Disclosure of Invention
The present invention has been made in view of the above-described problems.
Therefore, the technical problem solved by the invention is that the existing power plant data management method has the problems of incomplete data collection, low data processing efficiency, untimely abnormality detection and lack of scientificity in maintenance management, and the optimization problem of how to realize real-time monitoring of the running state of the power plant and optimization maintenance management.
In order to solve the technical problems, the invention provides a power plant data management method based on multi-source data, which comprises the following steps:
arranging a multi-source sensor to collect power plant data, transmitting the power plant data through an edge computing node, and preprocessing the data;
evaluating the current running state of the power plant by utilizing the preprocessed data, and identifying the change of the power plant data;
and detecting the equipment state according to the identification result, and generating a maintenance plan.
As a preferable scheme of the multi-source data-based power plant data management method, the power plant data comprises temperature data, pressure data, flow data, power output data, equipment log data and environment data;
the data preprocessing comprises the steps of cleaning data at an edge node, removing repeated, erroneous and missing data, and performing data standardization processing.
The method for managing the power plant data based on the multi-source data comprises the following steps of calculating the preprocessed data through a first-level judging model, a second-level judging model, a third-level judging model and a fourth-level judging model, and outputting a prediction result;
The first level judgment model is expressed as,
,
Wherein,Which represents a parameter of the thermodynamic state,The data of the temperature is represented and,Representing pressure data;
The secondary judgment model is expressed as,
,
Wherein,Which is indicative of a fluid dynamic state parameter,Representing flow data;
The three-level judgment model is expressed as,
,
Wherein,Representing the power output state parameter,Representing power output data;
The four-level judgment model is represented as,,
Wherein Env denotes an environmental impact status parameter,The data representing the temperature of the environment is displayed,Representing reference temperature data;
If it isAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndAnd (3) the mark is 1, otherwise, the mark is 0;
Classifying the power plant data change into a normal state, a thermodynamic parameter abnormality, a hydrodynamic parameter abnormality, an electric power output parameter abnormality, an environmental impact parameter abnormality and a power plant global fault according to the marking result;
The method comprises the steps of judging a normal state if four marked results are 1, judging according to the abnormal condition corresponding to the marked 0 if only one of the four marked results is 0, determining the marked four marked results as single abnormal, determining multiple abnormal if the number of marked 0 in the four marked results is more than or equal to 2 and less than or equal to 3, and judging the overall fault of the power plant if the four marked results are 0.
The method for managing the power plant data based on the multi-source data comprises the following steps of extracting a judging model output corresponding to single abnormality or multiple abnormalities, detecting the equipment state by using a random forest model, and detecting and diagnosing the equipment state by using the random forest model;
performing data cleaning, standardization and normalization on the extracted parameter values;
Inputting the parameter values extracted in real time into a trained random forest model, and predicting an output result;
if a single anomaly is detected, carrying out data cleaning, standardization and normalization on the extracted anomaly parameter data, carrying out multiple regression analysis, collecting historical data and real-time data of anomaly parameters and single anomaly related parameters, establishing a multiple regression model, carrying out regression analysis by taking the anomaly parameters as dependent variables and the single anomaly related parameters as independent variables, determining the influence degree of the related parameters on the anomaly parameters, identifying anomaly reasons and anomaly positions, carrying out time sequence analysis, collecting historical time sequence data of the anomaly parameters, establishing a time sequence model, analyzing the change trend of the anomaly parameters in the historical data, and predicting the change situation of the future anomaly parameters;
The method comprises the steps of detecting multiple anomalies, carrying out data cleaning, standardization and normalization on the extracted multiple anomaly parameter data, carrying out implicit relation analysis, collecting historical data and real-time data of multiple anomaly related parameters, establishing a multiple regression model, carrying out regression analysis by taking the multiple anomaly parameters as independent variables, determining the relation among related parameters, identifying possible core anomalies, carrying out clustering analysis on the multiple anomaly parameters by using a regression coefficient and residual error obtained by the multiple regression analysis, identifying an anomaly mode, judging whether hidden relations exist or not, carrying out causal analysis on each anomaly parameter if an analysis result shows that one core anomaly parameter causes other parameter anomalies, transferring the other anomaly parameter anomalies into a single anomaly processing flow, carrying out causal analysis on each anomaly parameter if multiple independent anomalies exist, determining the causal relation between each anomaly parameter and the multiple anomaly related parameters, identifying an anomaly reason, combining the causal analysis result, determining the specific position and the influence range of each anomaly, carrying out time sequence analysis on each independent anomaly parameter, collecting historical time sequence data, establishing a time sequence model, carrying out the change of each parameter analysis, and finally carrying out the analysis on the future change of each anomaly state, and the multiple anomaly state, and carrying out the multiple-related anomaly state analysis, and the multiple-state analysis.
As a preferable mode of the multi-source data-based power plant data management method of the invention, wherein the multiple regression model is expressed as,
,
Wherein,Represent the firstThe number of the abnormal parameters is set to be equal,,Represent the firstA plurality of the abnormality related parameters,,Represent the firstThe intercept of the individual anomaly parameter,Represent the firstAbnormal parameter(s)Regression coefficients for the multiple anomaly-related parameters,Represent the firstRegression error terms for the individual anomaly parameters;
The cluster analysis is expressed as,
,
Wherein,Represent the firstThe cluster to which the individual outlier parameters belong,,Indicating an indication function whenThe value is 1 when, and 0 otherwise,Represent the firstAbnormal parameter(s)The square of the euclidean distance between cluster centers,Represent the firstThe number of cluster centers is set up,;
Calculating a cluster center meanAnd the standard deviation of the clustering centerIf (if)Judging that other parameters are abnormal due to the abnormal core parameters, determining the abnormal parameters in the clusters as abnormal core parameters, classifying the abnormal core parameters as single abnormal, ifJudging that there are multiple independent anomalies, ifOr (b)The strength of the association between each anomaly parameter and multiple anomaly related parameters, expressed as,
,
Wherein,Represent the firstAbnormal parameter(s)Correlation strength of multiple anomaly related parameters;
regarding various abnormal parameters as network nodes and association strengthRegarding the weight of the edge, constructing a correlation network, analyzing the network structure, calculating degree centrality and betweenness centrality, expressed as,
,
,
Wherein,The degree of centrality is expressed as,The center of the medium number is represented by,Representing nodesTo the nodeIs used for the number of the shortest paths of the (a),Representing passing nodesIs the shortest path number of (a);
If it isOr (b)Judging that other parameters are abnormal due to the abnormal core parameters, determining the abnormal parameters in the clusters as abnormal core parameters, classifying the abnormal core parameters as single abnormal, ifOr (b)It is determined that there are a plurality of independent anomalies, wherein,Representing the mean of all node degree centrality,The standard deviation representing the centrality of all node degrees,Representing the mean of the centrality of all node bets,The standard deviation of the centrality of all node bets is represented.
The method comprises the steps of extracting single abnormality type, single abnormality position, single abnormality cause, future single abnormality change condition and single abnormality influence range from a single abnormality equipment state diagnosis report, analyzing the extracted single abnormality type, single abnormality position, single abnormality cause, future single abnormality change condition and single abnormality influence range by using a particle swarm optimization algorithm, optimizing configuration of maintenance resources, generating a preliminary single maintenance plan, wherein the preliminary single maintenance plan comprises equipment maintenance time, required spare parts and manpower resource configuration, performing time sequence analysis by using an ARIMA model based on the preliminary single maintenance plan and the extracted single abnormality parameters, predicting change trend of the future single abnormality parameters, evaluating influence of different maintenance schemes on the future operation state of the equipment according to a prediction result, selecting an optimal maintenance plan, combining historical maintenance data and real-time monitoring data, adjusting the maintenance plan, making maintenance measures, time arrangement, required resources and technical support.
The method for managing the power plant data based on the multi-source data comprises the steps of generating a maintenance plan, extracting multiple abnormality types, multiple abnormality positions, multiple abnormality reasons, correlation analysis results, future multiple abnormality change conditions and overall influences from multiple abnormality equipment state diagnosis reports, analyzing the extracted multiple abnormality types, the multiple abnormality positions, the multiple abnormality reasons, the correlation analysis results, the future multiple abnormality change conditions and the overall influences by using a genetic algorithm, optimizing maintenance resource configurations of a plurality of equipment and components, generating a preliminary multiple maintenance plan, wherein the preliminary multiple maintenance plan comprises maintenance time of the equipment, required spare parts and manpower resource configurations, performing time series analysis by using an LSTM model based on the preliminary multiple maintenance plan and the extracted multiple abnormality parameters, predicting the change trend of the future multiple abnormality parameters, evaluating the effects of different comprehensive maintenance plans according to the prediction results, selecting an optimal maintenance plan, combining historical maintenance data and real-time monitoring data, and determining specific maintenance measures, time schedules, required resources and technical supports.
Another object of the present invention is to provide a power plant data management system based on multi-source data, which can solve the problems of incomplete data collection, low data processing efficiency and untimely anomaly detection in the existing power plant data management by constructing the power plant data management system based on multi-source data.
The power plant data management system based on the multi-source data comprises an output acquisition module, a state identification module and a power plant maintenance module, wherein the output acquisition module is used for arranging a multi-source sensor to collect power plant data, transmitting the power plant data through an edge computing node and preprocessing the data, the state identification module is used for evaluating the current running state of the power plant and identifying the change of the power plant data by utilizing the preprocessed data, and the power plant maintenance module is used for detecting the equipment state according to the identification result to generate a maintenance plan.
A computer device comprising a memory storing a computer program and a processor implementing the steps of the multi-source data based power plant data management method as described above when the computer program is executed.
A computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of a plant data management method based on multi-source data as described above.
The multi-source data-based power plant data management method has the beneficial effects that the comprehensive and real-time monitoring of the running state of the power plant is realized by integrating the multi-source data. Compared with the traditional method, the method improves the real-time performance and accuracy of data processing and reduces the data transmission delay. The operation state of the power plant is evaluated, abnormal changes can be timely identified, and equipment state detection and fault diagnosis are provided. And generating an abnormality diagnosis report, optimizing maintenance resource allocation, improving maintenance efficiency and effect and reducing maintenance cost. Predictive maintenance is realized, potential faults are prevented, and safety and reliability of operation of the power plant are improved.
Detailed Description
So that the manner in which the above recited objects, features and advantages of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings. All other embodiments, which can be made by one of ordinary skill in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.
Embodiment 1, referring to fig. 1, for an embodiment of the present invention, there is provided a power plant data management method based on multi-source data, including:
arranging a multi-source sensor to collect power plant data, transmitting the power plant data through an edge computing node, and preprocessing the data;
evaluating the current running state of the power plant by utilizing the preprocessed data, and identifying the change of the power plant data;
and detecting the equipment state according to the identification result, and generating a maintenance plan.
The power plant data includes temperature data, pressure data, flow data, power output data, equipment log data, and environmental data;
the data preprocessing comprises the steps of cleaning data at an edge node, removing repeated, erroneous and missing data, and performing data standardization processing.
The step of identifying the power plant data change comprises the steps of calculating the preprocessed data through a first-level judging model, a second-level judging model, a third-level judging model and a fourth-level judging model, and outputting a prediction result;
The first level judgment model is expressed as,
,
Wherein,Which represents a parameter of the thermodynamic state,The data of the temperature is represented and,Representing pressure data;
The secondary judgment model is expressed as,
,
Wherein,Which is indicative of a fluid dynamic state parameter,Representing flow data;
The three-level judgment model is expressed as,
,
Wherein,Representing the power output state parameter,Representing power output data;
The four-level judgment model is represented as,,
Wherein Env denotes an environmental impact status parameter,The data representing the temperature of the environment is displayed,Representing reference temperature data;
In order to enhance the stability and robustness of the model while maintaining the data characteristics, the square root transformation coefficient was found to be 0.5 by analyzing the history data. In order to find a balance point, the power output data can reflect the actual situation, cannot excessively fluctuate, the stability of the model is guaranteed, and the power transformation coefficient is obtained by analyzing the historical data and is 0.3 th order.
If it isAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndBetween, marked 1, otherwise marked 0, ifAt the threshold valueAndAnd (3) the mark is 1, otherwise, the mark is 0;
Classifying the power plant data change into a normal state, a thermodynamic parameter abnormality, a hydrodynamic parameter abnormality, an electric power output parameter abnormality, an environmental impact parameter abnormality and a power plant global fault according to the marking result;
If the monitoring data are classified as normal states, continuously monitoring all parameters, ensuring that all parameters are always in normal ranges, and recording all monitoring data in a database in real time for subsequent analysis and reference.
If the power plant global fault is classified, immediately giving out fault alarm, notifying related maintenance personnel and management personnel, starting a shutdown protection program according to the fault type and severity, avoiding further damage and accident expansion of equipment, and simultaneously ensuring the safety and stability of non-fault parts by switching and isolating fault equipment.
The method comprises the steps of judging a normal state if four marked results are 1, judging according to the abnormal condition corresponding to the marked 0 if only one of the four marked results is 0, determining the marked four marked results as single abnormal, determining multiple abnormal if the number of marked 0 in the four marked results is more than or equal to 2 and less than or equal to 3, and judging the overall fault of the power plant if the four marked results are 0.
The method comprises the steps of extracting judging model output corresponding to single abnormality or multiple abnormalities, detecting the equipment state by using a random forest model, detecting and diagnosing the equipment state by using the random forest model, performing data cleaning, standardization and normalization processing on extracted parameter values, training the random forest model by using historical data in a normal running state, and setting the number of trees to be 100 and the depth to be 10. The random forest learns the characteristic mode in the normal state by constructing a plurality of decision trees and integrating the results. Inputting the parameter values extracted in real time into a trained random forest model, and predicting an output result;
if a single anomaly is detected, carrying out data cleaning, standardization and normalization on the extracted anomaly parameter data, carrying out multiple regression analysis, collecting historical data and real-time data of anomaly parameters and single anomaly related parameters, establishing a multiple regression model, carrying out regression analysis by taking the anomaly parameters as dependent variables and the single anomaly related parameters as independent variables, determining the influence degree of the related parameters on the anomaly parameters, identifying anomaly reasons and anomaly positions, carrying out time sequence analysis, collecting historical time sequence data of the anomaly parameters, establishing a time sequence model, analyzing the change trend of the anomaly parameters in the historical data, and predicting the change situation of the future anomaly parameters;
The method comprises the steps of detecting multiple anomalies, carrying out data cleaning, standardization and normalization on the extracted multiple anomaly parameter data, carrying out implicit relation analysis, collecting historical data and real-time data of multiple anomaly related parameters, establishing a multiple regression model, carrying out regression analysis by taking the multiple anomaly parameters as independent variables, determining the relation among related parameters, identifying possible core anomalies, carrying out clustering analysis on the multiple anomaly parameters by using a regression coefficient and residual error obtained by the multiple regression analysis, identifying an anomaly mode, judging whether hidden relations exist or not, carrying out causal analysis on each anomaly parameter if an analysis result shows that one core anomaly parameter causes other parameter anomalies, transferring the other anomaly parameter anomalies into a single anomaly processing flow, carrying out causal analysis on each anomaly parameter if multiple independent anomalies exist, determining the causal relation between each anomaly parameter and the multiple anomaly related parameters, identifying an anomaly reason, combining the causal analysis result, determining the specific position and the influence range of each anomaly, carrying out time sequence analysis on each independent anomaly parameter, collecting historical time sequence data, establishing a time sequence model, carrying out the change of each parameter analysis, and finally carrying out the analysis on the future change of each anomaly state, and the multiple anomaly state, and carrying out the multiple-related anomaly state analysis, and the multiple-state analysis.
The single anomaly related parameters include environmental factor data, plant operational parameters, historical maintenance records, and operational records.
The multiple anomaly related parameters include system integration parameters, interaction impact data, external environment data, system level operation records.
The multiple regression model is represented as,
,
Wherein,Represent the firstThe number of the abnormal parameters is set to be equal,,Represent the firstA plurality of the abnormality related parameters,,Represent the firstThe intercept of the individual anomaly parameter,Represent the firstAbnormal parameter(s)Regression coefficients for the multiple anomaly-related parameters,Represent the firstRegression error terms for the individual anomaly parameters;
The cluster analysis is expressed as,
,
Wherein,Represent the firstThe cluster to which the individual outlier parameters belong,,Indicating an indication function whenThe value is 1 when, and 0 otherwise,Represent the firstAbnormal parameter(s)The square of the euclidean distance between cluster centers,Represent the firstThe number of cluster centers is set up,;
Calculating regression coefficientsEstimated by the least squares method, denoted as,
,
Wherein,Is a design matrix of the device,Is a response vector.
Calculating residual errorsThe residual is the difference between the observed and predicted values, expressed as,
,
Computing cluster centerThe initial cluster center is randomly selected, iteratively updated, denoted as,
,
Wherein,Is the firstThe set of samples that the individual clusters contain,Is the firstThe number of samples in each cluster.
The euclidean distance, expressed as,
,
Calculating a cluster center meanAnd the standard deviation of the clustering centerIf (if)Judging that other parameters are abnormal due to the abnormal core parameters, determining the abnormal parameters in the clusters as abnormal core parameters, classifying the abnormal core parameters as single abnormal, ifJudging that there are multiple independent anomalies, ifOr (b)The strength of the association between each anomaly parameter and multiple anomaly related parameters, expressed as,
,
Wherein,Represent the firstAbnormal parameter(s)Correlation strength of multiple anomaly related parameters;
regarding various abnormal parameters as network nodes and association strengthRegarding the weight of the edge, constructing a correlation network, analyzing the network structure, calculating degree centrality and betweenness centrality, expressed as,
,
,
Wherein,The degree of centrality is expressed as,The center of the medium number is represented by,Representing nodesTo the nodeIs used for the number of the shortest paths of the (a),Representing passing nodesIs the shortest path number of (a);
If it isOr (b)Judging that other parameters are abnormal due to the abnormal core parameters, determining the abnormal parameters in the clusters as abnormal core parameters, classifying the abnormal core parameters as single abnormal, ifOr (b)It is determined that there are a plurality of independent anomalies, wherein,Representing the mean of all node degree centrality,The standard deviation representing the centrality of all node degrees,Representing the mean of the centrality of all node bets,The standard deviation of the centrality of all node bets is represented.
The method comprises the steps of extracting a single abnormality type, a single abnormality position, a single abnormality cause, a single future abnormality change condition and a single abnormality influence range from a single abnormality equipment state diagnosis report, analyzing the extracted single abnormality type, single abnormality position, single abnormality cause, single future abnormality change condition and single abnormality influence range by using a particle swarm optimization algorithm, optimizing the configuration of maintenance resources to generate a preliminary single maintenance plan, wherein the preliminary single maintenance plan comprises the maintenance time of equipment, required spare parts and manpower resource configuration, performing time series analysis by using an ARIMA model based on the preliminary single maintenance plan and the extracted single abnormality parameters, predicting the change trend of the single future abnormality parameters, evaluating the influence of different maintenance schemes on the future operation state of the equipment according to a prediction result, selecting an optimal maintenance scheme, combining historical maintenance data and real-time monitoring data, adjusting the maintenance plan, and formulating maintenance measures, time schedules, required resources and technical support.
The method comprises the steps of generating a maintenance plan, extracting multiple abnormality types, multiple abnormality positions, multiple abnormality reasons, correlation analysis results, future multiple abnormality change conditions and overall influences from multiple abnormal equipment state diagnosis reports, analyzing the extracted multiple abnormality types, multiple abnormality positions, multiple abnormality reasons, correlation analysis results, future multiple abnormality change conditions and overall influences by using a genetic algorithm, optimizing maintenance resource allocation of a plurality of equipment and components, generating a preliminary multiple maintenance plan, wherein the preliminary multiple maintenance plan comprises equipment maintenance time, required spare parts and manpower resource allocation, performing time sequence analysis by using an LSTM model based on the preliminary multiple maintenance plan and the extracted multiple abnormality parameters, predicting the change trend of the future multiple abnormality parameters, evaluating the effects of different comprehensive maintenance schemes according to the prediction results, determining the comprehensive maintenance plan by combining historical maintenance data and real-time monitoring data, and formulating specific maintenance measures, time schedule, required resources and technical support.
Embodiment 2 referring to fig. 2, for an embodiment of the present invention, there is provided a power plant data management system based on multi-source data, including:
The system comprises an output acquisition module, a state identification module and a power plant maintenance module;
the output acquisition module is used for arranging a multi-source sensor to collect power plant data, transmitting the power plant data through the edge computing node and preprocessing the data;
The state identification module is used for evaluating the current running state of the power plant by utilizing the preprocessed data and identifying the change of the power plant data;
And the power plant maintenance module is used for detecting the equipment state according to the identification result and generating a maintenance plan.
Example 3 an embodiment of the invention, which differs from the first two embodiments, is:
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. The storage medium includes a U disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes.
Logic and/or steps represented in the flowcharts or otherwise described herein, e.g., a ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium include an electrical connection (an electronic device) having one or more wires, a portable computer diskette (a magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). In addition, the computer readable medium may even be paper or other suitable medium on which the program is printed, as the program may be electronically captured, via, for instance, optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner, if necessary, and then stored in a computer memory.
It is to be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, the various steps or methods may be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, may be implemented using any one or combination of techniques known in the art, discrete logic circuits with logic gates for implementing logic functions on data signals, application specific integrated circuits with appropriate combinational logic gates, programmable Gate Arrays (PGAs), field Programmable Gate Arrays (FPGAs), and the like.
Embodiment 4 provides a power plant data management method based on multi-source data, and in order to verify the beneficial effects of the invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.
The experiment is conducted for comparison test of two sets of equipment management systems of a certain power plant, and the experiment period is 30 days continuously. Test data types include temperature, pressure, flow, power output, equipment log, and environmental data. The evaluation indexes comprise data collection coverage rate, data processing delay, abnormality detection timeliness, diagnosis accuracy rate, diagnosis time, daily maintenance cost and failure rate.
The traditional method adopts a single sensor and local automatic data collection, and performs data transmission and processing through a central server, wherein the data preprocessing comprises basic cleaning and standardization processing. The evaluation and anomaly detection of the equipment state adopts a rule judgment model, and the fault diagnosis uses a basic statistical analysis method. In terms of maintenance management, maintenance resource allocation lacks systematicness based on periodic maintenance and empirical adjustment.
The method of my invention performs data preprocessing by arranging multiple source sensors and computing nodes at the edges. And evaluating the running state of the power plant by using a multi-stage judging model, and identifying abnormal changes. And performing equipment state detection and fault diagnosis, and generating a scientific maintenance plan based on the diagnosis result and an optimization algorithm. The experimental results are shown in table 1.
Table 1 comparison of experimental results
,
The method adopts the multi-source sensor to collect the overall data, improves the coverage rate of the data, and avoids the defect that the single sensor and the local automatic data collection cannot be monitored overall. The introduction of edge compute nodes reduces the delay time of data processing so that the system can respond to data changes in more real time.
The rule judgment model in the traditional method is inferior to the multistage judgment model of the method in detection accuracy and real-time because of the limitation of the algorithm, so that the evaluation of the running state of the power plant is more accurate, and the abnormal detection is more timely.
Random forest models and multiple regression analysis provide higher diagnostic accuracy and faster diagnostic speeds. The traditional method adopts a basic statistical analysis method, and the diagnosis accuracy and the diagnosis speed are not equal to those of the algorithm adopted by the invention. Based on the generated maintenance plan, the maintenance cost is reduced, and the utilization rate of maintenance resources is improved. The traditional method is based on periodic maintenance and empirical adjustment, and lacks systematic resource allocation, so that resource waste and high maintenance cost are caused. The method optimizes the maintenance scheme through analysis and prediction, and realizes more efficient resource allocation.
It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.