CN116579768A - Power plant on-line instrument operation and maintenance management method and system - Google Patents
Power plant on-line instrument operation and maintenance management method and system Download PDFInfo
- Publication number
- CN116579768A CN116579768A CN202310851819.8A CN202310851819A CN116579768A CN 116579768 A CN116579768 A CN 116579768A CN 202310851819 A CN202310851819 A CN 202310851819A CN 116579768 A CN116579768 A CN 116579768A
- Authority
- CN
- China
- Prior art keywords
- state
- data
- instrument
- fault
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000012423 maintenance Methods 0.000 title claims abstract description 94
- 238000007726 management method Methods 0.000 title claims description 79
- 238000013210 evaluation model Methods 0.000 claims abstract description 45
- 238000000034 method Methods 0.000 claims abstract description 30
- 238000012098 association analyses Methods 0.000 claims abstract description 13
- 238000012549 training Methods 0.000 claims description 45
- 238000010276 construction Methods 0.000 claims description 41
- 238000012544 monitoring process Methods 0.000 claims description 26
- 238000003066 decision tree Methods 0.000 claims description 13
- 238000012795 verification Methods 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 10
- 238000007621 cluster analysis Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 4
- 238000012797 qualification Methods 0.000 claims description 3
- 238000010219 correlation analysis Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 12
- 238000004364 calculation method Methods 0.000 description 9
- 238000013528 artificial neural network Methods 0.000 description 7
- 238000004519 manufacturing process Methods 0.000 description 6
- 210000002569 neuron Anatomy 0.000 description 5
- DGAQECJNVWCQMB-PUAWFVPOSA-M Ilexoside XXIX Chemical compound C[C@@H]1CC[C@@]2(CC[C@@]3(C(=CC[C@H]4[C@]3(CC[C@@H]5[C@@]4(CC[C@@H](C5(C)C)OS(=O)(=O)[O-])C)C)[C@@H]2[C@]1(C)O)C)C(=O)O[C@H]6[C@@H]([C@H]([C@@H]([C@H](O6)CO)O)O)O.[Na+] DGAQECJNVWCQMB-PUAWFVPOSA-M 0.000 description 4
- 238000013527 convolutional neural network Methods 0.000 description 4
- 238000007405 data analysis Methods 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 239000011734 sodium Substances 0.000 description 4
- 229910052708 sodium Inorganic materials 0.000 description 4
- 238000012360 testing method Methods 0.000 description 3
- 230000003113 alkalizing effect Effects 0.000 description 2
- 239000003795 chemical substances by application Substances 0.000 description 2
- 239000003792 electrolyte Substances 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000010200 validation analysis Methods 0.000 description 2
- ZAMOUSCENKQFHK-UHFFFAOYSA-N Chlorine atom Chemical compound [Cl] ZAMOUSCENKQFHK-UHFFFAOYSA-N 0.000 description 1
- XUIMIQQOPSSXEZ-UHFFFAOYSA-N Silicon Chemical compound [Si] XUIMIQQOPSSXEZ-UHFFFAOYSA-N 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000000460 chlorine Substances 0.000 description 1
- 229910052801 chlorine Inorganic materials 0.000 description 1
- 238000005536 corrosion prevention Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 229910052710 silicon Inorganic materials 0.000 description 1
- 239000010703 silicon Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/20—Administration of product repair or maintenance
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
- G06Q10/103—Workflow collaboration or project management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Testing And Monitoring For Control Systems (AREA)
Abstract
The invention discloses a method and a system for managing operation and maintenance of an online instrument of a power plant, and relates to the field of instrument management, wherein the method comprises the following steps: acquiring operation control data of first equipment, inputting the operation control data into a state prediction model embedded in an operation and maintenance management system of an online instrument of a power plant, and outputting a state prediction result; judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; if not, inputting the instrument state data into a state evaluation model, and outputting a state calibration result; when the state calibration result is unqualified, performing fault association analysis according to the instrument state data to obtain a fault prediction type of the first equipment; and carrying out instrument operation and maintenance management according to the fault prediction type. The technical problems of low accuracy and low comprehensiveness of instrument operation and maintenance management aiming at the power plant in the prior art are solved, and the instrument operation and maintenance management quality of the power plant is low. The technical effects of improving the operation and maintenance management quality of the instrument of the power plant and the like are achieved.
Description
Technical Field
The invention relates to the field of instrument management, in particular to a method and a system for managing operation and maintenance of an online instrument of a power plant.
Background
The chemical supervision is an important link of corrosion prevention and scale prevention of a power plant and ensuring safe production, and the operation and maintenance management of an on-line instrument is a key link of the chemical supervision of the power plant. The state of the online instrument of the power plant is monitored, the reliability of the online instrument of the power plant is guaranteed, the management level of the online analysis instrument is improved, and the method has very important significance for safe production, energy conservation and consumption reduction of the power plant. The on-line meters are distributed in the whole system of the power plant, and have the characteristics of large quantity, multiple varieties, wide distribution, complex management work and the like. In the prior art, the technical problems of insufficient accuracy and low comprehensiveness of instrument operation and maintenance management aiming at a power plant and low quality of instrument operation and maintenance management of the power plant exist.
Disclosure of Invention
The application provides a power plant on-line instrument operation and maintenance management method and system. The technical problems of low accuracy and low comprehensiveness of instrument operation and maintenance management aiming at the power plant in the prior art are solved, and the instrument operation and maintenance management quality of the power plant is low.
In view of the above problems, the application provides a method and a system for managing the operation and maintenance of an on-line instrument of a power plant.
In a first aspect, the present application provides a method for managing operation and maintenance of an on-line meter of a power plant, where the method is applied to an on-line meter operation and maintenance management system of the power plant, and the method includes: receiving meter status data of a first meter; acquiring operation control data of first equipment, inputting the operation control data into a state prediction model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first equipment; judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; if not, inputting the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system, and outputting a state calibration result; when the state calibration result is unqualified, performing fault association analysis according to the instrument state data to obtain a fault prediction type of the first equipment; and carrying out instrument operation and maintenance management according to the fault prediction type.
In a second aspect, the present application also provides a power plant on-line instrument operation and maintenance management system, wherein the system comprises: the instrument data receiving module is used for receiving instrument state data of the first instrument; the state prediction module is used for acquiring operation control data of first equipment, inputting a state prediction model embedded in the power plant on-line instrument operation and maintenance management system and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first equipment; the state deviation judging module is used for judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; the state evaluation module is used for inputting the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system and outputting a state calibration result if the instrument state data is not satisfied; the fault association analysis module is used for carrying out fault association analysis according to the instrument state data when the state calibration result is unqualified, and obtaining a fault prediction type of the first equipment; and the instrument operation and maintenance management module is used for carrying out instrument operation and maintenance management according to the fault prediction type.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
analyzing instrument state data of the first instrument through a state prediction model to obtain a state prediction result; judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; if the data does not meet the requirement, inputting the instrument state data into a state evaluation model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state calibration result; and when the state calibration result is unqualified, performing fault association analysis according to the instrument state data, obtaining a fault prediction type of the first equipment, and performing instrument operation and maintenance management according to the fault prediction type. The method achieves the technical effects of adaptively predicting faults of equipment corresponding to the meters of the power plant by performing multidimensional data analysis on the meters of the power plant, improving the accuracy, comprehensiveness and practicability of the operation and maintenance management of the meters of the power plant, improving the quality of the operation and maintenance management of the meters of the power plant and providing guarantee for the safety production of the power plant.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments of the present disclosure will be briefly described below. It is apparent that the figures in the following description relate only to some embodiments of the present disclosure and are not limiting of the present disclosure.
FIG. 1 is a schematic flow chart of a method for managing operation and maintenance of an on-line instrument of a power plant;
FIG. 2 is a schematic flow chart of the output state calibration result in the power plant on-line instrument operation and maintenance management method of the application;
fig. 3 is a schematic structural diagram of an operation and maintenance management system for an on-line instrument of a power plant.
Reference numerals illustrate: the system comprises a meter data receiving module 11, a state predicting module 12, a state deviation judging module 13, a state evaluating module 14, a fault association analyzing module 15 and a meter operation and maintenance management module 16.
Detailed Description
The application provides a method and a system for managing operation and maintenance of an online instrument of a power plant. The technical problems of low accuracy and low comprehensiveness of instrument operation and maintenance management aiming at the power plant in the prior art are solved, and the instrument operation and maintenance management quality of the power plant is low. The method achieves the technical effects of adaptively predicting faults of equipment corresponding to the meters of the power plant by performing multidimensional data analysis on the meters of the power plant, improving the accuracy, comprehensiveness and practicability of the operation and maintenance management of the meters of the power plant, improving the quality of the operation and maintenance management of the meters of the power plant and providing guarantee for the safety production of the power plant.
Example 1
Referring to fig. 1, the application provides a method for managing operation and maintenance of an on-line instrument of a power plant, wherein the method is applied to an on-line instrument operation and maintenance management system of the power plant, and the method specifically comprises the following steps:
step S100: receiving meter status data of a first meter;
specifically, the system is connected with an on-line instrument operation and maintenance management system of the power plant, and the on-line instrument operation and maintenance management system of the power plant inquires state parameters of a first instrument of first equipment to obtain instrument state data of the first instrument. The power plant comprises a plurality of devices, and each device is provided with a plurality of state monitoring meters such as a sodium meter, a silicon meter, a dissolved oxygen meter, a conductivity meter, a pH meter, an ORP meter, a turbidity meter, a residual chlorine meter, an acid-base concentration meter and the like. The power plant on-line instrument operation and maintenance management system has the function of collecting, maintaining and managing state parameters of a plurality of state monitoring instruments in the power plant. Each device of the power plant is set as a first device, and each state monitoring instrument of the first device is set as a first instrument. The meter status data includes multiple sets of real-time status data for the first meter. Each set of real-time status data includes a plurality of real-time status parameters of the first meter. Illustratively, when the first meter is a sodium meter of the first device, each set of real-time status data includes an electrode response slope, an electrode preset response slope range, an electrode response zero, an electrode preset response zero range, a sodium electrode run time, a sodium electrode preset lifetime time, a reference electrode run time, a reference electrode preset lifetime time, an alkalizing agent consumed amount, an alkalizing agent pre-charge amount, an electrolyte consumed amount, an electrolyte pre-charge amount, and the like. The method achieves the technical effects of determining the instrument state data of the first instrument and laying a foundation for carrying out operation and maintenance management on the first instrument subsequently.
Step S200: acquiring operation control data of first equipment, inputting the operation control data into a state prediction model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first equipment;
further, step S200 of the present application further includes:
step S210: collecting a first construction data set of a state prediction model according to a first instrument model and a first equipment model, wherein the first construction data set of the state prediction model comprises operation control record data and instrument state identification data;
step S220: training a first base learner by taking the operation control record data as input data and the instrument state identification data as supervision data;
step S230: training an Nth base learner by taking the operation control record data as input data and the instrument state identification data as supervision data, wherein the model network structures of the N base learners are different;
specifically, a first device model is determined based on the first device. Based on the first meter, a first meter model is determined. And acquiring historical data based on the first instrument model and the first equipment model to obtain a first construction data set of the state prediction model. The first instrument model comprises type and model specification information corresponding to the first instrument. The first equipment model comprises type and model specification information corresponding to the first equipment. The first construction data set of the state prediction model comprises a plurality of groups of construction data corresponding to a plurality of same-family devices and a plurality of same-family meters. The plurality of peer devices includes a plurality of devices of the same type as the first device. The plurality of alien meters includes a plurality of meters of the same type as the first meter. Each group of construction data comprises operation control record data corresponding to the same group of equipment and instrument state identification data corresponding to the same group of instruments. The operation control record data comprises a plurality of historical operation control parameters corresponding to the same family of equipment. The meter status identification data includes a plurality of historical identification status parameters corresponding to meters of the same family. The plurality of historical identification status parameters includes a plurality of historical status parameters of corresponding peer meters in a healthy state when the peer devices are operating under the operation control record data.
Further, the operation control record data in the first construction data set of the state prediction model is used as input data, and the instrument state identification data in the first construction data set of the state prediction model is used as supervision data. Illustratively, a BP neural network is used as a first model network structure of the first base learner. The BP neural network is a multi-layer feedforward neural network trained according to an error back propagation algorithm. The BP neural network comprises an input layer, a plurality of layers of neurons and an output layer. The BP neural network can perform forward calculation and backward calculation. When calculating in the forward direction, the input information is processed layer by layer from the input layer through a plurality of layers of neurons and is turned to the output layer, and the state of each layer of neurons only affects the state of the next layer of neurons. If the expected output cannot be obtained at the output layer, the reverse calculation is carried out, the error signal is returned along the original connecting path, and the weight of each neuron is modified to minimize the error signal. The first model network structure comprises an input layer, an implicit layer and an output layer. The random 70% of the first set of build data of the state prediction model is partitioned into training data sets. The random 30% of the first build data set of the state prediction model is partitioned into test data sets. Based on the BP neural network, cross supervision training is carried out on the first model network structure through a training data set, and a first base learner is obtained. And taking the test data set as input information, inputting the input information into the first base learner, and updating parameters of the first base learner through the test data set.
Further, the operation control record data in the first construction data set of the state prediction model is used as input data, and the instrument state identification data in the first construction data set of the state prediction model is used as supervision data. And constructing model network structures of the N base learners, wherein the model network structures of the N base learners are different from each other. The model network structures of the N base learners include a second model network structure, a third model network structure … … nth model network structure. The second model network structure, the third model network structure … …, and the nth model network structure include a plurality of different model network structures such as decision trees, support vector machines, and the like. And then, according to the first construction data set of the state prediction model, respectively performing supervision training on the second model network structure and the third model network structure … … Nth model network structure to obtain a second base learner and a third base learner … … Nth base learner, and combining the first base learner to obtain N base learners. The N base learners include a first base learner, a second base learner, a third base learner … … nth base learner.
The technical effect of obtaining N base learners by training the first construction data set of the state prediction model is achieved, and therefore the accuracy of the state prediction model is improved.
Step S240: acquiring output data sets from the first base learner to the Nth base learner, and generating a second construction data set of a state prediction model;
further, step S240 of the present application further includes:
step S241: acquiring a plurality of N metadata sets according to the output data sets from the first base learner to the N base learner;
step S242: dividing the N metadata sets into k groups to obtain k groups of N metadata sets;
step S243: and (3) setting any group of the k groups of N metadata sets according to s-1:1, dividing, constructing a training sample data set according to the s-1 proportion data set, and constructing a verification sample data set according to the 1 proportion data set;
step S244: and adding the training sample data set and the verification sample data set into a second construction data set of the state prediction model.
Specifically, a plurality of N metadata sets are acquired based on the output data sets of the N base learners. Then, the plurality of N metadata sets are divided into k groups, and k groups of N metadata sets are acquired. According to s-1:1, respectively carrying out data division on k groups of N metadata sets to obtain an s-1 proportion data set and a 1 proportion data set. Setting the s-1 proportion data set as a training sample data set, setting the 1 proportion data set as a verification sample data set, and adding the training sample data set and the verification sample data set to a second construction data set of the state prediction model. Wherein each N-metadata set includes an output data set of each base learner. The output data set of each base learner is meter state identification data. The s-1 scale data set includes s-1 scale data corresponding to each set of N metadata sets. The 1 scale data set includes 1 scale data corresponding to each set of N metadata sets. And s is a positive integer greater than 2. Illustratively, s is 5, then 80% of the random data information in each set of N metadata is divided into s-1 scale data, and 20% of the random data information in each set of N metadata is divided into 1 scale data. The second set of construction data of the state prediction model includes a training sample data set and a validation sample data set.
Step S250: taking the second construction data set of the state prediction model as input data, and taking the instrument state identification data as supervision data to train a fitting learner;
step S260: merging the first base learner to the Nth base learner as parallel independent nodes to generate a data processing layer;
step S270: and merging an output layer of the data processing layer with an input layer of the fitting learner to generate the state prediction model.
Step S300: judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not;
specifically, the second set of construction data of the state prediction model is used as input data, and the instrument state identification data is used as supervision data. Based on the convolutional neural network, training sample data sets in the second construction data set of the state prediction model are continuously self-trained and learned to a convergence state, and a fitting learner is obtained. And verifying the fitting learner through a verification sample data set in the second construction data set of the state prediction model, and obtaining the fitting learner meeting the verification constraint condition when the output accuracy of the fitting learner meets the verification constraint condition. Among these, convolutional neural networks are a type of feedforward neural network that includes convolutional computation and has a deep structure. The convolutional neural network has characteristic learning capability and can carry out translation invariant classification on input information according to a hierarchical structure of the convolutional neural network. The fitting learner comprises an input layer, a convolution layer, a pooling layer and an output layer. The verification constraints include pre-setting a determined output accuracy threshold of the fitting learner,
Further, the first base learner, the second base learner, the third base learner … … and the nth base learner are set as N parallel independent nodes, and the N parallel independent nodes are combined to generate a data processing layer. And combining an output layer of the data processing layer with an input layer of the fitting learner to obtain a state prediction model, and embedding the state prediction model into the power plant on-line instrument operation and maintenance management system. The state prediction model comprises a data processing layer and a fitting learner. And, the data processing layer includes N parallel independent nodes. The N parallel independent nodes include a first base learner, a second base learner, a third base learner … … nth base learner. Then, each set of real-time status data in the meter status data is set as the first real-time status data, respectively. And inquiring the operation control parameters of the first equipment based on the first real-time state data to obtain the operation control data of the first equipment. And inputting the operation control data into a state prediction model to obtain a state prediction result. And carrying out deviation calculation on the state prediction result and the instrument state data to obtain the state deviation degree. Wherein the operation control data includes a plurality of real-time operation control parameters of the corresponding first device when the first meter is in the first real-time status data. The state prediction result includes a plurality of normal state parameters of the corresponding first meter in a healthy state when the first device is in the operation control data. The state deviation threshold comprises a state deviation threshold which is determined in advance.
In the case of calculating the deviation between the state prediction result and the meter state data, the difference between the normal state parameters in the state prediction result and the real-time state parameters in the corresponding first real-time state data is calculated to obtain a plurality of state parameter differences. And respectively carrying out ratio calculation on absolute values corresponding to the state parameter difference values and a plurality of normal state parameters in the state prediction result to obtain a plurality of state parameter deviation degrees. And outputting the average value corresponding to the state parameter deviation degrees as the state deviation degree.
Further, it is determined whether the state deviation degree satisfies a state deviation threshold. If the state deviation degree is smaller than the state deviation threshold, the state deviation degree does not meet the state deviation threshold, and the first real-time state data in the corresponding instrument state data of the first instrument are relatively real. If the state deviation degree is greater than/equal to the state deviation threshold, the state deviation degree meets the state deviation threshold, and the data authenticity of the first real-time state data in the corresponding instrument state data of the first instrument is low, so that the first instrument needs to be subjected to instrument verification management.
The method achieves the technical effects that the running control data of the first equipment are reliably predicted in the instrument state through the state prediction model, the state prediction result is determined, the instrument state deviation analysis is carried out by combining the instrument state data, the accurate state deviation degree is obtained, and therefore the accuracy of instrument operation and maintenance management of the power plant is improved.
Step S400: if not, inputting the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system, and outputting a state calibration result;
further, as shown in fig. 2, step S400 of the present application further includes:
step S410: acquiring state standard data, wherein the state standard data is a state characteristic value of normal operation of an instrument;
step S420: performing deviation analysis on the instrument state data and the state standard data to obtain state characteristic information;
specifically, when the state deviation degree is smaller than the state deviation threshold value, the state deviation degree does not satisfy the state deviation threshold value, and the meter state data corresponding to the state deviation degree is relatively true. The state prediction result corresponding to the state deviation degree that does not satisfy the state deviation threshold is set as the state standard data. And setting a plurality of state parameter deviation degrees corresponding to the state deviation degrees which do not meet the state deviation threshold as state characteristic information. The state standard data are state characteristic values of normal operation of the instrument. The state characteristic value of the instrument in normal operation comprises a plurality of normal state parameters in a state prediction result corresponding to the state deviation degree which does not meet the state deviation threshold value.
Step S430: and inputting the state characteristic information into the state evaluation model, and outputting the state calibration result.
Further, step S430 of the present application further includes:
step S431: collecting state monitoring record data and equipment state information according to a first instrument model and a first equipment model, wherein the equipment state information comprises qualification, availability and disqualification;
step S432: performing deviation analysis on the state monitoring record data and the state standard data to generate sample state characteristic information;
step S433: traversing the state standard data to perform deviation grading calibration to obtain deviation grading data;
step S434: constructing a state evaluation model construction dataset according to the sample state characteristic information, the deviation grading data and the equipment state information;
specifically, historical data acquisition is performed based on the first instrument model and the first equipment model, and state monitoring record data and equipment condition information are obtained. The state monitoring record data comprises a plurality of groups of historical state monitoring records corresponding to a plurality of same family devices and a plurality of same family meters. Each group of history state monitoring records comprises history instrument state data corresponding to the same group of instruments, history operation control data corresponding to the same group of equipment and history state prediction results. The equipment condition information comprises a plurality of historical equipment condition parameters corresponding to a plurality of groups of historical state monitoring records in the state monitoring record data. Each historical equipment condition parameter includes pass/available/fail.
Further, performing deviation analysis on a plurality of groups of historical state monitoring records in the state monitoring record data to obtain sample state characteristic information. The sample state characteristic information comprises a plurality of groups of sample state characteristic values corresponding to a plurality of groups of historical state monitoring records. Each set of sample state characteristic values includes a plurality of historical state parameter deviations between the historical operating control data and the historical state prediction results in each set of historical state monitoring records. The plurality of historical state parameter deviations are the same as the plurality of state parameter deviations, and are not described herein for brevity.
Further, traversing a plurality of groups of sample state characteristic values in the sample state characteristic information to perform deviation grading calibration, obtaining deviation grading data, and obtaining a state evaluation model construction data set by combining the sample state characteristic information and the equipment state information. The state evaluation model construction data set comprises sample state characteristic information, deviation grading data and equipment state information. The deviation ranking data includes deviation ranking information corresponding to each historical state parameter deviation degree in each set of sample state feature values of the sample state feature information. When the traverse state standard data is subjected to deviation grading calibration, each historical state parameter deviation degree in each group of sample state characteristic values of the sample state characteristic information is input into a pre-constructed deviation grading calibration table, and deviation grade information matching is performed on each historical state parameter deviation degree through the deviation grading calibration table, so that deviation grading data are obtained. The deviation grading calibration table comprises a plurality of preset state parameter deviation degrees and a plurality of preset deviation grade information which are preset and determined in advance. For example, in the deviation classification calibration table, when the deviation degree of the preset state parameter is 0 to 3%, the corresponding preset deviation class information is 1 class. When the deviation degree of the preset state parameter is 3-5%, the corresponding preset deviation grade information is grade 2. When the deviation degree of the preset state parameter is 5-10%, the corresponding preset deviation grade information is 3 grades. When the deviation degree of the preset state parameter is more than 10%, the corresponding preset deviation grade information is grade 4.
The technical effect of constructing a state evaluation model construction data set and providing data support for a subsequent state evaluation model construction is achieved.
Step S435: and constructing a data set according to the state evaluation model, and training the state evaluation model.
Further, step S435 of the present application further includes:
step S4351: taking the sample state characteristic information and the deviation grading data as input data, taking the equipment condition information as output identification data, and training based on a decision tree to obtain a first classifier;
step S4352: acquiring a first output residual coefficient of the first classifier, and constructing a first residual sample data set;
step S4353: taking the first residual error sample data set as input data, taking the equipment condition information as output identification data, training based on a decision tree, and obtaining a second classifier;
step S4354: obtaining a second output residual coefficient of the second classifier, and constructing a second residual sample data set;
step S4355: taking the second residual error sample data set as input data, taking the equipment condition information as output identification data, training based on a decision tree, and obtaining a third classifier;
Step S4356: repeating training until the M-th output residual error mean value is smaller than or equal to a residual error threshold value, merging the first classifier, the second classifier and the third classifier until the M-th classifier, and obtaining the state evaluation model, wherein the output of the state evaluation model is the sum of the outputs of the first classifier, the second classifier and the third classifier until the M-th classifier;
and when the output residual coefficient is a negative value during any training, setting the equipment condition information to be a negative value.
Specifically, the equipment condition information in the state evaluation model construction data set is digitally identified, a plurality of identification historical equipment condition parameters are obtained, and the equipment condition information in the state evaluation model construction data set is subjected to data updating according to the plurality of identification historical equipment condition parameters. Illustratively, when the equipment condition information in the state evaluation model construction dataset is digitally identified, and when the historical equipment condition parameter is qualified, the identified historical equipment condition parameter corresponding to the historical equipment condition parameter is marked as 1; when the historical equipment condition parameters are available, marking the identification historical equipment condition parameters corresponding to the historical equipment condition parameters as 2; and when the historical equipment condition parameter is unqualified, marking the identification historical equipment condition parameter corresponding to the historical equipment condition parameter as 3. And then, taking the sample state characteristic information and the deviation grading data as input data, taking the equipment condition information as output identification data, training based on a decision tree, and obtaining a first classifier and a first residual sample data set. The first residual sample data set includes a plurality of first output residual coefficients.
Illustratively, a plurality of first output residual coefficients are obtained by a first residual calculation formula. The first residual calculation formula isWherein->For the first output residual coefficient of the first classifier, < > for>Identifying device condition parameters for prediction of the first classifier, < >>Identifying an identified historical device condition parameter corresponding to the device condition parameter for the prediction of the first classifier.
Further, the first residual sample data set is used as input data, the equipment condition information is used as output identification data, training is performed based on a decision tree, and the second classifier and the second residual sample data set are obtained. The second set of residual sample data includes a plurality of second output residual coefficients.
Illustratively, a plurality of second output residual coefficients are obtained by a second residual calculation formula. The second residual calculation formula isWherein->For the second output residual coefficient of the second classifier, < > for>Identifying device condition parameters for prediction of the first classifier, < >>Is->The predictions of the corresponding second classifier identify the device condition parameters,identifying an identified historical device condition parameter corresponding to the device condition parameter for the prediction of the first classifier.
Further, the second residual error sample data set is used as input data, the equipment condition information is used as output identification data, training is performed based on a decision tree, and a third classifier is obtained. And repeating training until the M-th output residual error mean value is smaller than or equal to the residual error threshold value, merging the first classifier, the second classifier and the third classifier … … M-th classifier to obtain a state evaluation model, and embedding the state evaluation model into the power plant online instrument operation and maintenance management system. And, when the output residual coefficient is negative during training at any time, the equipment condition information is set to be negative. The state evaluation model includes a first classifier, a second classifier, a third classifier … … Mth classifier. The output of the state evaluation model is the sum of the outputs of the first classifier, the second classifier, the third classifier … … and the mth classifier. And then, inputting the state characteristic information into a state evaluation model, and outputting a state calibration result. The M-th output residual error mean value comprises an absolute value of a mean value of output residual error coefficients corresponding to the M-th classifier. The residual threshold comprises a preset determined output residual threshold. The status calibration result includes the first device being pass/available/fail. The method achieves the technical effects of obtaining an accurate state calibration result through the state evaluation model, thereby improving the reliability of instrument operation and maintenance management of the power plant.
Step S500: when the state calibration result is unqualified, performing fault association analysis according to the instrument state data to obtain a fault prediction type of the first equipment;
further, the step S500 of the present application further includes:
step S510: taking a first instrument model, a first equipment model and the instrument state data as scene constraint information, and collecting L pieces of fault record data;
step S520: performing cluster analysis on the L pieces of fault record data according to the fault types to obtain a plurality of fault types and a plurality of cluster frequencies;
step S530: traversing the cluster frequencies to be compared with L to generate a plurality of fault probabilities of the fault types;
step S540: and adding the plurality of fault types with the plurality of fault probabilities greater than or equal to a fault probability threshold into the fault prediction type.
Step S600: and carrying out instrument operation and maintenance management according to the fault prediction type.
Specifically, when the state calibration result belongs to disqualification, setting the first instrument model, the first equipment model and the instrument state data as scene constraint information, and carrying out fault record acquisition based on the scene constraint information to obtain L pieces of fault record data. Each fault record data comprises the same type of historical instrument state data corresponding to the same family instrument and the historical fault type information corresponding to the same family equipment. The same type of historical meter state data is the same as meter state data.
Further, cluster analysis is performed on the L pieces of fault record data based on the fault types, namely, fault record data corresponding to the same fault type are classified into one type, and a plurality of fault types and a plurality of cluster frequencies are obtained. Each fault type comprises a plurality of fault record data corresponding to the same fault type. Each cluster frequency includes a total number of pieces of fault record data within each fault type. And then, calculating the ratio of the clustering frequencies to L to obtain a plurality of fault probabilities of a plurality of fault types. And adding the fault type corresponding to the fault probability which is larger than or equal to the fault probability threshold value to the fault prediction type, and carrying out instrument operation and maintenance management according to the fault prediction type. Wherein the plurality of failure probabilities includes a plurality of ratios between the plurality of cluster frequencies and L. The fault probability threshold value comprises preset and determined fault probability threshold value information. The fault prediction type comprises a plurality of fault types corresponding to a plurality of fault probabilities greater than or equal to a fault probability threshold.
For example, when the operation and maintenance management of the instrument is performed according to the fault prediction type, maintenance resources such as maintenance parts, maintenance instruments, maintenance personnel and the like are configured according to the fault prediction type, periodic manual detection is not needed, and the efficiency is high. Thereby improving the practicability and comprehensiveness of the instrument operation and maintenance management of the power plant.
In summary, the method for managing the operation and maintenance of the online instrument of the power plant provided by the application has the following technical effects:
1. analyzing instrument state data of the first instrument through a state prediction model to obtain a state prediction result; judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; if the data does not meet the requirement, inputting the instrument state data into a state evaluation model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state calibration result; and when the state calibration result is unqualified, performing fault association analysis according to the instrument state data, obtaining a fault prediction type of the first equipment, and performing instrument operation and maintenance management according to the fault prediction type. The method achieves the technical effects of adaptively predicting faults of equipment corresponding to the meters of the power plant by performing multidimensional data analysis on the meters of the power plant, improving the accuracy, comprehensiveness and practicability of the operation and maintenance management of the meters of the power plant, improving the quality of the operation and maintenance management of the meters of the power plant and providing guarantee for the safety production of the power plant.
2. And carrying out reliable instrument state prediction on the operation control data of the first equipment through the state prediction model, determining a state prediction result, and carrying out instrument state deviation analysis by combining the instrument state data to obtain an accurate state deviation degree, thereby improving the accuracy of instrument operation and maintenance management of the power plant.
Example two
Based on the same inventive concept as the method for managing the operation and maintenance of the on-line meters of the power plant in the foregoing embodiment, the present invention further provides a system for managing the operation and maintenance of the on-line meters of the power plant, referring to fig. 3, the system includes:
a meter data receiving module 11, wherein the meter data receiving module 11 is used for receiving meter state data of a first meter;
the state prediction module 12 is configured to obtain operation control data of a first device, input a state prediction model embedded in an online instrument operation and maintenance management system of the power plant, and output a state prediction result, where the first instrument is a state monitoring instrument of the first device;
a state deviation judging module 13, wherein the state deviation judging module 13 is used for judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value;
the state evaluation module 14 is configured to input the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system, and output a state calibration result if the state evaluation module 14 does not meet the state evaluation model;
the fault association analysis module 15 is configured to perform fault association analysis according to the instrument state data when the state calibration result is unqualified, so as to obtain a fault prediction type of the first device;
The instrument operation and maintenance management module 16, the instrument operation and maintenance management module 16 is used for performing instrument operation and maintenance management according to the fault prediction type.
Further, the system further comprises:
the system comprises a data set obtaining module, a state prediction model obtaining module and a state prediction model obtaining module, wherein the data set obtaining module is used for collecting a first construction data set of the state prediction model according to a first instrument model and a first equipment model, and the first construction data set of the state prediction model comprises operation control record data and instrument state identification data;
the first execution module is used for training a first base learner by taking the running control record data as input data and the instrument state identification data as supervision data;
the second execution module is used for training an N-th base learner by taking the running control record data as input data and the instrument state identification data as supervision data, wherein the model network structures of the N-th base learners are different;
the third execution module is used for acquiring output data sets from the first base learner to the Nth base learner and generating a second construction data set of the state prediction model;
The fitting learner obtaining module is used for training the fitting learner by taking the second construction data set of the state prediction model as input data and the instrument state identification data as supervision data;
the fourth execution module is used for merging the first base learner to the Nth base learner as parallel independent nodes to generate a data processing layer;
and the fifth execution module is used for merging the output layer of the data processing layer with the input layer of the fitting learner to generate the state prediction model.
Further, the system further comprises:
the sixth execution module is used for acquiring a plurality of N metadata sets according to the output data sets from the first base learner to the N base learner;
the seventh execution module is used for dividing the plurality of N metadata sets into k groups and obtaining k groups of N metadata sets;
the data dividing module is used for dividing any group of the k groups of N metadata sets according to s-1:1, dividing, constructing a training sample data set according to the s-1 proportion data set, and constructing a verification sample data set according to the 1 proportion data set;
A second build data set obtaining module for adding the training sample data set and the validation sample data set to the state prediction model second build data set.
Further, the system further comprises:
the state standard data acquisition module is used for acquiring state standard data, wherein the state standard data is a state characteristic value of normal operation of the instrument;
the deviation analysis module is used for carrying out deviation analysis on the instrument state data and the state standard data to obtain state characteristic information;
and the state calibration result output module is used for inputting the state characteristic information into the state evaluation model and outputting the state calibration result.
Further, the system further comprises:
the recording data acquisition module is used for acquiring state monitoring recording data and equipment state information according to a first instrument model and a first equipment model, wherein the equipment state information comprises qualification, availability and disqualification;
the sample state characteristic information generation module is used for carrying out deviation analysis on the state monitoring record data and the state standard data to generate sample state characteristic information;
The deviation grading calibration module is used for traversing the state standard data to carry out deviation grading calibration and obtain deviation grading data;
an eighth execution module for constructing a state evaluation model construction dataset according to the sample state characteristic information, the deviation classification data, and the equipment condition information;
and the ninth execution module is used for constructing a data set according to the state evaluation model and training the state evaluation model.
Further, the system further comprises:
the first classifier acquisition module is used for taking the sample state characteristic information and the deviation grading data as input data, taking the equipment condition information as output identification data, and training based on a decision tree to acquire a first classifier;
the first residual sample data set construction module is used for acquiring a first output residual coefficient of the first classifier and constructing a first residual sample data set;
the second classifier acquisition module is used for taking the first residual error sample data set as input data, taking the equipment condition information as output identification data, training based on a decision tree and acquiring a second classifier;
The second residual sample data set construction module is used for acquiring a second output residual coefficient of the second classifier and constructing a second residual sample data set;
the third classifier acquisition module is used for taking the second residual error sample data set as input data, taking the equipment condition information as output identification data, and training based on a decision tree to acquire a third classifier;
a tenth execution module, configured to repeatedly train until an mth output residual error mean value is less than or equal to a residual error threshold value, and combine the first classifier, the second classifier, and the third classifier until the mth classifier to obtain the state evaluation model, where an output of the state evaluation model is a sum of outputs of the first classifier, the second classifier, and the third classifier until the mth classifier;
and when the output residual coefficient is a negative value during any training, setting the equipment condition information to be a negative value.
Further, the system further comprises:
the fault record acquisition module is used for acquiring L fault record data by taking a first instrument model, a first equipment model and the instrument state data as scene constraint information;
The cluster analysis module is used for carrying out cluster analysis on the L pieces of fault record data according to fault types to obtain a plurality of fault types and a plurality of cluster frequencies;
the fault probability generation module is used for traversing the cluster frequencies and the L ratios and generating a plurality of fault probabilities of the fault types;
and the fault prediction type determining module is used for adding the plurality of fault types with the fault probabilities being greater than or equal to a fault probability threshold value into the fault prediction type.
The power plant on-line instrument operation and maintenance management system provided by the embodiment of the application can execute the power plant on-line instrument operation and maintenance management method provided by any embodiment of the application, and has the corresponding functional modules and beneficial effects of the execution method.
All the included modules are only divided according to the functional logic, but are not limited to the above-mentioned division, so long as the corresponding functions can be realized; in addition, the specific names of the functional modules are only for distinguishing from each other, and are not used for limiting the protection scope of the present application.
The application provides a power plant on-line instrument operation and maintenance management method, wherein the method is applied to a power plant on-line instrument operation and maintenance management system, and comprises the following steps: analyzing instrument state data of the first instrument through a state prediction model to obtain a state prediction result; judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not; if the data does not meet the requirement, inputting the instrument state data into a state evaluation model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state calibration result; and when the state calibration result is unqualified, performing fault association analysis according to the instrument state data, obtaining a fault prediction type of the first equipment, and performing instrument operation and maintenance management according to the fault prediction type. The technical problems of low accuracy and low comprehensiveness of instrument operation and maintenance management aiming at the power plant in the prior art are solved, and the instrument operation and maintenance management quality of the power plant is low. The method achieves the technical effects of adaptively predicting faults of equipment corresponding to the meters of the power plant by performing multidimensional data analysis on the meters of the power plant, improving the accuracy, comprehensiveness and practicability of the operation and maintenance management of the meters of the power plant, improving the quality of the operation and maintenance management of the meters of the power plant and providing guarantee for the safety production of the power plant.
Note that the above is only a preferred embodiment of the present invention and the technical principle applied. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, while the invention has been described in connection with the above embodiments, the invention is not limited to the embodiments, but may be embodied in many other equivalent forms without departing from the spirit or scope of the invention, which is set forth in the following claims.
Claims (8)
1. The utility model provides a power plant on-line instrument operation and maintenance management method which is characterized in that the utility model is applied to a power plant on-line instrument operation and maintenance management system, and comprises the following steps:
receiving meter status data of a first meter;
acquiring operation control data of first equipment, inputting the operation control data into a state prediction model embedded in an online instrument operation and maintenance management system of the power plant, and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first equipment;
judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not;
If not, inputting the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system, and outputting a state calibration result;
when the state calibration result is unqualified, performing fault association analysis according to the instrument state data to obtain a fault prediction type of the first equipment;
and carrying out instrument operation and maintenance management according to the fault prediction type.
2. The method of claim 1, wherein obtaining operational control data of a first device, inputting a state prediction model embedded in an operation and maintenance management system of an on-line instrument of the power plant, and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first device, comprises:
collecting a first construction data set of a state prediction model according to a first instrument model and a first equipment model, wherein the first construction data set of the state prediction model comprises operation control record data and instrument state identification data;
training a first base learner by taking the operation control record data as input data and the instrument state identification data as supervision data;
training an Nth base learner by taking the operation control record data as input data and the instrument state identification data as supervision data, wherein the model network structures of the N base learners are different;
Acquiring output data sets from the first base learner to the Nth base learner, and generating a second construction data set of a state prediction model;
taking the second construction data set of the state prediction model as input data, and taking the instrument state identification data as supervision data to train a fitting learner;
merging the first base learner to the Nth base learner as parallel independent nodes to generate a data processing layer;
and merging an output layer of the data processing layer with an input layer of the fitting learner to generate the state prediction model.
3. The method of claim 2, wherein obtaining output data sets of the first base learner up to the nth base learner, generating a second build data set of a state prediction model, comprises:
acquiring a plurality of N metadata sets according to the output data sets from the first base learner to the N base learner;
dividing the N metadata sets into k groups to obtain k groups of N metadata sets;
and (3) setting any group of the k groups of N metadata sets according to s-1:1, dividing, constructing a training sample data set according to the s-1 proportion data set, and constructing a verification sample data set according to the 1 proportion data set;
And adding the training sample data set and the verification sample data set into a second construction data set of the state prediction model.
4. The method of claim 1, wherein inputting the meter status data into a status assessment model embedded in the power plant on-line meter operation and maintenance management system, outputting a status calibration result, comprises:
acquiring state standard data, wherein the state standard data is a state characteristic value of normal operation of an instrument;
performing deviation analysis on the instrument state data and the state standard data to obtain state characteristic information;
and inputting the state characteristic information into the state evaluation model, and outputting the state calibration result.
5. The method of claim 4, wherein inputting the state characteristic information into the state evaluation model and outputting the state calibration result comprises:
collecting state monitoring record data and equipment state information according to a first instrument model and a first equipment model, wherein the equipment state information comprises qualification, availability and disqualification;
performing deviation analysis on the state monitoring record data and the state standard data to generate sample state characteristic information;
Traversing the state standard data to perform deviation grading calibration to obtain deviation grading data;
constructing a state evaluation model construction dataset according to the sample state characteristic information, the deviation grading data and the equipment state information;
and constructing a data set according to the state evaluation model, and training the state evaluation model.
6. The method of claim 5, wherein constructing a dataset from the state evaluation model, training the state evaluation model, comprises:
taking the sample state characteristic information and the deviation grading data as input data, taking the equipment condition information as output identification data, and training based on a decision tree to obtain a first classifier;
acquiring a first output residual coefficient of the first classifier, and constructing a first residual sample data set;
taking the first residual error sample data set as input data, taking the equipment condition information as output identification data, training based on a decision tree, and obtaining a second classifier;
obtaining a second output residual coefficient of the second classifier, and constructing a second residual sample data set;
taking the second residual error sample data set as input data, taking the equipment condition information as output identification data, training based on a decision tree, and obtaining a third classifier;
Repeating training until the M-th output residual error mean value is smaller than or equal to a residual error threshold value, merging the first classifier, the second classifier and the third classifier until the M-th classifier, and obtaining the state evaluation model, wherein the output of the state evaluation model is the sum of the outputs of the first classifier, the second classifier and the third classifier until the M-th classifier;
and when the output residual coefficient is a negative value during any training, setting the equipment condition information to be a negative value.
7. The method of claim 1, wherein performing a fault correlation analysis based on the meter status data when the status calibration result is not acceptable, obtaining a fault prediction type for the first device, comprises:
taking a first instrument model, a first equipment model and the instrument state data as scene constraint information, and collecting L pieces of fault record data;
performing cluster analysis on the L pieces of fault record data according to the fault types to obtain a plurality of fault types and a plurality of cluster frequencies;
traversing the cluster frequencies to be compared with L to generate a plurality of fault probabilities of the fault types;
And adding the plurality of fault types with the plurality of fault probabilities greater than or equal to a fault probability threshold into the fault prediction type.
8. A power plant on-line meter operation and maintenance management system, the system comprising:
the instrument data receiving module is used for receiving instrument state data of the first instrument;
the state prediction module is used for acquiring operation control data of first equipment, inputting a state prediction model embedded in the power plant on-line instrument operation and maintenance management system and outputting a state prediction result, wherein the first instrument is a state monitoring instrument of the first equipment;
the state deviation judging module is used for judging whether the state deviation degree of the state prediction result and the instrument state data meets a state deviation threshold value or not;
the state evaluation module is used for inputting the instrument state data into a state evaluation model embedded in the power plant on-line instrument operation and maintenance management system and outputting a state calibration result if the instrument state data is not satisfied;
the fault association analysis module is used for carrying out fault association analysis according to the instrument state data when the state calibration result is unqualified, and obtaining a fault prediction type of the first equipment;
And the instrument operation and maintenance management module is used for carrying out instrument operation and maintenance management according to the fault prediction type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310851819.8A CN116579768B (en) | 2023-07-12 | 2023-07-12 | Power plant on-line instrument operation and maintenance management method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310851819.8A CN116579768B (en) | 2023-07-12 | 2023-07-12 | Power plant on-line instrument operation and maintenance management method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116579768A true CN116579768A (en) | 2023-08-11 |
| CN116579768B CN116579768B (en) | 2023-09-12 |
Family
ID=87543493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310851819.8A Active CN116579768B (en) | 2023-07-12 | 2023-07-12 | Power plant on-line instrument operation and maintenance management method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116579768B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116995280A (en) * | 2023-09-27 | 2023-11-03 | 苏州欣和智达能源科技有限公司 | Intelligent monitoring method for hydrogen fuel cell running state |
| CN117553840A (en) * | 2024-01-11 | 2024-02-13 | 深圳汉光电子技术有限公司 | Instrument based on intelligent management and system thereof |
| WO2024179001A1 (en) * | 2023-02-28 | 2024-09-06 | 华能山东发电有限公司白杨河发电厂 | Method for dynamic verification of online chemical instrument of thermal power plant |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105656990A (en) * | 2015-12-14 | 2016-06-08 | 重庆川仪自动化股份有限公司 | Instrument monitoring method and system |
| CN110375787A (en) * | 2019-07-25 | 2019-10-25 | 重庆市计量质量检测研究院 | A kind of measuring instrument operating status on-line evaluation method |
| CN114727166A (en) * | 2022-06-09 | 2022-07-08 | 南京天梯自动化设备股份有限公司 | Remote online metering instrument state monitoring method and system based on Internet of things |
| CN115081647A (en) * | 2022-05-19 | 2022-09-20 | 中国计量大学 | Industrial intelligent instrument fault pre-diagnosis method based on Bayesian network model |
| CN115392384A (en) * | 2022-08-31 | 2022-11-25 | 中铁第四勘察设计院集团有限公司 | Cable fault diagnosis method, system and equipment based on Stacking learning |
| CN116337135A (en) * | 2023-03-18 | 2023-06-27 | 北京嘉联优控科技有限公司 | Instrument fault diagnosis method, system, electronic equipment and readable storage medium |
-
2023
- 2023-07-12 CN CN202310851819.8A patent/CN116579768B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105656990A (en) * | 2015-12-14 | 2016-06-08 | 重庆川仪自动化股份有限公司 | Instrument monitoring method and system |
| CN110375787A (en) * | 2019-07-25 | 2019-10-25 | 重庆市计量质量检测研究院 | A kind of measuring instrument operating status on-line evaluation method |
| CN115081647A (en) * | 2022-05-19 | 2022-09-20 | 中国计量大学 | Industrial intelligent instrument fault pre-diagnosis method based on Bayesian network model |
| CN114727166A (en) * | 2022-06-09 | 2022-07-08 | 南京天梯自动化设备股份有限公司 | Remote online metering instrument state monitoring method and system based on Internet of things |
| CN115392384A (en) * | 2022-08-31 | 2022-11-25 | 中铁第四勘察设计院集团有限公司 | Cable fault diagnosis method, system and equipment based on Stacking learning |
| CN116337135A (en) * | 2023-03-18 | 2023-06-27 | 北京嘉联优控科技有限公司 | Instrument fault diagnosis method, system, electronic equipment and readable storage medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2024179001A1 (en) * | 2023-02-28 | 2024-09-06 | 华能山东发电有限公司白杨河发电厂 | Method for dynamic verification of online chemical instrument of thermal power plant |
| CN116995280A (en) * | 2023-09-27 | 2023-11-03 | 苏州欣和智达能源科技有限公司 | Intelligent monitoring method for hydrogen fuel cell running state |
| CN116995280B (en) * | 2023-09-27 | 2023-11-28 | 苏州欣和智达能源科技有限公司 | Intelligent monitoring method for hydrogen fuel cell running state |
| CN117553840A (en) * | 2024-01-11 | 2024-02-13 | 深圳汉光电子技术有限公司 | Instrument based on intelligent management and system thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116579768B (en) | 2023-09-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116579768B (en) | Power plant on-line instrument operation and maintenance management method and system | |
| CN105117602B (en) | A kind of metering device running status method for early warning | |
| CN111382542B (en) | Highway electromechanical device life prediction system facing full life cycle | |
| CN110879377B (en) | Metering device fault tracing method based on deep belief network | |
| Ferrucci et al. | Genetic programming for effort estimation: an analysis of the impact of different fitness functions | |
| CN109583520B (en) | State evaluation method of cloud model and genetic algorithm optimization support vector machine | |
| CN106651169A (en) | Fuzzy comprehensive evaluation-based distribution automation terminal state evaluation method and system | |
| CN110222991B (en) | Metering device fault diagnosis method based on RF-GBDT | |
| CN103793854A (en) | Multiple combination optimization overhead transmission line operation risk informatization assessment method | |
| CN110555477A (en) | municipal facility fault prediction method and device | |
| CN109492790A (en) | Wind turbines health control method based on neural network and data mining | |
| CN116432123A (en) | Electric energy meter fault early warning method based on CART decision tree algorithm | |
| CN115640969A (en) | Power grid operation and maintenance cost distribution method based on equipment state and operation age | |
| CN117708669A (en) | Multi-source data-based adaptive probability neural network high-voltage cable state evaluation method | |
| CN116738192A (en) | Digital twinning-based security data evaluation method and system | |
| CN117195505A (en) | Evaluation method and system for informatization evaluation calibration model of electric energy meter | |
| CN115600729A (en) | Grid load prediction method considering multiple attributes | |
| Li et al. | An optimized GRNN‐enabled approach for power transformer fault diagnosis | |
| Mao et al. | Comparative study on prediction of fuel cell performance using machine learning approaches | |
| CN114862267A (en) | Evaluation method and system of oil and gas pipeline alarm management system | |
| CN118052326A (en) | Campus water supply network water consumption prediction method based on neural network | |
| CN114240069A (en) | Power grid online risk calculation method based on regression learning and feature mining | |
| CN113591402A (en) | Digital power transformer health state fuzzy comprehensive evaluation method | |
| CN117523793A (en) | Power plant equipment fault early warning method and computer equipment | |
| CN117350146A (en) | GA-BP neural network-based drainage pipe network health evaluation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |