults, i = 1, 2, ..., n; then the severity of the i-th type of fault is represented by W. Ai The actual frequency of failures is expressed as R. Ai The difficulty of fault detection is represented by S. Ai ;Analyze the risk priority assessment value of the i-th type of failure based on the acquired data. ; Calculate the risk priority assessment value using the following formula: ； in, This indicates the severity of the i-th type of fault. This represents the actual frequency of occurrence of the i-th type of fault; This represents the difficulty of detecting the i-th type of fault.

6. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 5, characterized in that: During actual operation, the frequency of faults in the unit is affected by the unit's operating environment and load, and there will be differences between the values ​​obtained from real-time data collection and those obtained from historical operation and maintenance data. The actual failure frequency can be calculated using the following formula: ； in, This indicates the frequency of basic faults obtained from historical operation and maintenance data; T represents the ambient temperature of the unit operation, and T0 represents the optimal ambient temperature for unit operation; G represents the actual dust concentration in the unit operation environment; G0 represents the moderate dust concentration in the unit operation environment; P represents the actual load value of the unit operation; and P0 represents the optimal load value for unit operation.

7. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 5, characterized in that: Set threshold values ​​Q1 and Q2 for risk priority assessment; compare the obtained risk priority assessment values ​​with the set threshold values, and the analysis results are as follows: like If Q1 is greater than or equal to 1, and the risk level of the i-th type of failure is determined to be high risk, then the maintenance priority of the component that experiences the i-th type of failure is high priority. If Q2 ≤ <Q1, it is determined that the risk level of the i-th fault is medium risk, and the maintenance priority of the component where the i-th fault occurs is medium priority; If <Q2, if the risk level of the i-th fault is determined to be a low risk, the maintenance priority of the component with the i-th fault is a low priority.

8. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 1, characterized in that: The maintenance strategy generation layer sets differentiated maintenance strategies for unit components that experience faults of different risk levels: For components prone to high-risk failures, a predictive maintenance and real-time monitoring strategy is implemented: Based on a constructed behavioral twin model, the system predicts potential failures of actual unit components according to parameter changes in different components within the behavioral twin model, with a failure prediction lead time of [time period missing]. ; For components prone to medium-risk failures, implement a maintenance strategy of preventative maintenance and periodic inspections: set the interval to [duration to be specified]. The maintenance cycle is to regularly detect and maintain medium-risk faults; For components prone to low-risk failures, a maintenance strategy of post-failure repair and periodic spot checks is implemented: after a faulty component is detected, it is repaired or replaced, with a set time interval. The duration of regular spot checks.

9. The refined intelligent operation and maintenance management system for generating units based on digital twins and RCM as described in claim 8, characterized in that: The fault prediction allowance, maintenance cycle, and periodic inspection duration for unit components are dynamic values. The changes in the maintenance cycle and periodic inspection duration are affected by the vibration frequency and real-time temperature of the unit components. The correction index is calculated using the following formula: ； Where H represents the correction index for the fault prediction allowance, maintenance cycle, and periodic inspection duration of the unit components; X represents the collected component vibration frequency; X0 represents the vibration frequency value of the unit components during normal operation; Y represents the actual temperature of the unit components during operation; Y0 represents the temperature of the components during normal operation. The time allotted for fault prediction after adjustments to the core components of the unit, based on the operating status of the unit components, is: ; The maintenance cycle for key components of the unit has been adjusted based on the operating status of the unit's components. ; Based on the operating status of unit components, the periodic sampling inspection time for ordinary components of the unit is adjusted as follows: ; After adjusting the fault prediction allowance, maintenance cycle, and periodic inspection duration of the unit components based on their actual operating status, the adjusted duration is fed back to the maintenance strategy generation layer to update the maintenance strategy in real time.

Images