CN103576655A - Method and system for utility boiler combustion subspace modeling and multi-objective optimization - Google Patents
Method and system for utility boiler combustion subspace modeling and multi-objective optimization Download PDFInfo
- Publication number
- CN103576655A CN103576655A CN201310541803.3A CN201310541803A CN103576655A CN 103576655 A CN103576655 A CN 103576655A CN 201310541803 A CN201310541803 A CN 201310541803A CN 103576655 A CN103576655 A CN 103576655A
- Authority
- CN
- China
- Prior art keywords
- model
- optimization
- boiler
- burning
- subspace
- 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.)
- Granted
Links
Classifications
-
- 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
- Regulation And Control Of Combustion (AREA)
- Feedback Control In General (AREA)
- Control Of Steam Boilers And Waste-Gas Boilers (AREA)
Abstract
The invention discloses a method and system for utility boiler combustion subspace modeling and multi-objective optimization, belongs to the field of utility boiler optimization operation, and particularly relates to a method and system for setting up of a high-precision combustion model and solution of multi-objective optimization. The method includes the following steps of determining input variables of a combustion optimization model, determining the variable to be optimized from all the input variables, dividing a load into a plurality of sectors with the neighborhoods overlapped, setting up a combustion subspace ANFIS model for each sector in an off-line mode, collecting data in real time, conducting partial subspace model modification in an on-line mode, conducting on-line multi-objective optimization by taking the constraint of a unit load into consideration, taking maximizing the boiler efficiency and minimizing the nitrogen oxide emission as targets, and utilizing a mature global optimization algorithm and on the basis of comprehensive cost minimization, and separating and implementing the optimization result. The method and the system are suitable for combustion optimized operation of a coal powder utility boiler and have the advantages of being high in modeling accuracy and high in optimization speed.
Description
Technical field
The invention belongs to station boiler and optimize running technology field, the method and system that particularly a kind of high precision combustion model is set up and operation is optimized, specifically a kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system.
Background technology
Very important ingredient in coal-fired power plant China electrical production, its generated energy, considerably beyond the summation of other various generated energy, does not have large change within the quite a long time.Yet the whole world is faced with the exhausted crisis of serious primary energy, and burned coal price is high, and problem of environmental pollution is also subject to the extensive concern of countries in the world day by day.Therefore the strategy that national governments take is: greatly develop on the one hand gigawatt supercritical generating technology, be the optimisation technique research of carrying out station boiler coal combustion on the other hand, the energy-saving and emission-reduction of take improve boiler efficiency and reduce the research of technique of pollutant emission as target.
Burning optimization for high-efficiency low-pollution, the main difficulty facing is at present, the burning of coal in boiler is very complicated Dual-Phrase Distribution of Gas olid problem, combustion mechanism aspect is not very ripe, to the study general of its characteristic, be the mechanism model based on partial differential equation and algebraic equation composition, and the foundation of these models is to obtain under the condition of many hypothesis and simplification.For system optimization and control, it is mainly the method based on computational intelligence modeling.
Through the open source literature retrieval to prior art, find, document " Qiang Xu, Jia Yang and Yanqiu Yang. Identification and control of boiler combustion system based on neural networks and ant colony optimization algorithm. Proceedings of the 7th World Congress on Intelligent Control and Automation, June 25-27, 2008, Chongqing, China, pp.765-768 (boiler combustion system identification based on neural network and ant colony optimization algorithm and control, international conference: world's Conference of Intellectual Control and Automation collection of thesis, 2008:765-768) ", adopted the burn design of modeling and controller of the neural network in computational intelligence field and method that ant group optimization combines, can better improve the regulation quality of boiler combustion system, but can only solve boiler main and want the given setting value of controlled parameter tracking, and cannot make system remain the operation of Optimum Economic operating mode.Document " Hao Zhou, Kefa Cen, Jianren Fan. Modeling and optimization of the NOx emission characteristics of a tangentially fired boiler with artificial neural networks. Energy, 2004, 29:167 – 183 (the tangential firing boiler discharged nitrous oxides characteristic modeling and control based on artificial neural network, International Periodicals: the energy, 2004, 29:167 – 183) ", adopt Multilayer Feedforward Neural Networks to carry out global modeling and the control of tangential firing boiler discharged nitrous oxides characteristic, for reducing pollutant emission, have great importance, but do not consider boiler efficiency, thereby optimization is incomplete.Document (An Enke, Song Yao, Yang Xia. the coal-fired power station boiler multiple goal burning optimization based on support vector machine and genetic algorithm, energy-conservation, 2008 (10): 22 – 25), adopt support vector machine to carry out modeling, consider boiler efficiency and pollutant emission problem simultaneously, adopt genetic algorithm to carry out multiple-objection optimization calculating, but problem is that the first optimization time is long, what next obtained is Pareto optimal solution set, and finally put on on-the-spot solution, only has one group, and how obtaining being finally implemented on on-the-spot unique solution is to have problem to be solved.And the common issue with that above document exists is also: (1) is all to carry out overall state modeling by a neural network or support vector machine, for the complicated boiler combustion system with serious nonlinear characteristic, Model approximation performance in global scope is difficult to guarantee, the model training time can be very long, especially when ature of coal coal etc. changes, can there is series of problems in long and reliability aspect of time that model correction brings; (2) carrying out pollutant emission or/and do not consider unit load constraint during the optimization of boiler efficiency, the result of optimizing is like this likely to have reduced pollutant emission or/and improved boiler efficiency, but take, to reduce generated energy be cost, therefore considers that the constraint of unit load just becomes important aspect.
Therefore, the some subspaces that the present invention is based on load data are divided, set up respectively the adaptive neural network fuzzy system model of each subspace of combustion system, comprehensively obtain on this basis overall condition model, optimizing process is not only considered the constraint of unit load, introduce the integrate-cost function of boiler efficiency and oxides of nitrogen simultaneously, multi-objective optimization question is converted into single-object problem and processes.
Summary of the invention
The object of the present invention is to provide that a kind of step is clear and definite, clear concept, computing are simple, the modeling and optimization method and system of the boiler combustion of reliable results, the modeling method of inventing is all significantly improved in model training speed and extensive precision, the optimization method of inventing also has good improvement in optimization precision and optimal speed, the modeling and optimization that can be applicable to various Combustion System of Boiler Burning Fines is controlled, and has very strong practicality.
For achieving the above object, one aspect of the present invention provides a kind of power boiler burning subspace modeling and Multi objective optimization system, specifically by the following technical solutions:
It comprises boiler (1), the distributed monitoring control system (2) being connected with described boiler (1) and coal property test instrument (3), and the burning optimization workstation (4) being connected with coal property test instrument (3) with DCS (2), described burning optimization workstation (4) comprises interconnective communication module (5), interface module (6), optimize module (7) and model module (8), described communication module (5) obtains data from DCS (2) and coal property test instrument (3), optimum results is sent to DCS (2) and is further sent to boiler (1) by communication module (5).
The present invention on the other hand, provides a kind of power boiler burning subspace modeling and Multipurpose Optimal Method, adopts
Technical scheme realizes according to following steps:
Step (1). according to concrete boiler type and the relation that affects on burning, determine the input variable z of burning optimization model;
Step (2). from all input variable z, determine variable to be optimized
x;
Step (3). determine optimization method;
Step (4). off-line model is set up;
Step (5). real-time data acquisition;
Step (6). in line model correction;
Step (7). on-line optimization;
Step (8). optimum results is separated also to be implemented;
Step (9). repeating step (5)～(8).
In described step (1) the input variable z of burning optimization model include stove ature of coal, enter the total coal amount of stove, enter stove total air, each layer of secondary air register aperture, each grate firing throttle opening, bellows burner hearth differential pressure and oxygen content of smoke gas to the greatest extent, according to boiler concrete condition, optionally also comprise, enter each feeder aperture of stove, each coal pulverizer ventilation, burner pivot angle, each main steam temperature, each reheat steam temperature, main steam flow, total Feedwater Flow, each hot desuperheating water flow again, described in enter stove ature of coal and comprise coal net calorific value, volatile matter, ash content and total moisture.
Variable to be optimized in described step (2)
xcomprise each layer of secondary air register aperture, each grate firing throttle opening and oxygen content of smoke gas to the greatest extent, according to boiler concrete condition, optionally also comprise each feeder aperture, each coal pulverizer ventilation, burner pivot angle.
In described step (3), optimization method is ripe colony intelligence optimized algorithm, and wherein global search adopts differential evolution algorithm, and Local Search adopts simulated annealing, gives full play to the advantage of two kinds of optimized algorithms.
In described step (4), off-line model is set up and is comprised the steps:
Step (a1). from DCS or the SIS database of process industry widespread use, obtain unit operation historical data, comprise the input variable z of the burning optimization model that step (2) is described, and boiler efficiency E, oxides of nitrogen N and unit load L;
Step (a2). according to data mining Selection Model training sample from unit operation history data;
Step (a3). according to load, model training sample is divided into
iindividual section, i.e. L
1: [0,
a 2], L
2: [
a 1,
a 3] ..., L
i-1
: [
a i-2
,
a i ], L
i : [
a i-1
,
a i+ 1
],
Wherein,
a i be between 0 and unit rated load MCR between number, and meet relation: 0≤
a i <
a i+ 1
≤ 1.1MCR(
q=1,2 ...,
i);
Step (a4). according to above-mentioned
iindividual section sample data is set up respectively boiler efficiency Local Subspace model F
1,
i (
x,
y) (
i=1,2 ...,
i), set up discharged nitrous oxides Local Subspace model F
2,
i (
x,
y) (
i=1,2 ...,
i), set up unit load Local Subspace model F
3,
i (
x,
y) (
i=1,2 ...,
i).Modeling method is adaptive neural network fuzzy system (ANFIS).
Wherein,
yto remove variable to be optimized in whole input variable z of Optimized model
xthe rear remaining part input variable without optimization.
Step (a5). according to above-mentioned Local Subspace model, integrate respectively the F of boiler efficiency world model
1(
x,
y), the F of discharged nitrous oxides world model
2(
x,
y), the F of unit load world model
3(
x,
y), computing formula is as follows:
In formula,
Real-time data acquisition comprises the real time data of the signals such as input variable z, boiler efficiency E, oxides of nitrogen N, unit load L that gather burning optimization model in described step (5).
Described step comprises the steps: in line model correction in (6)
Step (b1). according to real time data, calculate respectively boiler efficiency model output E
m(k)=F
1(
x,
y), nitrogen oxide emission models output N
m(k)=F
2(
x,
y), unit load model output L
m(k)=F
3(
x,
y);
Step (b2). based on rolling window, calculate the model output E that nearest Q sampling calculates
m(k – Q+1:k), N
m(k – Q+1:k), L
mcomposition error J between (k – Q+1:k) and unit actual operating data E, N, L, if little this step of skipping of composition error J is without in line model correction, if composition error J revises model according to up-to-date unit data greatly; As preferably, the computing method of described composition error J are
In formula,
the mean value that represents respectively E (k-Q+1:k), N (k-Q+1:k), L (k-Q+1:k).
The criterion that composition error J is large or little is: if J>
think that composition error J is large; If J≤
think that composition error J is little.
generally get 0.01～10, concrete numeral is determined according to actual boiler parameter debugging.
In described step (6) in line model correction, each only need to revise described in step (a4)
iin individual boiler efficiency Local Subspace model 1 or 2,
iin individual discharged nitrous oxides Local Subspace model 1 or 2 and
iin individual unit load Local Subspace model 1 or 2, the pattern number that revise determines that method is: if unit load L meets L<
a 1, correction model 1; If unit load L meets
a i ≤ L<
a i+ 1
(
i=1,2 ...,
i-1), correction model
iwith
i+ 1; If unit load L meet L>=
a i , correction model
i.
In described step (7), on-line optimization adopts constraint Multipurpose Optimal Method to obtain best optimized variable
x=
x opt, target is to maximize boiler efficiency E, minimizes discharged nitrous oxides N, maintains unit load L almost constant simultaneously, objective function is
?Min?[E(k)–E
M(k)]E
﹩+[N
M(k)–N(k)]N
﹩
s.t.?E
M(k)=F
1(
x,?
y),
N
M(k)=F
2(
x,?
y),
L
M(k)=F
3(
x,?
y),
L
M(k)≥L(k)–△L,
x iL≤
x i≤
x iH.
In formula, E
﹩it is the cost that boiler efficiency reduces Yi Ge unit; N
﹩it is the cost of oxides of nitrogen rising Yi Ge unit; △ L is the constant interval that before and after optimizing, unit load is allowed, burning optimization can not, with the cost that is reduced to of generated energy, generally be got △ L=L
0* 5 ‰ (L
0represent unit rated load);
x iHwith
x iLit is respectively the bound constraint of variable to be optimized.
In described step (8), separated also enforcement of optimum results is by the optimum optimization variable obtaining
x=
x optin each layer of secondary air register aperture, each grate firing throttle opening, each feeder aperture, each coal pulverizer ventilation, burner pivot angle to the greatest extent, as new regulated quantity, send to on-the-spot topworks; By the optimum optimization variable obtaining
x=
x optin oxygen content of smoke gas send to oxygen content of smoke gas automatic control unit, as new oxygen content of smoke gas setting value.
The invention has the beneficial effects as follows: the training time of model and online correction time all shorten greatly, generalization ability also obviously strengthens, reliability increases, optimizing process is considered the constraint of unit load, introduces the integrate-cost function of boiler efficiency and oxides of nitrogen simultaneously, multi-objective optimization question is converted into single-object problem and processes, clear concept not only, optimum results is clear and definite, and optimal speed is fast, often can obtain in the short period of time the optimum solution of problem.
Accompanying drawing explanation
Fig. 1 is the schematic diagram of system.
Embodiment
Following examples are used for illustrating Implement methodology of the present invention, but are not used for limiting the scope of the invention.
For the coal-fired monoblock of certain 600MW, in conjunction with the specific embodiment of the present invention, a kind of power boiler burning subspace modeling and Multipurpose Optimal Method are described below.Concrete steps are as follows:
1. determine the input variable z of burning optimization model.Specifically include stove coal net calorific value, enter stove coal volatile matter, enter stove ash content of coal, enter stove coal total moisture, enter the total coal amount of stove, enter stove total air, each layer of secondary air register aperture, each grate firing throttle opening, bellows burner hearth differential pressure, oxygen content of smoke gas, each feeder aperture, each coal pulverizer ventilation, burner pivot angle to the greatest extent.
2. according to boiler situation, from all input variable z, determine variable to be optimized
x.Specifically comprise each layer of secondary air register aperture, each grate firing throttle opening, oxygen content of smoke gas to the greatest extent.According to boiler concrete condition, optionally also comprise burner pivot angle.
3. determine optimization method.Optimization method is ripe colony intelligence optimized algorithm, and wherein global search adopts differential evolution algorithm, and Local Search adopts simulated annealing.
4. off-line model is set up.Comprise the steps:
4.1. from historical data base, read unit operation history data, comprise the input variable z of all burning optimization models of definition in step 1, and boiler efficiency E, oxides of nitrogen N and unit load L;
4.2. adopt data digging method 600 groups of Selection Model training samples from unit operation history data;
4.3. according to load, model training sample is divided into 7 sections, i.e. L
1: [0,
a 2], L
2: [
a 1,
a 3], L
3: [
a 2,
a 4], L
4: [
a 3,
a 5], L
5: [
a 4,
a 6], L
6: [
a 5,
a 7], L
7: [
a 6,
a 8],
Wherein,
a 1=0.4,
a 2=0.5,
a 3=0.6,
a 4=0.7,
a 5=0.8,
a 6=0.9,
a 7=1.0,
a 8=1.1;
4.4. according to 7 section sample datas in step 4.3, adopt adaptive neural network fuzzy system (ANFIS) to set up respectively boiler efficiency Local Subspace model F
1,1(
x,
y), F
1,2(
x,
y), F
1,3(
x,
y), F
isosorbide-5-Nitrae(
x,
y), F
1,5(
x,
y), F
1,6(
x,
y), F
1,7(
x,
y), set up discharged nitrous oxides Local Subspace model F
2,1(
x,
y), F
2,2(
x,
y), F
2,3(
x,
y), F
2,4(
x,
y), F
2,5(
x,
y), F
2,6(
x,
y), F
2,7(
x,
y), set up unit load Local Subspace model F
3,1(
x,
y), F
3,2(
x,
y), F
3,3(
x,
y), F
3,4(
x,
y), F
3,5(
x,
y), F
3,6(
x,
y), F
3,7(
x,
y);
Wherein,
yto remove variable to be optimized in whole input variable z of Optimized model
xthe rear remaining part input variable without optimization.
4.5. according to the Local Subspace model in step 4.4, integrate respectively the F of boiler efficiency world model
1(
x,
y), the F of discharged nitrous oxides world model
2(
x,
y), the F of unit load world model
3(
x,
y), computing formula is as follows:
In formula,
5. real-time data acquisition.Obtain current data unit operation for entering stove coal net calorific value 25462kj/kg, enter stove coal volatile matter 29.22%, enter stove ash content of coal 15.75%, enter stove coal total moisture 11.31%, enter the total coal amount of stove 241.56t/h, enter stove total air 2572t/h, each layer of secondary air register aperture (54.3%, 56.2%, 57.7%, 54.1%, 50.7%, 44.6%), each grate firing is throttle opening (62.1% to the greatest extent, 52.2%), bellows burner hearth differential pressure 0.82kPa, oxygen content of smoke gas (3.32%), each feeder aperture (86.4%, 87.3%, 83.1%, 83.6%, 80.4%, 0.4%), each coal pulverizer ventilation (102.4t/h, 107.6t/h, 101.7t/h, 93.3t/h, 97.5t/h, 0.1t/h), burner pivot angle 60.5%, and boiler efficiency 92.5%, oxides of nitrogen 687.8mg/Nm
3, unit load 581.3MW.
6. in line model correction.Step is as follows:
6.1. according to real time data, calculate respectively boiler efficiency model output E
m(k)=F
1(
x,
y), nitrogen oxide emission models output N
m(k)=F
2(
x,
y), unit load model output L
m(k)=F
3(
x,
y);
6.2. based on rolling window, calculate the model output E that nearest 30 samplings calculate
m(k – 29:k), N
m(k – 29:k), L
mcomposition error J between (k – 29:k) and unit actual operating data E, N, L, as shown in the formula
In formula,
the mean value that represents respectively E (k-29:k), N (k-29:k), L (k-29:k).
Result of calculation obtains J=0.7.Getting the criterion that composition error J is large or little is
=1, result J<
without carrying out model, revise online, the adaptive neural network fuzzy system subspace model that is indicated as combustion system foundation has enough precision.
7. on-line optimization.According to actual conditions, get the cost E that boiler efficiency reduces Yi Ge unit
﹩=1000$, the cost N of oxides of nitrogen rising Yi Ge unit
﹩=0.003$, the constant interval △ L=600 * 5 ‰ that before and after optimizing, unit load is allowed, the bound constraint of variable to be optimized is respectively within the scope of topworks's physical restriction ± and 30%.Optimize following objective function
?Min?[E(k)–E
M(k)]E
﹩+[N
M(k)–N(k)]N
﹩
s.t.?E
M(k)=F
1(
x,?
y),
N
M(k)=F
2(
x,?
y),
L
M(k)=F
3(
x,?
y),
L
M(k)≥L(k)–△L,
x iL≤
x i≤
x iH.
The result that is finally optimized is each layer of secondary air register aperture (56.8%, 55.1%, 47.3%, 48.5%, 55.3%, 57.2%), each grate firing throttle opening (56.9%, 50.5%), oxygen content of smoke gas 2.98%, burner pivot angle 56.2% to the greatest extent.The boiler efficiency corresponding with this group optimum results is 93.3%, and discharged nitrous oxides is 498.5mg/Nm
3, unit load is 585.4MW.Visible, boiler efficiency has improved 0.8%, and discharged nitrous oxides has reduced 189.3mg/Nm
3, unit load has increased 4.1MW.This shows, by modeling provided by the invention and optimization method, the in the situation that of current fuel consumption, boiler efficiency can be improved, and discharged nitrous oxides can be reduced, and unit load slightly increases.
8. optimum results is separated also implements.By each layer of secondary air register aperture of the best obtaining, the most throttle opening of each grate firing, burner pivot angle, as new regulated quantity, send to DCS also and then send to on-the-spot topworks; The best oxygen content of smoke gas obtaining is sent to DCS thermal technology control loop, as new oxygen content of smoke gas setting value, by DCS, regulated, thereby make oxygen content of smoke gas actual value reach optimal value.
9. the new operating mode of 5～8 pairs of next rounds of repeating step continues to optimize, thereby remains that unit burns in optimum state.
Claims (10)
1. a power boiler burning subspace modeling and Multi objective optimization system, it is characterized in that: it comprises boiler (1), the distributed monitoring control system (2) being connected with described boiler (1) and coal property test instrument (3), and the burning optimization workstation (4) being connected with coal property test instrument (3) with DCS (2), described burning optimization workstation (4) comprises interconnective communication module (5), interface module (6), optimize module (7) and model module (8), described communication module (5) obtains data from DCS (2) and coal property test instrument (3), optimum results is sent to DCS (2) and is further sent to boiler (1) by communication module (5).
2. power boiler burning subspace modeling and the Multipurpose Optimal Method based on system claimed in claim 1, is characterized in that: specifically comprise the steps:
Step (1). according to concrete boiler type and the relation that affects on burning, determine the input variable z of burning optimization model;
Step (2). from all input variable z, determine variable x to be optimized;
Step (3). determine optimization method;
Step (4). off-line model is set up;
Step (5). real-time data acquisition;
Step (6). in line model correction;
Step (7). on-line optimization;
Step (8). optimum results is separated also to be implemented;
Step (9). repeating step (5)～(8).
3. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, it is characterized in that: in described step (1), the input variable z of burning optimization model includes stove ature of coal, enter the total coal amount of stove, enter stove total air, each layer of secondary air register aperture, each grate firing is throttle opening to the greatest extent, bellows burner hearth differential pressure and oxygen content of smoke gas, according to boiler concrete condition, optionally also include each feeder aperture of stove, each coal pulverizer ventilation, burner pivot angle, each main steam temperature, each reheat steam temperature, main steam flow, total Feedwater Flow, each is hot desuperheating water flow again, describedly enter stove ature of coal and comprise coal net calorific value, volatile matter, ash content and total moisture.
4. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: variable to be optimized in described step (2)
xcomprise each layer of secondary air register aperture, each grate firing throttle opening and oxygen content of smoke gas to the greatest extent, according to boiler concrete condition, optionally also comprise each feeder aperture, each coal pulverizer ventilation, burner pivot angle.
5. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, it is characterized in that: in described step (3), optimization method is ripe colony intelligence optimized algorithm, wherein global search adopts differential evolution algorithm, Local Search adopts simulated annealing, gives full play to the advantage of two kinds of optimized algorithms.
6. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: in described step (4), off-line model is set up and comprised the steps:
Step (a1). from DCS or the SIS database of process industry widespread use, obtain unit operation historical data, comprise the input variable z of the burning optimization model that step (2) is described, and boiler efficiency E, oxides of nitrogen N and unit load L;
Step (a2). according to data mining Selection Model training sample from unit operation history data;
Step (a3). according to load, model training sample is divided into
iindividual section, i.e. L
1: [0,
a 2], L
2: [
a 1,
a 3] ..., L
i-1
: [
a i-2
,
a i ], L
i : [
a i-1
,
a i+ 1
],
Wherein,
a i be between 0 and unit rated load MCR between number, and meet relation: 0≤
a i <
a i+ 1
≤ 1.1MCR(
q=1,2 ...,
i);
Step (a4). according to above-mentioned
iindividual section sample data is set up respectively boiler efficiency Local Subspace model F
1,
i (
x,
y) (
i=1,2 ...,
i), set up discharged nitrous oxides Local Subspace model F
2,
i (
x,
y) (
i=1,2 ...,
i), set up unit load Local Subspace model F
3,
i (
x,
y) (
i=1,2 ...,
i);
Modeling method is adaptive neural network fuzzy system (ANFIS);
Wherein,
yto remove variable to be optimized in whole input variable z of Optimized model
xthe rear remaining part input variable without optimization.
7. step (a5). according to above-mentioned Local Subspace model, integrate respectively the F of boiler efficiency world model
1(
x,
y), the F of discharged nitrous oxides world model
2(
x,
y), the F of unit load world model
3(
x,
y), computing formula is as follows:
In formula,
A kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: real-time data acquisition comprises the real time data of the signals such as input variable z, boiler efficiency E, oxides of nitrogen N, unit load L that gather burning optimization model in described step (5).
8. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: described step comprises the steps: in line model correction in (6)
Step (b1). according to real time data, calculate respectively boiler efficiency model output E
m(k)=F
1(
x,
y), nitrogen oxide emission models output N
m(k)=F
2(
x,
y), unit load model output L
m(k)=F
3(
x,
y);
Step (b2). based on rolling window, calculate the model output E that nearest Q sampling calculates
m(k – Q+1:k), N
m(k – Q+1:k), L
mcomposition error J between (k – Q+1:k) and unit actual operating data E, N, L, if little this step of skipping of composition error J is without in line model correction, if composition error J revises model according to up-to-date unit data greatly; As preferably, the computing method of described composition error J are
In formula,
the mean value that represents respectively E (k-Q+1:k), N (k-Q+1:k), L (k-Q+1:k);
Further preferred, the criterion that described composition error J is large or little is: if J>
think that composition error J is large; If J≤
think that composition error J is little;
generally get 0.01～10.
9. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: in described step (7), on-line optimization adopts constraint Multipurpose Optimal Method to obtain best optimized variable
x=
x opt, target is to maximize boiler efficiency E, minimizes discharged nitrous oxides N, maintains unit load L almost constant simultaneously, objective function is
?Min?[E(k)–E
M(k)]E
﹩+[N
M(k)–N(k)]N
﹩
s.t.?E
M(k)=F
1(
x,?
y),
N
M(k)=F
2(
x,?
y),
L
M(k)=F
3(
x,?
y),
L
M(k)≥L(k)–△L,
x iL≤
x i≤
x iH
In formula, E
﹩it is the cost that boiler efficiency reduces Yi Ge unit; N
﹩it is the cost of oxides of nitrogen rising Yi Ge unit; △ L is the constant interval that before and after optimizing, unit load is allowed, burning optimization can not, with the cost that is reduced to of generated energy, generally be got △ L=L
0* 5 ‰ (L
0represent unit rated load);
x iHwith
x iLit is respectively the bound constraint of variable to be optimized.
10. a kind of power boiler burning subspace modeling according to claim 2 and Multipurpose Optimal Method, is characterized in that: in described step (8), separated also enforcement of optimum results is by the optimum optimization variable obtaining
x=
x optin each layer of secondary air register aperture, each grate firing throttle opening, each feeder aperture, each coal pulverizer ventilation, burner pivot angle to the greatest extent, as new regulated quantity, send to on-the-spot topworks; By the optimum optimization variable obtaining
x=
x optin oxygen content of smoke gas send to oxygen content of smoke gas automatic control unit, as new oxygen content of smoke gas setting value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310541803.3A CN103576655B (en) | 2013-11-06 | 2013-11-06 | A kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310541803.3A CN103576655B (en) | 2013-11-06 | 2013-11-06 | A kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103576655A true CN103576655A (en) | 2014-02-12 |
| CN103576655B CN103576655B (en) | 2016-03-02 |
Family
ID=50048659
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310541803.3A Expired - Fee Related CN103576655B (en) | 2013-11-06 | 2013-11-06 | A kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103576655B (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103838216A (en) * | 2014-03-07 | 2014-06-04 | 华北电力大学（保定） | Power station boiler combustion optimization method based on data driven case matching |
| CN103870877A (en) * | 2014-03-28 | 2014-06-18 | 西安西热控制技术有限公司 | System and method for intelligently controlling boiler combustion based on neural network |
| CN104595924A (en) * | 2015-01-14 | 2015-05-06 | 河北省电力建设调整试验所 | Method and device for establishing boiler combustion process model |
| CN105590005A (en) * | 2016-01-22 | 2016-05-18 | 安徽工业大学 | Numerical simulation method for interaction of coal powder particles in burning |
| CN105823080A (en) * | 2016-03-24 | 2016-08-03 | 东南大学 | Model-free boiler combustion optical control method based on numerical optimization extremum searching |
| CN106019935A (en) * | 2016-04-28 | 2016-10-12 | 天津市职业大学 | Multi-target boiler combustion optimization based on constrained fuzzy association rules |
| CN106327021A (en) * | 2016-08-31 | 2017-01-11 | 西安艾贝尔科技发展有限公司 | Boiler combustion optimization air distribution method based on online model prediction |
| CN104534507B (en) * | 2014-11-18 | 2017-03-29 | 华北电力大学（保定） | A kind of boiler combustion optimization control method |
| CN106885228A (en) * | 2017-02-10 | 2017-06-23 | 青岛高校信息产业股份有限公司 | A kind of boiler coal-air ratio optimization method and system |
| CN108426266A (en) * | 2018-03-01 | 2018-08-21 | 中国神华能源股份有限公司 | Boiler combustion control system and method |
| CN109764327A (en) * | 2018-12-29 | 2019-05-17 | 浙江浙能技术研究院有限公司 | A kind of boiler imitates poor intelligent analysis system and method |
| CN109815594A (en) * | 2019-01-26 | 2019-05-28 | 新奥数能科技有限公司 | The method and apparatus that a kind of determining coal changes burner installation site in the chain furnace of gas |
| CN110017470A (en) * | 2019-02-27 | 2019-07-16 | 中国大唐集团科学技术研究院有限公司火力发电技术研究院 | Boiler combustion multiobjective optimization control method based on differential evolution algorithm |
| WO2020015466A1 (en) * | 2018-07-18 | 2020-01-23 | 厦门邑通软件科技有限公司 | Boiler coal saving control method |
| CN110848733A (en) * | 2020-01-15 | 2020-02-28 | 南京科远智慧科技集团股份有限公司 | Combustion optimization method based on coal quality on-line monitoring |
| 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 |
| CN111931436A (en) * | 2020-08-10 | 2020-11-13 | 哈尔滨锅炉厂有限责任公司 | Burner nozzle air quantity prediction method based on numerical simulation and neural network |
| CN112036091A (en) * | 2020-09-14 | 2020-12-04 | 哈尔滨锅炉厂有限责任公司 | Method for predicting air quantity of nozzle of four-corner tangential burner by using regression algorithm |
| CN112066411A (en) * | 2020-09-02 | 2020-12-11 | 常州工学院 | Optimization method for boiler combustion |
| CN112163380A (en) * | 2020-09-24 | 2021-01-01 | 哈尔滨锅炉厂有限责任公司 | System and method for predicting furnace oxygen concentration based on numerical simulation neural network |
| CN112628712A (en) * | 2021-01-11 | 2021-04-09 | 大唐黄岛发电有限责任公司 | Secondary air closed-loop optimization control system based on air door resistance coefficient |
| CN113533622A (en) * | 2021-07-19 | 2021-10-22 | 华能国际电力股份有限公司上海石洞口第二电厂 | Coal quality prediction method for coal mill based on neural network |
| CN114135899A (en) * | 2021-11-10 | 2022-03-04 | 吉林省电力科学研究院有限公司 | Device and method for improving combustion optimization rate of boiler |
| CN116049977A (en) * | 2022-12-26 | 2023-05-02 | 西南科技大学 | Parameter multi-objective optimization method for aero-engine combustion chamber |
| CN116736713A (en) * | 2023-06-13 | 2023-09-12 | 天津国能津能滨海热电有限公司 | Power plant combustion control system and method based on NARX prediction model |
| CN117057238A (en) * | 2023-08-15 | 2023-11-14 | 天津大学 | Combustor stable combustion blunt body design method based on physical information operator network model |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2641472Y (en) * | 2003-05-16 | 2004-09-15 | 浙江大学 | On-line optimization control device for boiler combustion based on computation intelligence |
| CN101634459A (en) * | 2009-08-24 | 2010-01-27 | 陶晓鹏 | Thermal power generation boiler intelligent combustion optimizing system and realizing method thereof |
| CN101799848A (en) * | 2010-03-09 | 2010-08-11 | 江西省电力科学研究院 | Method for obtaining energy loss analysis parameter answer value of furnace of thermal power set |
| CN102261671A (en) * | 2010-05-28 | 2011-11-30 | 王荣虎 | Boiler combustion multi-constraint and multi-object optimization expert system and optimization method thereof |
| CN102654768A (en) * | 2012-05-09 | 2012-09-05 | 北京华电天仁电力控制技术有限公司 | Rule-based power station boiler combustion coal-saving nitrogen-reducing control method and device |
| US20130166043A1 (en) * | 2011-07-20 | 2013-06-27 | Nec Laboratories America, Inc. | Optimal energy management of a microgrid system using multi-objective optimization |
-
2013
- 2013-11-06 CN CN201310541803.3A patent/CN103576655B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN2641472Y (en) * | 2003-05-16 | 2004-09-15 | 浙江大学 | On-line optimization control device for boiler combustion based on computation intelligence |
| CN101634459A (en) * | 2009-08-24 | 2010-01-27 | 陶晓鹏 | Thermal power generation boiler intelligent combustion optimizing system and realizing method thereof |
| CN101799848A (en) * | 2010-03-09 | 2010-08-11 | 江西省电力科学研究院 | Method for obtaining energy loss analysis parameter answer value of furnace of thermal power set |
| CN102261671A (en) * | 2010-05-28 | 2011-11-30 | 王荣虎 | Boiler combustion multi-constraint and multi-object optimization expert system and optimization method thereof |
| US20130166043A1 (en) * | 2011-07-20 | 2013-06-27 | Nec Laboratories America, Inc. | Optimal energy management of a microgrid system using multi-objective optimization |
| CN102654768A (en) * | 2012-05-09 | 2012-09-05 | 北京华电天仁电力控制技术有限公司 | Rule-based power station boiler combustion coal-saving nitrogen-reducing control method and device |
Cited By (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103838216B (en) * | 2014-03-07 | 2016-08-17 | 华北电力大学（保定） | Power boiler burning optimization method based on data-driven case coupling |
| CN103838216A (en) * | 2014-03-07 | 2014-06-04 | 华北电力大学（保定） | Power station boiler combustion optimization method based on data driven case matching |
| CN103870877A (en) * | 2014-03-28 | 2014-06-18 | 西安西热控制技术有限公司 | System and method for intelligently controlling boiler combustion based on neural network |
| CN103870877B (en) * | 2014-03-28 | 2016-09-28 | 西安西热控制技术有限公司 | A kind of boiler combustion intelligent control method based on neutral net |
| CN104534507B (en) * | 2014-11-18 | 2017-03-29 | 华北电力大学（保定） | A kind of boiler combustion optimization control method |
| CN104595924A (en) * | 2015-01-14 | 2015-05-06 | 河北省电力建设调整试验所 | Method and device for establishing boiler combustion process model |
| CN105590005A (en) * | 2016-01-22 | 2016-05-18 | 安徽工业大学 | Numerical simulation method for interaction of coal powder particles in burning |
| CN105590005B (en) * | 2016-01-22 | 2018-11-13 | 安徽工业大学 | The method for numerical simulation that combustion process interacts between a kind of pulverized coal particle |
| CN105823080A (en) * | 2016-03-24 | 2016-08-03 | 东南大学 | Model-free boiler combustion optical control method based on numerical optimization extremum searching |
| CN106019935A (en) * | 2016-04-28 | 2016-10-12 | 天津市职业大学 | Multi-target boiler combustion optimization based on constrained fuzzy association rules |
| CN106327021A (en) * | 2016-08-31 | 2017-01-11 | 西安艾贝尔科技发展有限公司 | Boiler combustion optimization air distribution method based on online model prediction |
| CN106885228A (en) * | 2017-02-10 | 2017-06-23 | 青岛高校信息产业股份有限公司 | A kind of boiler coal-air ratio optimization method and system |
| CN108426266B (en) * | 2018-03-01 | 2019-10-15 | 中国神华能源股份有限公司 | Boiler combustion control system and method |
| CN108426266A (en) * | 2018-03-01 | 2018-08-21 | 中国神华能源股份有限公司 | Boiler combustion control system and method |
| JP2021530669A (en) * | 2018-07-18 | 2021-11-11 | 厦門邑通軟件科技有限公司Xiamen Etom Software Technology Co., Ltd. | Boiler coal saving control method |
| US20210278078A1 (en) * | 2018-07-18 | 2021-09-09 | Xiamen Etom Software Technology Co., Ltd. | Boiler coal saving control method |
| WO2020015466A1 (en) * | 2018-07-18 | 2020-01-23 | 厦门邑通软件科技有限公司 | Boiler coal saving control method |
| CN109764327A (en) * | 2018-12-29 | 2019-05-17 | 浙江浙能技术研究院有限公司 | A kind of boiler imitates poor intelligent analysis system and method |
| CN109815594B (en) * | 2019-01-26 | 2022-12-27 | 新奥数能科技有限公司 | Method and device for determining installation position of burner in coal-to-gas chain furnace |
| CN109815594A (en) * | 2019-01-26 | 2019-05-28 | 新奥数能科技有限公司 | The method and apparatus that a kind of determining coal changes burner installation site in the chain furnace of gas |
| CN110017470A (en) * | 2019-02-27 | 2019-07-16 | 中国大唐集团科学技术研究院有限公司火力发电技术研究院 | Boiler combustion multiobjective optimization control method based on differential evolution algorithm |
| CN110848733A (en) * | 2020-01-15 | 2020-02-28 | 南京科远智慧科技集团股份有限公司 | Combustion optimization method based on coal quality on-line monitoring |
| 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 |
| CN111931436A (en) * | 2020-08-10 | 2020-11-13 | 哈尔滨锅炉厂有限责任公司 | Burner nozzle air quantity prediction method based on numerical simulation and neural network |
| CN112066411A (en) * | 2020-09-02 | 2020-12-11 | 常州工学院 | Optimization method for boiler combustion |
| CN112036091A (en) * | 2020-09-14 | 2020-12-04 | 哈尔滨锅炉厂有限责任公司 | Method for predicting air quantity of nozzle of four-corner tangential burner by using regression algorithm |
| CN112163380A (en) * | 2020-09-24 | 2021-01-01 | 哈尔滨锅炉厂有限责任公司 | System and method for predicting furnace oxygen concentration based on numerical simulation neural network |
| CN112628712A (en) * | 2021-01-11 | 2021-04-09 | 大唐黄岛发电有限责任公司 | Secondary air closed-loop optimization control system based on air door resistance coefficient |
| CN113533622A (en) * | 2021-07-19 | 2021-10-22 | 华能国际电力股份有限公司上海石洞口第二电厂 | Coal quality prediction method for coal mill based on neural network |
| CN114135899A (en) * | 2021-11-10 | 2022-03-04 | 吉林省电力科学研究院有限公司 | Device and method for improving combustion optimization rate of boiler |
| CN116049977A (en) * | 2022-12-26 | 2023-05-02 | 西南科技大学 | Parameter multi-objective optimization method for aero-engine combustion chamber |
| CN116049977B (en) * | 2022-12-26 | 2024-04-12 | 西南科技大学 | Parameter multi-objective optimization method for aero-engine combustion chamber |
| CN116736713A (en) * | 2023-06-13 | 2023-09-12 | 天津国能津能滨海热电有限公司 | Power plant combustion control system and method based on NARX prediction model |
| CN117057238A (en) * | 2023-08-15 | 2023-11-14 | 天津大学 | Combustor stable combustion blunt body design method based on physical information operator network model |
| CN117057238B (en) * | 2023-08-15 | 2024-04-19 | 天津大学 | Combustor stable combustion blunt body design method based on physical information operator network model |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103576655B (en) | 2016-03-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103576655B (en) | A kind of power boiler burning subspace modeling and Multipurpose Optimal Method and system | |
| Shi et al. | Combustion optimization of ultra supercritical boiler based on artificial intelligence | |
| Tang et al. | Adaptive nonlinear model predictive control of NOx emissions under load constraints in power plant boilers | |
| CN109325313A (en) | Based on improvement quantum telepotation boiler of power plant NOXPrediction model device | |
| CN103838216A (en) | Power station boiler combustion optimization method based on data driven case matching | |
| CN104613468A (en) | Circulating fluidized bedboiler combustion optimizing control method based on fuzzy adaptive inference | |
| CN111881554B (en) | Optimization control method for boiler changing along with air temperature | |
| CN115111601B (en) | Multi-target boiler combustion optimization control method based on embedded algorithm fusion under variable load | |
| CN113742997B (en) | Intelligent optimization setting method for air quantity in urban solid waste incineration process | |
| CN112628712A (en) | Secondary air closed-loop optimization control system based on air door resistance coefficient | |
| CN111881120B (en) | Intelligent operation optimization method for boiler | |
| Qiao | A novel online modeling for NOx generation prediction in coal-fired boiler | |
| CN110888401B (en) | Combustion control optimization method and device for thermal generator set and readable storage medium | |
| Wang et al. | NOx emission prediction of coal-fired power units under uncertain classification of operating conditions | |
| CN113177352A (en) | Boiler combustion optimization system and method based on numerical simulation and artificial intelligence | |
| Mikulandric et al. | Improvement of environmental aspects of thermal power plant operation by advanced control concepts | |
| CN114397813B (en) | Power generation boiler combustion continuous sliding mode control method based on slow time-varying disturbance observer | |
| Zhao et al. | Modeling and optimization of efficiency and NOx emission at a coal-fired utility boiler | |
| Balamurugan et al. | ANN-SQP Approach for NOx emission reduction in coal fired Boilers | |
| CN113887130B (en) | Industrial boiler operation optimization method based on ensemble learning | |
| Zhu et al. | Research on methods of forecasting unburned carbon content in the fly ash from coal-fired power plant | |
| Xu et al. | The balanced model and optimization of NOx emission and boiler efficiency at a coal-fired utility boiler | |
| Ilamathi et al. | A novel approach for modelling of NOx emission reduction in a tangentially fired coal boiler | |
| CN114754353B (en) | Circulating fluidized bed boiler combustion optimization method integrating neighborhood rough set machine learning | |
| CN111520740B (en) | Method for coordinately optimizing operation of multiple porous medium combustors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160302 Termination date: 20191106 |