CN109187515B - Combustion heat release rate measuring method based on dual-waveband chemiluminescence - Google Patents
Combustion heat release rate measuring method based on dual-waveband chemiluminescence Download PDFInfo
- Publication number
- CN109187515B CN109187515B CN201811477379.XA CN201811477379A CN109187515B CN 109187515 B CN109187515 B CN 109187515B CN 201811477379 A CN201811477379 A CN 201811477379A CN 109187515 B CN109187515 B CN 109187515B
- Authority
- CN
- China
- Prior art keywords
- chemiluminescence
- image
- heat release
- release rate
- iccd
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/75—Systems in which material is subjected to a chemical reaction, the progress or the result of the reaction being investigated
- G01N21/76—Chemiluminescence; Bioluminescence
- G01N21/766—Chemiluminescence; Bioluminescence of gases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Computer Hardware Design (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Plasma & Fusion (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By The Use Of Chemical Reactions (AREA)
Abstract
A combustion heat release rate measuring method based on dual-band chemiluminescence is characterized in that a test system consisting of a planar flame burner, an air supply system, a GG400 filter, an IF310 filter, a first ICCD camera, a second ICCD camera, a controller and a computer is built. The controller controls the two ICCD cameras to shoot respectively, firstly shoots an image of a burner furnace plate under the condition of no flame as a background image, then shoots images of OH and CH chemiluminescence under the condition of flame, and transmits the images shot by the two ICCD cameras to the computer to respectively obtain chemiluminescence intensity distribution of OH and CH from the chemiluminescence images. Meanwhile, a deep neural network model is established, and the obtained OH and CH chemiluminescence intensity distribution is input into the deep neural network model to predict the corresponding heat release rate distribution. The invention can predict the heat release rate only by collecting the chemiluminescence intensity through the camera, and the prediction result is highly consistent with the actual heat release rate, thus having high accuracy.
Description
Technical field
The present invention relates to combustion heat release rate field of measuring technique, in particular to a kind of to be based on the chemiluminescent burning of two waveband
Heat liberation rate, heat release rate measurement method.
Background technique
For most of combustion problem, the heat liberation rate, heat release rate of burning be often in combustion diagnosis most important evaluation index it
One.The distribution of heat liberation rate, heat release rate can determine combustion reaction region the most violent, judge whether burning is abundant, monitored and controlled with this
Combustion state processed.
The measurement method of heat liberation rate, heat release rate can be divided into contact and contactless two major classes.The common thermocouple for being of contact method
Thermometric obtains the temperature of measurand using the potential difference at conductor both ends, calculates heat release by the relationship of temperature and heat liberation rate, heat release rate
Rate, structure is simple, is convenient for operation；It can be difficult to adapting to complicated burning situation, dynamic response is poor, in flame gas and thermocouple
Between reach that thermally equilibrated speed is slow, and room and time resolution ratio is all very poor.Non-contacting measurement method mainly has laser spectrum
And radiation method.Laser-diagnose technology realizes the meticulous depiction to flame temperature and structure to a certain extent, is burnt
The affecting laws of phenomena such as heat release, extinguishing, backfire and Mechanical instability；But laser-diagnose technology not only system complex, but also
Requirement is stringent, is difficult to apply in the environment of having vibration, noise.Radiation method, instead of human eye, is had using photoelectric sensor
Faster response speed and bigger temperature-measuring range；But huge structure, involve great expense, technology it is complicated, and need to by fire detecting
Coloring agent is added in flame, therefore is easily destroyed the component and thermal equilibrium state of tested flame.
The chemiluminescence of flame refers to that the particle by chemical reaction excitation transits to lower level from higher energy level and launches
The phenomenon of the light of specific wavelength.Chemiluminescence occurs over just in the narrow space very close with flame reaction area, and the light that shines
Spectrum is related to specific chemical substance, is of great significance to the measurement of heat liberation rate, heat release rate.Chemiluminescence measuring technique does not need Gao Gong
The laser of consumption and complicated optical path, system is simple, requires to reduce to experimental enviroment.Recently it gets more and more people's extensive concerning.
The major defect that the research of heat liberation rate, heat release rate is measured currently based on flame chemiluminescence is that chemiluminescence light is not yet received
The accurate quantitative analysis relationship being distributed with heat liberation rate, heat release rate by force.Burning for hydrocarbon, excited state particle are mainly radiation wavelength
For the CH* of the OH* and 431nm of 308nm.Document (Lauer M. On the Adequacy of Chemiluminescence
as a Measure for Heat Release in Turbulent Flames With Mixture Gradients[J]
.Journal of Engineering for Gas Turbine and Power, 2010,132 (1): 1-8.) point out list
The relationship precision of wave band chemiluminescence and heat liberation rate, heat release rate is lower, and measurement accuracy is not high, merely cannot accurately be surveyed with OH* or CH*
The distribution of heat liberation rate, heat release rate is measured, conclusion is also without universality.
Summary of the invention
In view of the defects existing in the prior art, the present invention provides a kind of based on the chemiluminescent combustion heat release rate survey of two waveband
Amount method.
The present invention is built first by flat-flamed burner, air supply system, GG400 filter plate, IF310 filter plate, first
The pilot system that ICCD camera, the 2nd ICCD camera, controller and computer are constituted.Controller controls two ICCD cameras point
It is not shot, first shoots the burner stone or metal plate for standing a stove on as a precaution against fire image under no flame conditions as background image, then shot under flame conditions
OH* and CH* chemiluminescence image, and the image transmitting that two ICCD cameras are taken is to computer, chemically illuminated diagram
The chemiluminescence intensity distribution of OH* and CH* is respectively obtained as in.Meanwhile establish deep neural network model, will obtain OH* and
Corresponding heat liberation rate, heat release rate distribution can be predicted in CH* chemiluminescence intensity distribution input deep neural network model.
To realize the above-mentioned technical purpose, the technical solution adopted by the present invention is that:
Based on the chemiluminescent combustion heat release rate measurement method of two waveband comprising following steps:
The first step builds pilot system.
Pilot system includes flat-flamed burner, air supply system, ICCD camera, filter plate, controller and computer；
Air supply system provides hydrocarbon-air pre-mixing gas and protective gas, hydrocarbon-air for flat-flamed burner
Premixed gas, by plug ignition, generates stable Premixed Laminar Flow on the stone or metal plate for standing a stove on as a precaution against fire of flat-flamed burner.ICCD camera
It is two, respectively the first ICCD camera and the 2nd ICCD camera, the image shot is gray level image.Two ICCD phases
Machine is connect with controller, and controller is connect with computer；Two ICCD cameras are arranged in the different positions in the outside of stone or metal plate for standing a stove on as a precaution against fire
It sets and the stone or metal plate for standing a stove on as a precaution against fire of the equal alignment surface flame burner of camera lens, it is corresponding to install OH* excited state particle additional before the first ICCD camera lens
The filter plate of wave band installs the filter plate that CH* excited state particle corresponds to wave band additional before the 2nd ICCD camera lens.Such as first
It installs GG400 filter plate before ICCD camera lens additional, installs IF310 filter plate additional before the camera lens of the 2nd ICCD camera, respectively correspond
The central wavelength generated when hydrocarbon-air pre-mixing gas burning is respectively the CH* of the OH* and 431nm of 307nm.
Second step, shooting background image and chemiluminescence image.
The controller control image transmitting that two ICCD cameras are shot, and two ICCD cameras are taken to
Computer handles image.
Controller controls two ICCD cameras and shoots in present position to the stone or metal plate for standing a stove on as a precaution against fire under no flame conditions first.Two
ICCD camera acquires multiple image as background image respectively.If the first ICCD camera acquires n width without the stone or metal plate for standing a stove on as a precaution against fire under flame conditions
For image as background image, the 2nd ICCD camera acquires m width without the stone or metal plate for standing a stove on as a precaution against fire image under flame conditions as background image.
Then, hydrocarbon-air pre-mixing gas is lighted, controller controls two ICCD cameras in original present position
Flame is shot, if the chemiluminescence image of the first ICCD camera acquisition n width flame, the 2nd ICCD camera acquire m width fire
The chemiluminescence image of flame.
Third step, post processing of image
I) based on the n width background image and n width chemiluminescence figure acquired obtained in second step by the first ICCD camera
Picture, obtains the higher OH* chemiluminescence intensity matrix of confidence level, and method is as follows
(1) the n width background image acquired by the first ICCD camera obtained for second step, first in Matlab software
In, the gray value of each background image is extracted, and carry out handling averagely, is transformed into an ambient noise matrix, formula are as follows:
Wherein,For the gray value of background image,x,yFor image coordinate,nFor the quantity of background image.
(2) for the n width chemiluminescence image acquired obtained in second step by the first ICCD camera, in Matlab software
In, the gray value of each chemiluminescence image is extracted, and carry out handling averagely and be converted into gray scale value matrix, obtains preliminary OH*
Chemiluminescence intensity matrix；Formula are as follows:
Wherein,For the gray value of chemiluminescence image,x,yFor image coordinate,nFor the quantity of chemiluminescence image.
(3) chemiluminescence intensity matrix preliminary obtained in (2) is utilized to subtract ambient noise matrix obtained in (1),
Eliminate ambient noise influence, obtain confidence level it is higher, eliminate ambient noise after OH* chemiluminescence intensity matrix:
Ii) based on the m width background image and m width chemiluminescence figure acquired obtained in second step by the 2nd ICCD camera
Picture, using identical method in (1), (2) and (3), acquisition confidence level is higher, eliminates the CH* chemiluminescence after ambient noise
Intensity matrix.
4th step establishes deep neural network model prediction heat liberation rate, heat release rate.
Numerical simulation is carried out using chemical kinetics software Chemkin, simulation result is carried out based on deep learning method
The deep neural network mould that can use OH* and CH* two waveband chemiluminescence intensity to predict heat liberation rate, heat release rate distribution is established in analysis
Type.
Numerical-Mode is carried out to the hydrocarbon flame under thousand kinds or more of different operating conditions by chemical kinetics software Chemkin
Quasi-, that is, the chemiluminescence intensity distribution and heat liberation rate, heat release rate distribution for obtaining OH*, CH* excited state particle should as data set
Chemiluminescence intensity distribution and heat liberation rate, heat release rate point comprising OH*, CH* excited state particle under thousands of kinds of different operating conditions in data set
Cloth.
The method of deep neural network model after being optimized using data set training, can refer to Witten I H,
Frank E, Hall M A. Data Mining: Practical Machine Learning Tools and
Techniques [M] China Machine Press, 2005.
Data set is randomly divided into training set, verifying collection and test set by a certain percentage.Wherein, training set is deep for training
Neural network model is spent, weight therein and offset parameter are optimized；Verifying collection is used for the hyperparameter optimization of model, such as hidden
Hide the number of plies, every layer of number of nodes and form of nonlinear activation function etc.；Test set is then used for neural network model processing
Effect is tested.Such as Fig. 3, using 5 layers of deep neural network model, wherein including 1 input layer, 13,3 hidden layers 14
With 1 output layer 15.Neural network model hidden layer uses leaky-Rectified Liner Unit nonlinear activation function,
For output layer using line rectification function RELU as activation primitive, training method is adaptive moments estimation optimization algorithm.
To get the deep neural network model for arriving optimization after the training of several wheel numbers.
The OH* chemiluminescence intensity matrix and CH* chemiluminescence intensity Input matrix that third step is obtained to optimization
Deep neural network model obtains corresponding heat liberation rate, heat release rate using the deep neural network model of optimization and is distributed, that is, realizes and put
The measurement of heating rate.
Compared with prior art, the present invention can generate following technical effect:
1, the heat liberation rate, heat release rate measurement scheme that the present invention provides, it is easy to operate, it is only necessary to which that chemiluminescence intensity is acquired by camera；
The defect that conventional contact measurement method is difficult to adapt to severe burning situation is avoided, it is multiple to also overcome laser diagnostics method system
Disadvantage miscellaneous, environmental requirement is stringent.
2, since deep learning method can be fitted the complicated corresponding relationship of nonlinearity, Generalization Capability is good, this survey
Amount scheme has universality to different operating conditions, can be generalized to different combustion states.
3, the prediction result of this method is identical with actual heat liberation rate, heat release rate, and accuracy is high.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of flat-flamed burner；
Fig. 2 is the schematic diagram of flame assay system；
Fig. 3 is deep neural network model schematic diagram；
Fig. 4 is the comparison of prediction result and simulation value.
Each symbol indicates in figure:
1 is flat-flamed burner stone or metal plate for standing a stove on as a precaution against fire, and 2 be current stabilization flame retardant bed, and 3 be outer housing；4 be inner housing；5 be protection air cavity；6
For premix chamber；7 be ICCD camera and camera lens, and 8 be GG400 filter plate, and 9 be the flame generated, and 10 be IF310 filter plate, and 11 are
ST133 controller, 12 be computer, and 13 be the input layer of deep neural network, and 14 be hidden layer, and 15 be output layer, and 16 be work
1 simulation result of condition, 17 be 1 prediction result of operating condition, and 18 be 2 simulation result of operating condition, and 19 be 2 prediction result of operating condition, and 20 is imitative for operating condition 3
Very as a result, 21 be 3 prediction result of operating condition.
Specific embodiment
With reference to the accompanying drawing 1 to Fig. 4, embodiments of the present invention are described in further detail.
Based on the chemiluminescent combustion heat release rate measurement method of two waveband comprising following steps:
The first step builds pilot system.
Pilot system includes flat-flamed burner, air supply system, ICCD camera, controller and computer；Gas supply system
System provides hydrocarbon-air pre-mixing gas for flat-flamed burner, and hydrocarbon-air pre-mixing gas is in plane fire
By plug ignition on the stone or metal plate for standing a stove on as a precaution against fire of flame burner, stable flame front is generated.
It referring to Fig.1, is the simplified structure diagram of flat-flamed burner.Flat-flamed burner, including outer housing 3 with
And the inner housing 4 inside outer housing 3, the circular passage between the inner housing 4 and outer housing 3 are protection air cavity 5.Protection
The air supply system in 5 connection of the air cavity external world provides protective gas.Premix chamber 6, the premix chamber 6 are provided in the inner housing 4
The logical external world is used to provide hydrocarbon-air pre-mixing gas air supply system, and the upper end of premix chamber 6 is provided with current stabilization back-fire relief
Layer 2.The top of the inner housing 4 and outer housing 3 is provided with stone or metal plate for standing a stove on as a precaution against fire 1, and the through-hole of supplied gas outflow is offered on stone or metal plate for standing a stove on as a precaution against fire 1.Furnace
The middle section of disk 1 is corresponding with inner housing 4, and the periphery of stone or metal plate for standing a stove on as a precaution against fire 1 is corresponding with the protection air cavity 5 of annular.It is used in the present embodiment
Hydrocarbon-air pre-mixing gas is methane air premixed gas.The methane air premixed gas provided by air supply system,
It from the premix chamber 6 that the bottom of flat-flamed burner is passed into inner housing 4, is sufficiently mixed, flows through later steady in premix chamber 6
The middle part of stone or metal plate for standing a stove on as a precaution against fire 1 is flowed to after flow resistance fire bed 2.Current stabilization flame retardant bed 2 has two aspect effects to make first is that eliminating the vortex structure in flowing
Pre-mixing gas combustion keeps laminar condition, second is that preventing the generation of tempering phenomenon.Finally, the methane air premixed gas of laminar condition
On stone or metal plate for standing a stove on as a precaution against fire surface by plug ignition, stable flame is generated.Flat-flamed burner is also provided with protection air cavity 5, protects gas
Body such as nitrogen enters protection air cavity 5 from the side of flat-flamed burner, through protecting air cavity 5 to be equably passed into the outer of stone or metal plate for standing a stove on as a precaution against fire 1
It encloses, the combustion zone among stone or metal plate for standing a stove on as a precaution against fire 1 is separated with the external world, combustion flame is had an impact with reducing the disturbance of outer gas stream.Fire
The combustion parameter of flame is only vertical direction variation in one direction, it is believed that flame is one-dimensional.
Referring to Fig. 2, ICCD camera 7 is two, and respectively the first ICCD camera and the 2nd ICCD camera, camera shoots to obtain
Image be gray level image.Two ICCD cameras are connect with controller, and ST133 controller 11 is connect with computer 12；Two
Platform ICCD camera is arranged in the different location in the outside of stone or metal plate for standing a stove on as a precaution against fire and camera lens is directed at stone or metal plate for standing a stove on as a precaution against fire 1, in the first ICCD camera lens
The preceding filter plate for installing OH* excited state particle additional and corresponding to wave band, installs CH* excited state particle pair additional before the camera lens of the 2nd ICCD camera
Answer the filter plate of wave band.GG400 filter plate 8 is installed additional such as before the camera lens of the first ICCD camera, before the camera lens of the 2nd ICCD camera
Install IF310 filter plate 10 additional, the central wavelength for corresponding respectively to generate when hydrocarbon-air pre-mixing gas burning is
Two kinds of excited state particles of CH* of the OH* and 431nm of 307nm.
Second step, shooting background image and chemiluminescence image.
The ST133 controller 11 controls the figure that two ICCD cameras 7 are shot, and two ICCD cameras are taken
As being transferred to computer 12, image is handled by computer 12.
The controller first of ST133 controller 11 controls two ICCD cameras 7 in present position to the furnace under no flame conditions
Disk is shot.Since flame only changes along the vertical direction, the shooting angle of ICCD camera does not influence result.To guarantee essence
Degree, two ICCD cameras can be just in the position of its initial placement (two ICCD cameras i.e. above-mentioned in entire shooting process
It is arranged in the stone or metal plate for standing a stove on as a precaution against fire of the different location in the outside of stone or metal plate for standing a stove on as a precaution against fire and the equal alignment surface flame burner of camera lens) shooting, it is not necessarily to
Shift position.Two ICCD cameras acquire multiple image as background image respectively.If the first ICCD camera acquires n width without flame
In the case of stone or metal plate for standing a stove on as a precaution against fire image as background image, the 2nd ICCD camera acquires n width without the stone or metal plate for standing a stove on as a precaution against fire image under flame conditions as back
Scape image.
Then, hydrocarbon-air pre-mixing gas is lighted, controller controls two ICCD cameras in present position to fire
Flame 9 is shot, if the first ICCD camera acquisition n width has the stone or metal plate for standing a stove on as a precaution against fire image under flame conditions as chemiluminescence image, second
ICCD camera acquisition m width has the stone or metal plate for standing a stove on as a precaution against fire image under flame conditions as chemiluminescence image.
When two ICCD cameras are shot, to obtain the higher image of spatial resolution, the camera lens of two ICCD cameras should lean on as far as possible
Nearly flame.For the camera lens of protection ICCD camera and more comprehensive luminescent image is obtained simultaneously, it can not be too close to.Two ICCD phases
When machine is shot, guarantee that the camera lens central axes of each ICCD camera are central in stone or metal plate for standing a stove on as a precaution against fire, and, each ICCD camera vertical with stone or metal plate for standing a stove on as a precaution against fire leading edge face
Optical center height in the sharp side position of flame, the preferable chemiluminescence image of effect can be taken.
Third step, post processing of image
I) based on the n width background image and n width chemiluminescence figure acquired obtained in second step by the first ICCD camera
Picture obtains the higher OH* chemiluminescence intensity matrix of confidence level.
(1) the n width background image acquired by the first ICCD camera obtained for second step, first in Matlab software
In, the gray value of each background image is extracted, and carry out handling averagely, is transformed into an ambient noise matrix, formula are as follows:
Wherein,For the gray value of background image,x,yFor image coordinate,nFor the quantity of background image.
(2) for the n width chemiluminescence image acquired obtained in second step by the first ICCD camera, in Matlab software
In, the gray value of each chemiluminescence image is extracted, and carry out handling averagely and be converted into gray scale value matrix, obtains preliminary OH*
Excited state particle corresponds to chemiluminescence intensity matrix；Formula are as follows:
Wherein,For the gray value of chemiluminescence image,x,yFor image coordinate,nFor the quantity of chemiluminescence image.
(3) chemiluminescence intensity matrix preliminary obtained in (2) is utilized to subtract ambient noise matrix obtained in (1),
The influence that ambient noise can be eliminated as far as possible obtains the corresponding chemiluminescence intensity square of the higher OH* excited state particle of confidence level
Battle array:
Ii) based on obtained in second step by the 2nd ICCD camera acquisition m width background image and m width chemiluminescence
Image obtains the higher CH* chemiluminescence intensity matrix of confidence level using identical method in (1), (2) and (3).
4th step establishes deep neural network model prediction heat liberation rate, heat release rate.
Numerical simulation is carried out using chemical kinetics software Chemkin, simulation result is carried out based on deep learning method
The deep neural network mould that can use OH* and CH* two waveband chemiluminescence intensity to predict heat liberation rate, heat release rate distribution is established in analysis
Type.
Numerical-Mode is carried out to the hydrocarbon flame under thousand kinds or more of different operating conditions by chemical kinetics software Chemkin
Quasi-, the chemiluminescence intensity distribution and heat liberation rate, heat release rate distribution for obtaining OH*, CH* excited state particle are utilized as data set
Data set trains the deep neural network model after being optimized.
Data set is randomly divided into training set, verifying collection and test set by a certain percentage.Wherein, training set is deep for training
Neural network model is spent, weight therein and offset parameter are optimized；Verifying collection is used for the hyperparameter optimization of model, such as hidden
Hide the number of plies, every layer of number of nodes and form of nonlinear activation function etc.；Test set is then used for neural network model processing
Effect is tested.Such as Fig. 3, using 5 layers of deep neural network model, wherein including 1 input layer, 13,3 hidden layers 14
With 1 output layer 15.Neural network model hidden layer uses leaky-Rectified Liner Unit nonlinear activation function,
For output layer using line rectification function RELU as activation primitive, training method is adaptive moments estimation optimization algorithm.
After the training of several wheel numbers, the performance of the model be will be optimized, the deep neural network mould optimized
Type.
The corresponding chemiluminescence intensity matrix of the OH* excited state particle that third step is obtained and CH* excited state particle pair
The chemiluminescence intensity Input matrix answered is to the deep neural network model optimized, i.e., using the deep neural network mould of optimization
Type obtains corresponding heat liberation rate, heat release rate distribution, that is, realizes the measurement of heat liberation rate, heat release rate.
Three groups of heat liberation rate, heat release rate forecast of distribution values 17,19,21 obtained by deep neural network model are taken as shown in figure 4, appointing
It is compared with the simulation value 16,18,20 of corresponding heat liberation rate, heat release rate distribution, shows that the model can be according to OH* and CH* chemiluminescence
The distribution of intensity Accurate Prediction heat liberation rate, heat release rate.OH* and CH* chemiluminescence intensity obtained in the previous step is inputted, it can be real by the model
The accurate measurement of existing heat liberation rate, heat release rate.
The foregoing is merely a preferred embodiment of the present invention, are not intended to restrict the invention, for this field
For technical staff, the invention may be variously modified and varied.All within the spirits and principles of the present invention, made any
Modification, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (9)
1. one kind is based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the following steps are included:
The first step builds pilot system；
Pilot system includes flat-flamed burner, air supply system, ICCD camera, filter plate；Air supply system is flat flame combustion
Burner provides hydrocarbon-air pre-mixing gas, stone or metal plate for standing a stove on as a precaution against fire of the hydrocarbon-air pre-mixing gas in flat-flamed burner
On by plug ignition；
ICCD camera is two, respectively the first ICCD camera and the 2nd ICCD camera；It is installed additional before the first ICCD camera lens
OH* excited state particle corresponds to the filter plate of wave band, installs CH* excited state particle additional before the 2nd ICCD camera lens and corresponds to wave band
Filter plate；
Second step, shooting background image and chemiluminescence image；
Two ICCD cameras shoot the stone or metal plate for standing a stove on as a precaution against fire under no flame conditions first, and two ICCD cameras acquire multiple image respectively
As background image；
Then, hydrocarbon-air pre-mixing gas is lighted, two ICCD cameras shoot flame, two ICCD cameras
The chemiluminescence image of several flames is acquired respectively；
Third step obtains OH* chemiluminescence intensity matrix and CH* chemiluminescence intensity matrix；
I) based on obtained in second step by the first ICCD camera acquire n width background image and n width chemiluminescence image,
Obtain OH* chemiluminescence intensity matrix；
Ii) based on obtained in second step by the 2nd ICCD camera acquire m width background image and m width chemiluminescence image,
Obtain CH* chemiluminescence intensity matrix；
4th step establishes deep neural network model prediction heat liberation rate, heat release rate；
Numerical simulation is carried out to the hydrocarbon flame under thousand kinds or more of different operating conditions by chemical kinetics software Chemkin, is obtained
To the chemiluminescence intensity distribution and heat liberation rate, heat release rate distribution of OH*, CH* excited state particle, as data set, data set is utilized
Train the deep neural network model after being optimized；Deep neural network model after wherein optimizing obtains by the following method
: data set is randomly divided into training set, verifying collection and test set by a certain percentage；Wherein, training set is for training depth refreshing
Through network model, weight therein and offset parameter are optimized；Verifying collection is used for the hyperparameter optimization of model；Test set is then
For testing neural network model treatment effect；
Using 5 layers of deep neural network model, wherein including 1 input layer, 3 hidden layers and 1 output layer；Neural network
Model hidden layer uses leaky-Rectified Liner Unit nonlinear activation function, and output layer uses line rectification function
For RELU as activation primitive, training method is adaptive moments estimation optimization algorithm；
To get the deep neural network model to after optimizing after the training of several wheel numbers；
Depth of the OH* chemiluminescence intensity matrix and CH* chemiluminescence intensity Input matrix that third step is obtained to optimization
Neural network model obtains corresponding heat liberation rate, heat release rate using the deep neural network model of optimization and is distributed, i.e. realization heat liberation rate, heat release rate
Measurement.
2. described in claim 1 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the first step
In, pilot system further includes controller and computer, and the image that two ICCD cameras are shot is gray level image；Two ICCD
Camera is connect with controller, and controller is connect with computer, and controller controls two ICCD cameras and shot, and by two
The image transmitting that ICCD camera takes is handled image to computer, by computer.
3. according to claim 1 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
In one step, two ICCD cameras are arranged in the different location in the outside of stone or metal plate for standing a stove on as a precaution against fire and the equal alignment surface flame burner of camera lens
Stone or metal plate for standing a stove on as a precaution against fire.
4. according to claim 1 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
In one step, GG400 filter plate and IF310 filter plate are installed additional respectively before the camera lens of the first ICCD camera and the 2nd ICCD camera,
The central wavelength generated when corresponding to hydrocarbon-air pre-mixing gas burning is respectively the OH* and 431nm of 307nm
CH*。
5. according to claim 1 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
In one step, flat-flamed burner includes outer housing and the inner housing inside outer housing, the inner housing and outer housing
Between circular passage be protection air cavity；The protection air cavity connection external world is used to provide the air supply system of protective gas；In described
Premix chamber is provided in shell, the premix chamber connection external world is used to provide hydrocarbon-air pre-mixing gas gas supply
System, the upper end of premix chamber are provided with current stabilization flame retardant bed；The top of the inner housing and outer housing is provided with stone or metal plate for standing a stove on as a precaution against fire, on stone or metal plate for standing a stove on as a precaution against fire
Offer the through-hole of supplied gas outflow；The middle section of stone or metal plate for standing a stove on as a precaution against fire is corresponding with inner housing, the periphery of stone or metal plate for standing a stove on as a precaution against fire and the protection air cavity of annular
It is corresponding.
6. according to claim 3 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
In one step, hydrocarbon-air pre-mixing gas of use is methane air premixed gas；The methane-provided by air supply system
Air pre-mixing gas is sufficiently mixed in premix chamber from the premix chamber that the bottom of flat-flamed burner is passed into inner housing,
The middle part that stone or metal plate for standing a stove on as a precaution against fire is flowed to after current stabilization flame retardant bed is flowed through later；The methane air premixed gas of laminar condition is on stone or metal plate for standing a stove on as a precaution against fire surface by fire
Hua Sai ignites, and generates stable flame；
Flat-flamed burner is also provided with protection air cavity, and protective gas enters protection gas from the side of flat-flamed burner
Chamber separates the combustion zone among stone or metal plate for standing a stove on as a precaution against fire, with the external world through protecting air cavity to be equably passed into the periphery of stone or metal plate for standing a stove on as a precaution against fire to reduce outside
The disturbance of air-flow has an impact combustion flame.
7. according to claim 4 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: protect
Shield gas is nitrogen.
8. according to claim 1 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
The i of three steps) step implementation method it is as follows:
(1) the n width background image by the acquisition of the first ICCD camera obtained for second step mentions first in Matlab software
The gray value of each background image is taken, and carries out handling averagely, is transformed into an ambient noise matrix, formula are as follows:
Wherein,For the gray value of background image,x,yFor image coordinate,nFor the quantity of background image；
(2) for the n width chemiluminescence image acquired obtained in second step by the first ICCD camera, in Matlab software,
The gray value of each chemiluminescence image is extracted, and carries out handling averagely and is converted into gray scale value matrix, obtains preliminary OH* chemistry
Luminous intensity matrix；Formula are as follows:
Wherein,For the gray value of chemiluminescence image,x,yFor image coordinate,nFor the quantity of chemiluminescence image；
(3) it utilizes OH* chemiluminescence intensity matrix preliminary obtained in (2) to subtract ambient noise matrix obtained in (1), obtains
OH* chemiluminescence intensity matrix to after elimination ambient noise:
。
9. according to claim 8 be based on the chemiluminescent combustion heat release rate measurement method of two waveband, it is characterised in that: the
The ii) step of three steps obtains the CH* chemistry hair after eliminating ambient noise using (1), (2) and (3) identical method in i) step
Light intensity matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811477379.XA CN109187515B (en) | 2018-12-05 | 2018-12-05 | Combustion heat release rate measuring method based on dual-waveband chemiluminescence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811477379.XA CN109187515B (en) | 2018-12-05 | 2018-12-05 | Combustion heat release rate measuring method based on dual-waveband chemiluminescence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109187515A CN109187515A (en) | 2019-01-11 |
| CN109187515B true CN109187515B (en) | 2019-03-15 |
Family
ID=64938484
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811477379.XA Active CN109187515B (en) | 2018-12-05 | 2018-12-05 | Combustion heat release rate measuring method based on dual-waveband chemiluminescence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109187515B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111089850B (en) * | 2020-02-17 | 2021-09-28 | 北京航空航天大学 | Multi-component concentration estimation method based on single-component absorption spectrum |
| CN111473864B (en) * | 2020-05-26 | 2021-05-28 | 中国人民解放军国防科技大学 | Flame excited state particle radiation rate measuring method based on uniform light source |
| CN111610188A (en) * | 2020-05-26 | 2020-09-01 | 中国人民解放军国防科技大学 | CH chemiluminescence distribution acquisition method for eliminating soot influence |
| CN113252649B (en) * | 2021-06-28 | 2021-09-17 | 中国人民解放军国防科技大学 | NOx emission prediction method based on chemiluminescence |
| CN114357759B (en) * | 2021-12-31 | 2023-04-07 | 北京理工大学 | Diesel engine combustion heat release rate prediction method based on evaporation process |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104965001A (en) * | 2015-06-23 | 2015-10-07 | 南京航空航天大学 | Flame heat release rate pulsation measuring device |
| CN204789430U (en) * | 2015-06-23 | 2015-11-18 | 南京航空航天大学 | Flame heat release rate pulsation measuring device |
-
2018
- 2018-12-05 CN CN201811477379.XA patent/CN109187515B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN109187515A (en) | 2019-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109187515B (en) | Combustion heat release rate measuring method based on dual-waveband chemiluminescence | |
| Sick | High speed imaging in fundamental and applied combustion research | |
| Meier et al. | Investigations in the TECFLAM swirling diffusion flame: Laser Raman measurements and CFD calculations | |
| Dyer | Characterization of one-and two-dimensional homogeneous combustion phenomena in a constant volume bomb | |
| Vilfayeau et al. | Large eddy simulation of flame extinction in a turbulent line fire exposed to air-nitrogen co-flow | |
| Xuan et al. | Soot temperature characterization of spray a flames by combined extinction and radiation methodology | |
| Arndt et al. | Auto-ignition and flame stabilization of pulsed methane jets in a hot vitiated coflow studied with high-speed laser and imaging techniques | |
| Chomiak et al. | Flame quenching by turbulence | |
| CN215726809U (en) | Aeroengine combustion test device | |
| Blunck et al. | Spatial development and temperature of spark kernels exiting into quiescent air | |
| Turner et al. | Application of high-speed, species-specific chemiluminescence imaging for laminar flame speed and Markstein length measurements in spherically expanding flames | |
| Thomsen et al. | Soot measurements in candle flames | |
| Wang et al. | Time-resolved 3D investigation of the ignition process of a methane diffusion impinging flame | |
| Jocher et al. | Dual magnetic effects on soot production in partially premixed flames | |
| Fiala et al. | Heat release and OH* radiation in laminar non-premixed hydrogen-oxygen flames | |
| Morsy | Studies of laminar and turbulent combustion using particle image velocimetry | |
| Vicariotto | Water vapor addition in high concentrations to the fuel side of a two-dimensional methane/air diffusion flame | |
| Eigenbrod et al. | Formaldehyde lif-diagnostics of the autoignition of n-decane droplet pairs in microgravity | |
| Jiang | Imaging Based Studies of Premixed Flame Propagation in Tubes | |
| Ma | Two-colour temperature measurement based on high speed imaging | |
| Park et al. | Flow Field and Wall Temperature Measurements for Reacting Flow in a Lean Premixed Swirl Stabilized Can Combustor | |
| Potenza et al. | Two‐dimensional measurements of primary soot diameter in diffusion flames by two‐dimensional time resolved laser induced incandescence | |
| Park | Experimental Investigation of Flow and Wall Heat Transfer in an Optical Combustor for Reacting Swirl Flows | |
| Karpetis et al. | Miniature supersonic burner for the study of combustion at extreme conditions. i: Internal flow | |
| Zhang et al. | Imaging and radiation of laminar jet diffusion flames with low coflow air velocity in microgravity |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |