CN115861636A - Burner flame stability quantitative detection method based on digital image processing - Google Patents
Burner flame stability quantitative detection method based on digital image processing Download PDFInfo
- Publication number
- CN115861636A CN115861636A CN202211646796.9A CN202211646796A CN115861636A CN 115861636 A CN115861636 A CN 115861636A CN 202211646796 A CN202211646796 A CN 202211646796A CN 115861636 A CN115861636 A CN 115861636A
- Authority
- CN
- China
- Prior art keywords
- flame
- image
- characteristic
- parameter
- change
- 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
Images
Landscapes
- Photometry And Measurement Of Optical Pulse Characteristics (AREA)
Abstract
The invention discloses a burner flame stability quantitative detection method based on digital image processing, belonging to the technical field of flame detection. The method comprises the steps of acquiring a flame image in real time by using a camera and transmitting the flame image to an industrial personal computer, processing an original flame image in real time by the industrial personal computer to obtain a corresponding flame gray image and a flame binary image, and extracting various flame characteristic parameters according to the three flame images; and fusing flame stability information contained in the flame characteristic parameters to obtain a flame stability index, and quantitatively detecting the flame stability in real time. The method provided by the invention can adjust the fused parameters according to different combustion conditions so as to have wider application range and is sensitive to the change of the stable state of the flame.
Description
Technical Field
The invention belongs to the technical field of flame state detection, and particularly relates to a burner flame stability quantitative detection method based on digital image processing.
Background
Unstable flames are a recognized problem in the combustion process of fossil fuels, and in the case of low-quality fuels, mixed fuels, and co-combustion of biomass and fossil fuels, the flames are more difficult to maintain in a more desirable stable state. Unstable flames can cause many combustion problems, such as furnace vibration, low combustion efficiency, high NOx emissions, and even flame out or deflagration. These combustion problems are detrimental to the economics of the industrial process and can also pose a threat to the health and life safety of the operators.
Although a large number of flame monitoring technologies, such as planar laser-induced fluorescence, radical chemiluminescence imaging, photoelectric detectors, and the like, exist, these measurement methods have the problems of complicated devices, high cost, difficulty in adapting to severe industrial conditions, or single acquired information, and it is difficult to apply these technologies to realize quantitative detection of flame stability under practical industrial conditions. In contrast, the method for evaluating the flame stable state by using the flame image obtained by the camera is a non-invasive measuring method which is low in cost, relatively simple in structure and suitable for severe industrial environments. Most of the existing flame stability detection methods based on flame images are based on single flame characteristics, such as flame root area or brightness, flame outline, flame flicker frequency and the like. There are many limitations in the use of this method for detecting flame stability based on a single flame feature, such as the inability to extract flame roots or contours in the flame image due to the camera mounting location; the method only reflects the stable state of the flame from a certain angle, and may have a certain difference from the real stable state of the flame. The existing flame stability detection method based on the flame image also has a method for evaluating the flame stability state by fusing a plurality of parameters, but on one hand, the method has the problem of incomplete parameter selection, and on the other hand, the selected parameters may not reflect the flame stability state under specific combustion conditions.
Therefore, the invention provides a flame stability quantitative detection method based on a color camera and a digital image processing technology, and the method can fuse various characteristic parameters containing flame stability information according to the combustion working condition so as to enable the obtained flame stability result to be closer to the real stable state of flame. The flame stability quantitative detection method has wide application range and is sensitive to the change of the flame stability state.
Disclosure of Invention
The invention aims to provide a burner flame stability quantitative detection method based on digital image processing, which is characterized by comprising the following steps of:
1) Acquiring a flame image in real time by using a camera;
2) Converting the original flame image into a flame gray image and a flame binary image; extracting a flame characteristic parameter sequence from the flame original image, the flame gray level image and the flame binary image by using an image processing algorithm; the flame characteristic parameters comprise three types, namely geometric parameters, brightness parameters and thermodynamic parameters; the geometric parameters comprise ignition points, root areas, flare angles, flame lengths and flame areas; the brightness parameters include brightness and non-uniformity; thermodynamic parameters include maximum temperature, minimum temperature, average temperature, and scintillation frequency; when the characteristic parameters cannot be extracted from the flame image, only performing subsequent analysis and parameter fusion on the extractable characteristic parameters;
3) Determining a flame parameter change rate from the flame characteristic parameter sequence, determined from the standard deviation and the mean of the flame parameter sequence, i.e.,
wherein, delta x ，σ x And mu x Respectively, the variation characteristic, the standard deviation and the average value of the characteristic parameter x.
4) Analyzing the correlation between the flame parameter change rate and the flame flicker frequency, setting a threshold value according to the actual situation, and determining high correlation characteristic parameters for parameter fusion;
5) Determining a normal change rate and an abnormal change rate of the characteristic parameter from the high-correlation characteristic parameter sequence, both of which contain different flame stability information, wherein, for the flame characteristic parameter sequence, the change rate of the characteristic parameter within a 95% confidence interval is the normal change rate, i.e.,
wherein, delta x-nor Is the normal rate of change, σ, of the characteristic parameter x x-nor And mu x-nor The standard deviation and mean of the parameter x within a 95% confidence interval of the sequence of characteristic parameters x, respectively.
The rate of change of the characteristic parameter outside the 95% confidence interval is an abnormal rate of change, i.e.,
δ x-abnor ＝C x δ x-nor
X x-upper ＝μ x-upper -2σ x
X x-lower ＝μ x-lower +2σ x
wherein, delta x-abnor For abnormal rate of change of characteristic parameter x, by instability coefficient C x And normal rate of change δ x-n o r Jointly determining; mu.s x-upper And mu x-l o wer The mean values of the larger and smaller parts outside the 95% confidence interval of the characteristic parameter x sequence are respectively obtained.
6) Fusing the high correlation characteristic parameters and the normal change rate and the abnormal change rate of the flame flicker frequency to obtain a flame stability index for real-time detection of the flame stable state, wherein the fusing method comprises the following steps
Wherein, delta is a flame stability index, N e For high correlation of the number of characteristic parameters, C x(i) And δ x (i) -nor are the instability coefficient and normal rate of change, C, of the characteristic parameter x (i), respectively F And delta F-nor The instability coefficient and the normal rate of change of the flicker frequency.
The geometric parameters comprise an ignition point, a root area, an opening angle, a length and an area which are all determined by the flame binary image. The ignition point is the absolute distance between the outlet of the combustor and the nearest bright spot along the central axis direction of the combustor, the root area is the quotient of the sum of the number of bright pixel points in a preset boundary frame and the sum of the total pixels of the boundary frame, the flare angle is the included angle formed by fitting the bright spots on the two side edges of the flame along the combustion direction, the flame length is the absolute distance between the outlet of the combustor and the farthest bright spot, and the flame area is the quotient of the sum of the number of the bright spot pixels in the image and the total pixel sum of the image.
And the brightness parameters comprise brightness and nonuniformity which are determined by the flame binary image and the flame gray image together. The brightness is determined by the average gray value of the flame area in the flame gray image, and the nonuniformity is the contrast characteristic of the gray intensity in the flame area.
The thermodynamic parameters comprise a highest temperature, a lowest temperature, an average temperature and a flicker frequency, the three temperatures are determined by the original flame image, and the flicker frequency is determined by the gray flame image. Wherein the highest temperature, the lowest temperature and the average temperature are determined by the flame temperature distribution, and the flicker frequency is determined by the weighted average of the frequency components of the flame average gray level sequence determined by the flame gray level image sequence in the whole frequency range.
The quantitative detection method for the flame stability of the burner can fully utilize the flame image, extract the flame characteristic parameters and the change rate thereof on the basis of digital image processing algorithm and data statistical analysis, and obtain the flame stability index by fusing multi-parameter normal and abnormal change rates to quantitatively evaluate the flame stability. The flame stability index obtained by the method can be used for deeply understanding the combustion mechanism of the fuel, realizing the optimized control of boiler combustion and enhancing the safety and the economical efficiency of boiler operation.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
In the figure: 1-flame, 2-camera, 3-industrial control computer.
Detailed Description
The invention aims to provide a burner flame stability quantitative detection method based on digital image processing; the invention will be described with reference to fig. 1.
Continuously shooting combustion flames 1 by using a camera 2 to obtain flame images; transmitting the flame image to an industrial personal computer 3; the industrial personal computer 3 processes the flame image in real time, converts the original flame image into a flame gray image and a flame binary image, and extracts a flame characteristic parameter sequence from the three flame images. The flame characteristic parameters include three categories, namely geometric parameters, brightness parameters and thermodynamic parameters.
The geometric parameters comprise an ignition point, a root area, an opening angle, a flame length and a flame area; the geometric parameters are determined by the flame binary image, wherein the ignition point is the absolute distance between the outlet of the combustor and the nearest bright spot thereof along the central axis direction of the combustor, the root area is the quotient of the sum of the number of bright pixel points in a preset boundary frame and the sum of the total pixels of the boundary frame, the flare angle is the included angle formed by fitting the bright spots along the edges of the flame in the combustion direction to form two straight lines, the flame length is the absolute distance between the outlet of the combustor and the farthest bright spot, and the flame area is the quotient of the sum of the number of bright spot pixels in the image and the total pixel sum of the image.
The brightness parameters comprise brightness and nonuniformity, the brightness and the nonuniformity are jointly determined by the flame binary image and the flame gray image, the brightness is determined by the average gray value of a flame area in the flame gray image, and the nonuniformity is a contrast characteristic of the intensity of the gray in the flame area.
The thermodynamic parameters comprise a maximum temperature, a minimum temperature, an average temperature and a flicker frequency, the three temperatures are determined by a flame original image, the flicker frequency is determined by a flame gray level image, the maximum temperature, the minimum temperature and the average temperature are determined by flame temperature distribution, and the flicker frequency is determined by weighted average of frequency components of a flame average gray level sequence determined by the flame gray level image sequence in the whole frequency range.
And when the characteristic parameters cannot be extracted from the flame image, only performing subsequent analysis and parameter fusion on the extractable characteristic parameters.
Determining a flame parameter change rate from the flame characteristic parameter sequence, determined from the standard deviation and the mean of the flame parameter sequence, i.e.,
wherein, delta x ，σ x And mu x Respectively, the variation characteristic, the standard deviation and the average value of the characteristic parameter x.
Analyzing the correlation of the flame parameter rate of change with the flame flicker frequency, i.e.,
wherein, delta x(i) And F i The rate of change of the parameter x and the flame flicker frequency, N, in the ith measurement, respectively c In order to measure the number of times,
and &>
Is delta x(i) And F i Average value of (a). And setting a threshold according to the actual condition of the flame, and determining the high-correlation characteristic parameters for parameter fusion.
Determining from the sequence of high-correlation characteristic parameters a normal rate of change and an abnormal rate of change of the characteristic parameters, both of which contain different flame stability information, wherein for the sequence of flame characteristic parameters the rate of change of the characteristic parameters within a 95% confidence interval is the normal rate of change, i.e.,
wherein, delta x-nor Is the normal rate of change, σ, of the characteristic parameter x x-nor And mu x-nor The standard deviation and the mean value of the parameter x in the 95% confidence interval of the characteristic parameter x sequence are respectively.
The rate of change of the characteristic parameter outside the 95% confidence interval is an abnormal rate of change, i.e.,
δ x-abnor ＝C x δ x-nox
X x-upper ＝μ x-pper -2σ x
X x-lower ＝μ x-lower +2σ x
wherein, delta x-abnor Is the abnormal rate of change of the characteristic parameter x, determined by the instability coefficient C x And normal rate of change δ x-n o r Jointly determining; mu.s x-upper And mu x-l o wer The mean values of the larger and smaller parts outside the 95% confidence interval of the characteristic parameter x sequence are respectively obtained.
Fusing the high correlation characteristic parameters and the normal change rate and the abnormal change rate of the flame flicker frequency to obtain a flame stability index for real-time detection of the flame stable state, wherein the fusing method comprises the following steps
Wherein, delta is a flame stability index, N e For high correlation of the number of characteristic parameters, C x(i) And δ x (i) -nor are the instability coefficient and normal rate of change, C, of the characteristic parameter x (i), respectively F And delta F-nor The instability coefficient and the normal rate of change of the flicker frequency.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (4)
1. A burner flame stability quantitative detection method based on digital image processing is characterized by comprising the following steps:
1) Acquiring a flame image in real time by using a camera;
2) Converting the original flame image into a gray level image and a binary image; extracting a flame characteristic parameter sequence from the flame original image, the flame gray level image and the flame binary image by using an image processing algorithm; the flame characteristic parameters comprise three types, namely geometric parameters, brightness parameters and thermodynamic parameters; the geometric parameters comprise ignition points, root areas, flare angles, flame lengths and flame areas; the brightness parameters include brightness and non-uniformity; thermodynamic parameters include maximum temperature, minimum temperature, average temperature, and scintillation frequency; when the characteristic parameters cannot be extracted from the flame image, only performing subsequent analysis and parameter fusion on the extractable characteristic parameters;
3) Determining a flame parameter change rate from the flame characteristic parameter sequence, determined from the standard deviation and the mean of the flame parameter sequence, i.e.,
wherein, delta x ，σ x And mu x Respectively, the variation characteristic, the standard deviation and the average value of the characteristic parameter x.
4) Analyzing the correlation between the flame parameter change rate and the flame flicker frequency, setting a threshold value according to the actual situation, and determining high correlation characteristic parameters for parameter fusion;
5) Determining a normal change rate and an abnormal change rate of the characteristic parameter from the high-correlation characteristic parameter sequence, both of which contain different flame stability information, wherein, for the flame characteristic parameter sequence, the change rate of the characteristic parameter within a 95% confidence interval is the normal change rate, i.e.,
wherein, delta x-nor Is a normal change of the characteristic parameter xConversion rate, σ x-nor And mu x-nor The standard deviation and the mean value of the parameter x in the 95% confidence interval of the characteristic parameter x sequence are respectively.
The rate of change of the characteristic parameter outside the 95% confidence interval is an abnormal rate of change, i.e.,
δ x-abnor ＝C x· δ x - nor
X x-upper ＝μ x-upper -2σ x
X x-lower ＝μ x-lower +2σ x
wherein, delta x-abnor Is the abnormal rate of change of the characteristic parameter x, determined by the instability coefficient C x And normal rate of change δ x-nor Jointly determining; mu.s x-upper And mu x-lower The mean values of the larger and smaller parts outside the 95% confidence interval of the characteristic parameter x sequence are respectively obtained.
6) Fusing the high correlation characteristic parameters and the normal change rate and the abnormal change rate of the flame flicker frequency to obtain a flame stability index for real-time detection of the flame stable state, wherein the fusing method comprises the following steps
Wherein, delta is a flame stability index, N e For high correlation of the number of characteristic parameters, C x(i) And delta x(i)-nor Instability coefficient and normal rate of change, C, of the characteristic parameter x (i), respectively F And delta F-nor The instability coefficient and the normal rate of change of the flicker frequency.
2. The burner flame stability quantitative detection method based on digital image processing as claimed in claim 1, characterized in that: the geometric parameters comprise an ignition point, a root area, an opening angle, a flame length and a flame area which are all determined by the flame binary image, wherein the ignition point is the absolute distance between the outlet of the combustor and the nearest bright spot along the central axis direction of the combustor, the root area is the quotient of the sum of the number of bright pixels and the sum of the total pixels of the boundary frame in a preset boundary frame, the opening angle is the included angle formed by fitting the bright spots on the two side edges of the flame along the combustion direction, the flame length is the absolute distance between the outlet of the combustor and the farthest bright spot, and the flame area is the quotient of the sum of the number of bright spots in the image and the total pixel sum of the image.
3. The quantitative burner flame stability detection method based on digital image processing as claimed in claim 1, characterized in that: the brightness and the nonuniformity included by the brightness parameters are determined by the flame binary image and the flame gray image together, wherein the brightness is determined by the average gray value of a flame region in the flame gray image, and the nonuniformity is a contrast characteristic of the gray intensity in the flame region.
4. The quantitative burner flame stability detection method based on digital image processing as claimed in claim 1, characterized in that: the thermodynamic parameters comprise a maximum temperature, a minimum temperature, an average temperature and a flicker frequency, the three temperatures are determined by a flame original image, the flicker frequency is determined by a flame gray level image, the maximum temperature, the minimum temperature and the average temperature are determined by flame temperature distribution, and the flicker frequency is determined by weighted average of frequency components of a flame average gray level sequence determined by a flame gray level image sequence in the whole frequency range.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211646796.9A CN115861636A (en) | 2022-12-21 | 2022-12-21 | Burner flame stability quantitative detection method based on digital image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211646796.9A CN115861636A (en) | 2022-12-21 | 2022-12-21 | Burner flame stability quantitative detection method based on digital image processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115861636A true CN115861636A (en) | 2023-03-28 |
Family
ID=85674789
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211646796.9A Pending CN115861636A (en) | 2022-12-21 | 2022-12-21 | Burner flame stability quantitative detection method based on digital image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115861636A (en) |
-
2022
- 2022-12-21 CN CN202211646796.9A patent/CN115861636A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102538000B (en) | Combustion flame diagnostic method | |
| Lu et al. | A digital imaging based multifunctional flame monitoring system | |
| CN109614908B (en) | Flame combustion state detection system and detection method | |
| EP3080788B1 (en) | Flame detection system and method | |
| AU2010226995A1 (en) | Image Sensing System, Software, Apparatus and Method For Controlling Combustion Equipment | |
| CN109754019B (en) | Method for continuously monitoring combustion condition of boiler | |
| EP0947813A1 (en) | Interpretation of thermal paint | |
| JP7393511B2 (en) | Learning model generation method and program | |
| SE506185C2 (en) | Method and apparatus for monitoring the laser marking of a bar code | |
| CN110160661B (en) | Object surface temperature measuring method and device based on visible light photo | |
| CN115861636A (en) | Burner flame stability quantitative detection method based on digital image processing | |
| Panditrao et al. | Temperature estimation of visible heat sources by digital photography and image processing | |
| CN112053341A (en) | Surface temperature measurement method of alloy material based on light reflectivity and artificial intelligence | |
| JP7154537B2 (en) | Image feature extraction method and image feature extraction device | |
| CN217278938U (en) | Area array laser radar-based slag falling real-time detection system | |
| Matthes et al. | Camera based flame stability monitoring and control of multi-burner systems using deep learning based flame detection | |
| CN107860715B (en) | Method for detecting carbon content of boiler slag | |
| TWI381139B (en) | Vision-based method for combustion process monitoring, diagnosis, and computer-readable medium using the same | |
| CN114166120A (en) | Real-time slag falling detection method based on area array laser radar | |
| CN113506285A (en) | Boiler furnace three-dimensional temperature field detection method and device and computer equipment | |
| CN102023160B (en) | Measure method of combustible quality based on image | |
| Lu et al. | Advanced monitoring and characterisation of combustion flames | |
| KR100356272B1 (en) | A combustion method for boiler burner | |
| CN117612110B (en) | Hearth flame intelligent monitoring system and method based on computer vision technology | |
| KR100370348B1 (en) | Method and device for extracting flame image characteristics of burning diagnosis |
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 |