CN103759290A - Large coal-fired unit online monitoring and optimal control system and implementation method thereof - Google Patents

Large coal-fired unit online monitoring and optimal control system and implementation method thereof Download PDF

Info

Publication number
CN103759290A
CN103759290A CN201410021273.4A CN201410021273A CN103759290A CN 103759290 A CN103759290 A CN 103759290A CN 201410021273 A CN201410021273 A CN 201410021273A CN 103759290 A CN103759290 A CN 103759290A
Authority
CN
China
Prior art keywords
boiler
coal
control system
optimal control
fired power
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
Application number
CN201410021273.4A
Other languages
Chinese (zh)
Inventor
罗嘉
朱亚清
吴乐
刘小伟
叶向前
张曦
伍宇忠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huazhong University of Science and Technology
Electric Power Research Institute of Guangdong Power Grid Co Ltd
Original Assignee
Huazhong University of Science and Technology
Electric Power Research Institute of Guangdong Power Grid Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huazhong University of Science and Technology, Electric Power Research Institute of Guangdong Power Grid Co Ltd filed Critical Huazhong University of Science and Technology
Priority to CN201410021273.4A priority Critical patent/CN103759290A/en
Publication of CN103759290A publication Critical patent/CN103759290A/en
Pending legal-status Critical Current

Links

Images

Abstract

The invention provides a large coal-fired unit online monitoring and optimal control system and an implementation method thereof. The method includes the steps of analyzing combustion characteristics of a boiler of a large coal-fired unit, and performing boiler combustion optimization and adjustment test to acquire boiler operation key parameters; determining a relation of the boiler operation key parameters to boiler efficiency and NOx emission, and establishing an optimal control model according to the relation; training the optimal control model to obtain specific optimal values of boiler operation parameters under highest boiler efficiency. The large coal-fired unit online monitoring and optimal control system and the implementation method thereof have the advantages that boiler efficiency is improved on the premise of guaranteeing low NOx emission and operational economy of power plants is effectively improved.

Description

Large-sized Coal-fired Power group on-line monitoring and Optimal Control System and its implementation
Technical field
The present invention relates to field of power, particularly relate to the implementation method of a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System and a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System.
Background technology
High efficiency and low pollution are the eternal themes of electric power development.Along with technological progress, large capacity, high parameter and low emission stage have also been stepped in the development of the station boiler of China at present.But due to China's station boiler, to have ature of coal changeable, the feature of load wide variation, the feature that simultaneously boiler object has typical large inertia, postpones and become when model parameter is remarkable with operating mode greatly, therefore when load changes, fuel quantity, air inducing amount, air output coordination simultaneously, reach and should adapt to load variations, make fuel quantity, air output proportional again, also will make combustion chamber draft keep certain numerical value; When production load is relatively stable, should keeps the relatively stable of fuel quantity, air output, and can eliminate rapidly external interference to they impacts separately.
Due to NOx(nitrogen oxide in the combustion product of coal-burning boiler), the emission performance of unburned carbon in flue dust and flue gas peroxide amount is complicated, be achieved by a variety of factors, such as coal, boiler heat load, coal-air ratio, air distribution mode, fire box temperature etc., thereby be difficult to use the succinct clear and definite Mathematical Modeling based on mechanism to be described, often need to adopt real stove method of testing to be determined, and the method for groping gradually to fall low NOx drainage, improving boiler combustion efficiency by test result.But on-the-spot real stove test job amount is large, test operating mode is limited, and the various parameter that affects superposes mutually, causes in data analysis and has difficulties, and causes and can not obtain the estimation equation of related characteristics coefficient and clearer and more definite Mathematical Modeling according to test result.Thereby result of the test does not possess good versatility yet, it further can not be promoted in order to describe the boiler combustion process under different condition.When operating mode is different, the thermal efficiency of boiler is different, and smoke components, heat transfer situation also have a great difference, cause the difference of boiler combustion characteristic, have stronger non-linear, time variation, between each parameter of boiler, influence each other, and have very strong coupling.Boiler uses coal and operating parameter is ever-changing, cannot guarantee that it moves under operating condition of test, therefore cannot, according to the NOx emission performance under other operation operating modes of operating condition of test prognosis modelling, be unfavorable for that the combustion economization of low NOx drainage and raising boiler falls in coal-burning boiler by burning adjustment.In the burning optimization problem of large-sized boiler, (reduce the efficiency of combustion of NOx discharge capacity and raising boiler), need model description combustion process, provide the prediction to indexs such as NOx discharge, efficiencies of combustion in searching process, but the complexity of boiler combustion characteristic is but for modeling has brought very large difficulty.Generally speaking, research technique adds the calculation of rational model and DCS(Distributed Control System, Distributed Control System) data are very helpful to improving boiler efficiency with falling low NOx drainage.
Yet simple burning optimization adjustment test can not Automatic-searching burning optimal conditions, also accurately identification special parameter is big or small on the impact of burning for the model of setting up according to service data merely.Therefore, how to adjust better boiler operating parameter to reduce NOx discharge capacity and to improve boiler efficiency, become problem demanding prompt solution.
Summary of the invention
Based on this, the invention provides a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System and its implementation, can improve boiler efficiency and fall low NOx drainage, guarantee that boiler operates in minimum generating energy consumption and minimum pollution thing state automatically.
For achieving the above object, the present invention adopts following technical scheme:
An implementation method for Large-sized Coal-fired Power group on-line monitoring and Optimal Control System, comprises the following steps:
The boiler of Large-sized Coal-fired Power group is carried out to combustion characteristic analysis, and carry out boiler combustion optimization adjustment test, obtain boiler operatiopn key parameter;
Determine the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and set up optimizing control models according to described relation;
Described optimizing control models is trained, obtain the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
Large-sized Coal-fired Power group on-line monitoring and an Optimal Control System, comprising:
Parameter acquisition module, for the boiler of Large-sized Coal-fired Power group is carried out to combustion characteristic analysis, and carries out boiler combustion optimization adjustment test, obtains boiler operatiopn key parameter;
Relation is set up module, for determining the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and sets up optimizing control models according to described relation;
Optimize Numerical Simulation Module, for described optimizing control models is trained, obtain the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
By above scheme, can be found out, a kind of Large-sized Coal-fired Power group on-line monitoring of the present invention and Optimal Control System and its implementation, can set up the key parameter of boiler operatiopn and the relation between boiler efficiency and NOx discharge, and set up accordingly optimizing control models, then by the training of this optimizing control models being obtained to the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency, operations staff only need to carry out boiler combustion adjustment according to these concrete numerical value of optimizing, and just can on the basis that guarantees low NOx drainage, improve boiler efficiency.The present invention, no matter be all to guarantee that under open loop mode or closed loop mode boiler operates in minimum generating energy consumption and minimum pollution thing state automatically, effectively raises the economy of power plant's operation.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of the implementation method of a kind of Large-sized Coal-fired Power group on-line monitoring of the present invention and Optimal Control System;
Fig. 2 is whole implementation scheme schematic diagram of the present invention;
Fig. 3 is the relational model schematic diagram of NOx discharge of the present invention and boiler efficiency;
Fig. 4 is BP network model structural representation of the present invention;
Fig. 5 is BP network model algorithm steps flow chart of the present invention;
Fig. 6 is the structural representation of a kind of Large-sized Coal-fired Power group on-line monitoring of the present invention and Optimal Control System.
The specific embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
The invention provides that a kind of boiler operatiopn is stable, Automatic-searching lowest energy consumption and minimum on-line monitoring and the Optimal Control System that pollutes (NOx), even if make under atypia operating mode, do not need the staff of power plant rule of thumb to adjust boiler operating parameter yet, but directly by on-line monitoring analysis and control system, boiler manufacturing parameter under each operating mode can automatic analysis and adjusted to the best, guarantee that boiler operates in minimum generating energy consumption and minimum pollution thing state automatically.For achieving the above object, the invention provides the implementation method of a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System, shown in Figure 1, comprise the following steps:
Step S101, carries out combustion characteristic analysis to the boiler of Large-sized Coal-fired Power group, and carries out boiler combustion optimization adjustment test, obtains boiler operatiopn key parameter, then enters step S102.
As a good embodiment, described boiler operatiopn key parameter can comprise: excess air coefficient, wind/coal ratio, burn air quantity, fineness of pulverized coal, change coal pulverizer operation mode, load and coal characteristic etc.
In the present invention, the object of carrying out boiler combustion optimization adjustment test is as follows: on the one hand improve boiler efficiency and reduce on the basis of disposal of pollutants, draw this boiler operatiopn key parameter as fineness of pulverized coal, excess air coefficient, coal-air ratio (primary air flow), burn air quantity and the rule that affects on boiler efficiency and NOx discharge such as allocation of the amount of air and coal characteristic; On the other hand, by adjustment, test, can improve boiler operatiopn state, actual act parameter that grasp to adjust boiler operatiopn is prepared for on-line optimization expert system draws automatic adjustment scheme execution as secondary air register aperture, apparatus for separating fine powder from coarst powder baffle opening etc.
Step S102, determines the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and sets up optimizing control models according to described relation, then enters step S103.
Step S103, trains described optimizing control models, obtains the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
As a good embodiment, after obtaining described concrete optimization numerical value, can also comprise the steps: to adjust boiler operating parameter according to described concrete optimization numerical value.
As a good embodiment, the optimizing control models of setting up in the present invention can be the optimizing control models based on BP neural network theory.
In addition, as a good embodiment, the process that described optimizing control models is trained specifically can comprise as follows: adopt and with the gradient function of momentum and dynamic self-adapting learning rate, described optimizing control models is trained.Wherein as follows with the study formula of momentum:
(1) weights of output layer change
The weights that output to k output from i are had:
Δw 2 ki = - η ∂ E ∂ w 2 ki = - η ∂ E ∂ a 2 k ∂ a 2 k ∂ w 2 ki
= η ( t k - a 2 k ) f 2 ′ a 1 i = ηδ ki a 1 i ;
Wherein: δ ki=(t k-a 2k)=e kf2 ';
e k=t k-a2 k；
Δb 2 ki = - η ∂ E ∂ b 2 ki = - η ∂ E ∂ a 2 k · ∂ a 2 k ∂ b 2 ki
In like manner can obtain :=η (t k-a2 k) .f2 '=η. δ ki.
(2) hidden layer weights change
Weights to being input to i output from j, have:
Δw 1 ij = - η ∂ E ∂ W 1 ij = - η ∂ E ∂ a 2 k · ∂ a 2 k ∂ a 1 i · ∂ a 1 i ∂ W 1 ij
= η Σ k = 1 s 2 ( t k - a 2 k ) . f 2 ′ . w 2 ki . f 1 ′ . pj = η . δ ij . p j ;
Wherein: δ ij = e i . f 1 ′ , e i = Σ k = 1 s 2 δ ki w 2 ki ;
In like manner can obtain: Δ b1 i=η δ ij.
Meanwhile, the basic imagination that adopts self adaptation to adjust the improvement algorithm of parameter is that learning rate should be according to error change and self adaptation adjustment, so that the direction that weight coefficient adjustment reduces to error changes, its iterative process can be expressed as:
w(k+1)=w(k)-η▽f(w(k))。
With the gradient function of momentum and dynamic self-adapting learning rate, combine the advantage of two kinds of methods, both used factor of momentum, also used adaptive learning rate, the search of its next round iteration is automatically obvious near the effect of direction of steepest descent, thereby effectively avoided the generation of continuous little step-length, better astringency.
Set up after the forecast model of NOx emission performance and boiler efficiency, in conjunction with institute's established model and genetic algorithm, realize operating scheme Optimizing Search, optimization aim respectively: optimize separately NOx discharge capacity, optimize boiler efficiency and take into account NOx discharge capacity and boiler efficiency is implemented complex optimum separately.
Utilize constraint optimization to ask for the optimization operating scheme of low gross coal consumption rate and low NOx drainage, then according to this scheme, instruct operations staff's real time execution (open loop mode) or by Distributed Control System (DCS), optimization operating scheme be input to boiler automatically under certain condition, adjusting adjustable operational factor (closed loop mode) to realize low NOx combustion optimization operation.
According to above research technique, rational model calculation and DCS data, to improving boiler efficiency, form an automatic control Optimal Expert System with reducing NOx, assurance boiler can optimized running in each operating mode.
A kind of boiler operatiopn provided by the invention is stable, Automatic-searching lowest energy consumption and minimum on-line monitoring and the Optimal Control System that pollutes (NOx), as shown in Figure 2, the acquisition of whole control system embodiment, the first step is to obtain related data by burning optimization regulation experiment, second step is to set up Optimized model, and the 3rd step is to carry out modelling verification and optimization.
First by burning optimization regulation experiment, obtain needed all relevant parameters, as excess air coefficient, primary air flow (wind/coal ratio), OFA burn air quantity, fineness of pulverized coal, change coal pulverizer operation mode, load, coal (ature of coal) characteristic etc.
Then be that the parameter of download boiler operatiopn from station boiler Distributed Control System (DCS) is in combustion control system server, simultaneously by the data-interface of fineness of pulverized coal on-line measurement device, coal analysis instrument, boiler flyash carbon content on-line measurement device, fineness of pulverized coal value, coal analysis data and boiler flyash carbon content are downloaded in combustion on-line optimization control system server, set up combustion control system database.
Then according to the database of combustion control system, utilize BP neural network modeling approach, control system can obtain the operational factor of boiler and the relation between boiler efficiency and NOx discharge, object module is set up, now can utilize the data of artificial experiment measuring it is trained and optimize, according to predicated error, carry out parameter adjustment, model is finally optimized.
Realizing on the basis of boiler efficiency and NOx discharge modeling, burning on-line control system can utilize BP network model to solve and obtain the boiler operating parameter of pursuing under maximum boiler efficiency, comprise that excess air coefficient, primary air flow (wind/coal ratio), OFA burn the concrete optimization numerical value of the various boiler operating parameters such as air quantity, operations staff adopts the boiler operating parameter of optimization to carry out boiler combustion adjustment, just can obtain maximum boiler efficiency, thereby improve the economy of power plant's operation.
Below each step is elaborated:
One, burning optimization adjustment test
For actual coal-burning boiler, the affect rule of research key parameter on boiler efficiency and NOx discharge, design parameter is as follows: (a) excess air coefficient; (b) primary air flow (wind/coal ratio); (c) OFA burns air quantity; (d) fineness of pulverized coal; (e) become coal pulverizer operation mode; (f) load; (g) coal (ature of coal) characteristic etc.
Obtain the affect rule of key parameter on boiler efficiency and NOx discharge, and obtain related data, for model, set up and establish relevant rudimentary.
Specific descriptions and adjustment mode for each parameter are as follows:
(a) become excess air coefficient
Test is by regulating air blower inlet baffle opening, and maintains mill exhauster outlet blast (primary air pressure) and other operational factor is stablized under constant condition, and change boiler enters stove air quantity, to change boiler operatiopn oxygen amount.During test, by monitoring air preheater entrance oxygen amount, control, after operating mode is stable, measure boiler characteristics data, and weigh and send air-introduced machine electric current, superheated steam, reheat steam temperature, determine boiler optimum operation oxygen amount.
(b) become primary air flow (wind/coal ratio)
Becoming primary air ratio test is to keep suitable primary air flow in order to find boiler burner when different operating modes are moved.Its regulative mode is mainly on the basis of operating mode of knowing the real situation, and the power that changes primary air fan is adjusted primary air flow, simultaneously in order to guarantee that total blast volume is constant, synchronous change secondary air flow.When test is adjusted, furnace outlet excess air coefficient maintains necessarily, and Secondary Air air distribution mode adopts even type air distribution.
(c) become and burn air quantity
The difference of the upper and lower secondary air flow of test high spot review burner is divided the impact of pairing emission of NOx of boiler amount.The secondary wind pressure of combustion system, air quantity and baffle opening are demarcated simultaneously.
(d) become fineness of pulverized coal
Qualified fineness of pulverized coal is the basis of realizing coal dust uniform distribution and having lower flying dust carbon containing mass concentration (below 6%).Become fineness of pulverized coal test and mainly take into account pulverizer capacity and fineness of pulverized coal size, regulate fineness of pulverized coal mainly by regulating the angle of separator for coal mill baffle plate, meeting under the prerequisite of each coal pulverizer on-load ability, by regulating fineness of pulverized coal to improve the burnout rate of coal dust, and then raising boiler thermal output, but the adjusting of separator baffle plate also can be served impact to coal-grinding unit consumption band.
(e) become coal pulverizer operation mode
In Boiler Furnace, the quality of combustion conditions, is not only subject to the impact of the sharing of load of air distribution mode, burner, and relevant with coal pulverizer shutting mode.Becoming the test of coal pulverizer operation mode, is in order to grasp boiler under different load, the impact of coal pulverizer operation mode on boiler safety, stable, economical operation.Maintaining under the constant condition of other operational factors, only changing the operation mode of coal pulverizer, to measure the different impacts of mill mode on combustion efficiency in Boiler Furnace and economy of throwing.
(f) varying duty
Boiler load attribute testing is economy, the safety and stability of boiler when checking boiler to move under different load.Main project comprises combustion characteristics (efficiency of combustion, coking situation), main steam and reheated steam parameter, steam temperature control characteristic, boiler thermal output and the pollutant emission of boiler under different load.
(g) coal (ature of coal) characteristic
Because in coal-fired boiler NOx discharge, a big chunk is from the nitrogen in fuel, fuel bound nitrogen content is higher in general, and the discharge capacity of NOx is also just larger, and in fuel, the existence form of nitrogen is different, and NOx growing amount also just changes thereupon.Meanwhile, in coal volatile ingredient, various elements can affect NOx growing amount than also, and in coal, O/N and S/N ratio may have influence on the discharge capacity of NOx, comprise that in addition the coal characteristics such as volatile content, moisture, caloric value, grindability index also can affect the formation of NOx.
Two, model is set up
Result based on burning optimization adjustment test, sets up based on the theoretical optimizing control models of artificial neural network (ANN), and research produces the key parameter of material impact to boiler efficiency and NOx discharge, and carries out power sequence; Research burning optimization Dynamic Control Strategy.
Obtain the theoretical optimizing control models of artificial neural network (ANN), obtain the influence factor sequence of key parameter, and obtain burning optimization Dynamic Control Strategy.
(a), on the basis of above-mentioned result of the test, set up single factors vary and boiler efficiency and NOx variation model, and various factors is affected to strong and weak rank;
(b) optimum around boiler efficiency and NOx discharge, set up complex condition, the Dynamic Control Strategy of burning in real time, and set up model;
(c) the Adaptability Analysis model of coal characteristic variation and above-mentioned model;
Three, model optimization
Based on test data and the theoretical model set up, through the calculation of data input and model, can effectively obtain NOx and form the optimization relation with boiler efficiency, as shown in Figure 3.
The present invention adopts BP network for the structure of burning optimization model, and all analog operation is based on MATLAB computing platform.
(a) structure choice of BP network and design
As shown in Figure 4, BP neutral net is a kind of Multilayer Feedforward Neural Networks, input layer, hidden layer (being called again intermediate layer) and output layer, consists of.Hidden layer can be that one deck can be also multilayer (if figure below is typical Multi-Layer BP Neural network).This project adopts most widely used three-layer neural network, comprises input layer, one deck hidden layer and output layer.
Studies have shown that three-layer network can approach any non-linear continuous function, its iterative equation is:
y l+1＝f l+1(W l+1y l+b l+1) l＝0，1，2，...，L-1；
In formula: l is current layer number, L is total number of plies, f l+1transfer function, W l+1neuron connects weights, b l+1neuron threshold value.The transfer function of BP network requires must be micro-, any differentiable function all be can be used as the transfer function of BP network.The conventional logarithm that has Sigmoid type, tan or linear function.Because transfer function is everywhere can be micro-, so for BP network, on the one hand, the region of dividing is no longer a linear partition, but the region being formed by a non-linear hyperplane, it is smoother curved surface, thereby its minute analogy linear partition is more accurate, and fault-tolerance is also better than linear partition; On the other hand, network can strictly adopt gradient descent method to learn, and the analytic expression of weights correction is very clear and definite.
The problem that input layer can solve as required and data coding method are determined.In boiler low NOx combustion Optimized model, the selection of input parameter is that the relation of moving variable and output performance target is carried out based on each, theoretical research and field trial show that an input parameter can affect a plurality of performance objectives, an output parameter also can be subject to the impact of a plurality of input parameters, and often shows as the impact of vying each other.For guaranteeing accuracy, terseness and the sensitivity of model, should select performance objective to have the operational factor of critical impact; Output parameter is to represent that the parameter of boiler performance index is as boiler efficiency, NOx discharge etc.
For BP network, there is a very important theorem.For any continuous function in closed interval, can approach with the BP network of single hidden layer, thereby three layers of BP network just can complete the mapping that n arbitrarily ties up m dimension.It is a very complicated problem that hidden nodes object is selected, often need according to designer's experience and repeatedly numerical experiment determine, thereby do not exist a desirable analytic expression to represent.The number of the number of hidden unit and the requirement of problem, I/O unit has direct relation.Hidden unit number can cause learning time long, error not necessarily best too much, also probably causes " over-fitting " phenomenon to occur.Therefore, necessarily there is a best hidden unit number.This project is determined best hidden unit number by following formula:
n 1＝log 2n；
N in formula 1for hidden unit number, n is input block number.
(b) BP algorithm
BP algorithm is comprised of two parts: the forward transmission of information and the backpropagation of error.In forward transmittance process, input message is successively calculated and is passed to output layer through hidden layer from input layer.According to the output of output layer and desired output, make comparisons, calculate the error change value of output layer, then by back-propagation algorithm, error signal is returned to hidden layer along original connecting path anti-pass, then revise each layer of weights until reach expectation target according to this error.Along with the correction of this error back propagation is constantly carried out, network also constantly rises to the accuracy of input pattern response.
Specifically, to q group input sample X 1, X 2..., X q, the known output sample corresponding with it is Y 1, Y 2..., Y q.In the process of e-learning, use the actual output A of network 1, A 2..., A qwith target Y 1, Y 2..., Y qbetween error revise the weights of network, make network output A 1, A 2..., A qwith desired output Y 1, Y 2..., Y qapproaching as much as possible, even if the error sum of squares of network output reaches minimum.This is the common thought of nearly all use neural network algorithm of having teacher's training algorithm.BP algorithm is at this, to have added inwardly the backpropagation thought of error.
Following formula is deferred in the renewal of network weight and threshold value:
F(x)＝E[(t-y) 2];
In formula, x is weights and threshold vector, and mean square error function F (x) refers to the square error between target output value and actual prediction value when k iteration, that is:
F(x)＝(t(k)-y(k)) t(t(k)-y(k))=e t(k)e(k);
The training algorithm that upgrades network weight and threshold value has a variety of, and wherein conventional have gradient descent method, conjugate gradient method, Quasi-Newton method, Levenberg-Marquardt method and a normalized L-M method of Bayes etc.These training algorithms are all by following iterative equation, to search for the x value of minimum error function F (x):
x k+1＝x k+α kd k；
Vectorial d in formula kthe direction of search, α kit is learning rate.
(c) BP algorithm implementation step
BP algorithm according to initialization, given training dataset, calculate actual output y and network error e, adjust weights, judge whether to finish the steps such as training and carry out, its flow process is as shown in Figure 5.
(d) combustion control BP network model realize determining of input layer: be accuracy, terseness and the sensitivity that guarantees model, select to performance objective (boiler thermal output and NOx discharge) to have the operational factor of critical impact as input variable.And have the operational factor of critical impact to determine according to combustion in situ test to performance objective, may comprise: (a) excess air coefficient; (b) primary air flow (wind/coal ratio); (c) OFA burns air quantity; (d) fineness of pulverized coal; (e) become coal pulverizer operation mode; (f) load; (g) coal (ature of coal) characteristic etc.
Determining of hidden layer neuron and pass-algorithm: the neuron number of hidden layer is very doubt, and this affects the estimated performance of network to a great extent.This project is determined best hidden unit number by following formula:
n 1＝log 2n；
In formula, n1 is hidden unit number, and n is input block number.
Input layer is S shape tan tansig to hidden layer transfer function, and hidden layer is linear function purelin to output layer transfer function; Network training function is is trainbr function based on the normalized L-M algorithm of Bayes; Network performance evaluation function is amended performance function formula:
F = γ F e + ( 1 - γ ) F w = γ * 1 N Σ i = 1 N ( t i - a i ) 2 + ( 1 - γ ) * 1 n Σ j = 1 n w j 2 ;
F in formula efor network error all side and, F wfor network weight and threshold value all side and, γ is PR.
Input layer is S shape tan tansig to hidden layer transfer function, and hidden layer is linear function purelin to output layer transfer function; Network training function is that self adaptation lr momentum gradient descent method is traindx function;
When adopting momentum method, BP algorithm can find more excellent solution, and when adopting adaptive learning speed, BP algorithm can shorten the training time, and traindx adopts these two kinds of methods to train Multilayer Feedforward Neural Networks.Adaptive learning speed is first given an initial value, then utilize multiplication to make it to increase or reduce, to keep learning rate fast and stable, the momentum being defined between 0-1 is specified the momentum size of using, error ratio has limited the error that may increase in single training, if error rises, surpassed error ratio, given up new weights, and temporarily do not use momentum.Thereby adopt self adaptation lr momentum gradient descent method training pattern comparatively quickly and accurately.
Output layer neuron is determined: the output vector of model is corresponding with the output performance index of research object, and the performance indications that this place is paid close attention to are NOx discharge capacity and boiler efficiency, therefore chooses the two for final output layer neuron.
By peroxide amount, boiler slag carbon content, unburned carbon in flue dust, exhaust gas temperature, exhaust smoke level, reheater desuperheating water flow, be input variable, the boiler efficiency of take is set up the forecast model of boiler efficiency as output variable.And then to enter the total fuel quantity of stove, total air, each feeder aperture, each secondary air damper aperture, after-flame windshield plate aperture, oxygen content at economizer outlet, coal characteristic, each coal pulverizer ventilation is as input variable, and NOx concentration of emission is set up the forecast model of NOx concentration of emission as output variable.Concrete training is as shown in table 1 below with Monitoring and Controlling parameter.
The main Detection & Controling parameter of table 1 system
Sequence number Parameter Unit
1 Enter the total coal amount of stove t/h
2 Enter stove total air Kg/s
3 A layer fuel air secondary air register aperture %
4 B layer fuel air secondary air register aperture %
5 BC layer is assisted wind secondary air register aperture %
6 C layer fuel air secondary air register aperture %
7 D layer fuel air secondary air register aperture %
8 DE layer is assisted wind secondary air register aperture %
9 E layer fuel air secondary air register aperture %
10 F layer fuel air secondary air register aperture %
11 SOFA 1 throttle opening %
12 SOFA 2 throttle openings %
13 SOFA 3 throttle openings %
14 SOFA 4 throttle openings %
15 SOFA 5 throttle openings %
16 CCOFA 1 throttle opening %
17 CCOFA 2 throttle openings %
18 The pressure reduction of secondary air box and burner hearth Mpa
19 Mill A air quantity Kg/s
20 Mill B air quantity Kg/s
21 Mill C air quantity Kg/s
22 Mill D air quantity Kg/s
23 Mill E air quantity Kg/s
24 mill F air quantity kg/s
25 give A coal amount kg/s
26 give B coal amount kg/s
27 give C coal amount kg/s
28 give D coal amount kg/s
29 give E coal amount kg/s
30 give F coal amount kg/s
31 as-received carbon Car %
32 as-received hydrogen Har %
33 as-received oxygen Oar %
34 as-received nitrogen Nar %
35 low heat valve Qnet kJ/Kg
36 moisture Mar %
37 volatile matter Var %
38 burner pivot angle .
39 air preheater outlet temperature ℃
40 unburned carbon in flue dust %
41 oxygen content in exhaust gas %
42 exhaust gas temperature ℃
43 cold wind temperature ℃
Obtaining of training sample: on-the-spot hot test data and from the historical data of DCS of Power Plant (DCS) raw sample data as training BP neutral net, and retain some samples as test sample book.The transfer function of BP neutral net between [1,1], for guarantee network not the undue dispersion of factor data cause not restraining, before sample data is trained, need standardization pretreatment.So not only network convergence can be promoted, and the impact of the variation of input variable on output variable can be correctly reflected.This project has taked method for normalizing to carry out pretreatment to sample data, and formula is as follows:
x ^ = 2 * x - x min x max - x min - 1 ;
The Establishment and optimization of model: utilize the initial model training of training sample set pair and learn, when computing reaches convergence, object module is set up, and now can utilize test sample book to predict it, according to predicated error, carry out parameter adjustment, model is finally optimized.
The present invention can be good at solving actual power plant and improves boiler efficiency and the problem of falling low NOx drainage, there is good using value, can be effectively real-time according to parameter, provide optimization to adjust scheme, instruct operations staff to adjust, thereby improve the economy of power plant's operation.
Corresponding with the implementation method of above-mentioned a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System, the present invention also provides a kind of Large-sized Coal-fired Power group on-line monitoring and Optimal Control System, as shown in Figure 6, comprising:
Parameter acquisition module 101, for the boiler of Large-sized Coal-fired Power group is carried out to combustion characteristic analysis, and carries out boiler combustion optimization adjustment test, obtains boiler operatiopn key parameter;
Relation is set up module 102, for determining the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and sets up optimizing control models according to described relation;
Optimize Numerical Simulation Module 103, for described optimizing control models is trained, obtain the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
As a good embodiment, described Large-sized Coal-fired Power group on-line monitoring and Optimal Control System can also comprise:
Adjusting module, for after obtaining described concrete optimization numerical value, adjusts boiler operating parameter according to described concrete optimization numerical value.
As a good embodiment, described optimizing control models can be the optimizing control models based on BP neural network theory.
As a good embodiment, described optimization Numerical Simulation Module can comprise:
Training submodule, trains described optimizing control models for adopting with the gradient function of momentum and dynamic self-adapting learning rate.
As a good embodiment, described boiler operatiopn key parameter can comprise: excess air coefficient, wind/coal ratio, burn air quantity, fineness of pulverized coal, change coal pulverizer operation mode, load and coal characteristic etc.
Above-mentioned a kind of Large-sized Coal-fired Power group on-line monitoring is identical with the implementation method of a kind of Large-sized Coal-fired Power group on-line monitoring of the present invention and Optimal Control System with other technical characterictic of Optimal Control System, and it will not go into details herein.
By above scheme, can find out, a kind of Large-sized Coal-fired Power group on-line monitoring of the present invention and Optimal Control System and its implementation, can set up the key parameter of boiler operatiopn and the relation between boiler efficiency and NOx discharge, and set up accordingly optimizing control models, then by the training of this optimizing control models being obtained to the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency, operations staff only need to carry out boiler combustion adjustment according to these concrete numerical value of optimizing, and just can on the basis that guarantees low NOx drainage, improve boiler efficiency.The present invention, no matter be all to guarantee that under open loop mode or closed loop mode boiler operates in minimum generating energy consumption and minimum pollution thing state automatically, effectively raises the economy of power plant's operation.
Unless context separately has the description of specific distinct, the element in the present invention and assembly, the form that quantity both can be single exists, and form that also can be a plurality of exists, and the present invention does not limit this.Although the step in the present invention is arranged with label, and be not used in the precedence that limits step, unless expressly stated the order of step or the execution of certain step need other steps as basis, otherwise the relative order of step is adjustable.
The above embodiment has only expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (10)

1. an implementation method for Large-sized Coal-fired Power group on-line monitoring and Optimal Control System, is characterized in that, comprises the following steps:
The boiler of Large-sized Coal-fired Power group is carried out to combustion characteristic analysis, and carry out boiler combustion optimization adjustment test, obtain boiler operatiopn key parameter;
Determine the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and set up optimizing control models according to described relation;
Described optimizing control models is trained, obtain the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
2. the implementation method of Large-sized Coal-fired Power group on-line monitoring according to claim 1 and Optimal Control System, is characterized in that, after obtaining described concrete optimization numerical value, also comprises step: according to described concrete optimization numerical value, adjust boiler operating parameter.
3. the implementation method of Large-sized Coal-fired Power group on-line monitoring according to claim 1 and Optimal Control System, is characterized in that, described optimizing control models is the optimizing control models based on BP neural network theory.
4. the implementation method of Large-sized Coal-fired Power group on-line monitoring according to claim 3 and Optimal Control System, is characterized in that, the process that described optimizing control models is trained comprises:
Employing is trained described optimizing control models with the gradient function of momentum and dynamic self-adapting learning rate.
5. according to the Large-sized Coal-fired Power group on-line monitoring described in claim 1-4 any one and the implementation method of Optimal Control System, it is characterized in that, described boiler operatiopn key parameter comprises: excess air coefficient, wind/coal ratio, burn air quantity, fineness of pulverized coal, change coal pulverizer operation mode, load and coal characteristic.
6. Large-sized Coal-fired Power group on-line monitoring and an Optimal Control System, is characterized in that, comprising:
Parameter acquisition module, for the boiler of Large-sized Coal-fired Power group is carried out to combustion characteristic analysis, and carries out boiler combustion optimization adjustment test, obtains boiler operatiopn key parameter;
Relation is set up module, for determining the relation between described boiler operatiopn key parameter and boiler efficiency and NOx discharge, and sets up optimizing control models according to described relation;
Optimize Numerical Simulation Module, for described optimizing control models is trained, obtain the concrete optimization numerical value of the boiler operating parameter under the highest boiler efficiency.
7. Large-sized Coal-fired Power group on-line monitoring according to claim 6 and Optimal Control System, is characterized in that, also comprises:
Adjusting module, for after obtaining described concrete optimization numerical value, adjusts boiler operating parameter according to described concrete optimization numerical value.
8. Large-sized Coal-fired Power group on-line monitoring according to claim 6 and Optimal Control System, is characterized in that, described optimizing control models is the optimizing control models based on BP neural network theory.
9. Large-sized Coal-fired Power group on-line monitoring according to claim 8 and Optimal Control System, is characterized in that, described optimization Numerical Simulation Module comprises:
Training submodule, trains described optimizing control models for adopting with the gradient function of momentum and dynamic self-adapting learning rate.
10. according to Large-sized Coal-fired Power group on-line monitoring and Optimal Control System described in claim 6-9 any one, it is characterized in that, described boiler operatiopn key parameter comprises: excess air coefficient, wind/coal ratio, burn air quantity, fineness of pulverized coal, change coal pulverizer operation mode, load and coal characteristic.
CN201410021273.4A 2014-01-16 2014-01-16 Large coal-fired unit online monitoring and optimal control system and implementation method thereof Pending CN103759290A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410021273.4A CN103759290A (en) 2014-01-16 2014-01-16 Large coal-fired unit online monitoring and optimal control system and implementation method thereof

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410021273.4A CN103759290A (en) 2014-01-16 2014-01-16 Large coal-fired unit online monitoring and optimal control system and implementation method thereof

Publications (1)

Publication Number Publication Date
CN103759290A true CN103759290A (en) 2014-04-30

Family

ID=50526572

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410021273.4A Pending CN103759290A (en) 2014-01-16 2014-01-16 Large coal-fired unit online monitoring and optimal control system and implementation method thereof

Country Status (1)

Country Link
CN (1) CN103759290A (en)

Cited By (43)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104021315A (en) * 2014-06-27 2014-09-03 四川电力设计咨询有限责任公司 Method for calculating station service power consumption rate of power station on basis of BP neutral network
CN104331736A (en) * 2014-11-13 2015-02-04 国家电网公司 RBF (Radial Basis Function) neural network based supercritical boiler nitric oxide discharging dynamic predication method
CN104484749A (en) * 2014-12-04 2015-04-01 广东电网有限责任公司电力科学研究院 Method and system used for predicting carbon emission of coal-fired power plant
CN104504238A (en) * 2014-11-21 2015-04-08 国家电网公司 Method for predicting nitrogen oxide emission concentration of coal-fired unit based on neural network
CN104776446A (en) * 2015-04-14 2015-07-15 东南大学 Combustion optimization control method for boiler
CN105020705A (en) * 2015-03-04 2015-11-04 内蒙古瑞特优化科技股份有限公司 Method and system for optimizing and controlling combustion performance of circulating fluidized bed boiler in real time
CN105240822A (en) * 2015-09-17 2016-01-13 东方电气集团东方锅炉股份有限公司 Control method and system for three gas duct baffles of boiler based on neural network
CN105276611A (en) * 2015-11-25 2016-01-27 广东电网有限责任公司电力科学研究院 Adjustment and optimization method and system for combustion of power plant boiler
CN106168381A (en) * 2016-07-07 2016-11-30 神华集团有限责任公司 The control system of boiler combustion, method and apparatus
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
CN106556028A (en) * 2015-09-25 2017-04-05 上海上电漕泾发电有限公司 A kind of coal unit boiler combustion runs comprehensive optimization system with denitration
CN106765021A (en) * 2016-12-15 2017-05-31 吉林省电力科学研究院有限公司 A kind of method of the after-flame factor and bellows differential pressure factor corrdinated adjustment boiler performance
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
WO2017133316A1 (en) * 2016-02-02 2017-08-10 华北电力科学研究院有限责任公司 Method, device, and automatic control system for determining air intake amount for opposed firing
CN108019773A (en) * 2017-11-02 2018-05-11 华润电力（贺州）有限公司 A kind of boiler air-supply volume optimizing regulation technology based on DCS system
CN108613247A (en) * 2018-04-02 2018-10-02 上海航天智慧能源技术有限公司 A kind of Load Distribution method of steam-water dual-purpose gas fired-boiler group
CN109063359A (en) * 2018-08-16 2018-12-21 燕山大学 A kind of dynamic modelling method of Process of Circulating Fluidized Bed Boiler
CN109165789A (en) * 2018-09-04 2019-01-08 广东电网有限责任公司 The modeling method and device of emission of NOx of boiler amount prediction model based on LSTM
CN109174423A (en) * 2018-09-18 2019-01-11 国电龙源节能技术有限公司 Pulverizer outlet temperature control system and method based on fineness of pulverized coal on-line tuning
CN109237510A (en) * 2018-10-09 2019-01-18 沃森能源技术（廊坊）有限公司 A kind of Boiler Combustion Optimization System based on CO on-line monitoring
CN109376501A (en) * 2018-12-21 2019-02-22 浙江工业大学 A kind of electricity generation boiler fire coal oxidation of coal factor online soft sensor method
CN109376945A (en) * 2018-11-13 2019-02-22 华能国际电力股份有限公司上海石洞口第电厂 A kind of coal mixing combustion optimization system based on more coals
CN109882883A (en) * 2019-03-01 2019-06-14 北京慧辰资道资讯股份有限公司 A kind of method and device based on artificial intelligence optimization's boiler fired coal efficiency
CN110673485A (en) * 2019-10-21 2020-01-10 京东城市(南京)科技有限公司 Model training method, device, electronic apparatus, and medium for combustion control
CN110705881A (en) * 2019-10-08 2020-01-17 武汉市政工程设计研究院有限责任公司 Boiler efficiency online calculation method and system based on artificial neural network
CN110888401A (en) * 2018-09-11 2020-03-17 北京京东金融科技控股有限公司 Combustion control optimization method and device for thermal generator set and readable storage medium
CN111222685A (en) * 2019-11-21 2020-06-02 江苏方天电力技术有限公司 Boiler wide-load NOx emission concentration prediction method based on model migration
CN111272969A (en) * 2020-01-19 2020-06-12 西安热工研究院有限公司 Method for predicting NOx generation concentration of 300MW pulverized coal boiler
CN111639742A (en) * 2020-05-22 2020-09-08 安徽科技学院 System and method for diagnosing state fault of desulfurization and denitrification circulating pump
CN111853848A (en) * 2020-06-29 2020-10-30 东北电力大学 Optimization method for fuel quantity distribution among different-layer combustors of coal-fired boiler
CN112085353A (en) * 2020-08-21 2020-12-15 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 Boiler low-load stable combustion capability evaluation method based on degradation degree analysis
CN112361376A (en) * 2020-11-12 2021-02-12 广东工业大学 Control method and device for boiler combustion process and storage medium
CN112381289A (en) * 2020-11-13 2021-02-19 重庆电力高等专科学校 Method for improving characteristics of nitrogen oxides based on coal blending combustion boiler
CN112510703A (en) * 2020-11-26 2021-03-16 贵州电网有限责任公司 Multi-energy access power grid optimized scheduling method considering coal consumption curve correction
CN112651568A (en) * 2020-12-31 2021-04-13 新奥数能科技有限公司 Boiler load dynamic adjustment method and device, control terminal and storage medium
CN113074459A (en) * 2021-03-11 2021-07-06 国家能源集团新能源技术研究院有限公司 Boiler control optimization method and system
CN113339787A (en) * 2021-06-11 2021-09-03 华北电力大学（保定） Fluidized bed boiler operation optimization method and system based on digital twinning
CN113701186A (en) * 2021-07-19 2021-11-26 许昌龙岗发电有限责任公司 Development and application of secondary air optimization control system based on artificial neural network model
CN113834093A (en) * 2021-11-01 2021-12-24 西安热工研究院有限公司 Boiler oxygen content wide load optimization control system
CN114294637A (en) * 2022-01-04 2022-04-08 华润电力技术研究院有限公司 Low-temperature economizer state monitoring system and method based on machine learning
CN114415751A (en) * 2021-11-25 2022-04-29 内蒙古大唐国际托克托发电有限责任公司 Main and reheat steam temperature optimization system and method for thermal power generating unit
CN114791102A (en) * 2022-04-21 2022-07-26 中国矿业大学 Combustion optimization control method based on dynamic operation data analysis
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

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1945602A (en) * 2006-07-07 2007-04-11 华中科技大学 Characteristic selecting method based on artificial nerve network
US20090240636A1 (en) * 2003-09-30 2009-09-24 Reimar Hofmann Method, computer program with program code means and computer program product for analyzing variables influencing a combustion process in a combustion chamber, using a trainable statistical model
JP2011132915A (en) * 2009-12-25 2011-07-07 Honda Motor Co Ltd Device for estimating physical quantity
CN103324862A (en) * 2013-07-11 2013-09-25 中国石油大学(华东) Coal-fired boiler optimization method based on improved neural network and genetic algorithm

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090240636A1 (en) * 2003-09-30 2009-09-24 Reimar Hofmann Method, computer program with program code means and computer program product for analyzing variables influencing a combustion process in a combustion chamber, using a trainable statistical model
CN1945602A (en) * 2006-07-07 2007-04-11 华中科技大学 Characteristic selecting method based on artificial nerve network
JP2011132915A (en) * 2009-12-25 2011-07-07 Honda Motor Co Ltd Device for estimating physical quantity
CN103324862A (en) * 2013-07-11 2013-09-25 中国石油大学(华东) Coal-fired boiler optimization method based on improved neural network and genetic algorithm

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
孔亮等: "电站锅炉燃烧优化控制技术综述", 《电力设备》 *
张毅等: "电站锅炉运行性能综合预测模型", 《动力工程》 *
李向菊: "燃煤系统NO_x排放及锅炉效率平衡建模", 《中国优秀硕士学位论文全文数据库》 *
王培红等: "人工智能技术在电站锅炉燃烧优化中的应用研究", 《中国电机工程学报》 *
许程: "电站锅炉效率与NO_x排放混合建模优化研究", 《中国优秀硕士学位论文全文数据库》 *
赵欢等: "电站锅炉热效率与NO_x排放响应特性建模方法", 《中国电机工程学报》 *

Cited By (58)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104021315A (en) * 2014-06-27 2014-09-03 四川电力设计咨询有限责任公司 Method for calculating station service power consumption rate of power station on basis of BP neutral network
CN104331736A (en) * 2014-11-13 2015-02-04 国家电网公司 RBF (Radial Basis Function) neural network based supercritical boiler nitric oxide discharging dynamic predication method
CN104331736B (en) * 2014-11-13 2018-06-22 国家电网公司 Ultra-supercritical boiler NOx emission dynamic prediction method based on RBF neural
CN104504238A (en) * 2014-11-21 2015-04-08 国家电网公司 Method for predicting nitrogen oxide emission concentration of coal-fired unit based on neural network
CN104484749A (en) * 2014-12-04 2015-04-01 广东电网有限责任公司电力科学研究院 Method and system used for predicting carbon emission of coal-fired power plant
CN105020705A (en) * 2015-03-04 2015-11-04 内蒙古瑞特优化科技股份有限公司 Method and system for optimizing and controlling combustion performance of circulating fluidized bed boiler in real time
CN105020705B (en) * 2015-03-04 2017-06-09 内蒙古瑞特优化科技股份有限公司 Burning in circulating fluid bed boiler performance method for real-time optimization control and system
CN104776446A (en) * 2015-04-14 2015-07-15 东南大学 Combustion optimization control method for boiler
CN104776446B (en) * 2015-04-14 2017-05-10 东南大学 Combustion optimization control method for boiler
CN105240822A (en) * 2015-09-17 2016-01-13 东方电气集团东方锅炉股份有限公司 Control method and system for three gas duct baffles of boiler based on neural network
CN106556028A (en) * 2015-09-25 2017-04-05 上海上电漕泾发电有限公司 A kind of coal unit boiler combustion runs comprehensive optimization system with denitration
CN105276611A (en) * 2015-11-25 2016-01-27 广东电网有限责任公司电力科学研究院 Adjustment and optimization method and system for combustion of power plant boiler
CN105276611B (en) * 2015-11-25 2017-09-01 广东电网有限责任公司电力科学研究院 Power plant boiler firing optimization optimization method and system
WO2017133316A1 (en) * 2016-02-02 2017-08-10 华北电力科学研究院有限责任公司 Method, device, and automatic control system for determining air intake amount for opposed firing
CN106168381A (en) * 2016-07-07 2016-11-30 神华集团有限责任公司 The control system of boiler combustion, method and apparatus
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
CN106765021A (en) * 2016-12-15 2017-05-31 吉林省电力科学研究院有限公司 A kind of method of the after-flame factor and bellows differential pressure factor corrdinated adjustment boiler performance
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
CN108019773B (en) * 2017-11-02 2019-11-29 华润电力（贺州）有限公司 A kind of boiler air-supply volume optimizing regulation technology based on DCS system
CN108019773A (en) * 2017-11-02 2018-05-11 华润电力（贺州）有限公司 A kind of boiler air-supply volume optimizing regulation technology based on DCS system
CN108613247A (en) * 2018-04-02 2018-10-02 上海航天智慧能源技术有限公司 A kind of Load Distribution method of steam-water dual-purpose gas fired-boiler group
CN108613247B (en) * 2018-04-02 2022-09-20 上海航天智慧能源技术有限公司 Heat load distribution method of steam-water dual-purpose gas boiler group
CN109063359A (en) * 2018-08-16 2018-12-21 燕山大学 A kind of dynamic modelling method of Process of Circulating Fluidized Bed Boiler
CN109165789A (en) * 2018-09-04 2019-01-08 广东电网有限责任公司 The modeling method and device of emission of NOx of boiler amount prediction model based on LSTM
CN110888401A (en) * 2018-09-11 2020-03-17 北京京东金融科技控股有限公司 Combustion control optimization method and device for thermal generator set and readable storage medium
CN109174423A (en) * 2018-09-18 2019-01-11 国电龙源节能技术有限公司 Pulverizer outlet temperature control system and method based on fineness of pulverized coal on-line tuning
CN109174423B (en) * 2018-09-18 2023-08-11 国能龙源蓝天节能技术有限公司 Coal mill outlet temperature control system and method based on-line adjustment of pulverized coal fineness
CN109237510A (en) * 2018-10-09 2019-01-18 沃森能源技术（廊坊）有限公司 A kind of Boiler Combustion Optimization System based on CO on-line monitoring
CN109376945A (en) * 2018-11-13 2019-02-22 华能国际电力股份有限公司上海石洞口第电厂 A kind of coal mixing combustion optimization system based on more coals
CN109376501A (en) * 2018-12-21 2019-02-22 浙江工业大学 A kind of electricity generation boiler fire coal oxidation of coal factor online soft sensor method
CN109376501B (en) * 2018-12-21 2023-01-31 浙江工业大学 Online soft measurement method for coal-fired carbon oxidation factor of power generation boiler
CN109882883A (en) * 2019-03-01 2019-06-14 北京慧辰资道资讯股份有限公司 A kind of method and device based on artificial intelligence optimization's boiler fired coal efficiency
CN110705881A (en) * 2019-10-08 2020-01-17 武汉市政工程设计研究院有限责任公司 Boiler efficiency online calculation method and system based on artificial neural network
CN110673485A (en) * 2019-10-21 2020-01-10 京东城市(南京)科技有限公司 Model training method, device, electronic apparatus, and medium for combustion control
CN110673485B (en) * 2019-10-21 2020-11-24 京东城市(南京)科技有限公司 Model training method, device, electronic apparatus, and medium for combustion control
CN111222685B (en) * 2019-11-21 2022-07-08 江苏方天电力技术有限公司 Boiler wide-load NOx emission concentration prediction method based on model migration
CN111222685A (en) * 2019-11-21 2020-06-02 江苏方天电力技术有限公司 Boiler wide-load NOx emission concentration prediction method based on model migration
CN111272969A (en) * 2020-01-19 2020-06-12 西安热工研究院有限公司 Method for predicting NOx generation concentration of 300MW pulverized coal boiler
CN111272969B (en) * 2020-01-19 2022-02-22 西安热工研究院有限公司 Method for predicting NOx generation concentration of 300MW pulverized coal boiler
CN111639742A (en) * 2020-05-22 2020-09-08 安徽科技学院 System and method for diagnosing state fault of desulfurization and denitrification circulating pump
CN111853848A (en) * 2020-06-29 2020-10-30 东北电力大学 Optimization method for fuel quantity distribution among different-layer combustors of coal-fired boiler
CN111853848B (en) * 2020-06-29 2022-09-30 东北电力大学 Optimization method for fuel quantity distribution among different-layer combustors of coal-fired boiler
CN112085353A (en) * 2020-08-21 2020-12-15 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 Boiler low-load stable combustion capability evaluation method based on degradation degree analysis
CN112361376B (en) * 2020-11-12 2023-02-10 广东工业大学 Control method and device for boiler combustion process and storage medium
CN112361376A (en) * 2020-11-12 2021-02-12 广东工业大学 Control method and device for boiler combustion process and storage medium
CN112381289A (en) * 2020-11-13 2021-02-19 重庆电力高等专科学校 Method for improving characteristics of nitrogen oxides based on coal blending combustion boiler
CN112510703A (en) * 2020-11-26 2021-03-16 贵州电网有限责任公司 Multi-energy access power grid optimized scheduling method considering coal consumption curve correction
CN112651568A (en) * 2020-12-31 2021-04-13 新奥数能科技有限公司 Boiler load dynamic adjustment method and device, control terminal and storage medium
CN113074459A (en) * 2021-03-11 2021-07-06 国家能源集团新能源技术研究院有限公司 Boiler control optimization method and system
CN113339787A (en) * 2021-06-11 2021-09-03 华北电力大学（保定） Fluidized bed boiler operation optimization method and system based on digital twinning
CN113339787B (en) * 2021-06-11 2022-03-22 华北电力大学（保定） Fluidized bed boiler operation optimization method and system based on digital twinning
CN113701186A (en) * 2021-07-19 2021-11-26 许昌龙岗发电有限责任公司 Development and application of secondary air optimization control system based on artificial neural network model
CN113834093A (en) * 2021-11-01 2021-12-24 西安热工研究院有限公司 Boiler oxygen content wide load optimization control system
CN114415751A (en) * 2021-11-25 2022-04-29 内蒙古大唐国际托克托发电有限责任公司 Main and reheat steam temperature optimization system and method for thermal power generating unit
CN114294637A (en) * 2022-01-04 2022-04-08 华润电力技术研究院有限公司 Low-temperature economizer state monitoring system and method based on machine learning
CN114791102A (en) * 2022-04-21 2022-07-26 中国矿业大学 Combustion optimization control method based on dynamic operation data analysis
CN114791102B (en) * 2022-04-21 2023-09-22 中国矿业大学 Combustion optimization control method based on dynamic operation data analysis
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

Similar Documents

Publication Publication Date Title
CN103759290A (en) Large coal-fired unit online monitoring and optimal control system and implementation method thereof
Shi et al. Combustion optimization of ultra supercritical boiler based on artificial intelligence
CN107726358B (en) Boiler Combustion Optimization System based on CFD numerical simulations and intelligent modeling and method
CN102425807B (en) Combustion feedforward and feedback composite optimization controlling method for pulverized coal fired boiler
CN106681381A (en) SCR denitration system ammonia spraying quantity optimal control system and method based on intelligent feedforward signals
CN105629738A (en) SCR (Selective Catalytic Reduction) flue gas denitration system control method and apparatus
CN110263395A (en) The power plant's denitration running optimizatin method and system analyzed based on numerical simulation and data
CN105276611A (en) Adjustment and optimization method and system for combustion of power plant boiler
AU2019305721B2 (en) Boiler coal saving control method
CN103064289A (en) Multiple-target operation optimizing and coordinating control method and device of garbage power generator
CN103759277A (en) Intelligent soot blowing closed-loop control method, device and system for coal-fired power station boiler
CN101441442A (en) Control device for plant, control device for thermal power plant, and gas concentration estimation device of coal-burning boiler
CN104571022B (en) Power consumption analysis model experimental systems and method based on coal consumption Yu controllable factor relationship
CN104613468A (en) Circulating fluidized bedboiler combustion optimizing control method based on fuzzy adaptive inference
CN111306572B (en) Intelligent combustion optimizing energy-saving control system for boiler
CN112628712A (en) Secondary air closed-loop optimization control system based on air door resistance coefficient
CN110084717A (en) A kind of Utility Boiler moisture content of coal calculation method based on BP neural network
CN106527176A (en) MFOA (modified fruit fly optimization algorithm)-SVM (support vector machine)-based boiler thermal efficiency and NOX modeling method
CN110263452A (en) Flue gas Annual distribution characteristic analysis method, system and denitrating system in a kind of flue
CN105240868A (en) Boiler automatic combustion adjustment control method based on coal-air ratio coal quality correction
CN112016754A (en) Power station boiler exhaust gas temperature advanced prediction system and method based on neural network
CN103870878A (en) Power station boiler combustion characteristic neural network model
CN105808945B (en) A kind of hybrid intelligent boiler efficiency burning optimization method
CN104729888A (en) Flue fly ash constant-speed sampling device as well as flue fly ash constant-speed sampling system based on BP neural network and control method
CN106610589A (en) Online hardware closed-loop network source coordination linear active-disturbance-rejection control method

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20140430

RJ01 Rejection of invention patent application after publication