CN117648628A - Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit - Google Patents
Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit Download PDFInfo
- Publication number
- CN117648628A CN117648628A CN202311684207.0A CN202311684207A CN117648628A CN 117648628 A CN117648628 A CN 117648628A CN 202311684207 A CN202311684207 A CN 202311684207A CN 117648628 A CN117648628 A CN 117648628A
- Authority
- CN
- China
- Prior art keywords
- unit
- electromechanical
- data
- bulb
- state
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000011156 evaluation Methods 0.000 claims abstract description 54
- 238000013145 classification model Methods 0.000 claims abstract description 49
- 238000004088 simulation Methods 0.000 claims abstract description 40
- 230000015654 memory Effects 0.000 claims abstract description 29
- 238000013528 artificial neural network Methods 0.000 claims abstract description 23
- 238000012360 testing method Methods 0.000 claims description 51
- 238000013461 design Methods 0.000 claims description 42
- 238000012544 monitoring process Methods 0.000 claims description 30
- 238000005094 computer simulation Methods 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 27
- 238000012795 verification Methods 0.000 claims description 27
- 238000004458 analytical method Methods 0.000 claims description 20
- 238000004891 communication Methods 0.000 claims description 18
- 238000010248 power generation Methods 0.000 claims description 14
- 238000012423 maintenance Methods 0.000 claims description 13
- 238000010801 machine learning Methods 0.000 claims description 12
- 238000004519 manufacturing process Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 8
- 230000002159 abnormal effect Effects 0.000 claims description 6
- 230000007787 long-term memory Effects 0.000 claims description 6
- 230000006403 short-term memory Effects 0.000 claims description 6
- 230000009471 action Effects 0.000 claims description 3
- 230000005284 excitation Effects 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 10
- 238000000605 extraction Methods 0.000 abstract description 8
- 230000007547 defect Effects 0.000 abstract description 4
- 230000008569 process Effects 0.000 description 7
- 230000008439 repair process Effects 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000013179 statistical model Methods 0.000 description 2
- 241000282461 Canis lupus Species 0.000 description 1
- 206010027339 Menstruation irregular Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Abstract
The invention discloses a method, a device, equipment and a medium for evaluating the electromechanical state of a bulb tubular turbine, and relates to the technical field of equipment state evaluation. The method comprises the steps of firstly applying a bulb through-flow unit simulation model, dynamically simulating under a target working condition and different electromechanical states to obtain historical operation data and real-time operation data, then obtaining modeling samples corresponding to all electromechanical states based on simulation results, then applying the modeling samples to establish a electromechanical state classification model based on a long-short-period memory neural network, finally inputting sample data to be tested of the bulb through-flow target unit under the target working condition into the classification model to obtain electromechanical state classification results and current electromechanical state evaluation results, and thus overcoming the defects of difficult feature extraction, poor pertinence, dependence rules, experience and unstable evaluation effects of the traditional electromechanical state evaluation scheme when the bulb through-flow unit is used, and avoiding randomness and excessive dependence on experience.
Description
Technical Field
The invention belongs to the technical field of equipment state evaluation, and particularly relates to a method, a device, equipment and a medium for evaluating the electromechanical state of a bulb tubular machine set.
Background
Most of the existing electromechanical state evaluation methods are based on rules or statistical models, and less involve the construction of a system, and meanwhile, the existing methods also depend on more expert knowledge and experience, and the key features of the electromechanical equipment are required to be manually selected and extracted. However, because of the structural specificity and environmental complexity of the bulb through-flow unit (as the name implies, the hydroelectric generating unit that resembles a bulb), and the rule or statistical model based approach can make the evaluation index appear to be singular, the conventional approach suffers from the following drawbacks when dealing with large amounts of data and identifying complex patterns: (1) The feature extraction is difficult, namely the traditional method needs to manually select and extract the features, but the selection and the extraction of the features are very difficult and time-consuming for a bulb through-flow unit and a plurality of complex electromechanical systems; (2) The method has poor pertinence, namely the traditional method can only analyze and judge the known parameters of the equipment, the classification and grading of the states are basically divided into two states of health and damage, the classification and grading are single, the detection and prediction capability of unknown faults are weak, and the correlation analysis of the states is not strong; (3) Over-relying on rules and experience, i.e. the traditional method over-relying on manually formulated rules and expert experience, the result may be affected by subjective factors and cannot adapt to the change of the state of the unit; (4) The evaluation effect is unstable, namely, the accuracy and stability of the evaluation result are limited because the traditional method cannot fully utilize a large amount of real-time data and historical data.
In summary, the conventional electromechanical state evaluation method has the disadvantages of difficult feature extraction, poor pertinence, dependence on rules and experience, unstable evaluation effect and the like when being used for a bulb through-flow unit and a plurality of complex electromechanical systems. Therefore, how to provide an intelligent electromechanical state evaluation scheme which is suitable for a bulb through-flow unit and has more accurate and stable evaluation results so as to avoid randomness and excessive dependence on experience is a subject of urgent study for those skilled in the art.
Disclosure of Invention
The invention aims to provide a method, a device, computer equipment and a computer readable storage medium for evaluating the electromechanical state of a bulb through-flow unit, which are used for solving the problems of difficult feature extraction, poor pertinence, dependence rules and experience and unstable evaluation effect of the conventional electromechanical state evaluation scheme when the conventional electromechanical state evaluation scheme is used for the bulb through-flow unit.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
in a first aspect, a method for evaluating an electromechanical state of a bulb through-flow unit is provided, including:
dynamically simulating a bulb tubular unit simulation model built in advance under a target working condition and different electromechanical states to obtain historical operation data and real-time operation data;
Acquiring unit design data and unit attribute data of the bulb tubular unit simulation model;
aiming at each electromechanical state in the electromechanical states, according to the unit design data, the unit attribute data and the historical operation data and the real-time operation data obtained by dynamic simulation in the corresponding states, corresponding modeling samples are obtained by sorting in the following modes: taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2;
dividing all the modeling samples into a training set, a verification set and a test set;
training, verifying and testing the machine learning model based on the long-term and short-term memory neural network by using the training set, the verifying set and the testing set in sequence to obtain an electromechanical state classification model passing the verification and the test;
acquiring operation data of each unit period in the current M nearest unit periods of the bulb tubular target unit under the target working condition by a production monitoring system;
Inputting the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods into the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states;
and determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
Based on the above-mentioned invention, a new scheme for evaluating the electromechanical state of the bulb through-flow type machine set based on the long-short-period memory neural network is provided, namely, firstly, a bulb through-flow type machine set simulation model is applied, historical operation data and real-time operation data are obtained through dynamic simulation under the target working condition and different electromechanical states, then, modeling samples corresponding to all electromechanical states are obtained based on simulation results, then, a electromechanical state classification model based on the long-short-period memory neural network is built by using the modeling samples, finally, the sample data to be tested of the bulb through-flow type target machine set under the target working condition is input into the classification model to obtain the electromechanical state classification result and the current electromechanical state evaluation result, thus, the defects of difficult feature extraction, poor pertinence, dependence rules, experience and unstable evaluation effect of the traditional electromechanical state evaluation scheme when the traditional electromechanical state evaluation scheme is used for the bulb through-flow type machine set are overcome, random and over-dependence experience are avoided, more accurate, stable and intelligent electromechanical state evaluation can be provided, and practical application and popularization are facilitated.
In one possible design, the target operating condition includes an organic unit ready state operating condition, a unit idle state operating condition, a unit power generation state operating condition, a unit shutdown state operating condition, or a unit overhaul state operating condition.
In one possible design, when the target working condition includes an organic unit maintenance working condition, the model input data further includes unit power generation state data of the bulb through-flow unit simulation model before and after the unit maintenance state, and the sample data to be tested further includes unit power generation state data of the bulb through-flow target unit before and after the unit maintenance state.
In one possible design, the plurality of electromechanical states includes a healthy state, an abnormal state, a fault state, a condition to be serviced, and a damaged state.
In one possible design, the training set, the verification set and the test set are applied to train, verify and test the machine learning model based on the long-short-term memory neural network in sequence, so as to obtain a machine electric state classification model passing verification and test, which comprises the following steps:
training and verifying the machine learning model based on the long-term and short-term memory neural network in sequence by using the training set and the verification set to obtain a verified initial electromechanical state classification model;
Inputting corresponding model input data into the initial electromechanical state classification model aiming at each test sample in the test set to obtain a corresponding electromechanical state classification result;
obtaining a confidence analysis report of the initial electromechanical state classification model through confidence analysis according to the model output data of each test sample and electromechanical state classification results;
and when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model exceeds a preset reliability threshold, judging that the test is passed, and taking the initial electromechanical state classification model as a final electromechanical state classification model.
In one possible design, after obtaining a confidence analysis report of the initial electromechanical state classification model by confidence analysis, the method further comprises:
when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model does not exceed a preset reliability threshold, unit design data, unit attribute data, optimal operation data and/or operation setting parameters of the bulb through-flow unit simulation model are fed back and adjusted, wherein the operation setting parameters comprise a common reminding trigger value, an emergency alarm trigger value and/or a shutdown action trigger value;
And re-applying the bulb through-flow unit simulation model to perform dynamic simulation under the target working condition and the electromechanical states so as to re-model based on simulation results.
In one possible design, the production monitoring system comprises a bulb through-flow unit bearing oil monitoring subsystem, a governor oil monitoring subsystem, a technical water supply monitoring subsystem, an excitation monitoring subsystem, an axial flow fan monitoring subsystem and/or a unit on-line monitoring subsystem.
The second aspect provides an electromechanical state evaluation device of a bulb through-flow unit, which comprises a dynamic simulation module, a data acquisition module, a sample arrangement module, a sample division module, a model establishment module, a model application module and an evaluation determination module;
the dynamic simulation module is used for dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and different electromechanical states by applying a pre-built bulb tubular unit simulation model;
the data acquisition module is used for acquiring unit design data and unit attribute data of the bulb through-flow unit simulation model;
the sample arrangement module is respectively in communication connection with the dynamic simulation module and the data acquisition module, and is used for arranging and obtaining corresponding modeling samples according to the unit design data, the unit attribute data, and historical operation data and real-time operation data obtained by dynamic simulation in the corresponding states for each electromechanical state in the electromechanical states according to the following modes: taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2;
The sample dividing module is in communication connection with the sample sorting module and is used for dividing all the modeling samples into a training set, a verification set and a test set;
the model building module is in communication connection with the sample dividing module and is used for applying the training set, the verification set and the test set to train, verify and test the machine learning model based on the long-period memory neural network in sequence to obtain an electromechanical state classification model passing the verification and the test;
the data acquisition module is also used for acquiring the operation data of the bulb tubular target unit in each unit period in the current M most recent unit periods under the target working condition through the production monitoring system;
the model application module is respectively in communication connection with the model building module and the data acquisition module and is used for taking the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods as sample data to be tested, and inputting the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states;
And the evaluation determining module is in communication connection with the model application module and is used for determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
In a third aspect, the present invention provides a computer device, comprising a memory, a processor and a transceiver, which are in communication connection in turn, wherein the memory is used for storing a computer program, the transceiver is used for receiving and transmitting a message, and the processor is used for reading the computer program and executing the method for evaluating the electromechanical state of the bulb through-flow unit according to any possible design in the first aspect or the first aspect.
In a fourth aspect, the present invention provides a computer readable storage medium having instructions stored thereon which, when executed on a computer, perform a bulb through-flow unit electromechanical state evaluation method as described in the first aspect or any of the possible designs of the first aspect.
In a fifth aspect, the present invention provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the method for evaluating the electromechanical state of a bulb through-flow unit as described in the first aspect or any of the possible designs of the first aspect.
The beneficial effect of above-mentioned scheme:
(1) The invention creatively provides a new scheme for evaluating the electromechanical state of a bulb through-flow unit based on a long-period memory neural network, namely, a bulb through-flow unit simulation model is firstly applied, historical operation data and real-time operation data are obtained through dynamic simulation under a target working condition and different electromechanical states, then modeling samples corresponding to all electromechanical states are obtained based on simulation results, then an electromechanical state classification model based on the long-period memory neural network is established by using the modeling samples, finally, to-be-tested sample data of the bulb through-flow target unit under the target working condition is input into the classification model to obtain an electromechanical state classification result and a current electromechanical state evaluation result, so that the defects of difficult feature extraction, poor pertinence, dependence on rules and experience and unstable evaluation effect of the traditional electromechanical state evaluation scheme when the traditional electromechanical state evaluation scheme is used for the bulb through-flow unit are overcome, random and overdependent on experience are avoided, more accurate, stable and intelligent electromechanical state evaluation is provided, and practical application and popularization are facilitated;
(2) Compared with a general scheme for evaluating the electromechanical state of the bulb tubular turbine based on experience and special tests, the invention can effectively avoid the conditions that the special tests affect the power generation of the turbine and special conditions are required to be created to install special instruments and meters, and can effectively improve the evaluation efficiency;
(3) Compared with a general scheme for evaluating the electromechanical state of the bulb tubular turbine based on experience and special experiments, the method and the device can effectively avoid the conditions of single data source and single evaluation dimension by repeatedly utilizing the turbine design data, the turbine attribute data, the historical operation data, the real-time operation data, the pre-repair data and the post-repair data, so that various data are mutually corrected, and the evaluation accuracy is effectively improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic flow chart of an electromechanical state evaluation method of a bulb through-flow unit according to an embodiment of the present application.
Fig. 2 is a schematic diagram of a model training and verification process according to an embodiment of the present application.
Fig. 3 is a schematic structural diagram of an apparatus for evaluating the electromechanical state of a bulb through-flow unit according to an embodiment of the present application.
Fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present application.
Detailed Description
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly described below with reference to the accompanying drawings and the description of the embodiments or the prior art, and it is obvious that the following description of the structure of the drawings is only some embodiments of the present invention, and other drawings can be obtained according to these drawings without inventive effort to a person skilled in the art. It should be noted that the description of these examples is for aiding in understanding the present invention, but is not intended to limit the present invention.
It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, a first object may be referred to as a second object, and similarly a second object may be referred to as a first object, without departing from the scope of example embodiments of the invention.
It should be understood that for the term "and/or" that may appear herein, it is merely one association relationship that describes an associated object, meaning that there may be three relationships, e.g., a and/or B, may represent: three cases of A alone, B alone or both A and B exist; as another example, A, B and/or C, can represent the presence of any one of A, B and C or any combination thereof; for the term "/and" that may appear herein, which is descriptive of another associative object relationship, it means that there may be two relationships, e.g., a/and B, it may be expressed that: the two cases of A and B exist independently or simultaneously; in addition, for the character "/" that may appear herein, it is generally indicated that the context associated object is an "or" relationship.
Examples:
as shown in fig. 1, the method for evaluating the electromechanical state of the bulb through-flow unit according to the first aspect of the present embodiment may be performed by, but not limited to, a computer device with a certain computing resource, for example, a platform server, a personal computer (Personal Computer, PC, refer to a multipurpose computer with a size, price and performance suitable for personal use, a desktop computer, a notebook computer, a small notebook computer, a tablet computer, a super notebook, etc. all belong to a personal computer), a smart phone, a personal digital assistant (Personal Digital Assistant, PDA), or an electronic device such as a wearable device. As shown in FIG. 1, the method for evaluating the electromechanical state of the bulb through-flow unit can include, but is not limited to, the following steps S1 to S89.
S1, a pre-built bulb tubular machine set simulation model is applied, and historical operation data and real-time operation data are obtained through dynamic simulation under target working conditions and different electromechanical states.
In the step S1, the bulb through-flow unit simulation model is used for simulating the original data required for generating the model establishment, and can refer to the subsequent bulb through-flow target unit (the reference purpose is to ensure that the simulation model has similar basic data with the actual unit, such as unit design data, unit attribute data, pre-repair data, post-repair data, etc., wherein the pre-repair data refers to the unit power generation state data before the unit maintenance state, the post-repair data refers to the unit power generation state data after the unit maintenance state), and the raw data is conventionally built in advance through MATLAB software and a visual simulation tool Simulink in the MATLAB. Specifically, the target working conditions include, but are not limited to, an organic unit ready state working condition, a unit idle state working condition, a unit power generation state working condition, a unit shutdown state working condition or a unit maintenance state working condition, and the like; the plurality of electromechanical states include, but are not limited to, a healthy state, an abnormal state (i.e., a state in which an abnormal condition exists but operation can continue), a fault state (i.e., a state in which an abnormal condition exists but operation is affected), a state to be serviced (i.e., a state in which service is required), a damaged state, and the like. The foregoing historical operation data and real-time operation data obtained by dynamic simulation may be obtained by conventional simulation based on the existing simulation means, and include, but are not limited to: the method comprises the steps of dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and the healthy state, dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and the abnormal state, dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and the fault state, dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and the to-be-overhauled state, dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and the damaged state, and the like.
S2, unit design data and unit attribute data of the bulb tubular unit simulation model are obtained.
In step S2, the unit design data, that is, the system design data of the bulb through-flow unit, and the unit attribute data, that is, the system acceptance data of the bulb through-flow unit, are manually entered.
S3, aiming at each electromechanical state in the electromechanical states, according to the unit design data, the unit attribute data and the historical operation data and the real-time operation data obtained by dynamic simulation in the corresponding states, the corresponding modeling samples are obtained by arrangement according to the following modes: and taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2.
In the step S3, the unit period may be a regular period such as an hour, a day, or a week, or an irregular period such as an operation period from the unit power generation state to the next power generation state. Since the simulation process is not a real-going process, there are cases where the time required for the simulation is compressed in order to quickly obtain the simulation result, each of the unit periods among the consecutive M unit periods does not have a real duration (for example, not a real day, but a virtual day), and the value of M may be set in advance. The model input data are time sequence data, namely, each unit period corresponds to operation data, and the same unit design data, unit attribute data and other data. In addition, in order to enrich the characteristic data, preferably, when the target working condition includes an organic unit maintenance working condition, the model input data further includes, but is not limited to, unit power generation state data (i.e., the pre-repair data and the post-repair data) of the bulb through-flow unit simulation model before and after the unit maintenance state, which can be acquired by a conventional observation mode.
S4, dividing all the modeling samples into a training set, a verification set and a test set.
In step S4, all the modeling samples constitute a data set as shown in fig. 2. In particular, 60% of the modeled samples may be randomly extracted from the dataset to form a training set, and half of the modeled samples may be randomly extracted from all of the remaining modeled samples to form a validation set, and all of the modeled samples that are eventually remaining may also be formed into a test set.
S5, training, verifying and testing the machine learning model based on the long-term and short-term memory neural network by using the training set, the verifying set and the testing set in sequence to obtain an electromechanical state classification model passing the verification and the test.
In the step S5, the Long Short-Term Memory neural network (LSTM) is a time-loop neural network, which is specifically designed to solve the Long-Term dependency problem of the general loop neural network (Recurrent Neural Networks, abbreviated as RNN), so that the training set, the verification set and the test set can be applied to train, verify and test the machine learning model based on the Long-Term Memory neural network in sequence, to obtain an electromechanical state classification model passing the verification and test, specifically, but not limited to the following steps S51 to S54.
S51, training and verifying the machine learning model based on the long-short-period memory neural network by using the training set and the verification set in sequence to obtain an initial electromechanical state classification model passing verification.
In the step S51, specific training and verification processes may be as shown in fig. 2, and include, but are not limited to, a model building link, a forward propagation link, an error calculation link, an error judgment link, a gradient calculation link, a weight and deviation updating link, and the like, where the finally saved model is the initial electromechanical state classification model.
S52, inputting corresponding model input data into the initial electromechanical state classification model aiming at each test sample in the test set, and obtaining a corresponding electromechanical state classification result.
S53, obtaining a confidence analysis report of the initial electromechanical state classification model through confidence analysis according to the model output data of each test sample and electromechanical state classification results.
In the step S53, the specific manner of confidence analysis may be conventionally obtained by statistics based on the consistency comparison result of the model output data and the electromechanical state classification result, for example, if the model output data of 95 test samples are consistent with the electromechanical state classification result in 100 test samples, the confidence analysis report indicating that the classification reliability of the initial electromechanical state classification model reaches 95% may be obtained.
S54, when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model exceeds a preset reliability threshold, determining that the test is passed, and taking the initial electromechanical state classification model as a final electromechanical state classification model.
In the step S54, the preset confidence threshold may be exemplified by 95%. Furthermore, in order to be able to automatically re-model when the test is not passed, it is preferable that the method further includes, but is not limited to, the following steps S55 to S56 after obtaining a confidence analysis report of the initial electromechanical state classification model by a confidence analysis.
S55, when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model does not exceed a preset reliability threshold, unit design data, unit attribute data, optimal operation data, operation setting parameters and the like of the bulb through-flow unit simulation model are fed back and adjusted, wherein the operation setting parameters comprise, but are not limited to, common reminding trigger values, emergency alarm trigger values, shutdown action trigger values and the like.
In the step S55, the specific manner of the feedback adjustment is an existing conventional manner, for example, optimization adjustment is performed based on a genetic algorithm or a wolf algorithm. In addition, manual adjustment can also be performed by responding to a human-computer interaction mode.
S56, re-applying the bulb through-flow unit simulation model to perform dynamic simulation under the target working condition and the electromechanical states so as to re-model based on simulation results (namely returning to executing steps S1-S5).
S6, acquiring operation data of the bulb tubular target unit in each unit period of the current M nearest unit periods under the target working condition through a production monitoring system.
In the step S6, the production monitoring system is configured to acquire historical operation data and real-time operation data of the bulb through-flow type target unit through an existing monitoring manner; specifically, the production monitoring system comprises, but is not limited to, a bulb through-flow unit bearing oil monitoring subsystem, a governor oil monitoring subsystem, a technical water supply monitoring subsystem, an excitation monitoring subsystem, an axial flow fan monitoring subsystem, and/or a unit on-line monitoring subsystem, which can acquire corresponding data based on the existing monitoring mode so as to serve as the operation data of the whole target unit. Further, each of the current most recent M unit periods is of a real duration (e.g., a real day, not a virtual day).
S7, taking the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods as sample data to be tested, and inputting the sample data into the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise, but are not limited to, confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states.
In step S7, also for enriching the characteristic data, preferably, when the target working condition includes an organic unit maintenance working condition, the sample data to be tested further includes, but is not limited to, unit power generation state data (i.e., the pre-repair data and the post-repair data) of the bulb through-flow type target unit before and after the unit maintenance state, which may be acquired by a conventional observation manner. In addition, the sample data to be tested is also time sequence data, namely, each unit period corresponds to operation data and the same unit design data, unit attribute data and other data respectively.
S8, determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
The method for evaluating the electromechanical state of the bulb through-flow unit based on the long-period memory neural network is characterized in that a bulb through-flow unit simulation model is firstly applied, historical operation data and real-time operation data are obtained through dynamic simulation under a target working condition and different electromechanical states, modeling samples corresponding to all electromechanical states are obtained based on simulation results, then an electromechanical state classification model based on the long-period memory neural network is established by using the modeling samples, finally the sample data to be tested of the bulb through-flow target unit under the target working condition is input into the classification model to obtain an electromechanical state classification result and a current electromechanical state evaluation result, and therefore the defects of difficult feature extraction, poor pertinence, dependence rules and experience and unstable evaluation effect of the traditional electromechanical state evaluation scheme when the traditional electromechanical state evaluation scheme is used for the bulb through-flow unit are overcome, random and over-dependence experience are avoided, more accurate, stable and intelligent electromechanical state evaluation can be provided, and practical application and popularization are facilitated.
As shown in fig. 3, in a second aspect of the present embodiment, a virtual device for implementing the method for evaluating the electromechanical state of the bulb through-flow unit according to the first aspect is provided, where the virtual device includes a dynamic simulation module, a data acquisition module, a sample arrangement module, a sample division module, a model establishment module, a model application module, and an evaluation determination module;
the dynamic simulation module is used for dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and different electromechanical states by applying a pre-built bulb tubular unit simulation model;
the data acquisition module is used for acquiring unit design data and unit attribute data of the bulb through-flow unit simulation model;
the sample arrangement module is respectively in communication connection with the dynamic simulation module and the data acquisition module, and is used for arranging and obtaining corresponding modeling samples according to the unit design data, the unit attribute data, and historical operation data and real-time operation data obtained by dynamic simulation in the corresponding states for each electromechanical state in the electromechanical states according to the following modes: taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2;
The sample dividing module is in communication connection with the sample sorting module and is used for dividing all the modeling samples into a training set, a verification set and a test set;
the model building module is in communication connection with the sample dividing module and is used for applying the training set, the verification set and the test set to train, verify and test the machine learning model based on the long-period memory neural network in sequence to obtain an electromechanical state classification model passing the verification and the test;
the data acquisition module is also used for acquiring the operation data of the bulb tubular target unit in each unit period in the current M most recent unit periods under the target working condition through the production monitoring system;
the model application module is respectively in communication connection with the model building module and the data acquisition module and is used for taking the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods as sample data to be tested, and inputting the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states;
And the evaluation determining module is in communication connection with the model application module and is used for determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
The working process, working details and technical effects of the foregoing device provided in the second aspect of the present embodiment may refer to the method for evaluating the electromechanical state of the bulb through-flow unit described in the first aspect, which is not described herein again.
As shown in fig. 4, a third aspect of the present embodiment provides a computer device for executing the method for evaluating the electromechanical state of a bulb through-flow unit according to the first aspect, which includes a memory, a processor and a transceiver that are sequentially connected in communication, where the memory is used for storing a computer program, the transceiver is used for receiving and transmitting a message, and the processor is used for reading the computer program, and executing the method for evaluating the electromechanical state of a bulb through-flow unit according to the first aspect. By way of specific example, the Memory may include, but is not limited to, random-Access Memory (RAM), read-Only Memory (ROM), flash Memory (Flash Memory), first-in first-out Memory (First Input First Output, FIFO), and/or first-in last-out Memory (First Input Last Output, FILO), etc.; the processor may be, but is not limited to, a microprocessor of the type STM32F105 family. In addition, the computer device may include, but is not limited to, a power module, a display screen, and other necessary components.
The working process, working details and technical effects of the foregoing computer device provided in the third aspect of the present embodiment may refer to the method for evaluating the electromechanical state of the bulb through-flow unit described in the first aspect, which is not described herein again.
A fourth aspect of the present embodiment provides a computer readable storage medium storing instructions comprising the method for evaluating the electromechanical state of a bulb through-flow unit according to the first aspect, i.e. the computer readable storage medium has instructions stored thereon, which when run on a computer, perform the method for evaluating the electromechanical state of a bulb through-flow unit according to the first aspect. The computer readable storage medium refers to a carrier for storing data, and may include, but is not limited to, a floppy disk, an optical disk, a hard disk, a flash Memory, and/or a Memory Stick (Memory Stick), where the computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable devices.
The working process, working details and technical effects of the foregoing computer readable storage medium provided in the fourth aspect of the present embodiment may refer to the method for evaluating the electromechanical state of the bulb through-flow unit according to the first aspect, which is not described herein.
A fifth aspect of the present embodiment provides a computer program product comprising instructions which, when run on a computer, cause the computer to perform the bulb through-flow unit electromechanical state evaluation method according to the first aspect. Wherein the computer may be a general purpose computer, a special purpose computer, a computer network, or other programmable apparatus.
Finally, it should be noted that: the foregoing description is only of the preferred embodiments of the invention and is not intended to limit the scope of the invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The method for evaluating the electromechanical state of the bulb tubular turbine unit is characterized by comprising the following steps of:
dynamically simulating a bulb tubular unit simulation model built in advance under a target working condition and different electromechanical states to obtain historical operation data and real-time operation data;
acquiring unit design data and unit attribute data of the bulb tubular unit simulation model;
aiming at each electromechanical state in the electromechanical states, according to the unit design data, the unit attribute data and the historical operation data and the real-time operation data obtained by dynamic simulation in the corresponding states, corresponding modeling samples are obtained by sorting in the following modes: taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2;
Dividing all the modeling samples into a training set, a verification set and a test set;
training, verifying and testing the machine learning model based on the long-term and short-term memory neural network by using the training set, the verifying set and the testing set in sequence to obtain an electromechanical state classification model passing the verification and the test;
acquiring operation data of each unit period in the current M nearest unit periods of the bulb tubular target unit under the target working condition by a production monitoring system;
inputting the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods into the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states;
and determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
2. The bulb through-flow unit electromechanical state evaluation method according to claim 1, wherein the target operating condition comprises an organic unit ready state operating condition, a unit idle state operating condition, a unit power generation state operating condition, a unit shutdown state operating condition, or a unit overhaul state operating condition.
3. The method for evaluating the electromechanical state of a bulb through-flow unit according to claim 2, wherein when the target working condition comprises an organic unit maintenance working condition, the model input data further comprises unit power generation state data of the bulb through-flow unit simulation model before and after the unit maintenance working condition, and the sample data to be tested further comprises unit power generation state data of the bulb through-flow target unit before and after the unit maintenance working condition.
4. The method of evaluating the electromechanical state of a bulb through-flow unit according to claim 1, wherein the plurality of electromechanical states includes a healthy state, an abnormal state, a fault state, a condition to be serviced, and a damaged state.
5. The method for evaluating the electromechanical state of a bulb through-flow unit according to claim 1, wherein the training set, the verification set and the test set are applied to train, verify and test the machine learning model based on the long-short-term memory neural network in sequence to obtain an electromechanical state classification model passing the verification and the test, and the method comprises the following steps:
training and verifying the machine learning model based on the long-term and short-term memory neural network in sequence by using the training set and the verification set to obtain a verified initial electromechanical state classification model;
Inputting corresponding model input data into the initial electromechanical state classification model aiming at each test sample in the test set to obtain a corresponding electromechanical state classification result;
obtaining a confidence analysis report of the initial electromechanical state classification model through confidence analysis according to the model output data of each test sample and electromechanical state classification results;
and when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model exceeds a preset reliability threshold, judging that the test is passed, and taking the initial electromechanical state classification model as a final electromechanical state classification model.
6. The bulb through-flow unit electromechanical state evaluation method according to claim 5, characterized in that after obtaining a confidence analysis report of the initial electromechanical state classification model by confidence analysis, the method further comprises:
when the confidence analysis report indicates that the classification reliability of the initial electromechanical state classification model does not exceed a preset reliability threshold, unit design data, unit attribute data, optimal operation data and/or operation setting parameters of the bulb through-flow unit simulation model are fed back and adjusted, wherein the operation setting parameters comprise a common reminding trigger value, an emergency alarm trigger value and/or a shutdown action trigger value;
And re-applying the bulb through-flow unit simulation model to perform dynamic simulation under the target working condition and the electromechanical states so as to re-model based on simulation results.
7. The method for evaluating the electromechanical state of the bulb through-flow unit according to claim 1, wherein the production monitoring system comprises a bulb through-flow unit bearing oil monitoring subsystem, a governor oil monitoring subsystem, a technical water supply monitoring subsystem, an excitation monitoring subsystem, an axial flow fan monitoring subsystem and/or a unit on-line monitoring subsystem.
8. The electromechanical state evaluation device of the bulb through-flow unit is characterized by comprising a dynamic simulation module, a data acquisition module, a sample arrangement module, a sample division module, a model establishment module, a model application module and an evaluation determination module;
the dynamic simulation module is used for dynamically simulating to obtain historical operation data and real-time operation data under the target working condition and different electromechanical states by applying a pre-built bulb tubular unit simulation model;
the data acquisition module is used for acquiring unit design data and unit attribute data of the bulb through-flow unit simulation model;
The sample arrangement module is respectively in communication connection with the dynamic simulation module and the data acquisition module, and is used for arranging and obtaining corresponding modeling samples according to the unit design data, the unit attribute data, and historical operation data and real-time operation data obtained by dynamic simulation in the corresponding states for each electromechanical state in the electromechanical states according to the following modes: taking the unit design data, the unit attribute data and the running data which are obtained through dynamic simulation in the corresponding states and are in each unit period in the continuous M unit periods as model input data, and taking the corresponding states as model output data to obtain a modeling sample containing the model input data and the model output data, wherein M represents a positive integer not less than 2;
the sample dividing module is in communication connection with the sample sorting module and is used for dividing all the modeling samples into a training set, a verification set and a test set;
the model building module is in communication connection with the sample dividing module and is used for applying the training set, the verification set and the test set to train, verify and test the machine learning model based on the long-period memory neural network in sequence to obtain an electromechanical state classification model passing the verification and the test;
The data acquisition module is also used for acquiring the operation data of the bulb tubular target unit in each unit period in the current M most recent unit periods under the target working condition through the production monitoring system;
the model application module is respectively in communication connection with the model building module and the data acquisition module and is used for taking the unit design data, the unit attribute data and the operation data of each unit period in the current M unit periods as sample data to be tested, and inputting the electromechanical state classification model to obtain electromechanical state classification results, wherein the electromechanical state classification results comprise confidence degrees for dividing the current electromechanical state of the bulb through-flow type target unit into the electromechanical states;
and the evaluation determining module is in communication connection with the model application module and is used for determining a certain electromechanical state corresponding to the maximum confidence as a current electromechanical state evaluation result of the bulb tubular target unit.
9. A computer device comprising a memory, a processor and a transceiver in communication connection in sequence, wherein the memory is used for storing a computer program, the transceiver is used for receiving and transmitting messages, and the processor is used for reading the computer program and executing the method for evaluating the electromechanical state of the bulb through-flow unit according to any one of claims 1 to 7.
10. A computer-readable storage medium, wherein instructions are stored on the computer-readable storage medium, which when run on a computer, perform the bulb through-flow unit electromechanical state evaluation method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311684207.0A CN117648628A (en) | 2023-12-07 | 2023-12-07 | Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311684207.0A CN117648628A (en) | 2023-12-07 | 2023-12-07 | Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117648628A true CN117648628A (en) | 2024-03-05 |
Family
ID=90047609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311684207.0A Pending CN117648628A (en) | 2023-12-07 | 2023-12-07 | Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117648628A (en) |
-
2023
- 2023-12-07 CN CN202311684207.0A patent/CN117648628A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112596495B (en) | Industrial equipment fault diagnosis method and system based on knowledge graph | |
| CN111459700B (en) | Equipment fault diagnosis method, diagnosis device, diagnosis equipment and storage medium | |
| CN106354985B (en) | Power distribution system reliability assessment method considering distributed power supply | |
| CN107516279B (en) | Automatic early warning method for network public sentiment | |
| CN110417011A (en) | A kind of online dynamic secure estimation method based on mutual information Yu iteration random forest | |
| CN109800995A (en) | A kind of grid equipment fault recognition method and system | |
| CN112217207A (en) | Transient voltage stability margin prediction method and device | |
| Xiyun et al. | Wind power probability interval prediction based on bootstrap quantile regression method | |
| Zhang et al. | A new adaptive Mamdani-type fuzzy modeling strategy for industrial gas turbines | |
| CN116089231A (en) | Fault alarm method and device, electronic equipment and storage medium | |
| CN111767193A (en) | Server data anomaly detection method and device, storage medium and equipment | |
| CN108805419B (en) | Power grid node importance calculation method based on network embedding and support vector regression | |
| CN115600695B (en) | Fault diagnosis method for metering equipment | |
| CN117648628A (en) | Method, device, equipment and medium for evaluating electromechanical state of bulb through-flow unit | |
| CN116108749A (en) | Wind turbine flow field prediction method based on deep learning | |
| CN112712305B (en) | Aircraft system and health assessment method and device thereof | |
| CN113887101A (en) | Visualization method and device of network model, electronic equipment and storage medium | |
| CN110852544B (en) | Reliability evaluation method and device for wind generating set | |
| CN113919609A (en) | Power distribution network model quality comprehensive evaluation method and system | |
| Kumar et al. | Gas turbine engine operational data analysis for anomaly detection: Statistical vs. neural network approach | |
| Wang et al. | Fault Assessment of Hydraulic System Based on Gaussian Mixture Model | |
| CN113944801B (en) | Gas pressure regulator performance detection method and device based on data analysis | |
| Chaoyang et al. | Embedded fault diagnosis expert system on weapon equipment | |
| CN109960779A (en) | Repairable system analysis method for reliability based on uncertain service life and maintenance time | |
| Najar et al. | Comparative Machine Learning Study for Estimating Peak Cladding Temperature in AP1000 Under LOFW |
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 |