CN116339410A - Superheated steam temperature prediction control method based on deep learning model - Google Patents
Superheated steam temperature prediction control method based on deep learning model Download PDFInfo
- Publication number
- CN116339410A CN116339410A CN202310317470.XA CN202310317470A CN116339410A CN 116339410 A CN116339410 A CN 116339410A CN 202310317470 A CN202310317470 A CN 202310317470A CN 116339410 A CN116339410 A CN 116339410A
- Authority
- CN
- China
- Prior art keywords
- steam temperature
- superheated steam
- deep learning
- control
- learning 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
- 238000013136 deep learning model Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 25
- 238000004364 calculation method Methods 0.000 claims abstract description 14
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 23
- 239000011159 matrix material Substances 0.000 claims description 19
- 230000008859 change Effects 0.000 claims description 17
- 238000013500 data storage Methods 0.000 claims description 11
- 238000005070 sampling Methods 0.000 claims description 9
- 230000009467 reduction Effects 0.000 claims description 5
- 230000009471 action Effects 0.000 claims description 3
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 238000002485 combustion reaction Methods 0.000 claims description 3
- 238000010298 pulverizing process Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 239000000779 smoke Substances 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 4
- 239000003245 coal Substances 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000000227 grinding Methods 0.000 description 2
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000007664 blowing Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 239000003546 flue gas Substances 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000008121 plant development Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 239000004071 soot Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D23/00—Control of temperature
- G05D23/19—Control of temperature characterised by the use of electric means
- G05D23/20—Control of temperature characterised by the use of electric means with sensing elements having variation of electric or magnetic properties with change of temperature
-
- 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
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention relates to a superheated steam temperature prediction control method based on a deep learning model. Compared with the prior art, the method has the advantages of simple engineering application calculation, further improvement of overheat steam temperature control performance and the like.
Description
Technical Field
The invention relates to the field of thermal intelligent control and protection, in particular to a superheated steam temperature prediction control method based on a deep learning model.
Background
With the deep promotion of economic structure adjustment and transformation upgrading in China and the powerful implementation of electric power system reform, the intelligent power plant integrates the technologies of Internet, big data, artificial intelligence and the like under the driving of policies such as energy conservation, consumption reduction, emission reduction and the like, and the intelligent power plant has become the main trend of power plant development through the promotion of intelligent operation management, intelligent overhaul safety, intelligent control and the like, and the operation of the intelligent power plant can effectively promote the core competitiveness of the power plant and promote the sustainable development of the power plant.
The traditional superheated steam temperature control generally adopts control methods such as cascade PID, smith, state observer, multi-model predictive control and the like. Because the superheated steam temperature object has larger delay inertia, more disturbance factors and larger difference of object characteristics under different load working conditions, the control method can only solve the control problem of the superheated steam temperature under specific disturbance or specific working conditions, and can not solve the control problem of the full load working conditions and the comprehensive disturbance factors. With the rapid development of intelligent control technology and data mining technology, the application of deep learning technology to modeling and prediction of complex objects of thermal power generating units becomes an important research and application field.
The deep learning model is provided with the low-level features to be combined into more abstract high-level features, so that the deep learning model can be better applied to complex high-dimensional, nonlinear, time-varying and massive data application scenes of the thermal power generating unit. However, the deep learning model is a nonlinear mathematical model, so that the deep learning model cannot be directly used as a prediction model to carry out optimal control solution, and an intelligent optimizing algorithm is generally adopted, but the engineering application requirements cannot be met due to complex calculation of the deep learning model and the optimizing algorithm. At present, a deep learning model mainly adopts an early warning or intelligent feedforward control mode in actual engineering control of a complex object of a thermal power generating unit, so that the deep learning model cannot be conveniently embedded into control solution, and the prediction control function of the deep learning model is fully exerted.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a superheated steam temperature prediction control method based on a deep learning model.
The aim of the invention can be achieved by the following technical scheme:
according to one aspect of the invention, a method for controlling overheat steam temperature prediction based on a deep learning model is provided, the method comprises the steps of establishing a deep learning model of a boiler overheat steam temperature system, respectively calculating the change condition of overheat steam temperature at future time under different control quantities based on the model, selecting the control quantity corresponding to the optimal control result from the model as a controller to output, and controlling overheat steam temperature, so that the operation stability of the overheat steam temperature of a unit is improved.
As a preferred technical solution, the method specifically comprises the following steps:
step S1, taking relevant parameters of a boiler pulverizing system, a combustion system and an air and smoke system as characteristic input parameters, selecting historical operation data, and adopting a deep learning model for training and predicting future change of the overheat steam temperature of the boiler by using a cyclic neural network;
s2, reading parameter values of characteristic variables from a DCS in real time, and establishing a dynamic data storage matrix queue;
step S3, compensating the sampling period and the model time of training data according to the deep learning model in a dynamic data storage matrix queue, and extracting an input data matrix of the model at the current time;
s4, inputting the data matrix in the step S3 into a deep learning model to respectively calculate the change conditions of the boiler overheat steam temperature of 30S, 60S, 90S, 120S, 150S and 180S;
step S5, acquiring a future set value time sequence of the superheated steam temperature set value
Comparing the predicted temperature change data of the 6 future superheated steam with the predicted temperature change data of the 6 future superheated steam in the step S4, and performing prediction control;
step S6, determining the flow rate p of the desuperheating water 2 The variation limit value [ -deltap ] in one control period a ,Δp b ]In the overheat steam temperature prediction control, the newly set flow rate of the desuperheating water is used
Instead of the actual flow rate p of the desuperheating water at the current moment 2 (k) Model predictive calculation is carried out again, and overheat steam temperature change value in future time is obtained>
Step S7, calculating different values
Superheated steam temperature predictive control performance index J under input i ；
Step S8, selecting the minimum J i Corresponding to
And as a water supply flow command at the next moment, performing overheat steam temperature control.
As a preferred embodiment, in the step S1, 6 models for predicting future changes of the boiler superheated steam temperature of 30S, 60S, 90S, 120S, 150S and 180S are built together.
As a preferred technical solution, the dynamic data storage matrix queue in step S2 specifically includes:
where k represents the current time, p j (k) For the parameter of the jth characteristic variable at the time k, j=1, 2, …, n, the data t is the sampling period, and d is the time step.
In the step S2, after the real-time data of the feature variables are read, a shift operation is performed on the dynamic data storage matrix, and the real-time data are assigned to the parameters corresponding to the current time.
As a preferable technical solution, in the step S3, the input data matrix of the model current time is specifically:
where k represents the current time, p j (k) For the parameter of the jth characteristic variable at the time k, j=1, 2, …, n, the data t is the sampling period, and d is the time step.
As a preferable technical solution, in the step S6, a newly set flow rate of the desuperheating water
The specific calculation is as follows:
Δp i ＝min(-Δp a +i·Δp,Δp b ),i＝0,1,…,m
wherein Δp is a temperature reduction water flow control interval, i is the number of times of calculation, and Ceiling is an upward rounding operation.
In the step S7, as a preferable embodiment, the superheated steam temperature prediction control performance index J i The specific calculation is as follows:
wherein ω is a calculation coefficient of the flow rate variation of the desuperheating water;
is->
Wall temperature predicted value of future moment under action; />
Is the superheated steam temperature set value at the future moment.
As a preferable technical solution, the prediction model in the step S1 may be adjusted according to actual control requirements.
As an optimal technical scheme, related parameters in the step S1 comprise unit power, water supply flow, A-F coal grinding quantity and total coal quantity, total air quantity, grinding A-F primary air quantity, primary air pressure, opening of secondary air baffles of each layer of burner, water supply enthalpy value, swing valve positions of each corner burner, main steam pressure, middle point temperature, hearth area soot blowing signals, secondary A side desuperheating water flow, secondary A side desuperheater outlet steam temperature and secondary A side inlet overheat steam temperature.
Compared with the prior art, the invention provides a deep learning predictive control method which is convenient for practical engineering application aiming at the control problems of large delay inertia, more disturbance factors, nonlinear characteristics and the like of an overheat steam temperature control object.
Drawings
FIG. 1 is a flow chart of the superheated steam temperature prediction control method based on the deep learning model.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
The method is used for solving the control problems of large delay inertia, more disturbance factors, nonlinear characteristics and the like of an overheat steam temperature control object, and is convenient for practical engineering application.
As shown in fig. 1, the invention takes the prediction control of the superheated steam temperature at the second-stage a side of a boiler of a certain 1000MW ultra supercritical unit as an example, and the specific process is as follows:
1) And taking relevant parameters of a boiler pulverizing system, a combustion system and a flue gas system as characteristic input parameters, selecting historical operation data, adopting a circulating neural network to train a deep learning model for predicting future change of the superheated steam temperature of the secondary A side of the boiler, setting a data sampling period to be 20s, setting a time step to be 10 steps, and setting up 6 models for predicting future changes of the wall temperature of the boiler for 30s, 60s, 90s, 120s, 150s and 180 s. The model input and output characteristic parameters are shown in table 1.
TABLE 1
2) Reading parameter values of characteristic variables in the table 1 from a DCS system in real time, and establishing a dynamic data storage matrix queue for driving the model in real time in consideration of large difference between a model sampling period and a DCS data period;
wherein k represents the current time, p j (k) (j=1, 2, …, n) is the parameter of the jth feature variable at time k. After the real-time data of the characteristic variables are read, shifting operation is carried out on the dynamic data storage matrix, and the real-time data are assigned to the parameters corresponding to the current moment (k moment).
3) Extracting an input data matrix of the current time of the model from a dynamic data storage matrix queue according to the sampling period and model time compensation of training data of the deep learning model:
4) Inputting the data matrix in the step 3) into a deep learning model to respectively calculate the change conditions of the boiler superheated steam temperature of 30s, 60s, 90s, 120s, 150s and 180s in future;
5) Acquiring future set value time sequence of superheated steam temperature set value
Comparing the predicted temperature change data of the 6 future superheated steam with the predicted temperature change data of the 6 future superheated steam obtained in the step 4), and performing prediction control;
6) Determining the flow rate p of the desuperheating water 2 The variation limit value [ -deltap ] in one control period a ,Δp b ]In the overheat steam temperature prediction control, the newly set flow rate of the desuperheating water is used
Instead of the actual flow rate p of the desuperheating water at the current moment 2 (k) Model predictive calculation is carried out again, and overheat steam temperature change value in future time is obtained>
Δp i ＝min(-Δp a +i·Δp,Δp b ),i＝0,1,…,m
Wherein Deltap is the temperature reduction water flow control interval which is taken as 2t/h; i is the calculated times; ceiling is an upward rounding operation; Δp a And Δp b Taking-10 t/h and 10t/h respectively.
7) Calculating the difference
The overheat steam temperature prediction control performance index under the input:
wherein ω is a calculation coefficient of the flow rate variation of the desuperheating water, and is taken as 0.01;
is->
Wall temperature predicted value of future moment under action; />
Is the superheated steam temperature set value at the future moment.
8) Selecting the minimum J i Corresponding to
And as a water supply flow command at the next moment, performing overheat steam temperature control.
While the invention has been described with reference to certain preferred embodiments, it will be understood by those skilled in the art that various changes and substitutions of equivalents may be made and equivalents will be apparent to those skilled in the art without departing from the scope of the invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (9)
1. A superheated steam temperature prediction control method based on a deep learning model is characterized in that the method is characterized in that a deep learning model of a boiler superheated steam temperature system is established, the change condition of the superheated steam temperature at the future moment under different control quantities is calculated based on the model, and the control quantity corresponding to the optimal control result is selected from the control quantity to be output as a controller for superheated steam temperature control, so that the operation stability of the superheated steam temperature of a unit is improved.
2. The method for controlling the prediction of the superheated steam temperature based on the deep learning model according to claim 1, which is characterized by comprising the following steps:
step S1, taking relevant parameters of a boiler pulverizing system, a combustion system and an air and smoke system as characteristic input parameters, selecting historical operation data, and adopting a deep learning model for training and predicting future change of the overheat steam temperature of the boiler by using a cyclic neural network;
s2, reading parameter values of characteristic variables from a DCS in real time, and establishing a dynamic data storage matrix queue;
step S3, compensating the sampling period and the model time of training data according to the deep learning model in a dynamic data storage matrix queue, and extracting an input data matrix of the model at the current time;
s4, inputting the data matrix in the step S3 into a deep learning model to respectively calculate the change conditions of the boiler overheat steam temperature of 30S, 60S, 90S, 120S, 150S and 180S;
step S5, acquiring a future set value time sequence of the superheated steam temperature set value
Comparing the predicted temperature change data of the 6 future superheated steam with the predicted temperature change data of the 6 future superheated steam in the step S4, and performing prediction control;
step S6, determining the flow rate p of the desuperheating water 2 The variation limit value [ -deltap ] in one control period a ,Δp b ]In the overheat steam temperature prediction control, the newly set flow rate of the desuperheating water is used
Instead of the actual flow rate p of the desuperheating water at the current moment 2 (k) Model predictive calculation is carried out again, and overheat steam temperature change value in future time is obtained>
Step S7, calculating different values
Superheated steam temperature predictive control performance index J under input i ；
Step S8, selecting the minimum J i Corresponding to
And as a water supply flow command at the next moment, performing overheat steam temperature control.
3. The method according to claim 2, wherein in the step S1, 6 models for predicting the future changes of the boiler superheated steam temperature of 30S, 60S, 90S, 120S, 150S and 180S are built together.
4. The method for controlling superheated steam temperature prediction based on deep learning model according to claim 2, wherein the dynamic data storage matrix queue in step S2 is specifically:
where k represents the current time, p j (k) For the parameter of the jth characteristic variable at the time k, j=1, 2, …, n, the data t is the sampling period, and d is the time step.
5. The method for controlling the prediction of the superheated steam temperature based on the deep learning model according to claim 2, wherein in the step S2, after the real-time data of the characteristic variables are read, a shift operation is performed on the dynamic data storage matrix, and the real-time data are assigned to the parameters corresponding to the current time.
6. The method for controlling the prediction of the superheated steam temperature based on the deep learning model according to claim 2, wherein in the step S3, the input data matrix of the model at the moment is specifically:
where k represents the current time, p j (k) For the parameter of the jth characteristic variable at the time k, j=1, 2, …, n, the data t is the sampling period, and d is the time step.
7. The method for predictive control of superheated steam temperature based on deep learning model as claimed in claim 2, wherein in the step S6, the newly set flow rate of the desuperheating water is
The specific calculation is as follows:
Δp i ＝min(-Δp a +i·Δp,Δp b ),i＝0,1,…,m
wherein Δp is a temperature reduction water flow control interval, i is the number of times of calculation, and Ceiling is an upward rounding operation.
8. The method according to claim 2, wherein in the step S7, the superheated steam temperature prediction control performance index J is i The specific calculation is as follows:
wherein ω is a calculation coefficient of the flow rate variation of the desuperheating water;
is->
Wall temperature predicted value of future moment under action; />
Is the superheated steam temperature set value at the future moment.
9. The method according to claim 2, wherein the prediction model in step S1 is adjustable according to actual control requirements.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310317470.XA CN116339410A (en) | 2023-03-28 | 2023-03-28 | Superheated steam temperature prediction control method based on deep learning model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310317470.XA CN116339410A (en) | 2023-03-28 | 2023-03-28 | Superheated steam temperature prediction control method based on deep learning model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116339410A true CN116339410A (en) | 2023-06-27 |
Family
ID=86894553
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310317470.XA Pending CN116339410A (en) | 2023-03-28 | 2023-03-28 | Superheated steam temperature prediction control method based on deep learning model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116339410A (en) |
-
2023
- 2023-03-28 CN CN202310317470.XA patent/CN116339410A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108227488B (en) | Sliding mode prediction control-based ultra-supercritical thermal power generating unit coordination control method | |
| CN102494336A (en) | Combustion process multivariable control method for CFBB (circulating fluidized bed boiler) | |
| CN102841540A (en) | MMPC-based supercritical unit coordination and control method | |
| CN106292277B (en) | Subcritical thermal power generating unit coordination control method based on global sliding mode control | |
| CN111637444B (en) | Nuclear power steam generator water level control method based on Q learning | |
| CN105955210A (en) | Exhaust-heat boiler and industrial boiler power generation coordinated operation dynamic optimization method and system | |
| CN115409396A (en) | Comprehensive energy system multi-time scale scheduling method based on double-layer rolling optimization | |
| WO2019085446A1 (en) | Self-order-reducing multi-loop centralized prediction control system for reheat steam temperature of double reheat unit | |
| CN108361683B (en) | Full load section reheat temperature intelligent control system | |
| CN112016754A (en) | Power station boiler exhaust gas temperature advanced prediction system and method based on neural network | |
| Wang et al. | Flexible electric power control for coal-fired units by incorporating feedwater bypass | |
| CN106707756A (en) | Extended state observer-integrated supercritical thermal power unit turbine-boiler coordinated control method | |
| CN110673482B (en) | Power station coal-fired boiler intelligent control method and system based on neural network prediction | |
| Gao et al. | Research on coordinated control system of drum boiler units considering energy demand decoupling | |
| CN114721253A (en) | Heating furnace temperature fractional order PID control system and method based on artificial bee colony algorithm | |
| CN106855691A (en) | For the double-deck control system of supercritical thermal power unit machine furnace system Steam Generator in Load Follow | |
| JPH08339204A (en) | Autonomous adaptive optimization control system for thermal power station | |
| CN116339410A (en) | Superheated steam temperature prediction control method based on deep learning model | |
| CN111413864A (en) | 600MW supercritical thermal power generating unit modeling and control method | |
| CN114371619B (en) | MGT-CCHP variable working condition dynamic energy efficiency optimization control method | |
| CN116540798A (en) | Boiler wall temperature prediction control method based on deep learning model | |
| CN114562713A (en) | Main steam temperature control method and system for power generation boiler | |
| CN103363812A (en) | Control method of cement clinker grate cooler | |
| CN107102550A (en) | A kind of ultra supercritical coal-fired unit controls the forecast Control Algorithm of separator temperature | |
| Ma et al. | Intelligent Compensation for the Set Values of PID Controllers to Improve Boiler Superheated Steam Temperature Control |
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 |