CN114881207A - Steam pressure prediction method and device based on LSTM deep circulation neural network - Google Patents
Steam pressure prediction method and device based on LSTM deep circulation neural network Download PDFInfo
- Publication number
- CN114881207A CN114881207A CN202210446939.5A CN202210446939A CN114881207A CN 114881207 A CN114881207 A CN 114881207A CN 202210446939 A CN202210446939 A CN 202210446939A CN 114881207 A CN114881207 A CN 114881207A
- Authority
- CN
- China
- Prior art keywords
- steam pressure
- steam
- data
- neural network
- prediction 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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"
-
- 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
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P80/00—Climate change mitigation technologies for sector-wide applications
- Y02P80/10—Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier
- Y02P80/15—On-site combined power, heat or cool generation or distribution, e.g. combined heat and power [CHP] supply
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Control Of Steam Boilers And Waste-Gas Boilers (AREA)
Abstract
The application discloses a steam pressure prediction method and a steam pressure prediction device based on an LSTM deep circulation neural network, wherein a steam boiler system operation data is collected to obtain an operation parameter matrix, the steam boiler system operation parameter matrix is recombined according to an LSTM model to obtain a sample set, the sample amount in the sample set is randomly selected as a training data set and a verification data set, and the training data set and the verification data set are used as input to construct an initial steam pressure prediction model based on the LSTM deep circulation neural network; optimizing the super-parameter in the initial steam pressure prediction model, and training the LSTM deep circulation neural network steam pressure prediction model by taking the optimized super-parameter as a control quantity to obtain a steam pressure dynamic prediction model; inputting the real-time operation data of the steam boiler system to obtain a steam pressure predicted value. The characteristics of high coupling and large hysteresis of a steam boiler system are overcome, and a prediction model with strong applicability and high prediction precision is established.
Description
Technical Field
The application relates to the technical field of industrial steam boiler combustion, in particular to a steam pressure prediction method and device based on an LSTM deep circulation neural network.
Background
The steam boiler is a heat exchange device for producing steam, which utilizes the energy released by burning coal, oil, fuel gas, biomass and other fuels, and transfers the energy to water through a heat transfer process, so that the water is changed into steam, the steam is supplied to heat energy required by industrial production, or is converted into mechanical energy through a steam power machine, or is converted into electric energy through a steam turbine generator, or is directly supplied to downstream heat users. The steam pressure is too high, which may cause safety accidents such as boiler explosion; the steam pressure is too low, and the steam quality is inevitably reduced; the steam pressure must be in a moderate range. Therefore, boiler steam pressure is one of the important process indicators that is of interest in the production process.
Steam production generally involves three simultaneous processes: fuel combustion process, water vaporization process, flue gas to water conversion process. The factors influencing the steam pressure are many, various factors respond to the final steam pressure differently, the factors are mutually coupled, and the pure lag of the boiler system is large. In the existing production process, an operator/automatic control system controls steam pressure in a relatively controllable range through manual experience or a traditional control method, the steam pressure fluctuation is large, the product quality is difficult to guarantee, the production equipment efficiency cannot be maximized due to frequent load change, and the production economy is poor. Therefore, accurate prediction of steam pressure is significant for guiding actual production, improving automatic control effect, guaranteeing steam quality, improving production equipment efficiency and increasing the income of manufacturers.
Disclosure of Invention
Therefore, the steam pressure prediction method and device based on the LSTM deep circulation neural network are provided, and the problems that in the prior art, all factors influencing the steam pressure are mutually coupled, and the pure delay of a boiler system is large are solved.
In order to achieve the above purpose, the present application provides the following technical solutions:
in a first aspect, a method for predicting steam pressure based on an LSTM deep cycle neural network includes:
acquiring production operation data of a steam boiler, and storing the operation data into a historical database;
preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix;
recombining steam boiler system parameters according to an LSTM algorithm to obtain a model parameter matrix sample set;
obtaining an initial steam pressure prediction model of the LSTM deep circulation neural network according to the model parameter matrix sample set;
optimizing the super-parameters in the initial steam pressure prediction model, and training the initial steam pressure prediction model by taking the optimized super-parameters as control quantities to obtain a steam pressure dynamic prediction model;
and arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model, and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
Further, preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix, which specifically comprises:
sequencing all historical data according to a time sequence to obtain a historical data table;
removing missing values in the historical data table, and deleting the row of data if a certain position in the row vector of the steam boiler system is empty;
resampling the data according to the time variable to obtain an operation parameter matrix X of the steam boiler system;
carrying out standardized calculation on the operation parameter matrix X to obtain a parameter standardized matrix X * ；
The calculation formula of the normalized conversion is as follows:
wherein x is ik Representing the operation data of the ith row and the kth column in the parameter matrix X of the steam boiler system,
representing X in steam boiler system parameter matrix X ik The normalized running data is then compared to the standard running data,
is the mean value of the operation data of the k column in the steam boiler system parameter matrix X,
ε is a non-zero constant for the variance of the operating data in column k of steam boiler system parameter matrix X.
Further, the row vector of the operation parameter matrix X is the operation data of the steam boiler system at each moment, and the column vector of the operation parameter matrix X is the value of each operation attribute of the steam boiler system.
Further, the model parameter matrix is:
wherein T is the time step number in the super-parameter of the steam pressure initial prediction model based on the LSTM deep circulation neural network,
representing the normalized matrix X of the steam boiler system obtained at the time T- (T-a) * Data, wherein the value range of the parameter a is more than or equal to 1 and less than or equal to T, y t Is the output parameter of the LSTM model and represents the steam pressure at time t.
Further, obtaining an initial steam pressure prediction model of the LSTM deep cycle neural network according to the model parameter matrix sample set specifically includes:
randomly selecting 80% of samples in the sample set as a training data set, using the remaining 20% as a verification data set, using the training data set and the verification data set as model inputs, establishing an LSTM deep circulation neural network model, and setting the model input characteristic size, the number of network layers, the number of neurons in each layer, regularization parameters, the output characteristic size, the network iteration times and an optimization method to obtain an initial steam pressure prediction model based on the LSTM deep circulation neural network.
Further, optimizing the hyper-parameters in the initial steam pressure prediction model specifically includes:
firstly, optimizing the hyperparameter T to obtain the optimal hyperparameter T, and then optimizing the batch-size in the hyperparameter in the same optimization mode; and finally, optimizing the learning rate of the super-parameter lr.
Furthermore, the super parameter T is 15, the batch-size is 256, and lr is 0.0004.
Further, the steam boiler production operation data is obtained from the DCS system.
Further, the acquisition period of the operation data is 1 second.
In a second aspect, a steam pressure prediction device based on an LSTM deep cycle neural network includes:
the operation data acquisition module is used for acquiring the production operation data of the steam boiler and storing the operation data into a historical database;
the preprocessing module is used for preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix;
the parameter recombination module is used for recombining the steam boiler system parameters according to an LSTM algorithm to obtain a model parameter matrix sample set;
the initial steam pressure prediction model training module is used for obtaining an LSTM deep circulation neural network initial steam pressure prediction model according to the model parameter matrix sample set;
the steam pressure dynamic prediction model training module is used for optimizing the super-parameters in the initial steam pressure prediction model, taking the optimized super-parameters as control variables and training the initial steam pressure prediction model to obtain a steam pressure dynamic prediction model;
and the output module is used for arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
Compared with the prior art, the method has the following beneficial effects:
1. through historical operation data of the steam boiler system, a steam pressure prediction model based on an LSTM deep circulation neural network is constructed, the sequence dependence problem processing capacity of the LSTM deep circulation neural network is utilized, the characteristics of high coupling and large hysteresis of the steam boiler system are overcome, and the influence of interference factors such as frequent change of working conditions, mutual coupling of system parameters, large pure hysteresis of the system and the like on the steam pressure prediction accuracy can be ignored to a great extent.
2. The method has the characteristics of simplicity, strong adaptability and high prediction precision, and has the advantages of improving the automatic control effect, ensuring the quality of PVC (polyvinyl chloride), improving the efficiency of production equipment and increasing the income of manufacturers.
Drawings
To more intuitively illustrate the prior art and the present application, several exemplary drawings are given below. It should be understood that the specific shapes, configurations and illustrations in the drawings are not to be construed as limiting, in general, the practice of the present application; for example, it is within the ability of those skilled in the art to make routine adjustments or further optimizations based on the technical concepts disclosed in the present application and the exemplary drawings, for the increase/decrease/attribution of certain units (components), specific shapes, positional relationships, connection manners, dimensional ratios, and the like.
FIG. 1 is a first flowchart of a steam pressure prediction method based on an LSTM deep cycle neural network according to an embodiment of the present application;
fig. 2 is a second flowchart of a steam pressure prediction method based on an LSTM deep cycle neural network according to an embodiment of the present application.
Detailed Description
The present application will be described in further detail below with reference to specific embodiments thereof, with reference to the accompanying drawings.
In the description of the present application: "plurality" means two or more unless otherwise specified. The terms "first", "second", "third", and the like in this application are intended to distinguish the referenced objects without particular meaning in the technical meaning (e.g., emphasis on degree or order of importance, etc.) being construed). The terms "comprising," "including," "having," and the like, are intended to be inclusive and mean "not limited to" (some elements, components, materials, steps, etc.).
In the present application, terms such as "upper", "lower", "left", "right", "middle", and the like are generally used for easy visual understanding with reference to the drawings, and are not intended to absolutely limit the positional relationship in an actual product. Changes in these relative positional relationships are also considered to be within the scope of the present disclosure without departing from the technical concepts disclosed in the present disclosure.
Example one
Referring to fig. 1 and 2, the present embodiment provides a steam pressure prediction method based on an LSTM deep cycle neural network, including:
s1: acquiring production operation data of a steam boiler, and storing the operation data into a historical database;
specifically, according to preset input variables of a steam boiler steam pressure prediction model, operating data of a steam boiler system corresponding to the preset input variables are acquired from a DCS in real time, the acquisition period of the operating data is 1 second, and the acquired operating data comprise negative pressure of a hearth, temperature of the hearth, main steam pressure, main steam temperature, pipe network pressure, primary air quantity, secondary air quantity, boiler load, air supply pressure, smoke concentration, flue gas SO2, flue gas NOX, total fuel supply quantity, flue gas oxygen content and the like.
S2: preprocessing all historical data in a historical database to obtain a steam boiler system parameter standardization matrix;
specifically, the method comprises the following steps:
s21: sequencing all historical data according to a time sequence to obtain a historical data table;
s22: removing missing values in the historical data table, and if a certain position in the row vector of the steam boiler system is empty, deleting the row of data;
s23: resampling the data according to the time variable to obtain an operation parameter matrix X of the steam boiler system;
specifically, the data sheet with the abnormal values removed resamples data according to time variables, reorganizes the data in units of minutes, converts second-level data into minute-level data according to the average value, and stores the converted parameter values as the average value of the parameter in one minute in a matrix form, wherein the operation data of the steam boiler system at each moment is used as a row vector of the matrix, and the value of each operation attribute of the steam boiler system is used as a column vector of the matrix, so that a steam boiler system parameter matrix X is obtained.
S24: carrying out standardized calculation on the operation parameter matrix X to obtain a parameter standardized matrix X * ；
Specifically, the calculation formula of the normalized conversion is as follows:
wherein x is ik Representing the operation data of the ith row and the kth column in the parameter matrix X of the steam boiler system,
representing X in steam boiler system parameter matrix X ik The normalized running data is then compared to the standard running data,
is the mean value of the operation data of the k column in the steam boiler system parameter matrix X,
ε is a non-zero constant for the variance of the operating data in column k of steam boiler system parameter matrix X.
S3: recombining steam boiler system parameters according to an LSTM algorithm to obtain a model parameter matrix;
specifically, according to the data format requirement of the LSTM model, the parameters of the steam boiler system are subjected to data recombination to obtain a model parameter matrix X t The set of samples is then compared to the set of samples,
model parameter matrix X t Comprises the following steps:
wherein T is the time step number in the super-parameter of the steam pressure initial prediction model based on the LSTM deep circulation neural network,
representing the normalized matrix X of the steam boiler system obtained at the time T- (T-a) * Data, wherein the value range of the parameter a is more than or equal to 1 and less than or equal to T, y t Is the output parameter of the LSTM model and represents the steam pressure at time t.
S4: obtaining an initial steam pressure prediction model of the LSTM deep circulation neural network according to the model parameter matrix sample set;
specifically, 80% of samples in a sample set are randomly selected as a training data set, the rest 20% of samples are selected as a verification data set, the training data set and the verification data set are used as model inputs, an LSTM deep cycle neural network model is established, the model input characteristic size, the number of network layers, the number of neurons in each layer, regularization parameters, the output characteristic size, the network iteration times and an optimization method are set, and the initial steam pressure prediction model based on the LSTM deep cycle neural network is obtained.
S5: optimizing the super-parameters in the initial steam pressure prediction model, and training the initial steam pressure prediction model by taking the optimized super-parameters as control quantities to obtain a steam pressure dynamic prediction model;
s51: initializing a hyper-parameter in the initial steam pressure prediction model by using a hyper-parameter optimization subunit, and then optimizing the hyper-parameter to obtain an optimized hyper-parameter;
specifically, firstly, initializing the hyper-parameters to be optimized; judging whether the current parameter is an optimal value or not according to the training effect and the model loss in the verification data set by changing the value of the hyper-parameter; when the hyper-parameter optimization is carried out, firstly, the hyper-parameter T is optimized, and after the optimal T is obtained, the batch-size in the hyper-parameter is optimized in the same optimization mode; and finally, optimizing the learning rate of the super-parameter lr.
The super parameter T is 15, the batch-size is 256, and the lr is 0.0004, and the initialized value of the super parameter provided herein is only used as a reference value, and may be adjusted according to actual needs, and is not limited herein.
S52: and training the initial steam pressure prediction model by using the model training subunit and taking the optimized super-parameter as a control quantity to obtain a steam pressure dynamic prediction model based on the LSTM deep circulation neural network.
S6: and arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model, and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
It should be noted that the method for sorting the real-time data is the same as the implementation steps mentioned in S2-S5, and input variables are obtained; and then inputting the input variable into a prediction model to obtain a predicted value of the steam pressure.
According to the steam pressure prediction method based on the LSTM deep circulation neural network, a steam pressure prediction model based on the LSTM deep circulation neural network is constructed through historical operation data of a steam boiler system, the sequence dependence problem processing capacity of the LSTM deep circulation neural network is utilized, the characteristics of high coupling and large hysteresis of the steam boiler system are overcome, the influence of interference factors such as frequent change of working conditions, mutual coupling of system parameters, large system pure hysteresis and the like on the steam pressure prediction accuracy can be ignored to a great extent, and the steam pressure prediction method based on the LSTM deep circulation neural network has the advantages of being simple, strong in adaptability and high in prediction accuracy, and providing solid guarantee for improving the automatic control effect, guaranteeing the PVC quality, improving the efficiency of production equipment and increasing the income of a production manufacturer.
Example two
The embodiment provides a steam pressure prediction device based on an LSTM deep cycle neural network, which includes:
the operation data acquisition module is used for acquiring the production operation data of the steam boiler and storing the operation data into a historical database;
the preprocessing module is used for preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix;
the parameter recombination module is used for recombining the steam boiler system parameters according to an LSTM algorithm to obtain a model parameter matrix sample set;
the initial steam pressure prediction model training module is used for obtaining an LSTM deep circulation neural network initial steam pressure prediction model according to the model parameter matrix sample set;
the steam pressure dynamic prediction model training module is used for optimizing the super-parameters in the initial steam pressure prediction model, taking the optimized super-parameters as control quantities, and training the initial steam pressure prediction model to obtain a steam pressure dynamic prediction model;
and the output module is used for arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
For the specific definition of the steam pressure prediction device based on the LSTM deep circulation neural network, reference may be made to the above definition of the steam pressure prediction method based on the LSTM deep circulation neural network, and details are not repeated here.
All the technical features of the above embodiments can be arbitrarily combined (as long as there is no contradiction between the combinations of the technical features), and for brevity of description, all the possible combinations of the technical features in the above embodiments are not described; these examples, which are not explicitly described, should be considered to be within the scope of the present description.
The present application has been described in considerable detail with reference to certain embodiments and examples thereof. It should be understood that several conventional adaptations or further innovations of these specific embodiments may also be made based on the technical idea of the present application; however, such conventional modifications and further innovations may also fall within the scope of the claims of the present application as long as they do not depart from the technical idea of the present application.
Claims (10)
1. A steam pressure prediction method based on an LSTM deep cycle neural network is characterized by comprising the following steps:
acquiring production operation data of a steam boiler, and storing the operation data into a historical database;
preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix;
recombining steam boiler system parameters according to an LSTM algorithm to obtain a model parameter matrix sample set;
obtaining an initial steam pressure prediction model of the LSTM deep circulation neural network according to the model parameter matrix sample set;
optimizing the super-parameters in the initial steam pressure prediction model, and training the initial steam pressure prediction model by taking the optimized super-parameters as control quantities to obtain a steam pressure dynamic prediction model;
and arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model, and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
2. The LSTM deep-cycle neural network-based steam pressure prediction method of claim 1, wherein preprocessing all historical data in the historical database to obtain a steam boiler system parameter normalization matrix, specifically comprises:
sequencing all historical data according to a time sequence to obtain a historical data table;
removing missing values in the historical data table, and deleting the row of data if a certain position in the row vector of the steam boiler system is empty;
resampling the data according to the time variable to obtain an operation parameter matrix X of the steam boiler system;
carrying out standardized calculation on the operation parameter matrix X to obtain a parameter standardized matrix X * ；
The calculation formula of the normalized conversion is as follows:
wherein x is ik Representing the operation data of the ith row and the kth column in the parameter matrix X of the steam boiler system,
representing X in steam boiler system parameter matrix X ik The normalized running data is then compared to the standard running data,
is the mean value of the operation data of the k column in the steam boiler system parameter matrix X,
ε is a non-zero constant for the variance of the operating data in column k of steam boiler system parameter matrix X.
3. The LSTM deep-cycle neural network-based steam pressure prediction method of claim 2, wherein the row vector of the operation parameter matrix X is operation data of the steam boiler system at each time, and the column vector of the operation parameter matrix X is a value of each operation attribute of the steam boiler system.
4. The LSTM deep cyclic neural network-based steam pressure prediction method of claim 2, wherein the model parameter matrix is:
wherein T is the time step number in the super-parameter of the steam pressure initial prediction model based on the LSTM deep circulation neural network,
representing the steam obtained at time T- (T-a)Boiler system normalization matrix X * Data, wherein the value range of the parameter a is more than or equal to 1 and less than or equal to T, y t Is the output parameter of the LSTM model and represents the steam pressure at time t.
5. The LSTM deep-circulation neural network-based steam pressure prediction method of claim 1, wherein obtaining the LSTM deep-circulation neural network initial steam pressure prediction model according to the model parameter matrix sample set specifically includes:
randomly selecting 80% of samples in the sample set as a training data set, using the remaining 20% as a verification data set, using the training data set and the verification data set as model inputs, establishing an LSTM deep circulation neural network model, and setting the model input characteristic size, the number of network layers, the number of neurons in each layer, regularization parameters, the output characteristic size, the network iteration times and an optimization method to obtain an initial steam pressure prediction model based on the LSTM deep circulation neural network.
6. The LSTM deep-circulation neural network-based steam pressure prediction method of claim 1, wherein the optimization of the hyper-parameters in the initial steam pressure prediction model specifically comprises:
firstly, optimizing the hyperparameter T to obtain the optimal hyperparameter T, and then optimizing the batch-size in the hyperparameter in the same optimization mode; and finally, optimizing the learning rate of the super-parameter lr.
7. The LSTM deep-cycle neural network-based steam pressure prediction method of claim 6, wherein the super parameter T is 15, batch-size is 256, and lr is 0.0004.
8. The LSTM deep-cycle neural network-based steam pressure prediction method of claim 1, wherein the steam boiler production operation data is obtained from a DCS system.
9. The LSTM deep-cycle neural network-based steam pressure prediction method of claim 1, wherein the acquisition period of the operational data is 1 second.
10. A steam pressure prediction device based on an LSTM deep cycle neural network is characterized by comprising the following components:
the operation data acquisition module is used for acquiring the production operation data of the steam boiler and storing the operation data into a historical database;
the preprocessing module is used for preprocessing all historical data in the historical database to obtain a steam boiler system parameter standardization matrix;
the parameter recombination module is used for recombining the parameters of the steam boiler system according to an LSTM algorithm to obtain a model parameter matrix sample set;
the initial steam pressure prediction model training module is used for obtaining an LSTM deep circulation neural network initial steam pressure prediction model according to the model parameter matrix sample set;
the steam pressure dynamic prediction model training module is used for optimizing the super-parameters in the initial steam pressure prediction model, taking the optimized super-parameters as control quantities, and training the initial steam pressure prediction model to obtain a steam pressure dynamic prediction model;
and the output module is used for arranging the real-time operation data of the steam boiler system according to the specification of the steam pressure dynamic prediction model and then dynamically inputting the data into the steam pressure dynamic prediction model to obtain a steam pressure prediction value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210446939.5A CN114881207A (en) | 2022-04-26 | 2022-04-26 | Steam pressure prediction method and device based on LSTM deep circulation neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210446939.5A CN114881207A (en) | 2022-04-26 | 2022-04-26 | Steam pressure prediction method and device based on LSTM deep circulation neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114881207A true CN114881207A (en) | 2022-08-09 |
Family
ID=82671144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210446939.5A Pending CN114881207A (en) | 2022-04-26 | 2022-04-26 | Steam pressure prediction method and device based on LSTM deep circulation neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114881207A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117113886A (en) * | 2023-10-23 | 2023-11-24 | 北京中环信科科技股份有限公司 | Pressure prediction method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109165789A (en) * | 2018-09-04 | 2019-01-08 | 广东电网有限责任公司 | The modeling method and device of emission of NOx of boiler amount prediction model based on LSTM |
| CN113241129A (en) * | 2021-05-18 | 2021-08-10 | 北京和隆优化科技股份有限公司 | Prediction method of PVC (polyvinyl chloride) moisture content based on LSTM (localized surface plasmon resonance) deep circulation neural network |
| CN114139305A (en) * | 2021-11-16 | 2022-03-04 | 国网河北能源技术服务有限公司 | Single valve flow characteristic optimization method based on turbine regulating stage pressure prediction model |
-
2022
- 2022-04-26 CN CN202210446939.5A patent/CN114881207A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109165789A (en) * | 2018-09-04 | 2019-01-08 | 广东电网有限责任公司 | The modeling method and device of emission of NOx of boiler amount prediction model based on LSTM |
| CN113241129A (en) * | 2021-05-18 | 2021-08-10 | 北京和隆优化科技股份有限公司 | Prediction method of PVC (polyvinyl chloride) moisture content based on LSTM (localized surface plasmon resonance) deep circulation neural network |
| CN114139305A (en) * | 2021-11-16 | 2022-03-04 | 国网河北能源技术服务有限公司 | Single valve flow characteristic optimization method based on turbine regulating stage pressure prediction model |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117113886A (en) * | 2023-10-23 | 2023-11-24 | 北京中环信科科技股份有限公司 | Pressure prediction method and device |
| CN117113886B (en) * | 2023-10-23 | 2024-02-06 | 北京中环信科科技股份有限公司 | Pressure prediction method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103576655B (en) | A kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system | |
| CN104534507B (en) | A kind of boiler combustion optimization control method | |
| CN106991539B (en) | Energy system optimal scheduling method and device | |
| CN103064289A (en) | Multiple-target operation optimizing and coordinating control method and device of garbage power generator | |
| WO2024060488A1 (en) | Method based on deep recurrent neural network and evolutionary computation for optimizing combustion of industrial boiler | |
| CN110400018B (en) | Operation control method, system and device for coal-fired power plant pulverizing system | |
| CN110889638B (en) | Thermal power plant operation evaluation method and related equipment | |
| CN111457392B (en) | Intelligent setting method for air quantity in urban domestic garbage incineration process | |
| CN114881207A (en) | Steam pressure prediction method and device based on LSTM deep circulation neural network | |
| CN101598927B (en) | Control system of soda carbonization technique based on neural network and control method thereof | |
| CN111931436A (en) | Burner nozzle air quantity prediction method based on numerical simulation and neural network | |
| CN111401652A (en) | Boiler optimization method and system based on CO online detection | |
| CN106610589A (en) | Online hardware closed-loop network source coordination linear active-disturbance-rejection control method | |
| CN112950409A (en) | Production scheduling optimization method of gas and steam energy comprehensive utilization system | |
| CN116951398A (en) | Model optimization control method and system for reducing left and right steam temperature deviation of boiler reheater | |
| CN115111601B (en) | Multi-target boiler combustion optimization control method based on embedded algorithm fusion under variable load | |
| CN111428906B (en) | Industrial boiler steam volume prediction method based on image transformation | |
| CN111091251A (en) | Boiler operation optimization method and system based on big data technology | |
| CN111897203A (en) | Heater economic regulation method based on data driving and mechanism modeling state monitoring | |
| CN113110628A (en) | Water level control method of pressurized water reactor deaerator based on PSO | |
| CN113110033A (en) | Heat collection control system based on fuzzy PID algorithm ASHP | |
| CN112348696A (en) | Heat supply unit peak regulation upper limit evaluation method and system based on BP neural network | |
| CN114779722B (en) | Intelligent combustion optimization control system and method for coal-fired power station boiler | |
| Liu et al. | Research on Energy Management Strategy of Captive Power Plant Based on Hybrid Method. | |
| CN114413249B (en) | Data analysis method for power station boiler efficiency benchmarking optimization |
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 |