CN105631151A - Modeling method for pulverized coal fired boiler combustion optimization - Google Patents
Modeling method for pulverized coal fired boiler combustion optimization Download PDFInfo
- Publication number
- CN105631151A CN105631151A CN201610010165.6A CN201610010165A CN105631151A CN 105631151 A CN105631151 A CN 105631151A CN 201610010165 A CN201610010165 A CN 201610010165A CN 105631151 A CN105631151 A CN 105631151A
- Authority
- CN
- China
- Prior art keywords
- modeling
- boiler combustion
- model
- independent variable
- modeling method
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/36—Circuit design at the analogue level
- G06F30/367—Design verification, e.g. using simulation, simulation program with integrated circuit emphasis [SPICE], direct methods or relaxation methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/06—Power analysis or power optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Control Of Steam Boilers And Waste-Gas Boilers (AREA)
Abstract
The invention discloses a modeling method for pulverized coal fired boiler combustion optimization. The method concretely comprises the steps of (a) preprocessing data; (b) carrying out dimensionality reduction on independent variables: carrying out dimensionality reduction on the model independent variables in a boiler combustion process by an intelligent evolutionary genetic algorithm; (c) establishing models: based on the screening of the model independent variables, respectively establishing the soft measurement models taking nitrogen oxide (NOX) emission and unburned carbon in flue dust as targets by using a radial nerve network; (d) analyzing an error. According to the modeling method, screening and dimensionality reduction are carried out on the independent variables in the boiler combustion process by the intelligent evolutionary genetic algorithm, whether the independent variables take part in modeling or not can be selected according to an optimization result which is 0 or 1, and then modeling is carried out on the independent variables taking part in modeling by the radial nerve network; the modeling time and complexity are reduced while higher fitting precision and prediction accuracy are obtained, and a reliable model foundation is provided for boiler combustion optimization.
Description
[technical field]
The present invention relates to the technical field that pulverized-coal fired boiler controls the combustion process optimal control of technical field, particularly pulverized-coal fired boiler.
[background technology]
The burning optimization of pulverized-coal fired boiler is to realize power plant to improve the important measures of economy and the feature of environmental protection, then needs the combustion process of boiler is modeled before realizing burning optimization. In general, optimization aim is boiler thermal output and NOXDischarge capacity, but in modeling process, generally consider unburned carbon in flue dust and NOXDischarge is modeled, and then unburned carbon in flue dust is converted to boiler thermal output by recycling heat balance equation etc., on this basis, by intelligent algorithm etc., scalable auxiliary variable is carried out optimizing, is achieved that the whole process of boiler combustion modeling optimization.
Because of CFBB (CFB) and tradition coal-burning boiler, structurally there is some difference, so modeling method is also different, there is the related application of traditional coal dust boiler combustion optimization modeling method at present, such as CN200910096411, a kind of modeling method of boiler combustion optimization, boiler load is carried out segmentation, sets up the boiler combustion model under each load; CN201310541803, a kind of power boiler burning subspace modeling and multiple-objection optimization, be also several regions that load is divided into field overlapping, and off-line carries out the foundation of burning subspace model; CN201410122688, is suitable for the power boiler burning performance neural network model of different coal pulverizer combination, is make use of neutral net that coal pulverizer is combined with boiler combustion; CN201510064480, a kind of station boiler NOXDischarge dynamic soft-measuring method, using support vector machine as soft sensor modeling instrument, combines the thought of nonlinear auto-companding moving average, describes boiler combustion process; CN201510198128, the variable dimensionality reduction modeling method of a kind of boiler combustion optimization, make use of principal component analysis (PCA) that model variable is carried out feature extraction, utilizing minimum support vector machine (LSSVM), boiler combustion is modeled. But owing to the coupling of boiler combustion is big, variable impact is many, when not being independent from when the input independent variable of mathematical model is a lot, between independent variable, utilize neutral net that Expired Drugs easily occurs, thus causing the problems such as low, the modeling time length of the model accuracy set up, so modeling independent variable being screened a lot of necessity in modeling process, above-mentioned patent or do not carry out Variable Selection, map (PCA) just with higher dimensional space and carry out feature extraction, well not reducing the complexity of model, the modeling time is also longer. The independent variable of modeling is carried out dimensionality reduction by Intelligent evolution algorithm genetic algorithm (GA) by the present invention, take full advantage of the ability of genetic algorithm global optimization, it is then passed through radial base neural net (RBF) and sets up the model of boiler combustion process, while reducing the error of fitting and forecast error setting up model, accelerate the modeling time.
[summary of the invention]
The purpose of the present invention solves the problems of the prior art exactly, it is proposed to the modeling method that a kind of coal powder boiler combustion optimizes, and not only increases the precision of modeling, and reduces model complexity and modeling time, provides reliable guarantee for burning optimization.
For achieving the above object, the present invention proposes the modeling method that a kind of coal powder boiler combustion optimizes, and is on basis modeling independent variable being carried out dimensionality reduction based on genetic algorithm, utilizes radial base neural net that boiler combustion process is modeled, and concrete steps include:
(a) data prediction: utilize 3 �� criterions to reject the data with gross error gathered Distributed Control System (DCS);
(b) independent variable dimensionality reduction: utilize Intelligent evolution genetic algorithm that the model independent variable of boiler combustion process is carried out dimensionality reduction;
C () model is set up: on the basis of model independent variable screening, utilizes radial neural network to set up respectively with NOXDischarge capacity, unburned carbon in flue dust are the soft-sensing model of target;
(d) error analysis: the model set up is fitted the analysis of error and forecast error, calibration model parameter.
As preferably, in described (b) step, model needs the independent variable of screening to include coal-fired calorific value, boiler load, total air output, coal-supplying amount, absorbing quantity, coal property, exhaust gas temperature, main steam temperature, main steam flow, and coal property includes application base ash, carbon content, hydrogen content, nitrogen content.
As preferably, in described (b) step, genetic algorithm is utilized to be optimized calculating, it is necessary to solution space is mapped to space encoder, a solution of each coding correspondence problem, code length is designed as 12 by this patent, each corresponding independent variable chromosomal, the gene value of each can only be " 0 " or " 1 ", wherein, " 0 " represents the independent variable of correspondence not as final modeling independent variable, and " 1 " represents and participates in final modeling.
As preferably, in described (b) step, genetic algorithm being chosen the inverse of test set error in data quadratic sum as fitness function, namelySelecting operation selection ratio selection opertor, the operation that intersects selects single-point crossover operator, and mutation operation selects single-point mutation operator.
As optimization, in described (c) step, the function that radial base neural net adopts is Gaussian function, and asks for Basis Function Center c by K-means clustering method.
Beneficial effects of the present invention:
The independent variable of boiler combustion process is undertaken screening dimensionality reduction by the present invention by Intelligent evolution genetic algorithm, select whether independent variable participates in modeling according to optimum results " 0 " or " 1 ", then pass through the radial base neural net independent variable to participating in modeling to be modeled, or higher fitting precision and precision of prediction while, reduce modeling time and complexity, for boiler combustion optimization be provided reliable model basis.
[detailed description of the invention]
The modeling method that a kind of coal powder boiler combustion of the present invention optimizes, concrete steps include:
Step one, utilize 3 �� criterions reject Distributed Control System (DCS) collection the data with gross error;
Step 2, utilize Intelligent evolution genetic algorithm that the model independent variable of boiler combustion process is carried out dimensionality reduction;
Step 3, model independent variable screening basis on, utilize radial neural network to set up respectively with NOXDischarge capacity, unburned carbon in flue dust are the soft-sensing model of target;
Step 4, the model set up is fitted the analysis of error and forecast error, calibration model parameter.
Wherein, in the process that genetic algorithm optimization calculates, design procedure is:
1, the generation of initial population. Randomly generating N number of original string data structure, each string structure is body one by one, and the string structure data of each individuality only have " 0 " or " 1 " two kinds of values.
2, the calculating of fitness function. Choose the inverse of test set error in data quadratic sum as fitness function:
3, operation is selected. Selection ratio operator, probability that is namely individual selected and that be genetic in population of future generation is directly proportional to this individuality volume fitness size.
4, intersection operation. First the individuality in population is carried out random pair between two, and randomly selects a certain gene as cross point, transposition, produce new individuality.
5, mutation operation. Randomly generating change point, change genic value, namely " 0 " becomes " 1 ", and " 1 " becomes " 0 ".
6, result output. Through iterative computation, when meeting end condition, what the population in the last reign of a dynasty of output was corresponding is optimal solution or the suboptimal solution of problem, has filtered out the combination of most representational model independent variable.
The learning algorithm of radial base neural net specifically includes that
1, Basis Function Center c is asked for based on K-means clustering method. Including netinit, input training sample packet, readjust cluster centre etc.
2, variances sigma is solvedi. The basic function that the RBF neural of the present invention adopts is Gaussian function, variances sigmaiCan be solved by following formula:
3, the weights between hidden layer and output layer are calculated. Between hidden layer with output layer, the neuronic weights that are connected can directly calculate with method of least square and obtain:
Work process of the present invention:
The independent variable of boiler combustion process is undertaken screening dimensionality reduction by the present invention by Intelligent evolution genetic algorithm, select whether independent variable participates in modeling according to optimum results " 0 " or " 1 ", then pass through the radial base neural net independent variable to participating in modeling to be modeled, while obtaining higher fitting precision and precision of prediction, reduce modeling time and complexity, for boiler combustion optimization be provided reliable model basis.
Above-described embodiment is the description of the invention, is not limitation of the invention, any scheme after simple transformation of the present invention is belonged to protection scope of the present invention.
Claims (5)
1. the modeling method that a coal powder boiler combustion optimizes, it is characterised in that: described modeling method is based on genetic algorithm modeling independent variable to be carried out on the basis of dimensionality reduction, utilizing radial base neural net that boiler combustion process is modeled, and concrete steps include:
(a) data prediction: utilize 3 �� criterions to reject the data with gross error gathered Distributed Control System (DCS);
(b) independent variable dimensionality reduction: utilize Intelligent evolution genetic algorithm that the model independent variable of boiler combustion process is carried out dimensionality reduction;
C () model is set up: on the basis of model independent variable screening, utilizes radial neural network to set up respectively with NOXDischarge capacity, unburned carbon in flue dust are the soft-sensing model of target;
(d) error analysis: the model set up is fitted the analysis of error and forecast error, calibration model parameter.
2. the modeling method that a kind of coal powder boiler combustion as claimed in claim 1 optimizes, it is characterized in that: in described (b) step, model needs the independent variable of screening to include coal-fired calorific value, boiler load, total air output, coal-supplying amount, absorbing quantity, coal property, exhaust gas temperature, main steam temperature, main steam flow, and coal property includes application base ash, carbon content, hydrogen content, nitrogen content.
3. the modeling method that a kind of coal powder boiler combustion as claimed in claim 1 optimizes, it is characterized in that: in described (b) step, genetic algorithm is utilized to be optimized calculating, need solution space is mapped to space encoder, one solution of each coding correspondence problem, code length is designed as 12 by this patent, each corresponding independent variable chromosomal, the gene value of each can only be " 0 " or " 1 ", wherein, " 0 " represents the independent variable of correspondence not as final modeling independent variable, and " 1 " represents and participates in final modeling.
4. the modeling method that a kind of coal powder boiler combustion as claimed in claim 1 optimizes, it is characterised in that: in described (b) step, genetic algorithm is chosen the inverse of test set error in data quadratic sum as fitness function, namelySelecting operation selection ratio selection opertor, the operation that intersects selects single-point crossover operator, and mutation operation selects single-point mutation operator.
5. the modeling method that a kind of coal powder boiler combustion as claimed in claim 1 optimizes, it is characterised in that: in described (c) step, the function that radial base neural net adopts is Gaussian function, and asks for Basis Function Center c by K-means clustering method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610010165.6A CN105631151A (en) | 2016-01-06 | 2016-01-06 | Modeling method for pulverized coal fired boiler combustion optimization |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610010165.6A CN105631151A (en) | 2016-01-06 | 2016-01-06 | Modeling method for pulverized coal fired boiler combustion optimization |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105631151A true CN105631151A (en) | 2016-06-01 |
Family
ID=56046080
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610010165.6A Pending CN105631151A (en) | 2016-01-06 | 2016-01-06 | Modeling method for pulverized coal fired boiler combustion optimization |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105631151A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106442470A (en) * | 2016-08-31 | 2017-02-22 | 广州博谱能源科技有限公司 | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network |
CN107016176A (en) * | 2017-03-24 | 2017-08-04 | 杭州电子科技大学 | A kind of hybrid intelligent overall boiler burning optimization method |
CN109187914A (en) * | 2018-09-18 | 2019-01-11 | 哈尔滨锅炉厂有限责任公司 | The prediction technique of coal-burning power plant's NOx generation amount based on coal characteristic |
CN110299188A (en) * | 2019-05-09 | 2019-10-01 | 上海电力学院 | SCR flue gas denitrification system GRNN modeling method based on GA variables choice |
CN110751344A (en) * | 2019-10-30 | 2020-02-04 | 汉谷云智(武汉)科技有限公司 | Power plant boiler operation optimization system and method based on intelligent visualization technology |
CN112395924A (en) * | 2019-08-16 | 2021-02-23 | 阿里巴巴集团控股有限公司 | Remote sensing monitoring method and device |
CN113077041A (en) * | 2020-01-06 | 2021-07-06 | 大唐环境产业集团股份有限公司 | Deep neural network modeling independent variable dimension reduction method based on genetic algorithm |
CN113864814A (en) * | 2021-09-15 | 2021-12-31 | 华能国际电力股份有限公司上海石洞口第一电厂 | Boiler combustion optimization method, device and medium based on variable screening |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102679391A (en) * | 2012-05-21 | 2012-09-19 | 常州市新港热电有限公司 | Combustion online optimizing method of boiler |
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 |
-
2016
- 2016-01-06 CN CN201610010165.6A patent/CN105631151A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102679391A (en) * | 2012-05-21 | 2012-09-19 | 常州市新港热电有限公司 | Combustion online optimizing method of boiler |
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 (3)
Title |
---|
吴智群 等: "电站锅炉智能燃烧优化技术的应用研究", 《热力发电》 * |
王江荣 等: "基于遗传算法自变量降维的神经网络煤矿瓦斯涌出量预测模型", 《煤》 * |
陶钧 等: "基于遗传算法和径向基函数神经网络的转炉炼钢模型", 《系统仿真学报》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106442470A (en) * | 2016-08-31 | 2017-02-22 | 广州博谱能源科技有限公司 | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network |
CN107016176A (en) * | 2017-03-24 | 2017-08-04 | 杭州电子科技大学 | A kind of hybrid intelligent overall boiler burning optimization method |
CN109187914A (en) * | 2018-09-18 | 2019-01-11 | 哈尔滨锅炉厂有限责任公司 | The prediction technique of coal-burning power plant's NOx generation amount based on coal characteristic |
CN110299188A (en) * | 2019-05-09 | 2019-10-01 | 上海电力学院 | SCR flue gas denitrification system GRNN modeling method based on GA variables choice |
CN112395924A (en) * | 2019-08-16 | 2021-02-23 | 阿里巴巴集团控股有限公司 | Remote sensing monitoring method and device |
CN112395924B (en) * | 2019-08-16 | 2024-02-20 | 阿里巴巴集团控股有限公司 | Remote sensing monitoring method and device |
CN110751344A (en) * | 2019-10-30 | 2020-02-04 | 汉谷云智(武汉)科技有限公司 | Power plant boiler operation optimization system and method based on intelligent visualization technology |
CN113077041A (en) * | 2020-01-06 | 2021-07-06 | 大唐环境产业集团股份有限公司 | Deep neural network modeling independent variable dimension reduction method based on genetic algorithm |
CN113864814A (en) * | 2021-09-15 | 2021-12-31 | 华能国际电力股份有限公司上海石洞口第一电厂 | Boiler combustion optimization method, device and medium based on variable screening |
CN113864814B (en) * | 2021-09-15 | 2024-04-26 | 华能国际电力股份有限公司上海石洞口第一电厂 | Variable screening-based boiler combustion optimization method, device and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105631151A (en) | Modeling method for pulverized coal fired boiler combustion optimization | |
Saraswat et al. | Evaluation of energy alternatives for sustainable development of energy sector in India: An integrated Shannon’s entropy fuzzy multi-criteria decision approach | |
Bhave et al. | Screening and techno-economic assessment of biomass-based power generation with CCS technologies to meet 2050 CO2 targets | |
Rahat et al. | Data-driven multi-objective optimisation of coal-fired boiler combustion systems | |
CN111144609A (en) | Boiler exhaust emission prediction model establishing method, prediction method and device | |
Pourkiaei et al. | Modeling and experimental verification of a 25W fabricated PEM fuel cell by parametric and GMDH-type neural network | |
CN109670629A (en) | Coal-burning boiler thermal efficiency forecast method based on shot and long term Memory Neural Networks | |
CN104763999A (en) | Power plant pulverized coal boiler combustion performance online optimizing method and system | |
Ozkan et al. | Data mining-based upscaling approach for regional wind power forecasting: Regional statistical hybrid wind power forecast technique (RegionalSHWIP) | |
CN101802728A (en) | Gaseous emission is carried out system and method based on the virtual sensing of experience set | |
CN108985381B (en) | Method, device and equipment for determining nitrogen oxide emission prediction model | |
CN106527176A (en) | MFOA (modified fruit fly optimization algorithm)-SVM (support vector machine)-based boiler thermal efficiency and NOX modeling method | |
Arias-Rosales et al. | Wind turbine selection method based on the statistical analysis of nominal specifications for estimating the cost of energy | |
CN112330179B (en) | Fuzzy comprehensive evaluation method for coal blending combustion based on improved entropy weight method | |
CN104807039A (en) | Variable dimensionality reduction modeling method for boiler combustion optimization | |
CN113095556A (en) | Medium-and-long-term load prediction method and system considering carbon neutralization tendency | |
Castrejón et al. | Analysis of the contribution of CCS to achieve the objectives of Mexico to reduce GHG emissions | |
CN102982390A (en) | Boiler combustion efficiency predicting method based on support vector machine incremental algorithm | |
CN103870878B (en) | A kind of Combustion Characteristics in Utility Boiler neural network model | |
Ma et al. | Three-objective optimization of boiler combustion process based on multi-objective teaching–learning based optimization algorithm and ameliorated extreme learning machine | |
Tang et al. | Theory-guided deep neural network for boiler 3-D NOx concentration distribution prediction | |
CN107203140A (en) | A kind of sintering production process performance estimating method based on fuzzy synthesis | |
CN113239612A (en) | Boiler combustion state diagnosis method and system and storage medium | |
CN101692270B (en) | Multivariate flame monitor-based on-line judgment method for fuel type | |
CN101701951B (en) | Fuel type on-line identification method based on support vector machine |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |