CN109507910A - A kind of fired power generating unit modeling and control method based on data-driven - Google Patents
A kind of fired power generating unit modeling and control method based on data-driven Download PDFInfo
- Publication number
- CN109507910A CN109507910A CN201811290516.9A CN201811290516A CN109507910A CN 109507910 A CN109507910 A CN 109507910A CN 201811290516 A CN201811290516 A CN 201811290516A CN 109507910 A CN109507910 A CN 109507910A
- Authority
- CN
- China
- Prior art keywords
- model
- power generating
- generating unit
- fired power
- data
- 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
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B17/00—Systems involving the use of models or simulators of said systems
- G05B17/02—Systems involving the use of models or simulators of said systems electric
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The present invention provides a kind of fired power generating unit modeling and control method based on data-driven, according to unit real-time running data, chooses input/output variable of the variable having a direct impact to system model as system first, it is determined that data preprocessing method；And fired power generating unit model is recognized using Subspace Identification algorithm, the discrete model of fired power generating unit is obtained, and be combined to discrete model, obtains the overall model of fired power generating unit；Then fired power generating unit is controlled using model predictive control method, has reached good control performance, the Load Regulation ability of fired power generating unit is predicted, play the role of instructing production scheduling well.
Description
Technical field
The present invention relates to fired power generating unit systems technology fields, and in particular to a kind of thermal motor establishment based on data-driven
Mould and control method.
Background technique
Currently, generating set has evolved into the monoblock of large capacity, high parameter, various process equipments and automatic control
The dynamic property of control equipment also becomes increasingly complex.The dynamic characteristic of their each sections is only grasped, and their reasonably structures
At control system, higher reliability and automatization level just can guarantee.Therefore System Discrimination has received generally weighing for people
Depending on.Recognizing model is a kind of approximation of the input-output characteristic of object under certain criterion meaning, and approximate degree depends on
In understanding and selected identification side of the people to the understanding in-depth degree of process priori knowledge and to data collectivity matter
Whether method is reasonable.It is also fewer in terms of engineer application based on the fired power generating unit identification of Model Parameters of Site for Unit operation.
For Large-scale fire-electricity unit, Steam Temperature System and vapour pressure system are the core of entire unit, and maintain main temperature
Stabilization with main vapour pressure is one of most important task of unit control system.The control strategy of fired power generating unit at present, mainly
PI D method, parameter tuning is difficult, inevitable after hyperharmonic.The model of fired power generating unit is obtained by System Discrimination,
To be controlled using advanced control algorithm such as PREDICTIVE CONTROL unit.Model Predictive Control about fired power generating unit is ground
Study carefully, it is domestic constantly to be explored there are many scientific research institution, but it is concentrated mainly on the denitration of fired power generating unit, the control of Stream temperature degree
Equal parts, there is no use PREDICTIVE CONTROL to whole system.
It is to disclose a kind of fired power generating unit dcs control object in CN102446233B in Authorization Notice No.
The method of simulation modeling, the emulation modelling method to main analog amount control object include: a, are simplified according to speed of feedwater and counted
Calculate feed pressure, feedwater flow, separator pressure parameter；B, drum/separator water level is reduced to disengaging working medium flow difference
Integral；C, steam flow is determined by calculating working medium mass dryness fraction；D, it is realized using the integrating function of PI D-module by pressure, enthalpy
Value calculates temperature, then simulation calculation goes out main steam temperature；E, simulation calculation exhaust gas volumn and induced draft blower withdraws amount, and then simplify imitative
True flake hearth-tapping negative pressure；F, according to feeder revolving speed simulation calculation boiler oil amount；G, imitative according to fuel quantity or main steam flow
Genuine computer group load.But it does not carry out operation by truthful data, and and uses PREDICTIVE CONTROL to whole system.Together
Sample is that there is also ask accordingly in the control method for disclosing a kind of fired power generating unit in CN106483870A in application publication number
Topic.
It is therefore desirable to provide whole strategy of the fired power generating unit from modeling to control, on the one hand guarantee modeling method
It is simple and be easily achieved, on the other hand guarantee that control strategy control performance is satisfied, can be good at guarantee system satisfaction again
In the case where constraint condition, and guarantee that system has certain robustness.
Summary of the invention
In view of this, the present invention provides a kind of fired power generating unit modeling and control method based on data-driven, according to unit
Real-time running data chooses input/output variable of the variable having a direct impact to system model as system first, it is determined that
Data preprocessing method；And fired power generating unit model is recognized using Subspace Identification algorithm, obtain the discrete of fired power generating unit
Model, and discrete model is combined, obtain the overall model of fired power generating unit；Then model predictive control method pair is utilized
Fired power generating unit is controlled, and good control performance has been reached, and is predicted the Load Regulation ability of fired power generating unit, is risen
To the effect for instructing production scheduling well.
To solve the above problems, the present invention provides a kind of fired power generating unit modeling and control method based on data-driven, use
In the mathematical model for establishing fired power generating unit and fired power generating unit is controlled, is included the following steps:
S1: it fired power generating unit modeling: in conjunction with thermal power unit operation mechanism and fired power generating unit field adjustable historical data, chooses
Steam turbine tune valve opening, Coal-fired capacity, confluent, total blast volume, coal-air ratio, desuperheating water are input quantity, choose unit load, main vapour pressure
Power, Stream temperature degree are output quantity；
S2: S1 fired power generating unit one day input quantity and output quantity are chosen, and the data of input quantity and output quantity are returned
One change, elimination of burst noise, the data prediction for going mean value, filtering；
S3: establishing the discrete model of fired power generating unit by subspace model identification methods, utilizes number pretreated in S2
According to decomposing the order for having found out each model by QR or SVD, established respectively with fuel quantity, confluent, steam turbine valve opening, cold
But water, coal-air ratio are input quantity, using centrum's temperature, main steam temperature, main steam pressure as the thermal motor system of output quantity
System model, the state-space expression for having obtained each model are as follows:
X (k+1)=Ax (k)+Bu (k) (1)
Y (k)=Cx (k)+Du (k) (2)
Wherein u (k) ∈ R5、y(k)∈R4It is the inputoutput data vector at k moment, x (k) ∈ R respectivelynIt is the k moment
State variable；The A in model above, B, C, D matrix can be obtained using subspace state space system identification.
The main steam temperature modelling effect obtained for direct identification modeling, which is paid no attention to, to think over a problem, and analysis is found: utilizing fuel
Amount, confluent and steam turbine valve opening can be very good to establish the mathematical model of centrum's temperature；And pass through mid-point position
Therefore the temperature that steam flow has just obtained system main steam after sectionalized superheater first models centrum's temperature, so
Attemperation water flow models system main steam temperature in recycling centrum's temperature, coal-air ratio and superheater afterwards, this
The available fairly accurate main steam temperature model of sample.
S4: obtained discrete model is combined by method of linearization, is obtained the overall model of fired power generating unit, is found out
The state-space expression of unit model；
S5: by obtained predictive control model, model predictive controller is designed；To each variable of fired power generating unit
Initial value, bound, rate of change bound are configured, and to the control time domain of controller, control room every, prediction time domain
It is configured；
S6: fired power generating unit load up and load down ability and each change are obtained by predictive control model and predictive controller
Measure the variation tendency of each variable in the case where meeting constraint condition.
Further, subspace model identification methods are adoption status space forms to describe the system to be recognized:
xk+1=Axk+Buk
yk=Cxk+Duk
The main problem of Subspace Identification model is the Input and output measurements u generated according to systemk、 yk, determine system
Order n, sytem matrix A, B, C, D.
Further, steam turbine tune valve opening is the mode of simple and quick adjusting unit load and main vapour pressure；Coal-fired capacity
It is to adjust the decision amount of unit load with confluent；Total blast volume has larger impact to unit load；Coal-air ratio is to guarantee that coal is abundant
The key variables of burning, desuperheating water are the supplementary means for controlling main steam temperature；
Output quantity includes unit load, main vapour pressure and main steam temperature as the variation of load can also change therewith, and
Main vapour pressure and main steam temperature are the parameters with restriction range, will have a direct impact on the operational safety of unit.
Further, the control time domain is 3-7, and the control room is every the sampling interval for referring to each control instruction output
Number, preferably 1, the prediction time domain is that can cover time required for system dynamic course, preferably 200.
Further, by obtained predictive control model, model predictive controller is designed；That is, using obtaining
After PREDICTIVE CONTROL module, using the target function of predictive control algorithm building system, the constraint item that each variable needs to meet is provided
Part solves the optimization problem of this belt restraining, further according to it is obtaining as a result, using rolling optimization thought, to design bid
Quasi- model predictive controller.
Further, when being configured to the initial value of each variable of fired power generating unit, bound, rate of change bound,
Middle each variable initial value of fired power generating unit, which is subject to, actual value of current steady when running, bound and rate of change and practical to be controlled
The characteristic of amount processed is related；That is, main steam temperature is no more than some temperature range, fuel quantity and water in entire control process
The variable quantity of amount within a certain period of time is also to have the upper limit, it is impossible to which unlimited to increase, these settings all control for normative forecast
Optimization problem in algorithm gives the hard constraint condition for needing to meet.
Beneficial effects of the present invention: the present invention establishes fired power generating unit mould using unit real-time running data as data-driven
Type；The premise for obtaining unit model is to obtain the actual operating data of fired power generating unit first, with fired power generating unit actual operating data
Premised on seek fired power generating unit model, using the real-time running data of fired power generating unit, data are true and reliable, the pre- place of data
Reason method is simple.
In addition, use fired power generating unit modeling method be Subspace Identification algorithm, based on the truthful data of unit into
Row operation, when Identification Data, do not need to be iterated optimization, and calculation amount is few, utilize simple linear algebra tool, Ke Yiwei
Geometric projection, is preferably decomposed using QR, SVD decomposes this kind of mathematical tool with extraordinary robustness, is applicable in this
In the operation of fired power generating unit this large data sets and big system；And this method is a kind of step, non-iterative calculation method,
Spatial model can be directly obtained, therefore it does not have constringent problem, can obtain the model of preferable fired power generating unit.
In addition, it is accurate by the model that Subspace Model Identification algorithm is established, and greatly simplify and calculate, it ensure that one
Fixed precision is effectively reduced the complexity of identification, remains the dynamic property of links, overcomes traditional modeling side
The deficiency of method.
In addition, controlled using forecast Control Algorithm fired power generating unit, and to fired power generating unit Load Regulation ability into
Row prediction, is configured the input and output bound, initial value, change rate bound of each variable of fired power generating unit model, and right
The control time domain of controller, control room are configured every, prediction time domain, forecast Control Algorithm strong robustness, and can effectively be located
Reason constraint carries out control to fired power generating unit and achieves preferable control performance, and reached good prediction unit load tune
The function of energy saving power has fully considered the constraint of each variable, and the static state and dynamic control performance for keeping system to have；Phase
Compare compared with conventional PID controllers control effect: there is not hyperharmonic oscillation adjustment process in system, and can be in limiting time
It is interior, more accurately make it；By data-driven, the model of system is obtained using subspace state space system identification, it is right
System is controlled using forecast Control Algorithm；This method give a global solutions from modeling to control for fired power generating unit
Certainly scheme has good directive function to the control strategy of fired power generating unit.
Detailed description of the invention
Fig. 1 is the part screenshot of unit real-time running data of the invention；
Sectional view when Fig. 2 is importing data of the invention；
Sectional view when Fig. 3 is data prediction of the invention；
Fig. 4 is Subspace Model Identification algorithm schematic diagram of the invention；
Fig. 5 is that model order of the invention selects schematic diagram；
Fig. 6 is power module schematic diagram of the invention；
Fig. 7 is main steam pressure model schematic of the invention；
Fig. 8 is centrum's temperature model schematic of the invention；
Fig. 9 is main steam temperature model schematic of the invention；
Figure 10 is discrete model combination diagram of the invention；
Figure 11 is parameter setting schematic diagram of the invention；
Figure 12 is each variable change range of the invention, rate of change setting schematic diagram；
Input quantity curve synoptic diagram when Figure 13 is increasing power of the invention；
Output quantity curve synoptic diagram when Figure 14 is increasing power of the invention；
Input quantity curve synoptic diagram when Figure 15 is drop power of the invention；
Output quantity curve synoptic diagram when Figure 16 is drop power of the invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
Attached drawing 1-16, the technical solution of the embodiment of the present invention is clearly and completely described.Obviously, described embodiment is
A part of the embodiments of the present invention, instead of all the embodiments.Based on described the embodiment of the present invention, this field is general
Logical technical staff's every other embodiment obtained, shall fall within the protection scope of the present invention.
Subspace model identification methods
Subspace state space system identification is adoption status space form to describe the system to be recognized:
xk+1=Axk+Buk
yk=Cxk+Duk
Wherein uk∈R5、yk∈R4It is the inputoutput data vector at k moment, x respectivelyk∈RnIt is the state change at k moment
Amount.Utilize the A, B, C in the available model above of subspace state space system identification, D matrix.
The main problem of Subspace Identification model is the Input and output measurements u generated according to systemk、yk, determine system
Order n, sytem matrix A, B, C, D.Specific implementation step is described further below.
Actual operating data
Unit real-time running data such as Fig. 1；Unit is that whole day runs without interruption, and chooses one day unit every time here
Operation data is recognized.Before carrying out Model Distinguish firstly the need of selection to the key variables that have a direct impact of modeling simultaneously
Variable data is pre-processed.
Variable is modeled to choose
In conjunction with thermal power unit operation mechanism, by consulting relative literature data, and fired power generating unit field adjustable is combined
Experience, steam turbine tune valve opening are the methods of simple and quick adjusting load and main vapour pressure, and Coal-fired capacity and confluent are to adjust to bear
The decision amount of lotus, total blast volume have larger impact to load, and coal-air ratio is to guarantee that the clean-burning key variables of coal, desuperheating water are
Control the supplementary means of main steam temperature；Output quantity necessarily includes unit load, and main vapour pressure and Stream temperature degree are with load
Variation can also change therewith, and main vapour pressure and Stream temperature degree are the parameters with restriction range, will have a direct impact on machine
The operational safety of group.
Data prediction then is carried out as Fig. 2 imports data, after having chosen Correlation Identification parameter, needs the reality to unit
When operation data screened, pre-processed.The magnitude of each parameter of unit differs greatly, in order to effectively embody the influence of each parameter,
Normalized is needed, all parameters are transformed into same order.There may be outlier in a whole segment data, need to pick
It removes.It may include that the influence disturbed needs to filter to handle, for big data so that data are not smooth enough in data
Speech, small variation is not obvious enough, needs mean value to show and change；Go trend that can eliminate long period trend term；Enter Fig. 3
Carry out data prediction.
Load, main steam pressure model foundation
Fig. 4 is Subspace Model Identification algorithm, utilizes the available load main vapour pressure model of the algorithm.Input quantity is
Steam turbine tune valve opening, Coal-fired capacity, confluent export as power, main vapour pressure.
Fig. 5 is that model order selects schematic diagram, using the inputoutput data of system, constructs Hankel data matrix, benefit
The projection of Hankel Row rank is sought with QR decomposition, singular value decomposition is carried out to the projection result, obtains the rank of system
It is secondary；Third height is the model order sought in figure；
Load, the main vapour pressure model obtained it can be seen from Fig. 6, Fig. 7 by Subspace Model Identification algorithm is integrally intended
Conjunction trend is preferable, and the output of model can be good at the reality output of fitting system.
Centrum's temperature model
Fig. 8 be centrum's temperature model identification effect, input quantity be steam turbine tune valve opening, Coal-fired capacity, confluent, it is defeated
It is centrum's temperature out.Centrum's temperature is the saturated steam temperature at steam-water separator, centrum's temperature and main steam temperature
It there is certain relationship between degree, first passes through and seeks centrum's temperature model, then the centrum's temperature model by obtaining obtains
The main steam temperature model of system.As can be seen from the figure centrum's temperature overall fit trend compared with number.
If Fig. 9 is main steam model, input quantity is centrum's temperature, coal-air ratio, desuperheating water, and output quantity is main steam pressure
Power.The centrum's temperature obtained using identification is as one of input quantity, to recognize main steam temperature.Pass through main steam temperature
Degree comparison diagram can be seen that the model overall fit works well, and there are also some gaps in amplitude, can preferably track reality
System achieves good identification effect compared to other modeling methods.
As shown in Figure 10, carry out discrete model combination, the power obtained using above-mentioned identification, main steam pressure model, in
Between put temperature model, the relationship between main steam temperature model, discrete model is combined, and to each input variable into
The processing for going mean value is gone, after combining discrete model, system becomes one 5 and enters 3 systems gone out, by the model of linearisation
The characteristics of meeting master mould, and it can be fitted the dynamic characteristic of actual set well.
A, B, C, D parameter matrix of state-space model
By the tool box Simulink of MATLAB, by the load of above-mentioned identification, main steam pressure model, intermediate point temperature
Degree model, Stream temperature degree model carry out linearisation combination, have obtained a 6 rank state-space models, are below state space mould
The sytem matrix of type:
System is controlled using forecast Control Algorithm
The true Load Regulation ability of system in order to obtain, integrally regard unit as control object.The pre- observing and controlling of application model
The model prediction of the rolling optimization thought and PREDICTIVE CONTROL of algorithm processed, the display processing feature to constraint, according to input and it is defeated
The constraint condition of output carries out the prediction of Load Regulation ability using forecast Control Algorithm, the PREDICTIVE CONTROL of fired power generating unit is converted
The problem of to solve load prediction ability.
Load Regulation ability is divided into the ability for increasing power capability and dropping power, increases power to fired power generating unit first here
Ability predicted.Before carrying out output power prediction using Model Predictive Control, data have all carried out pretreatment and magnitude
It is uniformly processed.
In the case where increasing power, 1 parameter setting is carried out referring to Fig.1, the initial value of each variable is set, and needs to set
Power step parameter is positive number, and without modification, step scale is 10 (values after change of scale) to other parameters, this value can be with
It is big as far as possible because we it is required to determine that Load Regulation ability maximum upper limit.
As shown in figure 12, the constraint bound of input/output variable and the change rate bound of performance variable are set.
As shown in figure 13, the control room that predictive controller is arranged is divided into 5s, and prediction time domain length is 250s, controls time domain
For 4s；Wherein each input quantity is respectively steam turbine tune valve opening (%) in figure；Coal-fired capacity (t/h)；Total water supply (t/h)；Coal-air ratio；
Desuperheating water (t/h).In restriction range, in order to make full use of the accumulation of heat of unit, accelerate the load responding speed at unit initial stage,
Unit valve opening increases rapidly, and valve opening is as quick regulating measure, and next time is adjusted for convenience, cannot in higher or
Reduced levels are arranged its penalty for this, and valve opening must be restored to reset condition.In order to promote unit load,
Coal amount, the water of unit rapidly increase.Coal and water, are the energy sources of unit generation, and coal and water will determine unit load
End-state.Coal-air ratio is the important parameter for influencing unit load, in order to guarantee coal full combustion and less, the coal-air ratio that radiates
Theoretically speaking should be a constant, provide based on experience value.Desuperheating water is last procedure for adjusting Stream temperature degree,
It is the supplementary means of system, it is contemplated that its adjustment effect to Stream temperature degree does not do stringent constraint processing.
Output quantity curve when increasing power with reference to Figure 14, wherein each input quantity is respectively unit load (MW) in figure；Main vapour pressure
Power (MPa)；Stream temperature degree (DEG C).When system loading needs to quickly increase, Stream temperature degree and main vapour pressure can all become therewith
It is dynamic.Dividing from figure can obtain, and coal and water influence have certain hysteresis to the varying duty of unit.Stream temperature degree and main vapour pressure
Power is that unit load adjusting must satisfy constraint condition as output constraint.Under constraint condition, system maximum load up energy
Power is a limited ascending curve.
When dropping power such as Figure 15 shown in input quantity curve synoptic diagram, in the case where dropping power, need to set power step
Parameter is negative, and other parameters are without modification；In restriction range, in order to reduce rapidly unit load, unit valve opening is rapid
Reduce, next time is adjusted for convenience, cannot be in higher or lower level, its penalty is arranged for this, valve is opened
Degree must be restored to reset condition.In order to be finally reached the target for reducing unit load, coal amount, the water of unit are very fast to be subtracted
It is few.Coal-air ratio is the important parameter for influencing unit load, and in order to guarantee that coal full combustion and heat dissipation are less, coal-air ratio is theoretically
For should be a constant, provide based on experience value.Desuperheating water is last procedure for adjusting Stream temperature degree, is system
Supplementary means, it is contemplated that its adjustment effect to Stream temperature degree does not do stringent constraint processing.
Output quantity curve synoptic diagram when dropping power such as Figure 16；When system loading needs rapid decrease, the Stream temperature of unit
Degree and main vapour pressure also will receive influence.It can be obtained from figure, coal and water influence that there is certain lag to make on the varying duty of unit
With；In order to guarantee that unit safety operation, Stream temperature degree, main vapour pressure must limit in a certain range；That is, system
Maximum load down ability is a limited decline curve.
Certain domestic 600MW unit of present invention combination establishes the state-space expression of unit model in S4:
Load, main vapour pressure model be 2 ranks, obtained system parameter matrix is as follows:
Centrum's temperature model is 2 ranks, and system parameter matrix is as follows:
C2=[16.21-12.66] D2=[0 0 0]
Stream temperature degree model is 2 ranks, and system parameter matrix is as follows:
C3=[16.41 20.84] D3=[0 0 0]
It by model linearization combined method, combines discrete model, the unit for foring five inputs three output is integrally marked
Claim model as follows:
The above is a preferred embodiment of the present invention, it is noted that for those skilled in the art
For, without departing from the principles of the present invention, it can also make several improvements and retouch, these improvements and modifications
Also it should be regarded as protection scope of the present invention.
Claims (5)
1. a kind of fired power generating unit modeling and control method based on data-driven, for establishing the mathematical model of fired power generating unit and right
Fired power generating unit is controlled, and is included the following steps:
S1: fired power generating unit modeling: in conjunction with thermal power unit operation mechanism and fired power generating unit field adjustable historical data, steam turbine tune is chosen
Valve opening, Coal-fired capacity, confluent, coal-air ratio, spray water flux are input quantity, choose unit load, main vapour pressure, Stream temperature degree and are
Output quantity；
S2: multiple workaday input quantities and output quantity in one period of S1 fired power generating unit are chosen, and to input quantity and output quantity
Data be normalized, elimination of burst noise, the data prediction for going mean value and filtering；
S3: establishing the discrete model of fired power generating unit by subspace model identification methods, using data pretreated in S2, leads to
It crosses QR or SVD and decomposes the order for having found out each model, established respectively with fuel quantity, confluent, steam turbine valve opening, cooling water
Amount, coal-air ratio are input quantity, using centrum's temperature, main steam temperature, main steam pressure as the fired power generating unit system mould of output quantity
Type, the state-space expression for having obtained each model are as follows:
X (k+1)=Ax (k)+Bu (k) (1)
Y (k)=Cx (k)+Du (k) (2)
Wherein u (k) ∈ R5、y(k)∈R4It is the inputoutput data vector at k moment, x (k) ∈ R respectivelynIt is the state change at k moment
Amount；
S4: obtained discrete model is combined by method of linearization, is obtained the overall model of fired power generating unit, has been found out unit
The state-space expression of model；
S5: by obtained predictive control model, being designed model predictive controller,
The initial value of each variable of fired power generating unit, bound, rate of change bound are configured,
And to the control time domain of controller, control room every, prediction time domain be configured；
S6: fired power generating unit load up and load down ability and each change are obtained by the predictive controller designed in unit model and S5
Measure the variation tendency of each variable in the case where meeting constraint condition.
2. fired power generating unit modeling and control method based on data-driven as described in claim 1, it is characterised in that: subspace
Identification Method is adoption status space form to describe the system to be recognized:
xk+1=Axk+Buk
yk=Cxk+Duk
The main problem of Subspace Identification model is the Input and output measurements u generated according to systemk、yk, determine the order of system
N, sytem matrix A, B, C, D.
3. fired power generating unit modeling and control method based on data-driven as claimed in claim 2, it is characterised in that: steam turbine tune
Valve opening is the mode of simple and quick adjusting unit load and main vapour pressure；Coal-fired capacity and confluent are to adjust unit load
Decision amount；Total blast volume has larger impact to unit load；Coal-air ratio is to guarantee the clean-burning key variables of coal, and desuperheating water is control
The supplementary means of main steam temperature processed；
Output quantity includes unit load, main vapour pressure and main steam temperature as the variation of load can also change therewith, and main vapour
Pressure and main steam temperature are the parameters with restriction range, will have a direct impact on the operational safety of unit.
4. fired power generating unit modeling and control method based on data-driven as claimed in claim 3, it is characterised in that: the control
Time domain processed is 3-7, and the control room is every the sampling interval number for referring to the output of each control instruction, preferably 1, the prediction time domain is
Time required for system dynamic course, preferably 200 can be covered.
5. fired power generating unit modeling and control method based on data-driven as claimed in claim 4, it is characterised in that: by obtaining
Predictive control model, model predictive controller is designed；That is, after using obtained PREDICTIVE CONTROL module, using prediction
Control algolithm constructs the target function of system, provides the constraint condition that each variable needs to meet, solves the optimization of this belt restraining
Problem, further according to it is obtaining as a result, using rolling optimization thought, to design the model predictive controller of standard.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811290516.9A CN109507910A (en) | 2018-10-31 | 2018-10-31 | A kind of fired power generating unit modeling and control method based on data-driven |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811290516.9A CN109507910A (en) | 2018-10-31 | 2018-10-31 | A kind of fired power generating unit modeling and control method based on data-driven |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109507910A true CN109507910A (en) | 2019-03-22 |
Family
ID=65747309
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811290516.9A Pending CN109507910A (en) | 2018-10-31 | 2018-10-31 | A kind of fired power generating unit modeling and control method based on data-driven |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109507910A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110941186A (en) * | 2019-12-26 | 2020-03-31 | 华润电力技术研究院有限公司 | Steam temperature control optimization method based on neural network and universal gravitation search algorithm |
| CN111191881A (en) * | 2019-12-13 | 2020-05-22 | 大唐东北电力试验研究院有限公司 | Thermal power generating unit industrial equipment state monitoring method based on big data |
| CN111413864A (en) * | 2020-02-27 | 2020-07-14 | 国网河南省电力公司电力科学研究院 | 600MW supercritical thermal power generating unit modeling and control method |
| CN111623369A (en) * | 2020-06-28 | 2020-09-04 | 华电潍坊发电有限公司 | Control method for adjusting boiler fuel feeding quantity by using smoke oxygen content signal |
| CN113267994A (en) * | 2021-04-23 | 2021-08-17 | 湖南省湘电试验研究院有限公司 | Thermal power generating unit main steam pressure control method and system based on three-level control series connection |
| CN113295587A (en) * | 2021-05-19 | 2021-08-24 | 山东鲁能控制工程有限公司 | High-parameter thermal power generating unit flue gas emission pollution parameter remote measurement system and device |
| CN113433987A (en) * | 2021-07-20 | 2021-09-24 | 山信软件股份有限公司 | Denitration reactor temperature control system and control method thereof |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102840571A (en) * | 2012-09-20 | 2012-12-26 | 贵州电力试验研究院 | Subspace identification based forecasting method for superheated steam output of boiler of firepower power station |
| CN103293955A (en) * | 2013-05-17 | 2013-09-11 | 浙江大学 | Method for hybrid system modeling and coordinated optimization control of blast-furnace hot blast stove |
| CN104122797A (en) * | 2014-07-22 | 2014-10-29 | 东南大学 | Novel thermal power generating unit load multivariable predictive control method |
| EP2919079A2 (en) * | 2014-03-14 | 2015-09-16 | Trillary S.r.l. | Optimization and control method for a distributed micro-generation energy plant |
| CN105278507A (en) * | 2015-11-03 | 2016-01-27 | 广东电网有限责任公司电力科学研究院 | Process optimization control method and process optimization control system for main steam system of ultra-supercritical unit |
-
2018
- 2018-10-31 CN CN201811290516.9A patent/CN109507910A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102840571A (en) * | 2012-09-20 | 2012-12-26 | 贵州电力试验研究院 | Subspace identification based forecasting method for superheated steam output of boiler of firepower power station |
| CN103293955A (en) * | 2013-05-17 | 2013-09-11 | 浙江大学 | Method for hybrid system modeling and coordinated optimization control of blast-furnace hot blast stove |
| EP2919079A2 (en) * | 2014-03-14 | 2015-09-16 | Trillary S.r.l. | Optimization and control method for a distributed micro-generation energy plant |
| CN104122797A (en) * | 2014-07-22 | 2014-10-29 | 东南大学 | Novel thermal power generating unit load multivariable predictive control method |
| CN105278507A (en) * | 2015-11-03 | 2016-01-27 | 广东电网有限责任公司电力科学研究院 | Process optimization control method and process optimization control system for main steam system of ultra-supercritical unit |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111191881A (en) * | 2019-12-13 | 2020-05-22 | 大唐东北电力试验研究院有限公司 | Thermal power generating unit industrial equipment state monitoring method based on big data |
| CN111191881B (en) * | 2019-12-13 | 2024-05-14 | 大唐东北电力试验研究院有限公司 | Thermal power generating unit industrial equipment state monitoring method based on big data |
| CN110941186A (en) * | 2019-12-26 | 2020-03-31 | 华润电力技术研究院有限公司 | Steam temperature control optimization method based on neural network and universal gravitation search algorithm |
| CN111413864A (en) * | 2020-02-27 | 2020-07-14 | 国网河南省电力公司电力科学研究院 | 600MW supercritical thermal power generating unit modeling and control method |
| CN111623369A (en) * | 2020-06-28 | 2020-09-04 | 华电潍坊发电有限公司 | Control method for adjusting boiler fuel feeding quantity by using smoke oxygen content signal |
| CN111623369B (en) * | 2020-06-28 | 2022-01-25 | 华电潍坊发电有限公司 | Control method for adjusting boiler fuel feeding quantity by using smoke oxygen content signal |
| CN113267994A (en) * | 2021-04-23 | 2021-08-17 | 湖南省湘电试验研究院有限公司 | Thermal power generating unit main steam pressure control method and system based on three-level control series connection |
| CN113267994B (en) * | 2021-04-23 | 2023-05-05 | 湖南省湘电试验研究院有限公司 | Main steam pressure control method and system of thermal power generating unit based on three-stage control series connection |
| CN113295587A (en) * | 2021-05-19 | 2021-08-24 | 山东鲁能控制工程有限公司 | High-parameter thermal power generating unit flue gas emission pollution parameter remote measurement system and device |
| CN113433987A (en) * | 2021-07-20 | 2021-09-24 | 山信软件股份有限公司 | Denitration reactor temperature control system and control method thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109507910A (en) | A kind of fired power generating unit modeling and control method based on data-driven | |
| CN111443681B (en) | Multi-model predictive control design method for supercritical thermal power generating unit coordinated control system | |
| CN108227488B (en) | Sliding mode prediction control-based ultra-supercritical thermal power generating unit coordination control method | |
| Havlena et al. | Application of model predictive control to advanced combustion control | |
| CN101509656B (en) | Supercritical DC furnace synthesis type coordinating control method | |
| CN102778880B (en) | Energy-balance-based coordination control method of integrated gasification combined cycle power station | |
| CN110285403A (en) | Main Steam Temperature Control method based on controlled parameter prediction | |
| CN108227500A (en) | A kind of control method for coordinating and system of the quick peak regulation of fired power generating unit | |
| CN103225799A (en) | Method for controlling main steam temperature in thermal power plant | |
| US10240775B2 (en) | Multi-objective steam temperature control | |
| CN107065518A (en) | A kind of coordinated algorithm of predictive functional control | |
| CN106933202A (en) | The feedforward control reinitialized using the interval based on estimated status information | |
| Wang et al. | Multivariable constrained predictive control of main steam temperature in ultra-supercritical coal-fired power unit | |
| CN109668139B (en) | Engine-boiler coordination control method for supercritical thermal power generating unit | |
| Zhang et al. | Optimization control of a 330 MW drum boiler unit based on DMC algorithm and DEB strategy | |
| CN114326387A (en) | Denitration control device and method for thermal power generating unit | |
| Mohamed et al. | Predictive control strategy for a supercritical power plant and study of influences of coal mills control on its dynamic responses | |
| Hou et al. | Application of multi-agent EADRC in flexible operation of combined heat and power plant considering carbon emission and economy | |
| Hultgren et al. | Integrated control and process design for improved load changes in fluidized bed boiler steam path | |
| Han et al. | A L1-LEMPC hierarchical control structure for economic load-tracking of super-critical power plants | |
| Poncia et al. | Multivariable model predictive control of a thermal power plant with built-in classical regulation | |
| Spinelli et al. | An optimal hierarchical control scheme for smart generation units: an application to combined steam and electricity generation | |
| Wójcik et al. | Optimal control for combustion process | |
| Ma et al. | ANN and PSO based intelligent model predictive optimal control for large-scale supercritical power unit | |
| CN114562713A (en) | Main steam temperature control method and system for power generation boiler |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190322 |
|
| RJ01 | Rejection of invention patent application after publication |