A variable feature extraction method for NOx prediction model of thermal power plant based on contrast method

A predictive model and feature extraction technology, applied in neural learning methods, biological neural network models, complex mathematical operations, etc., can solve problems such as complex formation mechanism and inconspicuous extraction effect, achieve wide application range, reduce calculation time consumption, The effect of dimensionality reduction

Pending Publication Date: 2019-02-26
国家电投集团河南电力有限公司 +2
View PDF5 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Due to the influence of NO during boiler operation X There are many factors generated, and its formation mechanism is also very complicated, and the effect of feature extraction using one method is not obvious

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • A variable feature extraction method for NOx prediction model of thermal power plant based on contrast method
  • A variable feature extraction method for NOx prediction model of thermal power plant based on contrast method
  • A variable feature extraction method for NOx prediction model of thermal power plant based on contrast method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0048] see Figure 1~4 , a kind of variable feature extraction method of thermal power plant NOx prediction model based on the comparison method, comprising the following steps:

[0049] S1) Collect the design data of the target boiler, analyze the mechanism of the combustion process of pulverized coal in combination with the actual operating environment, and find out the influence of NO X The main factors of emissions, determine the input variables of the model, and export a certain amount of historical data for denoising and filtering preprocessing. The determined input variables include: 1-5 layers of SOFA opening, ammonia escape, ammonia injection flow, ammonia flow valve opening, pipeline flow regulating valve setting value, regulating valve adjustment command, reactor inlet pressure, flue gas flow, furnace chamber Outlet flue gas temperature, total coal volume, total air volume, secondary air temperature at air preheater outlet of the unit, instantaneous flow rate of co...

Embodiment 2

[0064] This embodiment is a further elaboration on the basis of Embodiment 1, and S2 comprises the following steps:

[0065] First, the final evaluation function of the S-MIFS algorithm is obtained:

[0066]

[0067] In the above formula, f i ∈F is the candidate variable, c is the leading variable, S j ∈S is the selected variable. β is the penalty factor, and |S| is the norm of S or the number of selected variables of the selected set S.

[0068] S2-1, initialization setting, F is the variable set to be selected, S is the selected variable set (blank), and c is the leading variable;

[0069] S2-2. Calculate f i Mutual information with the leading variable c, take out the variable f corresponding to the maximum value i Stored in S; its mathematical expression is as follows:

[0070]

[0071] where p(f i ) for f i The marginal probability distribution of c, p(c) is the marginal probability distribution of c, p(f i , c) is f i and the joint probability distributio...

Embodiment 3

[0077] This embodiment is a further elaboration on the basis of Embodiment 1, and S3 comprises the following steps:

[0078] S3-1. Arranging the original data in rows to form a matrix X;

[0079]

[0080] S3-2. Perform data standardization on X to make its mean value zero;

[0081] S3-3. Find the covariance matrix C of X;

[0082] S3-4. Arrange the eigenvectors according to the eigenvalues ​​from large to small, and take the first k to form a matrix P by row;

[0083] S3-5. By calculating Y=PX, the dimensionally reduced data Y is obtained;

[0084] S3-6. Calculate the contribution rate of each characteristic root.

[0085] Obtain NO after screening through the calculation of Example 1-Example 3 X The root mean square error between the actual concentration value and the test value is 3.3204, which is slightly lower than the accuracy before screening, but the time is obviously reduced by about 30% than before. The method of the invention does have positive effects in the...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a variable characteristic extraction method for NOx prediction model of thermal power plant based on a comparison method, comprising the following steps: S1) collecting designdata of a target boiler, analyzing the combustion process of pulverized coal according to the actual operation environment, and finding out the main factors influencing NOx emission; S2) the data processed by S1 is screened based on mutual information, and the S-MIFS algorithm is used to select the input variables which have great influence on NOx concentration; S3) performing principal componentanalysis on the data obtained from S2 to select a component whose cumulative contribution reaches 95%; S4) dividing the result obtained by S3 into two groups, one group is sent to the neural network for training, and the other group is sent to the neural network for testing; S5) calculating the root mean square error between the output value of the neural network and the original test value; compared with the prior art, the invention can be used for extracting the input variable characteristics of the NOx prediction model of the thermal power plant, and can reduce the dimension of the input variable under the premise of slightly reducing the accuracy, and reduce the calculation time consumption.

Description

technical field [0001] The invention relates to the technical field of a variable feature extraction method of a NOx prediction model of a thermal power plant, in particular to a method for extracting a variable feature of a NOx prediction model of a thermal power plant based on a comparison method. Background technique [0002] Nitrogen oxides (mainly NO, and a small amount of NO 2 , collectively referred to as NO X ) is one of the main air pollutants produced by fossil fuel combustion. According to survey data, the thermal power industry is still NO X industries with the highest emissions. With the country's emphasis on environmental protection, the new environmental protection regulations have increasingly stringent requirements on the environmental protection of power production. NO X The emissions become a key indicator to measure the quality of boiler combustion, and it is the basis for judging whether the boiler is running green. Therefore NO X Accurate measureme...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06F17/10G06N3/08
CPCG06F17/10G06N3/08Y04S10/50
Inventor 王晓峰李海军夏静史恒惠刘长良张丛
Owner 国家电投集团河南电力有限公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap