CN112365065A - WFGD self-adaptive online optimization scheduling method - Google Patents
WFGD self-adaptive online optimization scheduling method Download PDFInfo
- Publication number
- CN112365065A CN112365065A CN202011279200.7A CN202011279200A CN112365065A CN 112365065 A CN112365065 A CN 112365065A CN 202011279200 A CN202011279200 A CN 202011279200A CN 112365065 A CN112365065 A CN 112365065A
- Authority
- CN
- China
- Prior art keywords
- data
- circulating pump
- slurry circulating
- model
- wfgd
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000005457 optimization Methods 0.000 title claims abstract description 18
- 239000002002 slurry Substances 0.000 claims abstract description 59
- 238000006477 desulfuration reaction Methods 0.000 claims abstract description 29
- 230000023556 desulfurization Effects 0.000 claims abstract description 29
- 238000013528 artificial neural network Methods 0.000 claims abstract description 23
- 238000012216 screening Methods 0.000 claims abstract description 7
- 241000282461 Canis lupus Species 0.000 claims description 21
- 238000012549 training Methods 0.000 claims description 21
- 238000004422 calculation algorithm Methods 0.000 claims description 12
- 230000008859 change Effects 0.000 claims description 12
- 238000003062 neural network model Methods 0.000 claims description 10
- 230000003044 adaptive effect Effects 0.000 claims description 9
- 238000010276 construction Methods 0.000 claims description 6
- 239000007788 liquid Substances 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000007423 decrease Effects 0.000 claims description 2
- 238000005070 sampling Methods 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 6
- RAHZWNYVWXNFOC-UHFFFAOYSA-N sulfur dioxide Inorganic materials O=S=O RAHZWNYVWXNFOC-UHFFFAOYSA-N 0.000 description 17
- 238000004458 analytical method Methods 0.000 description 3
- 230000007246 mechanism Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000003245 coal Substances 0.000 description 2
- 238000002485 combustion reaction Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000003344 environmental pollutant Substances 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 231100000719 pollutant Toxicity 0.000 description 2
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000003546 flue gas Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012821 model calculation Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/067—Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
The invention relates to the technical field of coal-fired power plant desulfurization, and particularly discloses a WFGD (WFGD) self-adaptive online optimization scheduling method, which comprises the following steps: step one, collecting historical data of measuring points from a unit, wherein the data information comprises unit load, concentration of SO2 at an inlet/outlet of a desulfurization system and current parameters of a slurry circulating pump; secondly, acquiring each time node when the slurry circulating pump is started/stopped according to the current parameter of the slurry circulating pump; thirdly, performing steady-state screening on the obtained data samples; fourthly, establishing an independent WFGD online optimization scheduling model based on an GWO-BP neural network to form a prediction model group; fifthly, guiding the power station to carry out optimized dispatching on the slurry circulating pump; and sixthly, carrying out the purpose of model self-adaptive updating. According to the method, based on the historical operation data of the unit, data samples before and after the slurry circulating pump is started and stopped are extracted, and the advantage of big data is brought into play.
Description
Technical Field
The disclosure relates to the technical field of coal-fired power plant desulfurization, in particular to a WFGD self-adaptive online optimization scheduling method.
Background
WFGD is coal-fired unit desulfurization system, and SO2Is one of the most main pollutants of the thermal power plant, and 90 percent of SO is in the atmosphere2From coal combustion, whose excessive emissions cause a great pollution to the environment and therefore to SO2Also, government interest has increased. In order to reduce pollutant emission and purify air, many power generation enterprises have carried out ultralow emission modification to meet the environmental protection requirement. At present, a desulfurization system of a coal-fired unit is lack of operation and operation specifications under ultra-low emission, SO in the actual operation process, operators generally use SO to ensure standard emission2The emission concentration is controlled in a safe area which is lower than the emission limit value to a certain degree, so that the desulfurization energy consumption is improved undoubtedly, the desulfurization operation economy is reduced, and the desulfurization cost of a power plant is greatly increased.
The optimization operation technology of the desulfurization system of the coal-fired power plant has attracted the attention of researchers. At present, the optimized operation of the desulfurization system mainly comprises mechanism analysis methods based on process mechanism analysis, process object mathematical models and the like, and data driving methods of a neural network and a support vector machine. However, as an important component of a thermal power generating unit, the actual operation state of the flue gas desulfurization system is affected by various external conditions, including unit load, coal quality, boiler combustion mode, and the like, and is difficult to be expressed by a general mechanism model. In addition, due to the continuous change of external conditions, the desulfurization system is difficult to be in a long-term stable operation state, the actual operation data fluctuation of the system is large, and a reliable optimization scheme is difficult to be provided based on a stable target working condition.
Disclosure of Invention
The invention aims to solve the problems that the existing desulfurization system is unstable in operation and large in actual operation data fluctuation.
In order to achieve the above object, the basic scheme of the present invention provides a WFGD adaptive online optimization scheduling method, which includes the following steps:
step one, collecting historical data of measuring points from a unit, wherein the data information comprises unit load, concentration of SO2 at an inlet/outlet of a desulfurization system and current parameters of a slurry circulating pump;
secondly, acquiring time nodes when the slurry circulating pump is started/stopped according to current parameters of the slurry circulating pump, and acquiring historical operation data of the parameters 1 hour before the slurry circulating pump is started/stopped and 1 hour after the slurry circulating pump is started/stopped;
thirdly, performing steady-state screening on the obtained data samples to obtain a slurry circulating pump start/stop data sample set, and dividing the data samples into corresponding sample libraries according to the running condition of the slurry circulating pump;
the fourth step, average load of the unit before and after starting/stopping and average inlet SO2Concentration, average outlet SO before start/stop2The concentration of the model is used as an input parameter of the model, and the concentration variation of an outlet SO2 after starting/stopping is used as an output parameter of the model, and an independent WFGD online optimization scheduling model based on a GWO-BP neural network is respectively established to form a prediction model group;
fifthly, dividing real-time operation data of the unit by the number of operation units of the slurry circulating pump, entering the corresponding model, and calculating an outlet SO after starting/stopping the pump2The change of the concentration of the (A) can obtain the outlet SO after the pump is started/stopped2The concentration of the slurry circulating pump is controlled, and the power station is guided to carry out optimized dispatching on the slurry circulating pump;
and sixthly, acquiring the starting/stopping data of the slurry circulating pump from the real-time data, dividing the data according to the running condition of the slurry circulating pump, and storing the data into corresponding model sample libraries to achieve the purpose of self-adaptive updating of the model.
Further, the first step is specifically: of slave unitsAn SIS data source collects historical data of time span required by each measuring point, and the historical data comprises unit load and SO at the inlet/outlet of a desulfurization system at intervals of 1min2Concentration, slurry circulation pump current, etc. The data to be collected is arranged in time sequence, wherein the time span of the data and the collection interval can be customized.
Further, in the second step, the time when the current of only one slurry circulating pump changes or decreases from 0 to 0 is recorded, and the current changes only once within 1 hour before and after the time, and historical operation data of relevant parameters within one hour before and after the time is recorded.
Further, the conditions of the steady-state screening in the third step are as follows: the average load change of the unit before and after the slurry circulating pump is started and stopped is 13-15MW, and the average inlet SO2The concentration change is 30-50mg/m3。
Further, the construction of the GWO-BP neural network model in the fourth step specifically comprises the following steps:
determining input and output parameters of a BP neural network, and carrying out normalization processing on input and output data;
step two, GWO algorithm population initialization, initialization of the wolf population position X at randomj(j ═ 1, 2, …, n), each individual wolf containing a set of weights αiIn which α is1+α2+…+αiI is the number of corresponding input parameters, and parameters a, A and C are initialized;
thirdly, constructing a fitness function, and multiplying each normalized input parameter by the corresponding weight alphaiAs new input parameters, training BP neural network, and predicting output y by using trained modelkAnd expected outputR of (A) to (B)2Constructing a fitness function;
step four, executing the BP neural network training in the step three to each wolf individual in the population, calculating the fitness value of each wolf individual through a fitness function, and selecting 3 wolfs with the highest fitness value as the current optimal solution XαSub-optimal solution XβAnd the third best solution Xδ;
And step five, updating the positions of other omega wolf individuals and updating the parameters a, A and C. Reconstructing a new BP neural network according to the step three, training the network, recalculating the fitness value of each wolf individual, and updating Xα、XβAnd Xδ;
Step six, judging GWO whether the algorithm reaches the maximum iteration first three, if so, stopping iteration and outputting the optimal result Xα(ii) a Otherwise, repeating the third step to the fifth step until the maximum iteration times is reached;
step seven, obtaining the weight alpha of the optimal input parameter by utilizing GWO algorithmiAnd training the BP neural network again to obtain a final BP neural network model with the highest training precision.
Further, the sixth step specifically includes: and (4) acquiring the start-stop data of the liquid circulating pump from the real-time data in the second step and the real-time data in the third step, dividing the data according to the running condition of the liquid circulating pump, storing the data into a corresponding model sample library, and training and updating the GWO-BP model again according to the construction of the GWO-BP neural network model in the fourth step.
The principle and the effect of the invention are as follows:
1. compared with the prior art, on the premise of ensuring that the discharge concentration at the outlet of the desulfurization system reaches the standard, the operation condition of the desulfurization system is divided by taking the operation condition of the slurry circulating pump as a condition, and an independent WFGD online optimization scheduling model based on an GWO-BP neural network is respectively established to form a model group, so that the problems of unstable operation and large actual operation data fluctuation of the conventional desulfurization system are solved.
2. The influence of the start/stop of a single slurry circulating pump on the concentration change of the SO2 at the outlet of the desulfurization system is predicted through the established model group, SO that the online scheduling of the slurry circulating pump is realized, the aim of reducing the operation cost of the desulfurization system is fulfilled, and meanwhile, the start/stop data of the slurry circulating pump is continuously acquired from the real-time data of the unit, and the online self-adaptive updating of the model is realized.
3. According to the method, based on the historical operation data of the unit, data samples before and after the slurry circulating pump is started and stopped are extracted, and the advantage of big data is brought into play.
4. An GWO-BP modeling method is provided, an artificial neural network is combined with an optimization algorithm, and meanwhile, a multi-model fusion modeling method is adopted for the problem of multiple target working conditions, so that the prediction accuracy of a model is improved.
5. And the real-time operation data of the unit is used, so that the model is updated in a self-adaptive manner, and the method is more suitable for practical engineering application.
6. The method can better predict the concentration variation of SO2 at the outlet of the desulfurization system after the slurry circulating pump is started and stopped, and provides a guide basis for the online optimization scheduling of the slurry circulating pump. And meanwhile, a slurry circulating pump in real-time data can be used for starting and stopping a sample, so that the model can be updated in a self-adaptive manner. The method meets the requirements of engineering and fully exerts the advantages of big data of the unit. The aim is to optimize the operation cost of the desulfurization system and improve the economic benefit of a power plant.
Drawings
FIG. 1 is a flowchart of a WFGD adaptive online optimization scheduling method according to an embodiment of the present invention.
Detailed Description
The following is further detailed by the specific embodiments:
example (b):
a WFGD adaptive online optimization scheduling method includes the following steps:
data were collected after long blow or overhaul in the SIS system at 15min intervals (as the case may be). The framework of the invention mainly comprises core modules of data acquisition, data preprocessing, data processing, model construction, monitoring analysis, pollution positioning and the like, and a detailed flow chart is shown in figure 1:
firstly, historical data of measuring points are collected. Collecting each from a unit's SIS data sourceMeasuring historical data of a time span required by points, and taking SO at an inlet of a desulfurization system including a load at intervals of 1min2Concentration, etc. The data to be collected is arranged in time sequence, wherein the time span of the data and the collection interval can be customized.
And secondly, acquiring each time node when the slurry circulating pump is started/stopped according to the current parameter of the slurry circulating pump, recording the time when the current of only one slurry circulating pump starts to change from 0 or is reduced to 0, and only once change within 1 hour before and after the time, and recording historical operating data of relevant parameters within one hour before and after the time.
Thirdly, performing steady-state screening on the obtained data sample, wherein the conditions of the steady-state screening are as follows: average load change of the unit before and after starting/stopping of the slurry circulating pump is not more than 15MW, average inlet SO2The concentration variation is not more than 50mg/m3And obtaining a slurry circulating pump start/stop data sample set. And dividing the data samples into corresponding sample libraries according to the operation condition of the slurry circulating pump.
Fourthly, the average load and the average inlet SO of the unit before and after the slurry circulating pump is started/stopped2Average outlet SO before starting/stopping of concentration and slurry circulating pump2The concentration of the slurry circulating pump is used as an input parameter of the model, and the outlet SO is formed after the slurry circulating pump is started/stopped2The concentration variation of the model is used as an output parameter of the model, and normalization processing is carried out on input data and output data. Multiplying each input parameter after normalization by corresponding weight alphaiAs new input parameters, training BP neural network, and predicting output y by using trained modelkAnd expected outputR of (A) to (B)2And constructing a fitness function. Utilizing GWO algorithm to find weight alpha which enables BP algorithm training accuracy to be highestiAnd retraining the BP model to obtain a final BP neural network model with the highest training precision. Respectively establishing independent WFGD online optimization scheduling models based on GWO-BP neural networks by using the method to form a prediction model group;
the GWO-BP neural network model is constructed by the following specific steps:
determining input and output parameters of a BP neural network, and carrying out normalization processing on input and output data;
step two, GWO algorithm population initialization, initialization of the wolf population position X at randomj(j ═ 1, 2, …, n), each individual wolf containing a set of weights αiIn which α is1+α2+…+αiI is the number of corresponding input parameters, and parameters a, A and C are initialized;
thirdly, constructing a fitness function, and multiplying each normalized input parameter by the corresponding weight alphaiAs new input parameters, training BP neural network, and predicting output y by using trained modelkAnd expected outputR of (A) to (B)2Constructing a fitness function;
step four, executing the BP neural network training in the step three to each wolf individual in the population, calculating the fitness value of each wolf individual through a fitness function, and selecting 3 wolfs with the highest fitness value as the current optimal solution XαSub-optimal solution XβAnd the third best solution Xδ;
And step five, updating the positions of other omega wolf individuals and updating the parameters a, A and C. Reconstructing a new BP neural network according to the step three, training the network, recalculating the fitness value of each wolf individual, and updating Xα、XβAnd Xδ;
Step six, judging GWO whether the algorithm reaches the maximum iteration first three, if so, stopping iteration and outputting the optimal result Xα(ii) a Otherwise, repeating the third step to the fifth step until the maximum iteration times is reached;
step seven, obtaining the weight alpha of the optimal input parameter by utilizing GWO algorithmiAnd training the BP neural network again to obtain a final BP neural network model with the highest training precision.
Fifthly, dividing real-time operation data of the unit by the number of operation units of the slurry circulating pump, entering the corresponding model, and calculating an outlet SO after starting/stopping the pump2The change of the concentration of the (A) can obtain the outlet SO after the pump is started/stopped2The concentration of (c). If the slurry circulating pump is started/stopped, the SO at the outlet of the desulfurization system is obtained by model calculation2The concentration of the sulfur dioxide can meet the national emission standard, and the power station desulfurization system can be guided to reduce the pump to operate.
And sixthly, acquiring the start-stop data of the liquid circulating pump from the real-time data in the second step and the real-time data in the third step, dividing the data according to the running condition of the slurry circulating pump, storing the data into a corresponding model sample library, and training and updating the GWO-BP model again according to the construction of the GWO-BP neural network model in the fourth step.
The method can better predict the concentration variation of SO2 at the outlet of the desulfurization system after the slurry circulating pump is started and stopped, and provides a guide basis for the online optimization scheduling of the slurry circulating pump. And meanwhile, a slurry circulating pump in real-time data can be used for starting and stopping a sample, so that the model can be updated in a self-adaptive manner. The method meets the requirements of engineering and fully exerts the advantages of big data of the unit. The aim is to optimize the operation cost of the desulfurization system and improve the economic benefit of a power plant.
The foregoing is merely an example of the present invention and common general knowledge of known specific structures and features of the embodiments is not described herein in any greater detail. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.
Claims (6)
1. A WFGD adaptive online optimization scheduling method is characterized by comprising the following steps:
step one, collecting historical data of measuring points from a unit, wherein the data information comprises unit load, concentration of SO2 at an inlet/outlet of a desulfurization system and current parameters of a slurry circulating pump;
secondly, acquiring time nodes when the slurry circulating pump is started/stopped according to current parameters of the slurry circulating pump, and acquiring historical operation data of the parameters 1 hour before the slurry circulating pump is started/stopped and 1 hour after the slurry circulating pump is started/stopped;
thirdly, performing steady-state screening on the obtained data samples to obtain a slurry circulating pump start/stop data sample set, and dividing the data samples into corresponding sample libraries according to the running condition of the slurry circulating pump;
the fourth step, average load of the unit before and after starting/stopping and average inlet SO2Concentration, average outlet SO before start/stop2The concentration of the model is used as an input parameter of the model, and the concentration variation of an outlet SO2 after starting/stopping is used as an output parameter of the model, and an independent WFGD online optimization scheduling model based on a GWO-BP neural network is respectively established to form a prediction model group;
fifthly, dividing real-time operation data of the unit by the number of operation units of the slurry circulating pump, entering the corresponding model, and calculating an outlet SO after starting/stopping the pump2The change of the concentration of the (A) can obtain the outlet SO after the pump is started/stopped2The concentration of the slurry circulating pump is controlled, and the power station is guided to carry out optimized dispatching on the slurry circulating pump;
and sixthly, acquiring the starting/stopping data of the slurry circulating pump from the real-time data, dividing the data according to the running condition of the slurry circulating pump, and storing the data into corresponding model sample libraries to achieve the purpose of self-adaptive updating of the model.
2. The adaptive online optimal WFGD scheduling method of claim 1, wherein: the first step is specifically: historical data of time span required by each measuring point is collected from an SIS data source of the unit, and the time span including unit load and SO at the inlet/outlet of a desulfurization system is taken at intervals of 1min2Concentration, slurry circulation pump current, etc. Data to be collected is time-sequencedAn ordering in which the time span and the sampling interval of the data are customizable.
3. The adaptive online optimized scheduling method for WFGD of claim 1 or 2, wherein: in the second step, the time when the current of only one slurry circulating pump changes or decreases from 0 to 0 is recorded, and the current changes only once within 1 hour before and after the time, and historical operation data of relevant parameters within one hour before and after the time is recorded.
4. The adaptive online optimized scheduling method for WFGD of claim 3, wherein: the conditions of the steady-state screening in the third step are as follows: the average load change of the unit before and after the slurry circulating pump is started and stopped is 13-15MW, and the average inlet SO2The concentration change is 30-50mg/m3。
5. The adaptive online optimized scheduling method for WFGD of claim 4, wherein: the construction of the GWO-BP neural network model in the fourth step comprises the following specific steps:
determining input and output parameters of a BP neural network, and carrying out normalization processing on input and output data;
step two, GWO algorithm population initialization, initialization of the wolf population position X at randomj(j ═ 1, 2, …, n), each individual wolf containing a set of weights αiIn which α is1+α2+…+αiI is the number of corresponding input parameters, and parameters a, A and C are initialized;
thirdly, constructing a fitness function, and multiplying each normalized input parameter by the corresponding weight alphaiAs new input parameters, training BP neural network, and predicting output y by using trained modelkAnd expected outputR of (A) to (B)2Constructing a fitness function;
step four, executing the BP neural network training in the step three to each wolf individual in the population, calculating the fitness value of each wolf individual through a fitness function, and selecting 3 wolfs with the highest fitness value as the current optimal solution XαSub-optimal solution XβAnd the third best solution Xδ;
And step five, updating the positions of other omega wolf individuals and updating the parameters a, A and C. Reconstructing a new BP neural network according to the step three, training the network, recalculating the fitness value of each wolf individual, and updating Xα、XβAnd Xδ;
Step six, judging GWO whether the algorithm reaches the maximum iteration first three, if so, stopping iteration and outputting the optimal result Xα(ii) a Otherwise, repeating the third step to the fifth step until the maximum iteration times is reached;
step seven, obtaining the weight alpha of the optimal input parameter by utilizing GWO algorithmiAnd training the BP neural network again to obtain a final BP neural network model with the highest training precision.
6. The adaptive online optimized scheduling method for WFGD of claim 5, wherein: the sixth step is specifically as follows: and (4) acquiring the start-stop data of the liquid circulating pump from the real-time data in the second step and the real-time data in the third step, dividing the data according to the running condition of the liquid circulating pump, storing the data into a corresponding model sample library, and training and updating the GWO-BP model again according to the construction of the GWO-BP neural network model in the fourth step.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011279200.7A CN112365065A (en) | 2020-11-16 | 2020-11-16 | WFGD self-adaptive online optimization scheduling method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202011279200.7A CN112365065A (en) | 2020-11-16 | 2020-11-16 | WFGD self-adaptive online optimization scheduling method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN112365065A true CN112365065A (en) | 2021-02-12 |
Family
ID=74515751
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202011279200.7A Pending CN112365065A (en) | 2020-11-16 | 2020-11-16 | WFGD self-adaptive online optimization scheduling method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN112365065A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112906967A (en) * | 2021-02-24 | 2021-06-04 | 大唐环境产业集团股份有限公司 | Desulfurization system slurry circulating pump performance prediction method and device |
CN114749006A (en) * | 2022-03-28 | 2022-07-15 | 浙江浙能兰溪发电有限责任公司 | Method for optimizing wet desulphurization slurry circulating pump |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20010021883A1 (en) * | 1998-03-24 | 2001-09-13 | Lang Fred D. | L factor method for determining heat rate and emission rates of a fossil-fired system |
JP2003126650A (en) * | 2001-10-25 | 2003-05-07 | Kawasaki Heavy Ind Ltd | Control method and apparatus for wet flue gas desulfurization plant |
CN109948869A (en) * | 2019-04-22 | 2019-06-28 | 东南大学 | Desulphurization system SO based on orderly cluster discretization2Exit concentration prediction technique |
CN109978048A (en) * | 2019-03-22 | 2019-07-05 | 大唐环境产业集团股份有限公司 | A kind of Desulfurization tower slurry circulating pump malfunction analysis and problem shpoting method |
CN111723945A (en) * | 2020-06-03 | 2020-09-29 | 杭州欣禾圣世科技有限公司 | BP neural network optimization method based on improved wolf algorithm |
-
2020
- 2020-11-16 CN CN202011279200.7A patent/CN112365065A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20010021883A1 (en) * | 1998-03-24 | 2001-09-13 | Lang Fred D. | L factor method for determining heat rate and emission rates of a fossil-fired system |
JP2003126650A (en) * | 2001-10-25 | 2003-05-07 | Kawasaki Heavy Ind Ltd | Control method and apparatus for wet flue gas desulfurization plant |
CN109978048A (en) * | 2019-03-22 | 2019-07-05 | 大唐环境产业集团股份有限公司 | A kind of Desulfurization tower slurry circulating pump malfunction analysis and problem shpoting method |
CN109948869A (en) * | 2019-04-22 | 2019-06-28 | 东南大学 | Desulphurization system SO based on orderly cluster discretization2Exit concentration prediction technique |
CN111723945A (en) * | 2020-06-03 | 2020-09-29 | 杭州欣禾圣世科技有限公司 | BP neural network optimization method based on improved wolf algorithm |
Non-Patent Citations (2)
Title |
---|
傅蔚阳;刘以安;薛松;: "基于灰狼算法与小波神经网络的目标威胁评估", 浙江大学学报(工学版), no. 04, pages 680 - 686 * |
曹建宗;林宸雨;陈文通;刘琦;周权;要亚坤;樊帅军;马彩妮;马双忱;: "现代预测和优化算法在脱硫系统运行中的应用", 华电技术, no. 03, pages 59 - 66 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112906967A (en) * | 2021-02-24 | 2021-06-04 | 大唐环境产业集团股份有限公司 | Desulfurization system slurry circulating pump performance prediction method and device |
CN114749006A (en) * | 2022-03-28 | 2022-07-15 | 浙江浙能兰溪发电有限责任公司 | Method for optimizing wet desulphurization slurry circulating pump |
CN114749006B (en) * | 2022-03-28 | 2023-08-15 | 浙江浙能兰溪发电有限责任公司 | Optimization method for wet desulfurization slurry circulating pump |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112365065A (en) | WFGD self-adaptive online optimization scheduling method | |
CN105930916A (en) | Parallel modular neural network-based byproduct gas real-time prediction method | |
CN114191953A (en) | Flue gas desulfurization and denitrification control method based on convolutional neural network and XGboost | |
CN112465239B (en) | Desulfurization system operation optimization method based on improved PSO-FCM algorithm | |
CN112085196A (en) | Ammonia injection adjusting method for SCR denitration system based on SCA algorithm optimization BP neural network | |
CN113094988A (en) | Data-driven slurry circulating pump operation optimization method and system | |
CN112364562A (en) | Cooperative control method and system for flue gas environmental protection island | |
Hao et al. | Prediction of nitrogen oxide emission concentration in cement production process: a method of deep belief network with clustering and time series | |
CN114119273A (en) | Park comprehensive energy system non-invasive load decomposition method and system | |
CN113440990A (en) | EMD-LSTM based outlet SO2Concentration prediction method | |
Fu et al. | Control strategy for denitrification efficiency of coal-fired power plant based on deep reinforcement learning | |
Shao et al. | Predictive scheduling of wet flue gas desulfurization system based on reinforcement learning | |
CN111815039A (en) | Weekly scale wind power probability prediction method and system based on weather classification | |
CN112613237B (en) | CFB unit NOx emission concentration prediction method based on LSTM | |
CN106647259A (en) | Coking gas desulfurization and denitrification integrated equipment centralized management and control system | |
CN109766666A (en) | Boiler smoke based on low nitrogen burning and SNCR-SCR Collaborative Control discharges NOxConcentration prediction method | |
Fu et al. | Prediction model of desulfurization efficiency of coal-fired power plants based on long short-term memory neural network | |
Gao et al. | A prediction approach on energy consumption for public buildings using mind evolutionary algorithm and bp neural network | |
CN115713029A (en) | Wind power plant stochastic model prediction optimization control method considering delay | |
CN113468813B (en) | Desulfurization system inlet SO2Concentration prediction method and device and electronic equipment | |
Lan et al. | Urban air quality prediction based on space-time optimization LSTM model | |
CN114139785A (en) | Boiler combustion thermal efficiency prediction method and device | |
CN113947013A (en) | Boiler short-term NO based on hybrid deep neural network modelingxEmission prediction method | |
Qing et al. | Prediction model of the NOx emissions based on long short-time memory neural network | |
Xia et al. | Forecasting clean energy power generation in China based on a novel fractional discrete grey model with a dynamic time-delay function |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |