A power plant unit state evaluation method and system based on a knowledge graph

Abstract

The application discloses a power plant unit state evaluation method and system based on a knowledge graph, relates to the field of thermal power generation technology, and comprises the following steps: collecting and processing historical operation data of a thermal power generating unit, screening out key operation state parameters of the thermal power generating unit based on a MIC correlation analysis method and establishing a key operation state parameter database of the thermal power generating unit, constructing a key operation state parameter database of the thermal power generating unit after normalizing data processing of past key operation state parameter data of the thermal power generating unit, determining a health benchmark interval, obtaining past thermal power generating unit operation state key parameter evaluation standards through a machine learning model, mapping the health interval and normal and abnormal signals of the thermal power generating unit according to preset rules, and writing the mapping results into a knowledge graph database, establishing a thermal power generating unit operation state response strategy database, and setting response strategies corresponding to past fault handling experience based on the health benchmark interval.

Description

Technical Field

[0001] This invention relates to the field of power plant unit technology, specifically to a power plant unit status assessment method and system based on knowledge graphs. Background Technology

[0002] Existing power plants, such as thermal power plants, have thermal generator units that play a crucial role in power generation. These systems are large, generating complex status data. The data feedback from different unit states varies, and the correlation between data points is poor. In production, methods often rely on combined data states or single-data processing followed by comparison with thresholds to generate abnormal signals. This approach is prone to errors in judging the constantly changing unit operating system, leading to false alarms in normal states and frequent alarms in abnormal states, resulting in a waste of human and material resources.

[0003] The current methods for assessing the condition of thermal power generating units often combine data evaluation with technical experience to handle abnormal conditions. However, existing technologies do not adequately integrate data, thermal power generating units, and response strategies, which can easily lead to misjudgments of unit condition and delayed fault handling, resulting in reduced production efficiency and even serious accidents.

[0004] Therefore, this invention provides a knowledge graph-based power plant unit status assessment method that comprehensively judges the multivariate status of generator units and performs multi-level judgments on the degree of anomalies. It combines knowledge graphs to provide timely feedback on the unit status based on thermal power generator unit parameter data and gives corresponding strategies. Summary of the Invention

[0005] To address the aforementioned technical problems, a knowledge graph-based method for assessing the status of power plant units is provided. This technical solution resolves the issues raised in the background section regarding the susceptibility of existing variable unit operation systems to errors in judgment, leading to false alarms in normal conditions and frequent alarms in abnormal conditions, resulting in excessive work orders and over-maintenance.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] A knowledge graph-based method for power plant unit condition assessment includes the following steps:

[0008] Collect and process historical operating data of thermal power generating units, and use the MIC correlation analysis method to screen out key operating status parameters of thermal power generating units from the historical operating data;

[0009] Noise reduction and discrete regression processing are performed on key operating status parameters of thermal power generating units, and a database of key operating status parameters of thermal power generating units is constructed simultaneously.

[0010] Based on the database of key operating status parameters of thermal power generating units, an evaluation standard for key operating status parameters of thermal power generating units is constructed.

[0011] Based on the evaluation standards for key operating status parameters of thermal power generating units, a health benchmark interval is determined, which includes a healthy interval, a sub-healthy interval, and an abnormal interval.

[0012] Obtain real-time operating data of thermal power generating units, and combine the database of key operating status parameters of thermal power generating units with the health benchmark interval to construct a knowledge graph of key operating status parameters of thermal power generating units;

[0013] The initial index of the real-time operating status of the thermal power generating unit is determined, and the status of the thermal power generating unit is evaluated simultaneously based on the initial index of the real-time operating status and the health benchmark interval. The resulting response strategy for the operating status of the thermal power generating unit includes routine inspection, shortening the monitoring cycle, formulating maintenance plans, and emergency shutdown.

[0014] The key operating status parameters of thermal power generating units are extracted sequentially from the database of key operating status parameters of thermal power generating units. The key operating status parameters of thermal power generating units are evaluated by the analytic hierarchy process (AHP) to obtain auxiliary parameters for key status of thermal power generating units, and the status of thermal power generating units is then re-examined.

[0015] In an optional embodiment, the collection and processing of historical operating data of thermal power generating units, based on the MIC correlation analysis method, to screen out key operating status parameters of thermal power generating units from the historical operating data, specifically includes:

[0016] i operating status parameters of thermal power generating units are obtained after preprocessing based on historical operating data of thermal power generating units.

[0017] Obtain historical status data of thermal power generating units, and extract j historical status indices of thermal power generating units from the historical status data of thermal power generating units.

[0018] Using i thermal power generating unit status parameters and j thermal power generating unit historical status indices as input values ​​for the MIC formula, the correlation index between the thermal power generating unit status parameter Xi and the j thermal power generating unit historical status indices Yj is measured, and key operating status parameters of the thermal power generating unit are selected from the historical operating data of the thermal power generating unit based on the correlation index.

[0019] The expression for the MIC formula is:

[0020] ;

[0021] In the formula, ) represents the correlation index between the state parameter Xi and the historical state parameter Yi of the thermal power generating unit. The degree of distribution dependence of the state parameters Xi and historical state parameters Yi of the thermal power generating unit. As a normalization factor, the original MIC values ​​are scaled proportionally to eliminate dimensional differences, making it easier to focus on key state parameters. and the status of thermal power generating units Unified unit.

[0022] In an optional embodiment, the step of constructing an evaluation standard for key operating status parameters of thermal power generating units based on a database of key operating status parameters specifically includes:

[0023] By applying the principle of subjective empowerment, the key operating status parameters of thermal power generating units are scored using the experience evaluation of skilled workers at different levels, and a regression model is used to derive the weighted scoring standard for the key operating status parameters of thermal power generating units.

[0024] Based on the principles of objective and subjective weighting, weights are generated through model coefficients, and then manually fine-tuned according to business needs to obtain the evaluation standards for key operating status parameters of thermal power generating units.

[0025] In an optional embodiment, determining the health benchmark range based on the evaluation criteria for key operating status parameters of thermal power generating units specifically includes:

[0026] Based on historical operating data of thermal power generating units, unit fault records are obtained, including Level 1, Level 2, Level 3, and Level 4 faults. Based on these fault records, key operating status parameter fault data of thermal power generating units are extracted through machine learning. Based on the evaluation criteria for key operating status parameters of thermal power generating units, a healthy baseline interval is generated using a normal cloud model.

[0027] In an optional embodiment, the step of acquiring real-time operating data of thermal power generating units, and combining it with a database of key operating status parameters of thermal power generating units and a health benchmark interval to construct a knowledge graph of key operating status parameters of thermal power generating units, specifically includes:

[0028] S1.1. Define knowledge graph node A, where knowledge graph node A is real-time operating data of thermal power generating units;

[0029] S1.2. Define knowledge graph node B, which is a database of key operating status parameters of thermal power generating units;

[0030] S1.3. Define knowledge graph node C, where knowledge graph node C is the health baseline interval;

[0031] S1.4. Conduct auxiliary re-examination of key operating status parameters of thermal power generating units in a sub-healthy state and write them into the knowledge graph;

[0032] S1.5. Define the mapping relationship AB between knowledge graph node A and knowledge graph node B. The mapping relationship AB includes a first mapping relationship AB1 and a second mapping relationship AB2. The first mapping relationship AB is the time difference between the real-time operation data acquisition timestamp of the thermal power generating unit and the acquisition timestamp of the key operation status parameter database of the thermal power generating unit. The second mapping relationship AB2 is the ratio of the number of parameter types in the key operation status parameter database of the thermal power generating unit to the number of parameter types in the real-time operation data of the thermal power generating unit.

[0033] S1.6. Define the mapping relationship AC between knowledge graph node A and knowledge graph node C, wherein the mapping relationship AC is the deviation of the parameter in the real-time operation data of the thermal power generating unit from the average value of the two endpoints of the health benchmark interval;

[0034] S1.7. Define the mapping relationship BC between knowledge graph node B and knowledge graph node C. The mapping relationship BC is the deviation of the average value of the parameters in the database of key operating status parameters of thermal power generating units from the values ​​at both ends of the health benchmark interval.

[0035] In an optional embodiment, the step of determining the initial index of the real-time operating status of the thermal power generating unit, and simultaneously evaluating the status of the thermal power generating unit based on the initial index of the real-time operating status and the health benchmark interval, to obtain a response strategy for the operating status of the thermal power generating unit, specifically includes:

[0036] Based on the knowledge graph of key operating status parameters of thermal power generating units, the real-time operating data of thermal power generating units, the mapping relationship AB between knowledge graph node A and knowledge graph node B, and the mapping relationship AC between knowledge graph node A and knowledge graph node C are obtained.

[0037] Extract the time difference between the real-time operation data acquisition timestamp of the thermal power generating unit and the acquisition timestamp of the key operating status parameter database of the thermal power generating unit from the mapping relationship AB between knowledge graph node A and knowledge graph node B. Simultaneously extract the ratio of the number of parameter types in the key operating status parameter database of the thermal power generating unit to the number of parameter types in the real-time operation data of the thermal power generating unit.

[0038] Extract the deviation values ​​of parameters from the average values ​​of the two endpoints of the health baseline interval in the real-time operation data of thermal power generating units from the mapping relationship AC between knowledge graph nodes A and C.

[0039] Based on the real-time operation data of thermal power generating units, and combined with the time difference between the real-time operation data collection timestamp of the thermal power generating units and the collection timestamp of the database of key operating status parameters of thermal power generating units, as well as the deviation of the parameters in the real-time operation data of thermal power generating units from the average value of the two endpoints of the health benchmark interval, the initial index of the real-time operating status of thermal power generating units is determined.

[0040] The status of thermal power generating units is evaluated based on the initial index of real-time operating status and the health benchmark interval, and strategies for dealing with the operating status of thermal power generating units are obtained.

[0041] The formula for calculating the initial index of the real-time operating status of the thermal power generating unit is as follows:

[0042] ;

[0043] In the formula, For thermal power generating units in The initial index of the real-time operating status of the thermal power generating unit at any given moment. For thermal power generating units in The quantized value of the timestamp of the real-time running data collection. Quantified values ​​of timestamps collected for the database of key operating status parameters of thermal power generating units. For thermal power generating units in The first real-time running data at time 1 One running parameter, This is the quantized value of the left endpoint of the health baseline interval. This is the quantized value of the right endpoint of the health baseline interval. The number of parameter types in the database of key operating status parameters for thermal power generating units. In order to be in The number of parameter types in the real-time operating data of thermal power generating units at any given moment.

[0044] In an optional embodiment, the auxiliary re-inspection of the thermal power generating unit's status specifically includes:

[0045] The original key status auxiliary parameters of the thermal power generating unit are obtained by collecting data from sensors and processing the initial index of the real-time operating status of the thermal power generating unit.

[0046] The original key status auxiliary parameters of the thermal power generating unit are preprocessed, and the key status auxiliary parameter data of the thermal power generating unit are obtained by standard deviation and normalization.

[0047] Based on the analytic hierarchy process (AHP) to decompose a complex index system and the MIC analysis to determine the weight of the deviation of key auxiliary parameters of thermal power generating units on the impact of unit failures, the objective contribution rate of auxiliary parameters is obtained.

[0048] The key status auxiliary parameters of thermal power generating units are evaluated based on expert experience weights, and the subjective contribution rate of the auxiliary parameters is obtained using a regression model.

[0049] Based on the objective contribution rate and subjective contribution rate of the auxiliary parameters, weights are generated through model coefficients, and the auxiliary parameter contribution rate evaluation standard is obtained by manual fine-tuning in combination with business needs.

[0050] Historical data of key auxiliary parameters of thermal power generating units are collected, and deviation calculation or similarity measurement is performed to assess the degree of deviation. Standard deviation intervals are set, including abnormal intervals, normal intervals, and undetermined intervals.

[0051] Real-time data of key auxiliary parameters of thermal power generating units are obtained, and a knowledge graph of key auxiliary parameters of thermal power generating units is constructed by combining the evaluation criteria for contribution rate of auxiliary parameters and the standard deviation range.

[0052] The auxiliary review is embedded into a knowledge graph to obtain a multi-level knowledge graph.

[0053] In an optional embodiment, the evaluation criteria for the contribution rate of the auxiliary parameters specifically include:

[0054] The average weight of the auxiliary parameters is obtained by comprehensively scoring the key status auxiliary parameters of thermal power generating units by multiple groups of experts. The subjective contribution rate of the key status auxiliary parameters of thermal power generating units is obtained by performing regression model processing on the average weight of the auxiliary parameters.

[0055] The calculation equation used for the expert experience weight is as follows:

[0056] ;

[0057] In the formula, Subjective contribution rate of key auxiliary parameters for thermal power generating units. These are the various auxiliary parameters. This is the average of the weights of each auxiliary parameter. It is the regression coefficient.

[0058] Based on the aforementioned key state auxiliary parameters, the objective contribution rate of the key state auxiliary parameters of thermal power generating units is obtained through a random forest machine learning model, specifically including:

[0059] Based on historical data of key auxiliary parameters of thermal power generating units, a random forest classifier is constructed to calculate the importance score of key auxiliary parameters of thermal power generating units. The objective contribution rate of key auxiliary parameters of thermal power generating units is obtained through the importance score of key auxiliary parameters of thermal power generating units.

[0060] Weights are generated by model coefficients, and the contribution rate evaluation standard of key status auxiliary parameters is obtained by manual fine-tuning based on business needs.

[0061] Furthermore, a knowledge graph-based power plant unit status assessment system is proposed to implement the assessment method described above, including:

[0062] The data acquisition module includes a multi-source data synchronous acquisition unit, a data quality control unit, and an edge computing and cloud-edge collaboration unit. These units are distributed at selected locations on the generator set and acquire, store, and synchronize real-time status parameters through electrical connections with other units.

[0063] The data processing module, based on the acquired, stored, and synchronized real-time status parameters, obtains a database of key operating status parameters of thermal power generating units through data noise reduction, selection, and normalization processing.

[0064] The status assessment module obtains evaluation results based on the key operating status parameters of thermal power generating units and the evaluation standards for key operating status parameters of thermal power generating units;

[0065] The re-examination module reprocesses the data parameters of the sub-healthy state.

[0066] The display module shows the unit status in a time series and provides feedback on real-time status, periodic status, and long-term forecast status.

[0067] Furthermore, a computer-readable storage medium is proposed, on which a computer-readable program is stored, which, when invoked, executes the knowledge graph-based thermal power generator state assessment method as described above.

[0068] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0069] This invention proposes a knowledge graph-based method and system for assessing the status of power plant units. By constructing a knowledge graph of the status of thermal power generating units, abnormal states can be quickly captured during real-time monitoring. The system analyzes the unit status corresponding to the abnormal parameters and provides corresponding response strategies. Simultaneously, it assists in re-inspection, improving the accuracy of unit status judgment and reducing abnormal alarm signals. This reduces the efficiency and manpower costs of detecting and monitoring thermal power generating units. Attached Figure Description

[0070] Figure 1 This is a flowchart of a power plant unit condition assessment method based on knowledge graphs proposed in this invention.

[0071] Figure 2 This is a flowchart of the method for constructing a knowledge graph of thermal power generating units in this invention;

[0072] Figure 3 This is a schematic diagram of the power plant unit condition assessment system in this invention. Specific Implementation

[0073] The following description is intended to disclose the present invention so that those skilled in the art can implement it. The preferred embodiments described below are merely examples, and other obvious variations will be apparent to those skilled in the art.

[0074] Reference Figures 1-3 As shown, a knowledge graph-based method for power plant unit condition assessment includes the following steps:

[0075] Collect and process historical operating data of thermal power generating units, and use the MIC correlation analysis method to screen out key operating status parameters of thermal power generating units from the historical operating data;

[0076] Noise reduction and discrete regression processing are performed on key operating status parameters of thermal power generating units, and a database of key operating status parameters of thermal power generating units is constructed simultaneously.

[0077] Based on the database of key operating status parameters of thermal power generating units, an evaluation standard for key operating status parameters of thermal power generating units is constructed.

[0078] Based on the evaluation standards for key operating status parameters of thermal power generating units, a health benchmark interval is determined, which includes a healthy interval, a sub-healthy interval, and an abnormal interval.

[0079] Obtain real-time operating data of thermal power generating units, and combine the database of key operating status parameters of thermal power generating units with the health benchmark interval to construct a knowledge graph of key operating status parameters of thermal power generating units;

[0080] Determine the initial index of the real-time operating status of the thermal power generating unit, and simultaneously assess the status of the thermal power generating unit based on the initial index of the real-time operating status and the health benchmark interval, so as to obtain the response strategy for the operating status of the thermal power generating unit. The response strategy includes routine inspection, shortening the monitoring cycle, formulating maintenance plans and emergency shutdown.

[0081] The key operating status parameters of thermal power generating units are extracted sequentially from the database of key operating status parameters of thermal power generating units. The key operating status parameters of thermal power generating units are evaluated by the analytic hierarchy process (AHP) to obtain auxiliary parameters for key status of thermal power generating units, and the status of thermal power generating units is then re-examined.

[0082] Specifically, the key operating status parameters of thermal power generating units are evaluated using the analytic hierarchy process (AHP) to obtain key auxiliary parameters for the thermal power generating units, including:

[0083] i operating status parameters of thermal power generating units are obtained after preprocessing based on historical operating data of thermal power generating units.

[0084] Obtain historical status data of thermal power generating units, and extract j historical status indices of thermal power generating units from the historical status data of thermal power generating units.

[0085] Using i thermal power generating unit state parameters and j thermal power generating unit historical state indices as input values ​​for the MIC formula, the correlation index between the thermal power generating unit state parameter Xi and the j thermal power generating unit historical state indices Yj is measured, and key operating state parameters of thermal power generating units are selected from the historical operating data of thermal power generating units based on the correlation index.

[0086] The expression for the MIC formula:

[0087] ;

[0088] In the formula, X represents the state parameter of the thermal power generating unit. i Historical state parameters Y of thermal power generating units i The relevant index, State parameter X of thermal power generating unit i Historical state parameters Y of thermal power generating units i The degree of distribution dependence, As a normalization factor, the original MIC values ​​are scaled proportionally to eliminate dimensional differences, making it easier to focus on key state parameters. and the status of thermal power generating units Unified unit;

[0089] The MIC correlation analysis method was used to select key state parameters to characterize the operating status of the generator sets, and the key auxiliary state parameters of the thermal power generator sets were determined according to the MIC correlation model.

[0090] After preprocessing, the key state parameters are used as input values ​​for the MIC formula to obtain relevant results.

[0091] According to MIC-related modes:

[0092] Strong correlation: MIC ≥ 0.7, directly included in the core library;

[0093] Moderate correlation: 0.4 ≤ MIC < 0.7, retained based on expert experience;

[0094] Weak correlation: MIC < 0.4, discard;

[0095] A parameter is considered a critical status parameter when the MIC value between the parameter value and the unit status is greater than 0.4.

[0096] Furthermore, based on the database of key operating status parameters of thermal power generating units, an evaluation standard for key operating status parameters of thermal power generating units is constructed, specifically including:

[0097] The scoring criteria for key state parameters are derived from the obtained parameter state database through subjective and objective weighting principles.

[0098] Based on the principle of objective weighting, the unit fault records are obtained. The unit fault records are defined according to the severity of the faults as Level 1 faults (faults that directly threaten personal safety, cause serious damage to main equipment, or cause systemic shutdowns), Level 2 faults (faults that directly threaten personal safety, cause serious damage to main equipment, or cause systemic shutdowns), Level 3 faults (potential faults that do not affect the safe operation of the unit but require planned maintenance), and Level 4 faults (non-critical anomalies that can be eliminated through routine maintenance and do not affect current operation). The key status parameter data of the thermal power generating unit corresponding to different unit fault records are also obtained.

[0099] After analyzing and evaluating the unit fault records and processing the data, the correlation and contribution value between the operating status of the thermal power generating unit and each parameter are established. The key operating status parameters of the thermal power generating unit that affect the equipment efficiency are screened out, and the weight of the key status parameters is determined according to the MIC value.

[0100] A scoring standard with subjective weighting parameters was obtained based on the experience evaluation of skilled workers at different levels.

[0101] Based on the principle of objective weighting, the comparative strength (standard deviation) and conflict (negative correlation with other variables) of comprehensive variables are obtained to generate objective weighting parameter scoring criteria;

[0102] The scoring criteria for thermal power generating unit parameters are derived by using objective weighted parameter scoring standards and subjective weighted parameter scoring standards. The weights are generated through model coefficients and then manually fine-tuned according to business needs.

[0103] Specifically, real-time operating data of thermal power generating units is acquired, and combined with a database of key operating status parameters and health baseline intervals of thermal power generating units, a knowledge graph of key operating status parameters of thermal power generating units is constructed, including:

[0104] S1.1. Define knowledge graph node A, where knowledge graph node A is the real-time operating data of thermal power generating units;

[0105] S1.2. Define knowledge graph node B, which is a database of key operating status parameters of thermal power generating units;

[0106] S1.3. Define knowledge graph node C, where knowledge graph node C is the health baseline interval;

[0107] S1.4. Conduct auxiliary re-examination of key operating status parameters of thermal power generating units in a sub-healthy state and write them into the knowledge graph;

[0108] S1.5. Define the mapping relationship AB between knowledge graph node A and knowledge graph node B. The mapping relationship AB includes the first mapping relationship AB1 and the second mapping relationship AB2. The first mapping relationship AB is the time difference between the real-time operation data acquisition timestamp of the thermal power generating unit and the acquisition timestamp of the thermal power generating unit operation key status parameter database. The second mapping relationship AB2 is the ratio of the number of parameter types in the thermal power generating unit operation key status parameter database to the number of parameter types in the thermal power generating unit real-time operation data.

[0109] S1.6. Define the mapping relationship AC between knowledge graph node A and knowledge graph node C. The mapping relationship AC is the deviation of the parameter in the real-time operation data of the thermal power generating unit from the average value of the two endpoints of the health benchmark interval.

[0110] S1.7. Define the mapping relationship BC between knowledge graph node B and knowledge graph node C. The mapping relationship BC is the deviation of the average value of the parameters in the database of key operating status parameters of thermal power generating units from the values ​​at both ends of the health benchmark interval.

[0111] Specifically, the initial index of the real-time operating status of the thermal power generating unit is determined, and the status of the thermal power generating unit is simultaneously evaluated based on the initial index of the real-time operating status and the health benchmark interval, resulting in a response strategy for the operating status of the thermal power generating unit, including:

[0112] Based on the knowledge graph of key operating status parameters of thermal power generating units, the real-time operating data of thermal power generating units, the mapping relationship AB between knowledge graph node A and knowledge graph node B, and the mapping relationship AC between knowledge graph node A and knowledge graph node C are obtained.

[0113] Extract the time difference between the real-time operation data acquisition timestamp of the thermal power generating unit and the acquisition timestamp of the key operating status parameter database of the thermal power generating unit from the mapping relationship AB between knowledge graph node A and knowledge graph node B. Simultaneously extract the ratio of the number of parameter types in the key operating status parameter database of the thermal power generating unit to the number of parameter types in the real-time operation data of the thermal power generating unit.

[0114] Extract the deviation values ​​of parameters from the average values ​​of the two endpoints of the health baseline interval in the real-time operation data of thermal power generating units from the mapping relationship AC between knowledge graph nodes A and C.

[0115] Based on the real-time operation data of thermal power generating units, and combined with the time difference between the real-time operation data collection timestamp of the thermal power generating units and the collection timestamp of the database of key operating status parameters of thermal power generating units, as well as the deviation of the parameters in the real-time operation data of thermal power generating units from the average value of the two endpoints of the health benchmark interval, the initial index of the real-time operating status of thermal power generating units is determined.

[0116] The status of thermal power generating units is evaluated based on the initial index of real-time operating status and the health benchmark interval, and strategies for dealing with the operating status of thermal power generating units are obtained.

[0117] The formula for calculating the initial index of the real-time operating status of the thermal power generating unit is as follows:

[0118] ;

[0119] In the formula, For thermal power generating units in The initial index of the real-time operating status of the thermal power generating unit at any given moment. For thermal power generating units in The quantized value of the timestamp of the real-time running data collection. Quantified values ​​of timestamps collected for the database of key operating status parameters of thermal power generating units. For thermal power generating units in The first real-time running data at time 1 One running parameter, This is the quantized value of the left endpoint of the health baseline interval. This is the quantized value of the right endpoint of the health baseline interval. The number of parameter types in the database of key operating status parameters for thermal power generating units. In order to be in The number of parameter types in the real-time operating data of thermal power generating units at any given moment.

[0120] Based on preset mode: when An emergency shutdown was initiated and a maintenance plan was developed.

[0121] when Shorten the monitoring cycle;

[0122] when Routine inspections.

[0123] Specifically, the auxiliary re-inspection process for the status of thermal power generating units includes:

[0124] The original key status auxiliary parameters of the thermal power generating unit are obtained by collecting data from sensors and processing the initial index of the real-time operating status of the thermal power generating unit.

[0125] The key auxiliary parameters of the original thermal power generating unit are preprocessed, and the data of key auxiliary parameters of the thermal power generating unit are obtained by standard deviation and normalization.

[0126] Based on the analytic hierarchy process (AHP) to decompose a complex index system and the MIC analysis to determine the weight of the deviation of key auxiliary parameters of thermal power generating units on the impact of unit failures, the objective contribution rate of auxiliary parameters is obtained.

[0127] The key status auxiliary parameters of thermal power generating units are evaluated based on expert experience weights, and the subjective contribution rate of the auxiliary parameters is obtained using a regression model.

[0128] Based on the objective contribution rate and subjective contribution rate of auxiliary parameters, weights are generated through model coefficients, and the auxiliary parameter contribution rate evaluation criteria are obtained by manual fine-tuning in combination with business needs.

[0129] Collect historical data of key auxiliary parameters of thermal power generating units, perform deviation calculations or similarity measurements to assess the degree of deviation, and set standard deviation intervals, including abnormal intervals, normal intervals, and undetermined intervals.

[0130] Real-time data of key auxiliary parameters of thermal power generating units are obtained, and a knowledge graph of key auxiliary parameters of thermal power generating units is constructed by combining the evaluation criteria for contribution rate of auxiliary parameters and the standard deviation range.

[0131] By embedding auxiliary review into the knowledge graph, a multi-level knowledge graph can be obtained.

[0132] Furthermore, the evaluation criteria for the contribution rate of key auxiliary parameters of thermal power generating units specifically include:

[0133] The average weight of the auxiliary parameters is obtained by comprehensively scoring the key status auxiliary parameters of thermal power generating units by multiple groups of experts. The subjective contribution rate of the key status auxiliary parameters of thermal power generating units is obtained by processing the average weight of the auxiliary parameters through a regression model.

[0134] The calculation equation used for the expert experience weight is as follows:

[0135] ;

[0136] In the formula, Subjective contribution rate of key auxiliary parameters for thermal power generating units. These are the various auxiliary parameters. This is the average of the weights of each auxiliary parameter. It is the regression coefficient.

[0137] Based on key state auxiliary parameters, the objective contribution rate of key state auxiliary parameters of thermal power generating units is obtained through a random forest machine learning model, specifically including:

[0138] Based on historical data of key auxiliary parameters of thermal power generating units, a random forest classifier is constructed to calculate the importance score of key auxiliary parameters of thermal power generating units. The objective contribution rate of key auxiliary parameters of thermal power generating units is obtained through the importance score of key auxiliary parameters of thermal power generating units.

[0139] The subjective contribution rate of key auxiliary parameters of thermal power generating units is obtained by generating weights through model coefficients and manually fine-tuning them according to business needs.

[0140] Furthermore, a power plant unit condition assessment system based on knowledge graphs is proposed, including:

[0141] The data acquisition module includes a multi-source data synchronous acquisition unit, a data quality control unit, and an edge computing and cloud-edge collaboration unit. These units are located at selected locations on the generator set and acquire, store, and synchronize real-time status parameters through electrical connections with other units.

[0142] The data processing module, based on the acquired, stored, and synchronized real-time status parameters, obtains a database of key operating status parameters of thermal power generating units through data noise reduction, selection, and normalization processing.

[0143] The status assessment module obtains evaluation results based on the key operating status parameters of thermal power generating units and the evaluation standards for key operating status parameters of thermal power generating units;

[0144] The re-examination module processes the data parameters that are in a sub-healthy state again.

[0145] The display module shows the unit status in a time series and provides feedback on real-time status, periodic status, and long-term forecast status.

[0146] The status of the thermal power generating unit, as reflected by the status parameter data, is transformed into a real-time control diagram according to the results obtained from fixed time periods, real-time monitoring, and random monitoring. Based on the visualization module, the status of the thermal power generating unit can be visualized.

[0147] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for assessing the condition of power plant units based on knowledge graphs, characterized in that, Includes the following steps: Collect and process historical operating data of thermal power generating units, and use the MIC correlation analysis method to screen out key operating status parameters of thermal power generating units from the historical operating data; Noise reduction and discrete regression processing are performed on key operating status parameters of thermal power generating units, and a database of key operating status parameters of thermal power generating units is constructed simultaneously. Based on the database of key operating status parameters of thermal power generating units, an evaluation standard for key operating status parameters of thermal power generating units is constructed. Based on the evaluation standards for key operating status parameters of thermal power generating units, a health benchmark interval is determined, which includes a healthy interval, a sub-healthy interval, and an abnormal interval. Obtain real-time operating data of thermal power generating units, and combine the database of key operating status parameters of thermal power generating units with the health benchmark interval to construct a knowledge graph of key operating status parameters of thermal power generating units; The initial index of the real-time operating status of the thermal power generating unit is determined, and the status of the thermal power generating unit is evaluated simultaneously based on the initial index of the real-time operating status and the health benchmark interval. The resulting response strategy for the operating status of the thermal power generating unit includes routine inspections, shortening the monitoring cycle, and developing maintenance plans and emergency shutdowns. The key operating status parameters of thermal power generating units are extracted sequentially from the database of key operating status parameters of thermal power generating units. The key operating status parameters of thermal power generating units are evaluated by the analytic hierarchy process to obtain auxiliary parameters of key status of thermal power generating units, and auxiliary re-inspection of the status of thermal power generating units is carried out. Specifically, determining the initial index of the real-time operating status of the thermal power generating unit, and simultaneously evaluating the status of the thermal power generating unit based on the initial index of the real-time operating status and the health benchmark interval, to obtain a response strategy for the operating status of the thermal power generating unit, includes: Based on the knowledge graph of key operating status parameters of thermal power generating units, the real-time operating data of thermal power generating units, the mapping relationship AB between knowledge graph node A and knowledge graph node B, and the mapping relationship AC between knowledge graph node A and knowledge graph node C are obtained. Extract the time difference between the real-time operation data acquisition timestamp of the thermal power generating unit and the acquisition timestamp of the key operating status parameter database of the thermal power generating unit from the mapping relationship AB between knowledge graph node A and knowledge graph node B. Simultaneously extract the ratio of the number of parameter types in the key operating status parameter database of the thermal power generating unit to the number of parameter types in the real-time operation data of the thermal power generating unit. Extract the deviation values ​​of parameters from the average values ​​of the two endpoints of the health baseline interval in the real-time operation data of thermal power generating units from the mapping relationship AC between knowledge graph nodes A and C. Based on the real-time operation data of thermal power generating units, and combined with the time difference between the real-time operation data collection timestamp of the thermal power generating units and the collection timestamp of the database of key operating status parameters of thermal power generating units, as well as the deviation of the parameters in the real-time operation data of thermal power generating units from the average value of the two endpoints of the health benchmark interval, the initial index of the real-time operating status of thermal power generating units is determined. The status of thermal power generating units is evaluated based on the initial index of real-time operating status and the health benchmark interval, and strategies for dealing with the operating status of thermal power generating units are obtained. The formula for calculating the initial index of the real-time operating status of the thermal power generating unit is as follows: ； In the formula, For thermal power generating units in The initial index of the real-time operating status of the thermal power generating unit at any given moment. For thermal power generating units in The quantized value of the timestamp of the real-time running data collection. Quantified values ​​of timestamps collected for the database of key operating status parameters of thermal power generating units. For thermal power generating units in The first real-time running data at time 1 One running parameter, This is the quantized value of the left endpoint of the health baseline interval. This is the quantized value of the right endpoint of the health baseline interval. The number of parameter types in the database of key operating status parameters for thermal power generating units. In order to be in The number of parameter types in the real-time operating data of thermal power generating units at any given moment.

2. The power plant unit condition assessment method based on knowledge graphs according to claim 1, characterized in that, The process of collecting and processing historical operating data of thermal power generating units involves using the MIC correlation analysis method to filter out key operating status parameters of the thermal power generating units from the historical operating data, specifically including: i operating status parameters of thermal power generating units are obtained after preprocessing based on historical operating data of thermal power generating units. Obtain historical status data of thermal power generating units, and extract j historical status indices of thermal power generating units from the historical status data of thermal power generating units. Using the state parameters of i thermal power generating units and the historical state indices of j thermal power generating units as input values ​​for the MIC formula, the state parameter X of the thermal power generating units is measured. i and the historical state index Y of j thermal power generating units j The relevant indices are used to select key operating status parameters of thermal power generating units from historical operating data of thermal power generating units; The expression for the MIC formula is: ； In the formula, X represents the state parameter of the thermal power generating unit. i Historical state parameters Y of thermal power generating units i The relevant index, State parameter X of thermal power generating unit i Historical state parameters Y of thermal power generating units i The degree of distribution dependence, As a normalization factor, the original MIC values ​​are scaled proportionally to eliminate dimensional differences, making it easier to focus on key state parameters. and the status of thermal power generating units Unified unit.

3. The power plant unit condition assessment method based on knowledge graphs according to claim 1, characterized in that, The evaluation criteria for key operating status parameters of thermal power generating units, based on the database of key operating status parameters, specifically include: By applying the principle of subjective empowerment, the key operating status parameters of thermal power generating units are scored using the experience evaluation of skilled workers at different levels, and a regression model is used to derive the weighted scoring standard for the key operating status parameters of thermal power generating units. Based on the principles of objective and subjective weighting, weights are generated through model coefficients, and then manually fine-tuned according to business needs to obtain the evaluation standards for key operating status parameters of thermal power generating units.

4. The power plant unit condition assessment method based on knowledge graphs according to claim 1, characterized in that, The determination of the health benchmark range based on the evaluation standards for key operating status parameters of thermal power generating units specifically includes: Based on historical operating data of thermal power generating units, unit fault records are obtained, including Level 1, Level 2, Level 3, and Level 4 faults. Based on these fault records, key operating status parameter fault data of thermal power generating units are extracted through machine learning. Based on the evaluation criteria for key operating status parameters of thermal power generating units, a healthy baseline interval is generated using a normal cloud model.

5. The power plant unit condition assessment method based on knowledge graphs according to claim 1, characterized in that, The process of acquiring real-time operating data of thermal power generating units, combining it with a database of key operating status parameters of thermal power generating units and health benchmark intervals, and constructing a knowledge graph of key operating status parameters of thermal power generating units specifically includes: S1.

1. Define knowledge graph node A, where knowledge graph node A is real-time operating data of thermal power generating units; S1.

2. Define knowledge graph node B, which is a database of key operating status parameters of thermal power generating units; S1.

3. Define knowledge graph node C, where knowledge graph node C is the health baseline interval; S1.

4. Conduct auxiliary re-examination of key operating status parameters of thermal power generating units in a sub-healthy state and write them into the knowledge graph; S1.

5. Define the mapping relationship AB between knowledge graph node A and knowledge graph node B. The mapping relationship AB includes a first mapping relationship AB1 and a second mapping relationship AB2. The first mapping relationship AB1 is the time difference between the real-time operation data acquisition timestamp of the thermal power generator set and the acquisition timestamp of the key operation status parameter database of the thermal power generator set. The second mapping relationship AB2 is the ratio of the number of parameter types in the key operation status parameter database of the thermal power generator set to the number of parameter types in the real-time operation data of the thermal power generator set. S1.

6. Define the mapping relationship AC between knowledge graph node A and knowledge graph node C, wherein the mapping relationship AC is the deviation of the parameter in the real-time operation data of the thermal power generating unit from the average value of the two endpoints of the health benchmark interval; S1.

7. Define the mapping relationship BC between knowledge graph node B and knowledge graph node C. The mapping relationship BC is the deviation of the average value of the parameters in the database of key operating status parameters of thermal power generating units from the values ​​at both ends of the health benchmark interval.

6. The power plant unit condition assessment method based on knowledge graphs according to claim 1, characterized in that, The auxiliary re-inspection of the thermal power generating unit status specifically includes: The original key status auxiliary parameters of the thermal power generating unit are obtained by collecting data from sensors and processing the initial index of the real-time operating status of the thermal power generating unit. The original key status auxiliary parameters of the thermal power generating unit are preprocessed, and the key status auxiliary parameter data of the thermal power generating unit are obtained by standard deviation and normalization. Based on the analytic hierarchy process (AHP) to decompose a complex index system and the MIC analysis to determine the weight of the deviation of key auxiliary parameters of thermal power generating units on the impact of unit failures, the objective contribution rate of auxiliary parameters is obtained. The key status auxiliary parameters of thermal power generating units are evaluated based on expert experience weights, and the subjective contribution rate of the auxiliary parameters is obtained using a regression model. Based on the objective contribution rate and subjective contribution rate of the auxiliary parameters, weights are generated through model coefficients, and the auxiliary parameter contribution rate evaluation standard is obtained by manual fine-tuning in combination with business needs. Historical data of key auxiliary parameters of thermal power generating units are collected, and deviation calculation or similarity measurement is performed to assess the degree of deviation. Standard deviation intervals are set, including abnormal intervals, normal intervals, and undetermined intervals. Real-time data of key auxiliary parameters of thermal power generating units are obtained, and a knowledge graph of key auxiliary parameters of thermal power generating units is constructed by combining the evaluation criteria for contribution rate of auxiliary parameters and the standard deviation range. The auxiliary review is embedded into a knowledge graph to obtain a multi-level knowledge graph.

7. The power plant unit condition assessment method based on knowledge graphs according to claim 6, characterized in that, The evaluation criteria for the contribution rate of the auxiliary parameters specifically include: The average weight of the auxiliary parameters is obtained by comprehensively scoring the key status auxiliary parameters of thermal power generating units by multiple groups of experts. The subjective contribution rate of the key status auxiliary parameters of thermal power generating units is obtained by performing regression model processing on the average weight of the auxiliary parameters. The calculation equation used for the expert experience weight is as follows: ； In the formula, Subjective contribution rate of key auxiliary parameters for thermal power generating units. k These are the various auxiliary parameters. This is the average of the weights of each auxiliary parameter. It is the regression coefficient; Based on the aforementioned key state auxiliary parameters, the objective contribution rate of the key state auxiliary parameters of thermal power generating units is obtained through a random forest machine learning model, specifically including: Based on historical data of key auxiliary parameters of thermal power generating units, a random forest classifier is constructed to calculate the importance score of key auxiliary parameters of thermal power generating units. The objective contribution rate of key auxiliary parameters of thermal power generating units is obtained through the importance score of key auxiliary parameters of thermal power generating units. Weights are generated by model coefficients, and auxiliary parameter contribution rate evaluation criteria are obtained by manual fine-tuning based on business needs.

8. A knowledge graph-based power plant unit condition assessment system, used to implement the knowledge graph-based power plant unit condition assessment method as described in any one of claims 1-7, characterized in that, include: The data acquisition module includes a multi-source data synchronous acquisition unit, a data quality control unit, and an edge computing and cloud-edge collaboration unit. These units are distributed at selected locations on the generator set and acquire, store, and synchronize real-time status parameters through electrical connections with other units. The data processing module, based on the acquired, stored, and synchronized real-time status parameters, obtains a database of key operating status parameters of thermal power generating units through data noise reduction, selection, and normalization processing. The status assessment module obtains evaluation results based on the key operating status parameters of thermal power generating units and the evaluation standards for key operating status parameters of thermal power generating units; The re-examination module reprocesses the data parameters of the sub-healthy state. The display module shows the unit status in a time series and provides feedback on real-time status, periodic status, and long-term forecast status.

9. A computer-readable storage medium having a computer-readable program stored thereon, characterized in that, When the computer-readable program is invoked, it executes the knowledge graph-based thermal power generator state assessment method as described in any one of claims 1-7.

Images