CN102840571B - Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam - Google Patents
Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam Download PDFInfo
- Publication number
- CN102840571B CN102840571B CN201210349737.5A CN201210349737A CN102840571B CN 102840571 B CN102840571 B CN 102840571B CN 201210349737 A CN201210349737 A CN 201210349737A CN 102840571 B CN102840571 B CN 102840571B
- Authority
- CN
- China
- Prior art keywords
- boiler
- data
- energy
- model
- identification
- 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.)
- Active
Links
Landscapes
- Control Of Steam Boilers And Waste-Gas Boilers (AREA)
- Feedback Control In General (AREA)
Abstract
The forecasting procedure that the invention discloses a kind of firepower power station boiler based on Subspace Identification output superheated steam comprises the steps: that 1 timing gathers creation data from level of factory DCS configuration software; 2 revise bad point, filtering and noise reduction; 3 calculate when the input/output variable needed for Model Distinguish; 4 utilize continuous acquisition and data configuration Hankel matrix after treatment, and carry out rollings renewal; Boiler process is approximated to linear state-space model by 5; Matrix { A, B, C, the D} of 6 use subspace theory identification state-space models; 7 utilize up-to-date input energy information and identification to obtain the output energy of the following some bats of model calculating; 8 utilize history predicted value and current forecast result averaging in the same time mutually and and utilizing the deviation of forecast in the past to revise.9 roll upgrades the data of identification model.The present invention is used for the online forecasting of the superheated steam output energy of boiler for producing process, and the optimal control for boiler combustion provides guidance and reference.
Description
Technical field
The present invention relates to the forecasting procedure that a kind of firepower power station boiler based on Subspace Identification exports superheated steam.
Background technology
The target that boiler combustion optimization controls is under certain unit load, obtains by operational factors such as adjustment boiler coal feeding, air distribution, air inducing the running status that efficient superheated steam exports energy and stability and safety.It is a very complicated production procedure that energy passes to superheated steam from coal supply, adopts during modelling by mechanism and is difficult to Confirming model structure and model parameter, and therefore more research concentrates on the black-box model of data-driven or the grey-box model of half mechanism half data at present.Boiler modeling method at present based on data-driven mainly contains artificial neural network and SVMs etc.Such as Electronic University Of Science & Technology Of Hangzhou Xue An gram, the modeling method (patent publication No. CN 101498459A) of boiler combustion optimization that proposes of the people such as Wang Chunlin, modeling is carried out in height segmentation according to unit operation load, radial basis neural network is set up in high load capacity section, and set up supporting vector machine model in underload section, namely set up one based on the mixed model under different unit load section.But, the method only establishes the static association between each service data of boiler, in production scene as run into load lifting special operation condition time, mixed model may constantly switch between neutral net and SVMs, cannot ensure the continuity of process data needed for Model Distinguish, dynamic accuracy is lower.In addition, First air/measuring point such as secondary air flow and draught flow rate that said method adopts, do not consider the impact of temperature and pressure information, therefore the application of institute's established model also can be subject to season and climatic influences.
Summary of the invention
The technical problem to be solved in the present invention is, proposes a kind of based on the online rolling modeling of data-driven and the energy forecasting procedure based on institute's established model; The Data Physical meaning that the energy dynamics model of the present invention's foundation and Model Distinguish are used is clear and definite, and without the need to carrying out cluster analysis process to data, model on-line identification speed is fast, introduces model rolling identification scheme simultaneously, improves the generalization ability of model.The dynamic model of multistep identification is averaged in the output forecast of following synchronization, and utilizes a upper moment forecast departure to revise, improve the forecast precision that boiler exports superheated steam energy further.According to the predicting condition of the superheated steam energy to 60 seconds futures, can judge the combustion case in boiler future, the final optimal control for burning provides guidance and reference.
Technical scheme of the present invention is: based on the boiler overheating steam energy forecasting procedure of data-driven, comprise following steps:
Step 1: set up data communication with level of factory DCS configuration software, obtain creation data online, comprises boiler operating parameter and characterizes the measuring point information of boiler combustion situation.
Step 2: judgement and the process of according to unit load, collection measuring point information being carried out to abnormity point, and adopt forward direction weighted filtering method to remove the high-frequency noise of measuring point.
Step 3: according to the calculation of thermodynamics formula of steam and air, calculates the energy of the air-supply of current time boiler, air inducing and superheated steam.Utilize the as-fired coal matter information of off-line analysis to calculate the thermal discharge of coal dust firing simultaneously.Determine the input data of boiler model identification and export data, and data are normalized.
Step 4: utilize continuous acquisition and the energy record calculated as the data of boiler condition spatial model identification, the rolling simultaneously completing Model Distinguish data upgrades, structure input and output
hankelmatrix.
Step 5: the boiler for producing in short time period is approximated to a linear process, system separate manufacturing firms expression formula is as follows:
In formula
,
,
that system exists respectively
tthe state vector in moment, input observation vector and export observation vector,
nfor the order of unidentified system,
it is the white noise sequence of a zero-mean.
Step 6: utilize open loop subspace method identification boiler model sytem matrix
a, B, C, D.
Step 7: utilize up-to-date boiler input energy datum, as the input variable of the state-space model that identification obtains, the boiler overheating steam (main steam and reheat heat steam) calculating future time instance exports energy.
Step 8: the superheated steam energy forecast calculated with the model of current identification, with a upper moment, phase in future predicted value was in the same time averaged, go up the deviation of a moment to the forecast of current output energy and the actual energy of current calculating simultaneously and revise the current output energy forecasting following 5 seconds, obtain the forecast result that steam generator system exports superheated steam energy for following 60 seconds.
Step 9: substitute the most original legacy data with the new data that current time collection also obtains after pretreatment, the rolling completing Model Distinguish data upgrades.Return Model Distinguish and forecast that step 4 carries out subsequent time.
The creation data gathered is primarily of boiler operating parameter and the measuring point information composition characterizing boiler combustion situation, wherein boiler operating parameter comprises unit load, feeder coal supply speed, one/secondary air flow, pressure and temperature, air-introduced machine section flue gas flow, pressure and temperature etc.; The measuring point information characterizing boiler combustion situation comprises main steam and reheated steam flow, pressure and temperature etc.
The judgement of described abnormity point and process and online denoising, specific practice occurs that the measuring point of saltus step distinguishes in the load relatively stable stage, utilizes the measuring point information of front 2 bats to carry out export-oriented difference and replace:
x n =2x n-1 -x n-2 .Online employing forward direction 5 weighted filtering methods, to remove high-frequency noise, Filtering Formula:
x n =0.04x n-4 -0.08x n-3 + 0.13x n-2 + 0.25x n-1 + 0.5x n .
Described data normalization process is as follows:
Wherein
xfor the initial data of input and output,
x min with
x max be respectively existing record
xin minimum of a value and maximum,
x * for the value after normalization.
The present invention is compared with data-driven modeling method in the past, and the data that the dynamic model of foundation and identification use all possess clear and definite physical meaning.Adopt rolling discrimination method to improve model accuracy, also reflects the objective fact that boiler process parameter can change with unit load in time simultaneously.In addition this method on-line identification model velocity is fast, and the superheated steam energy accuracy of measurement utilizing model prediction boiler to export is high.The dynamic model of multistep identification is averaged in the output forecast of following synchronization, and utilizes a upper moment forecast departure to revise, improve the forecast precision that boiler exports superheated steam energy further.According to the predicting condition of the superheated steam energy to 60 seconds futures, can judge the combustion case in boiler future, the final optimal control for burning provides guidance and reference.
Accompanying drawing explanation
The state-space model input/output variable schematic diagram of Fig. 1 steam generator system;
Fig. 2 is based on the online rolling identification value of forecasting figure of data-driven;
Wherein, solid line representative utilizes the model prediction boiler overheating steam of identification (main steam and reheat heat steam) to export energy, and open circles is that the superheated steam of the boiler reality utilizing the measuring point information characterizing boiler combustion situation to calculate exports energy.
Detailed description of the invention
By the following examples the present invention be further described and describe; the present embodiment is implemented under premised on technical solution of the present invention; give detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.
Embodiment
The following condition of firepower power station boiler operatiopn demand fulfillment in the present embodiment:
A. boiler for producing is approximately a linear process at short notice;
B. one/secondary air flow, absorbing quantity and burner hearth powder amount of delivering coal all is in manual open loop control mode;
C. all control signals are Persistent Excitations;
D. control signal and process random noise are incoherent;
E. the measurement data amount of collection in worksite storage is abundant;
The present embodiment step is as follows:
Step 1: set up OPC data communication with level of factory DCS configuration software, a creation data is gathered online every 5 seconds, the measuring point information spinner obtained will comprise: unit load, feeder coal supply speed, one/secondary air flow, pressure and temperature, the boiler operating parameters such as air-introduced machine section flue gas flow, pressure and temperature, and main steam and reheated steam flow, pressure and temperature etc. characterize the measuring point information of boiler combustion situation.
Step 2: judge that the current moment gathers measuring point information and whether there is abnormity point according to unit load, particularly to the measuring point occurring saltus step in the load relatively stable stage, utilizes the front 2 measuring point information of clapping to carry out the abnormity point that export-oriented difference replaces saltus step:
x n =2
x n-1 -x n-2 .Online employing forward direction 5 weighted filtering methods, to remove high-frequency noise, Filtering Formula:
x n =0.04
x n-4 -0.08
x n-3 +0.13
x n-2 +0.25
x n-1 +0.5
x n .
Step 3: according to the calculation of thermodynamics formula of steam and air, calculates the energy of current time to the air-supply of boiler, air inducing and superheated steam.Utilize the as-fired coal matter information of off-line analysis to calculate the thermal discharge of coal dust firing simultaneously.Using one/Secondary Air energy, smoke evacuation energy and cold reheated steam energy as the input data of boiler model identification, using main steam energy and the reheat heat steam energy output data as Model Distinguish.The data of I/O are normalized as follows respectively:
Wherein
xfor the energy datum that steam generator system inputs or outputs,
x min with
x max be respectively existing
xminimum of a value in record and maximum,
x * for the value after normalization.
Step 4: utilize continuous acquisition and calculate half an hour energy record as the Identification Data needed for boiler modeling, wherein First air energy, Secondary Air energy, smoke evacuation energy, coal supply combustion heat release amount and cold reheated steam energy form input vector
u k , output vector
y k be made up of main steam energy and reheat heat steam energy, output data comparatively input data and are set to 60 seconds lag time.The input vector of steam generator system
u k structure
ioK
jrow
hankelmatrix is as follows:
According to the needs of Model Distinguish, if identification needs (2 altogether
i+j-1) data are organized, front
ipoint data forms first row
hankelmatrix, uses subscript
prepresentative " past "; From
i+ 1 to 2
ithe first row that data are formed
hankelmatrix, uses subscript
frepresentative " future ".Similarly, to input
y k vector structure
hankelmatrix
with
; And to noise e
k vector structure
hankelmatrix
with
.Generally
irelevant to system order to be identified, unsuitable too small or excessive, and
jshould select enough large, usually ensure
carry out identification system, simultaneously can the impact of noise decrease, the present embodiment is got
i=12,
j=700.System " past " and " future " status switch are defined as follows in addition:
，
Step 5: the boiler for producing in half an hour is approximated to a linear process, as shown in Figure 1, the separate manufacturing firms expression formula of system is as follows for its five inputs, two output variable schematic diagrames:
In formula
,
,
that system exists respectively
tthe state vector in moment, input observation vector and export observation vector,
nfor the order of unidentified system,
it is the white noise sequence of a zero-mean.
Step 6: identification system matrix
a, B, C, D.
1) the other definition expansion ornamental matrix of define system
as follows:
Definition controllable matrix
with
expansion ornamental matrix
with
as follows:
The lower triangle of another definition input and noise
toeplitzmatrix
with
as follows:
2) conversion of matrix identification problem
Current time can be derived by the state-space expression of system
tstate vector
x t with past state
x t-1 , input
y t-1 and output
u t-1 between relational expression:
To state vector
x t-1 further expansion has:
Above formula variable subscript is respectively
t=N, N+1 ... N+j-1time expression formula combine and can obtain:
In above formula
, due to
kkalman filtering gain,
represent the dynamic error of Kalman filtering.Therefore when
ntime enough large, Kalman filtering is stable, has
.So when Identification Data is abundant, namely
ntime very large, above formula converges on:
Can be released by the output equation of system state space again:
In formula
the subspace matrices relevant to the input and output in past;
input relevant subspace matrices to the certainty in future;
input (noise) relevant subspace matrices to the randomness in future.And have
,
,
.
When certainty input
with randomness noise
mutual uncorrelated time, this problem can be reduced to searching one and export future
a linear predictor
:
The analytic solutions of this problem are:
3) expansion ornamental matrix is solved
or to to-be sequence estimation
Following matrix is carried out
qRdecompose:
Then subspace matrices
with
can be by
rmatrix in block form in matrix calculates:
In above formula
representing matrix
moore-Penrosegeneralized inverse, by
sVDdecomposition solves.Again to subspace matrices
carry out
sVDdecomposition has:
In formula
the enough I of diagonal element are approximately zero, and
dimension can be used as by the dimension of identification state space matrices A.By
known
column space should be with
column space equal, namely its dimension is systematic education to be identified, is defined before recycling
can determine respectively
matrix and to-be
matrix:
4) determine unidentified system matrix
a, B, C, D.
By expansion ornamental matrix
definition, front m is capable is exactly the C matrix of system state space model, order
for
remove the matrix that front m is capable,
for
remove the matrix that rear m is capable, have
Thus obtain the A matrix of steam generator system state-space model and C matrix is:
Recycling formula
with
can derive:
In formula
be
orthocomplement, orthogonal complement subset, above formula has been launched item by item:
Carry out algebraic transformation to above formula can obtain about sytem matrix
bwith
dlinear equation, and then try to achieve unidentified system
bmatrix and
dmatrix:
The steam generator system separate manufacturing firms model expression of trying to achieve is as follows:
Step 7: the boiler input energy datum utilizing up-to-date 60 seconds, as the input variable of the state-space model that identification obtains, namely calculated value is boiler overheating steam (main steam and the reheat heat steam) energy that steam generator system exports for following 60 seconds.
Step 8: the model after each rolling identification can be used for forecasting the energy that following 60 seconds boilers export, if the predicted value of previous step be [
y a1
,
y a2
...,
y a12
], the predicted value that current step model calculates be [
y b1
,
y b2
...,
y b12
] (wherein
ybe 2 dimensional vectors, comprise main steam energy and heat at vapours energy), be then [0.5 to the predicted value of the superheated steam energy of following 60 seconds
y b1
+ 0.5
y a2
+ r, 0.5
y b2
+ 0.5
y a3
..., 0.5
y b11
+ 0.5
y a12
,
y b12
], wherein r is that previous step forecast exports the deviation exported with current step system Practical Calculation in following 5 seconds, revises the current forecast output energy of following 5 seconds.
Step 9: substitute the most original legacy data with the new data that current time collection pretreatment obtains, the rolling completing Model Distinguish data upgrades.Return Model Distinguish and forecast that step 4 carries out subsequent time.
This example demonstrates the method for the real process data collection of DCS, pretreatment, calculating and renewal, and utilize the online data after process to the rolling discrimination method of boiler condition spatial model and forecasting procedure, the energy forecasting the superheated steam that boiler produces can be used for, the value of forecasting is shown in accompanying drawing 2, and the optimal control that predicted value can be firepower power station boiler combustion provides reference and guidance.
Claims (1)
1., based on a forecasting procedure for the firepower power station boiler output superheated steam of Subspace Identification, its feature exists: the method includes the steps of:
Step 1: set up data communication with level of factory DCS configuration software, obtain creation data online, comprises boiler operating parameter and characterizes the measuring point information of boiler combustion situation;
Step 2: judgement and the process of according to unit load, collection measuring point information being carried out to abnormity point, and adopt forward direction weighted filtering method to remove the high-frequency noise of measuring point;
Step 3: according to the calculation of thermodynamics formula of steam and air, calculates the energy of the air-supply of current time boiler, air inducing and superheated steam; Utilize the as-fired coal matter information of off-line analysis to calculate the thermal discharge of coal dust firing simultaneously; Determine the input data of boiler model identification and export data, and data are normalized;
Step 4: utilize continuous acquisition and the energy record calculated as the data of boiler condition spatial model identification, the rolling simultaneously completing Model Distinguish data upgrades, structure input and output
hankelmatrix;
Step 5: the boiler for producing in short time period is approximated to a linear process, system separate manufacturing firms expression formula is as follows:
In formula
,
,
that system exists respectively
tthe state vector in moment, input observation vector and export observation vector,
nfor the order of unidentified system,
it is the white noise sequence of a zero-mean;
Step 6: utilize open loop subspace method identification boiler model sytem matrix
a, B, C, D;
Step 7: utilize up-to-date boiler input energy datum, as the input variable of the state-space model that identification obtains, the boiler overheating steam calculating future time instance exports energy;
Step 8: the superheated steam energy forecast calculated with the model of current identification, with a upper moment, phase in future predicted value was in the same time averaged, go up the deviation of a moment to the forecast of current output energy and the actual energy of current calculating simultaneously and revise the current output energy forecasting following 5 seconds, obtain the forecast result that steam generator system exports superheated steam energy for following 60 seconds;
Step 9: substitute the most original legacy data with the new data that current time collection also obtains after pretreatment, the rolling completing Model Distinguish data upgrades; Return Model Distinguish and forecast that step 4 carries out subsequent time;
The creation data gathered is primarily of boiler operating parameter and the measuring point information composition characterizing boiler combustion situation, boiler operating parameter comprises unit load, feeder coal supply speed, one/secondary air flow, pressure and temperature, air-introduced machine section flue gas flow, pressure and temperature; The measuring point information characterizing boiler combustion situation comprises main steam and reheated steam flow, pressure and temperature;
The specific practice of the judgement of abnormity point and process and online denoising occurs that the measuring point of saltus step distinguishes in the load relatively stable stage, utilizes the front 2 measuring point information of clapping to carry out export-oriented difference and replace:
x n =2x n-1 -x n-2 ; Online employing forward direction 5 weighted filtering methods, to remove high-frequency noise, Filtering Formula:
x n =0.04x n-4 -0.08x n-3 + 0.13x n-2 + 0.25x n-1 + 0.5x n ,
Data normalization process is as follows:
Wherein
xfor the initial data of input and output,
x min with
x max be respectively existing record
xin minimum of a value and maximum,
x * for the value after normalization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210349737.5A CN102840571B (en) | 2012-09-20 | 2012-09-20 | Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210349737.5A CN102840571B (en) | 2012-09-20 | 2012-09-20 | Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102840571A CN102840571A (en) | 2012-12-26 |
| CN102840571B true CN102840571B (en) | 2015-08-12 |
Family
ID=47368178
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210349737.5A Active CN102840571B (en) | 2012-09-20 | 2012-09-20 | Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102840571B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105629736B (en) * | 2016-03-22 | 2018-03-20 | 东南大学 | The fired power generating unit SCR denitration Disturbance Rejection forecast Control Algorithm of data-driven |
| CN106339788B (en) * | 2016-08-02 | 2020-04-24 | 华电电力科学研究院 | Thermal power generating unit performance index calculation device and method based on DCS |
| CN108255063B (en) * | 2018-01-26 | 2021-03-19 | 深圳禾苗通信科技有限公司 | Small rotor unmanned aerial vehicle system modeling method based on closed-loop subspace identification |
| CN109507910A (en) * | 2018-10-31 | 2019-03-22 | 国网河南省电力公司电力科学研究院 | A kind of fired power generating unit modeling and control method based on data-driven |
| CN110095982B (en) * | 2019-04-22 | 2023-05-23 | 江苏中科云控智能工业装备有限公司 | Automatic identification method for load characteristics, state space model and control method for power supply |
| CN112036088A (en) * | 2020-09-03 | 2020-12-04 | 华能国际电力股份有限公司玉环电厂 | Superheated steam temperature prediction method based on multi-innovation random gradient optimization |
| CN113359441B (en) * | 2021-05-26 | 2022-09-09 | 东南大学溧阳研究院 | Uncertainty compensation based combustion chamber thermal acoustic oscillation model prediction control method |
| CN114492956B (en) * | 2022-01-07 | 2024-03-12 | 上海烁恒科技有限责任公司 | Working condition prediction method, equipment and storage medium based on historical data search |
| CN115234936B (en) * | 2022-09-22 | 2022-12-13 | 海安玻克超硬材料有限公司 | Identification processing method for combustion efficiency of biomass combustor |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101709863A (en) * | 2009-12-18 | 2010-05-19 | 杭州电子科技大学 | Hybrid control method for furnace pressure system of coal-fired boiler |
| CN101825869A (en) * | 2010-05-13 | 2010-09-08 | 上海交通大学 | Method for identifying superheater model parameters based on data drive |
| CN102103376A (en) * | 2010-12-16 | 2011-06-22 | 浙江大学 | Method for evaluating control performance of PID (Proportion Integration Differentiation) loop based on subspace identification |
| JP2011196658A (en) * | 2010-03-23 | 2011-10-06 | Miura Co Ltd | Program, controller and boiler system |
-
2012
- 2012-09-20 CN CN201210349737.5A patent/CN102840571B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101709863A (en) * | 2009-12-18 | 2010-05-19 | 杭州电子科技大学 | Hybrid control method for furnace pressure system of coal-fired boiler |
| JP2011196658A (en) * | 2010-03-23 | 2011-10-06 | Miura Co Ltd | Program, controller and boiler system |
| CN101825869A (en) * | 2010-05-13 | 2010-09-08 | 上海交通大学 | Method for identifying superheater model parameters based on data drive |
| CN102103376A (en) * | 2010-12-16 | 2011-06-22 | 浙江大学 | Method for evaluating control performance of PID (Proportion Integration Differentiation) loop based on subspace identification |
Non-Patent Citations (1)
| Title |
|---|
| 子空间预测控制及其在CFB锅炉燃烧***的研究;李经昊;《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》;20111231;第15至40页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102840571A (en) | 2012-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102840571B (en) | Firepower power station boiler based on Subspace Identification exports the forecasting procedure of superheated steam | |
| CN112200433B (en) | Power plant thermal performance analysis and optimization system | |
| US11232376B2 (en) | System and method for optimizing combustion of boiler | |
| EP1921280B1 (en) | Systems and methods for multi-level optimizing control systems for boilers | |
| CN108052007B (en) | Thermal power generating unit operation optimization method, device and equipment and computer storage medium | |
| CN102566551B (en) | Data mining-based method for analyzing thermal power plant operation index optimal target value | |
| EP2729891B1 (en) | Automatic identification of operating parameters for power plants | |
| CN103440528B (en) | Thermal power unit operation optimization method and device based on power consumption analysis | |
| CN110532674A (en) | A kind of coal-fired power station boiler fire box temperature measurement method | |
| US11262065B2 (en) | System and method for optimizing combustion of boiler | |
| Hou et al. | Coordinated control system modeling of ultra-supercritical unit based on a new fuzzy neural network | |
| CN106765052A (en) | A kind of intelligence computation forecast Control Algorithm of station boiler vapor (steam) temperature | |
| CN103455635A (en) | Thermal process soft sensor modeling method based on least squares and support vector machine ensemble | |
| CN103235986A (en) | Operation and consumption optimization method based on boiler safety analysis | |
| CN110941186B (en) | Steam temperature control optimization method based on neural network and universal gravitation search algorithm | |
| CN103729569A (en) | Soft measurement system for flue gas of power-station boiler on basis of LSSVM (Least Squares Support Vector Machine) and online updating | |
| CN103390211A (en) | Thermal generator set overall process energy management and cost analysis system | |
| Xu et al. | Data based online operational performance optimization with varying work conditions for steam-turbine system | |
| Xia et al. | An online case-based reasoning system for coal blends combustion optimization of thermal power plant | |
| JP2016537729A (en) | Method, system, and computer program product for analyzing generation processes and / or process engineering processes and / or process steps in a plant | |
| Xu et al. | A new online optimization method for boiler combustion system based on the data-driven technique and the case-based reasoning principle | |
| US11287126B2 (en) | System and method for controlling operation of boiler | |
| CN117057573A (en) | Comprehensive energy system toughness scheduling method considering comprehensive demand response based on deep reinforcement learning | |
| Wang et al. | A two-tier approach to the data-driven modeling on thermal efficiency of a BFG/coal co-firing boiler | |
| Chaibakhsh | Modelling and long-term simulation of a heat recovery steam generator |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |