A flame image-based industrial furnace combustion diagnosis method

Abstract

The application discloses a kind of industrial furnace combustion diagnostic methods based on flame image, including the arrangement of camera: a pair of side walls of boiler flue above the secondary air nozzle of industrial boiler is respectively provided with one camera device;Image acquisition and processing: two flame cameras are respectively named as camera one and camera two, the video signal of camera one and camera two is connected into the same computer, every certain time respectively extracts a frame of RGB image, then according to the area rate of flame and gray value, left and right combustion deviation index Dp1, front and back combustion deviation index Dp2 and combustion fluctuation index Db are calculated, and the combustion state of flame is diagnosed according to the size of Dp1, Dp2 and Db.The method provides diagnostic information of the stability and uniformity of industrial boiler combustion, with the advantages of low latency, large monitoring range, good adaptability, high precision, low cost, provides high-quality diagnostic information for combustion control and automatic operation.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of waste incineration, and particularly relates to an industrial furnace combustion diagnosis method based on a flame image. BACKGROUND

[0002] Industrial furnaces such as waste incinerators, biomass boilers and coal-fired chain furnaces generally lack real-time and accurate combustion diagnosis methods during combustion control, resulting in low operation efficiency. The performance results are large load fluctuations, high excess air coefficients and low automation levels of such furnaces, which are difficult to achieve the effects of stable combustion, uniform flame distribution and less manual intervention in the combustion control field. The difficulty of industrial furnace combustion diagnosis lies in the harsh environment around the furnace, high time delay of conventional diagnosis methods and high cost sensitivity to instruments. If low-delay diagnosis that can achieve long-term stable operation at low cost is not used, it is difficult to promote and apply it on a large scale.

[0003] For example, the thermocouple for monitoring the temperature in the furnace can only measure a small range of local temperature, and the time delay of the response temperature change after ash accumulation is about 30-60 seconds. The zirconia oxygen analyzer for monitoring O2 in the waste heat boiler is limited to measuring temperature and corrosive gas environment, and is generally far away from the combustion source. The response time of the oxygen content change is about 40-60 seconds. These conventional detection methods either have limited detection range or have large time delay, which greatly affects the combustion diagnosis and control of industrial furnaces. The industry urgently needs non-destructive, large-scale and low-delay diagnosis technology. SUMMARY

[0004] The technical problem solved by the present application is to provide an industrial furnace combustion diagnosis method based on a flame image, which can effectively solve the problems of limited detection range, large time delay and great influence on the combustion diagnosis and control of industrial furnaces.

[0005] Technical scheme: An industrial furnace combustion diagnosis method based on a flame image, comprising the following steps:

[0006] S1, arrangement of a camera: a pair of side walls of a boiler flue above a secondary air nozzle of an industrial furnace are respectively provided with a camera device, a pre-embedded pipe is arranged in the middle of the side wall of the boiler flue, a flame camera of the camera device is arranged in the pre-embedded pipe, the pre-embedded pipe is in the shape of a conical enlargement, the cone angle of the conical enlargement is the same as the viewing angle of the flame camera, and the upper edge is horizontally arranged;

[0007] S2, image acquisition and processing: two flame cameras are respectively named as camera one and camera two, the video signals of the camera one and the camera two are connected to the same computer, one frame of RGB image is extracted every certain time, and then the left and right combustion deviation indexes D are calculated according to the flame area rate and the gray value. p1Before and after combustion deviation index D p2 and combustion fluctuation index D b ,according to D p1 , D p2 and D b The size of the flame is used to diagnose the combustion state, that is, when D p1 When the value is greater than 0, the burning area is biased towards one side of camera one; the larger the value, the greater the degree of bias in burning. D p1 When the value is less than 0, the burning area is biased towards one side of camera two; the smaller the value, the greater the degree of bias in burning. D p2 When the value is greater than 0, the combustion zone is biased towards the front wall; the larger the value, the greater the degree of bias. D p2 When the value is less than 0, the burning area is biased towards the rear wall; the smaller the value, the greater the degree of bias in burning. D b The smaller the value, the smoother the combustion process; D b The larger the value, the greater the fluctuation in combustion state.

[0008] Preferably, the first and second cameras are of the same specifications, and the center lines of their viewing angles converge on the center line of the boiler. The height of the convergence point is greater than the height of the secondary air nozzle, and the height difference is significant. H The range is 0.5~2 m.

[0009] Preferably, the flame camera is an air-cooled camera or a water-cooled camera, and the front end of the flame camera does not exceed the range of the pre-embedded pipe.

[0010] Furthermore, a compressed air purging pipe is provided between the flame camera and the pre-embedded pipe to prevent ash and coking from accumulating at the front end of the pre-embedded pipe.

[0011] Preferably, the formula for grayscale conversion of the RGB image in step S2 is shown in equation (1) below:

[0012] (1)

[0013] In the formula, (x,y) These are the pixel coordinates. The pixel coordinates are (x,y) grayscale value, R(x,y) The coordinates in the R channel are (x,y) The pixel value of the pixel, G(x,y) The coordinates in the G channel are (x,y) The pixel value of the pixel, B(x,y) the pixel value of the pixel point with coordinates (x,y) in the B channel;

[0014] The average flame gray value of the flame camera at different times is as follows:

[0015] (2)

[0016] In the formula, Z is a judgment function, when ≥0, Z =1; when <0, Z =0; ∑ is an accumulated summation symbol, indicating that the corresponding function is calculated once for all (x, y) pixel points and then the results are accumulated and summed up;

[0017] The average flame gray value of the camera one in the previous n seconds g f1 The formula is as follows, wherein n is 5-20:

[0018] (3)

[0019] The average flame gray value of the camera two in the previous n seconds is calculated by the same method g f2 , and the formula is as follows:

[0020] (4)

[0021] The combustion deviation index is calculated according to the average flame gray value of the camera one and the camera two in the previous n seconds D p1 , and the formula is as follows:

[0022] (5)

[0023] D p1 If it is positive, the combustion area deviates to the side of the camera one, and the greater the value, the greater the degree of deviation; D p1 If it is negative, the combustion area deviates to the side of the camera two, and the smaller the value, the greater the degree of deviation.

[0024] Further, the images of the camera one and the camera two close to the 50% picture of the front wall are extracted and horizontally combined into an image, and the average flame gray value in the previous n seconds is calculated g f3 ; the images of the camera one and the camera two close to the 50% picture of the back wall are extracted and horizontally combined into an image, and the average flame gray value in the previous n seconds is calculated g f4 The formula is as follows:

[0025] (6)

[0026] (7)

[0027] Further calculate the combustion deviation index D p2 , the formula is as follows:

[0028] (8)

[0029] D p2 If positive, the combustion area deviates to the front wall, and the larger the value, the greater the degree of deviation; D p2 If negative, the combustion area deviates to the back wall, and the smaller the value, the greater the degree of deviation.

[0030] Preferably, the calculation process of the flame area rate in step S2 is as follows:

[0031] The images of the camera one and the camera two are merged horizontally, that is, the images of the two flame cameras are horizontally spliced and then processed as an image, at this time, the horizontal pixel number of the image is i = i1 + i2 , the vertical pixel number is still j, the average area rate variance of the flame in the previous m seconds is calculated, m is recommended to be 3-10,

[0032] First, the average area rate of the flame of each extracted frame in the previous m seconds S v The calculation formula is as follows:

[0033] (9)

[0034] In the formula, Z is a judgment function, when ≥0, Z=1, when <0, Z=0; in the formula, the subscript x,y is the summation index, x varies from 1 to i , varies from 1 to y , and j varies from 1 to i,j , respectively, the horizontal pixel number and the vertical pixel number of the merged image;

[0035] Secondly, the calculation formula of the average area rate variance of the flame in the previous m seconds D s is as follows:

[0036] (10)

[0037] In the formula, The average area rate of the flame in the screenshot image at each time point in the previous m seconds S v The average value of the average area rate of the flame in the screenshot image at each time point in the previous m seconds

[0038] Next, the high-temperature flame average gray value variance of the images of camera one and camera two in the previous m seconds is calculated, the high-temperature flame average gray value of each extracted frame in the previous m seconds is calculated first The judgment basis of the high-temperature flame ≥0, and the formula is as follows:

[0039] (11)

[0040] In the formula, is a judgment function, when ≥0, =1, when <0, =0;

[0041] Then, the flame average gray value variance in the previous m seconds is calculated D g The formula is as follows:

[0042] (12)

[0043] In the formula, is the average value of the average gray value of the flame in the screenshot image at each time point in the previous m seconds

[0044] Finally, the combustion fluctuation index is calculated D b The formula is as follows:

[0045] (13)

[0046] The smaller the combustion fluctuation index D b is, the more stable the combustion process is; D b The larger the combustion fluctuation index

[0047] Beneficial effects: The method can calculate corresponding indexes according to the flame shape to reflect the combustion deviation and fluctuation in the industrial furnace, diagnose the combustion state in time, and be used for the control of the combustion process. The left-right deviation index can be used to intuitively judge whether the flame in the furnace deviates to the left or right side, and the front-back deviation index can be used to judge whether the flame in the furnace deviates to the front or back wall, so as to adjust the combustion state in time. The identification index can be used as an input parameter of automatic combustion control ACC. The combustion fluctuation index can dynamically reflect whether the combustion is stable, indirectly measure the unburned condition of gaseous combustible, and be used as a basis for combustion adjustment. The method has the following advantages:

[0048] ​The diagnostic method of this invention has a small delay, and the response time of the camera is generally less than 1 second. The index algorithm provided by this invention can be updated in real time, providing timely combustion diagnosis.

[0049] With a large detection range, cameras are placed on the two side walls above the secondary air of the industrial furnace at a considerable distance, which can completely cover the entire field of view below the monitoring point.

[0050] It has good adaptability, can achieve combustion diagnosis without the need for machine learning, has good transferability, and can flexibly set the calculation time range of indicators according to needs and application scenarios;

[0051] High precision: The flame dynamic morphology-based index provided by this invention can accurately reflect the flame deviation and fluctuation in the furnace, and has the advantage of high precision.

[0052] It is low-cost, with the main equipment consisting of two sets of high-temperature resistant cameras inside the furnace, cooling and purging equipment, computers, etc. There are no high-cost equipment and instruments, which has the advantages of low cost and easy installation. Attached Figure Description

[0053] Figure 1 This is a schematic diagram showing the positional relationship between the camera and the industrial boiler in this invention;

[0054] Figure 2 This is a schematic diagram showing the specific location where the camera is installed in this invention;

[0055] Figure 3 This is a schematic diagram of the camera video signal processing flow in this invention;

[0056] Figure 4 In Example 2 D b The operating curve of the index and the O2 of the waste heat boiler;

[0057] Figure 5 In Example 2 D b Operating curves of index and CO in waste heat boiler flue;

[0058] Figure 6 In Example 2 D b Operating curves of index and main steam output of waste heat boiler;

[0059] The numbers in the diagram are: 1. Industrial boiler, 2. Secondary air nozzle, 3. Boiler flue, 4. Camera device, 41. Embedded pipe, 42. Flame camera, 43. Center line of view. Detailed Implementation

[0060] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments: Example 1

[0061] A method for industrial furnace combustion diagnosis based on flame images, such as Figure 3 As shown, it includes the following steps:

[0062] S1. Camera placement: such as Figures 1-2 As shown, a camera device 4 is installed on a pair of side walls of the boiler flue 3 above the secondary air nozzle 2 of the industrial boiler 1. A pre-embedded pipe 41 is provided in the middle of the side wall of the boiler flue 3. The height of the two pre-embedded pipes is the same, and the two pre-embedded pipes are arranged opposite each other. The flame camera 42 of the camera device 4 is installed in the pre-embedded pipe 41. The pre-embedded pipe 41 is conical and enlarged. The cone angle of the cone is the same as the viewing angle of the flame camera 42. The conical pre-embedded pipe is inclined downward and the outer contour of the upper end face is at a horizontal angle. The two flame cameras are the same size. The center lines 43 of the viewing angles of the two flame cameras meet on the center line of the boiler. The height of the meeting point is greater than the height of the secondary air nozzle, and the height difference H is 0.5~2 m. The flame camera 42 is an air-cooled camera or a water-cooled camera, and the front end of the flame camera 42 does not exceed the range of the pre-embedded pipe 41. A compressed air purging pipe is provided between the flame camera 42 and the pre-embedded pipe 41 to prevent ash accumulation and coking at the front end of the pre-embedded pipe 41.

[0063] S2. Image Acquisition and Processing: Two flame cameras 42 are named Camera 1 and Camera 2 respectively. The video signals of Camera 1 and Camera 2 are connected to the same computer. At regular intervals, one frame of RGB image is extracted from each camera. Then, the left and right combustion deviation index is calculated based on the flame area ratio and grayscale value. D p1 Before and after combustion deviation index

[0064] D p2 and combustion fluctuation index D b ,according to D p1 , D p2 and D b The size of the flame is used to diagnose the combustion state, that is, when D p1 When the value is greater than 0, the burning area is biased towards one side of camera one; the larger the value, the greater the degree of bias in burning. D p1 When the value is less than 0, the burning area is biased towards one side of camera two; the smaller the value, the greater the degree of bias in burning. D p2 When the value is greater than 0, the combustion zone is biased towards the front wall; the larger the value, the greater the degree of bias. D p2When less than 0, the combustion area is deviated to the rear wall, and the smaller the value is, the greater the degree of deviation is; D b The smaller the value is, the more stable the combustion process is; D b The greater the value is, the greater the fluctuation of the combustion state is.

[0065] The formula of the RGB image gray value is shown as formula (1) below:

[0066] (1)

[0067] In the formula, (x,y) is the pixel point coordinate, g (x,y) is the pixel point coordinate is (x,y) the gray value of the pixel point, R(x,y) is the pixel value of the pixel point with the coordinate (x,y) in the R channel, G(x,y) is the pixel value of the pixel point with the coordinate (x,y) in the G channel, B(x,y) is the pixel value of the pixel point with the coordinate (x,y) in the B channel;

[0068] The formula of the flame gray mean value of the flame camera at different moments is as follows:

[0069] (2)

[0070] In the formula, Z is a judgment function, when ≥0, Z=1; when <0, Z=0;

[0071] The flame gray value mean of the camera one in the previous n seconds g f1 The formula is as follows, wherein n is 5-20:

[0072] (3)

[0073] The flame gray value mean of the camera two in the previous n seconds is calculated by the same method g f2 , and the formula is as follows:

[0074] (4)

[0075] The combustion deviation index is calculated according to the flame gray value mean of the camera one and the camera two in the previous n seconds D p1 , and the formula is as follows:

[0076] (5)

[0077] D p1 If positive, the combustion area is deviated to the side of camera one, the greater the value, the greater the degree of deviation; D p1 If negative, the combustion area is deviated to the side of camera two, the smaller the value, the greater the degree of deviation;

[0078] The images of camera one and camera two close to the 50% of the front wall are extracted and horizontally combined into an image, and the average value of the flame gray value in the previous n seconds is calculated g f3 The images of camera one and camera two close to the 50% of the back wall are extracted and horizontally combined into an image, and the average value of the flame gray value in the previous n seconds is calculated g f4 The formula is as follows:

[0079] (6)

[0080] (7)

[0081] Further calculate the combustion deviation index D p2 The formula is as follows:

[0082] (8)

[0083] D p2 If positive, the combustion area is deviated to the front wall, the greater the value, the greater the degree of deviation; D p2 If negative, the combustion area is deviated to the back wall, the smaller the value, the greater the degree of deviation.

[0084] The calculation process of the flame area rate is as follows:

[0085] Horizontally combine the images of camera one and camera two, that is, horizontally splice the images of the two flame cameras and process as an image, at this time the number of horizontal pixels of the image i = i 1 +i 2 The number of vertical pixels is still j, calculate the average area rate variance of the flame in the previous m seconds, m is 3-10,

[0086] First, the average area rate of the flame of each extracted frame in the previous m seconds S v The calculation formula is as follows:

[0087] (9)

[0088] In the formula, Z is a judgment function, when Z = 1 when ≥ 0 Z = 0 when < 0

[0089] Secondly, the variance of the average area ratio of the flame in the previous m seconds D s The formula is as follows:

[0090] (10)

[0091] In the formula, is the average value of the average area ratio of the flame in the previous m seconds; S v

[0092] Next, the variance of the average high-temperature flame gray value of the image in the previous m seconds is calculated by combining the images of camera one and camera two. First, the average high-temperature flame gray value of each extracted frame in the previous m seconds is calculated The judgment basis for the high-temperature flame is ≥ 0, and the formula is as follows:

[0093] (11)

[0094] In the formula, = 1 when ≥ 0 = 0 when < 0

[0095] Then, the variance of the average gray value of the flame in the previous m seconds is calculated D g The formula is as follows:

[0096] (12)

[0097] In the formula, is the average value of the average gray value of the flame in the previous m seconds;

[0098] Finally, the combustion fluctuation index D b is calculated, and the formula is as follows:

[0099] (13)

[0100] The smaller the combustion fluctuation index D b , the more stable the combustion process is; D b The larger the combustion fluctuation index, the greater the fluctuation in the combustion state.

[0101] In summary, the most ideal state of flame combustion is: D ​​​​​​p1 =0, D p2 =0 and D b =0, the closer the value to the condition, the better the flame burning state.

[0102] The above indexes have guiding significance for combustion adjustment, when D p1 >0, it indicates that the flame is deviated to the side of camera one, the greater the value, the more serious the deviation, the combustion speed of the side with too intense combustion needs to be adjusted to suppress, and the combustion speed of the side with weak combustion needs to be strengthened; on the contrary, if D p1 <0, it indicates that the flame is deviated to the side of camera two, the smaller the value, the more serious the deviation, the combustion speed of both sides needs to be balanced through operation adjustment;

[0103] When D p2 >0, it indicates that the flame is deviated to the front wall, the greater the value, the more serious the deviation, intervention adjustment can be performed according to the need when the index exceeds a certain limit value to prevent the flame from being too close to the front wall; when D p2 <0, it indicates that the flame is deviated to the back wall, the smaller the value, the more serious the deviation, intervention adjustment can be performed according to the need when the index exceeds a certain limit value to prevent the flame from being too close to the back wall;

[0104] According to the definition, the index D b >0, the smaller the value, the smaller the flame fluctuation in the previous m seconds, equal to 0 when the flame has no fluctuation, which is an ideal state, when D b The greater the value, the more intense the flame fluctuation in the previous m seconds, the fluctuation includes the changes of the flame area and the flame intensity, a plurality of regulation threshold values can be set according to the actual operation condition, when D b a corresponding adjustment strategy is taken to suppress the fluctuation of the combustion in the furnace when the index exceeds a certain threshold value, so that the combustion process is stable, such as reducing the combustion section air distribution to suppress the excessive combustion caused by too fast combustion speed increase, to reduce the probability of excessive CO and load change. Example 2

[0105] In order to detect and diagnose the combustion fluctuation and predict CO, two flame cameras 42 are arranged on the left and right side walls in a flue of a 300 t / d incinerator according to the method of the present application, D b index is obtained according to the algorithm of the present application, and a laser gas analyzer is arranged at the outlet of the flue to measure CO, and the operation result is shown as follows: Figures 4-6 b ​The index has good early warning effect on fluctuation of O2 and CO, and has strong guidance on the sharp fluctuation of load; for example Figure 4 As shown in the figure, when O2 has sharp fluctuation in a short time, D b Index will have larger fluctuation; for example Figure 5 As shown in the figure, when CO has instantaneous fluctuation, D b Index will have corresponding fluctuation in advance, and the accuracy is high, the running time in the figure is 11 h; for example Figure 6 As shown in the figure, every time the running personnel adjusts the feed and the grate combustion to adjust the main steam volume, D b Index often has larger fluctuation, and the accuracy is high, the running time in the figure is 11 h.

[0106] The above only describes the preferred embodiments of the present application, and is not used to limit the present application, any modification, equivalent replacement and improvement made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A flame image-based industrial furnace combustion diagnostic method, characterized by, It comprises the following steps: S1, arrangement of camera: a pair of side walls of the boiler flue (3) above the secondary air nozzle (2) of the industrial boiler (1) is respectively provided with a camera device (4), the middle part of the side wall of the boiler flue (3) is provided with a pre-embedded pipe (41), the flame camera (42) of the camera device (4) is arranged in the pre-embedded pipe (41), the pre-embedded pipe (41) is in the shape of a conical enlargement, the cone angle is the same as the angle of view of the flame camera (42), and the upper edge is horizontally arranged; S2, image acquisition and processing: two flame cameras (42) are named camera one and camera two, the video signals of camera one and camera two are connected to the same computer, every certain time, one frame of RGB image is extracted respectively, then the left and right combustion deviation index is calculated according to the flame area rate and the gray value D p1 , the front and back combustion deviation index D p2 and the combustion fluctuation index D b , according to D p1 , D p2 and D b , the size of the flame combustion state is diagnosed, that is, when D p1 is greater than 0, the combustion area is deviated to the side of camera one, the larger the value, the greater the degree of bias burning; when D p1 is less than 0, the combustion area is deviated to the side of camera two, the smaller the value, the greater the degree of bias burning; when D p2 is greater than 0, the combustion area is deviated to the front wall, the larger the value, the greater the degree of bias burning; When D p2 If less than 0, the combustion area is deviated to the rear wall, and the smaller the value, the greater the degree of deviation; D b The smaller, the more stable the combustion process; D b The greater, the greater the fluctuation of the combustion state.

2. A flame image based industrial furnace combustion diagnostic method according to claim 1, characterized in that: The first and second cameras are identical in specifications. The center lines (43) of the viewing angles of the two cameras converge on the center line of the boiler. The height of the convergence point is greater than the height of the secondary air nozzle, and the height difference is greater than that of the second camera. H The range is 0.5~2 m.

3. A flame image based industrial furnace combustion diagnostic method as claimed in claim 1, wherein: The flame camera (42) is a forced air cooling camera or a water cooling camera, and the front end of the flame camera (42) does not exceed the range of the pre-embedded pipe (41).

4. A flame image based industrial furnace combustion diagnostic method according to claim 3, characterized in that: A compressed air blowing pipe is arranged between the flame camera (42) and the pre-embedded pipe (41) to prevent dust and coking at the front end of the pre-embedded pipe (41).

5. A flame image based industrial furnace combustion diagnostic method as claimed in claim 1, wherein: The formula of the RGB image gray value in step S2 is shown in the following formula (1): （1） In the formula, (x,y) is a pixel point coordinate, g(x,y) is a pixel point coordinate is (x,y) a gray value of R(x,y) is a pixel value of a pixel point in the R channel with a coordinate of (x,y) is a pixel value of a pixel point in the G channel with a coordinate of G(x,y) is a pixel value of a pixel point in the B channel with a coordinate of (x,y) is a pixel value of a pixel point in the B channel with a coordinate of B(x, y) is a pixel value of a pixel point in the B channel with a coordinate of (x,y) is a pixel value of a pixel point in the B channel with a coordinate of The formula of the flame gray mean value of the flame camera at different moments is as follows: （2） where Z is a decision function, Z = 1 when ≥ 0; and Z = 0 when < 0. Camera one front n seconds flame gray value mean g f1 The formula is as follows, where n is 5-20: （3） The average value of the first n seconds of the flame gray value of camera two is calculated by the same method g f2 The formula is as follows: （4） A combustion deviation index is calculated based on the average of the flame gray value of the first n seconds of camera one and camera two D p1 The formula is as follows: （5） D p1 If positive, the combustion area is deviated to the side of camera one, the greater the value, the greater the degree of deviation; D p1 If negative, the combustion area is deviated to the side of camera two, the smaller the value, the greater the degree of deviation.

6. A flame image based industrial furnace combustion diagnostic method according to claim 5, characterized in that: The images of camera one and camera two are extracted near 50% of the front wall and horizontally combined into one image, and the average value of the flame gray value in the previous n seconds is calculated g f3 ; The images of the camera one and the camera two are extracted close to the 50% picture of the back wall and horizontally combined into an image, and the average value of the flame gray value in the previous n seconds is calculated g f4 The formula is as follows: （6） （7） Further, a combustion deviation index is calculated D p2 The formula is as follows: （8） D p2 If positive, the combustion zone is biased toward the front wall, the greater the value, the greater the degree of bias; D p2 If negative, the combustion zone is biased toward the rear wall, the smaller the value, the greater the degree of bias.

7. A flame image based industrial furnace combustion diagnostic method according to claim 1, characterized in that, The calculation process of the flame area rate in step S2 is as follows: The images of camera one and camera two are merged horizontally, that is, the images of the two flame cameras are horizontally spliced and then processed as an image, at this time the horizontal pixel number of the image is i=i 1 +i 2 The vertical pixel number is still j, the variance of the average area rate of the flame in the previous m seconds is calculated, and m is recommended to be 3-10, First, the average area ratio of the flame of each extracted frame for m seconds S v The calculation formula is as follows: （9） wherein Z is a judging function, when ≥ 0, Z = 1, when < 0, Z = 0; Secondly, the variance of the average area ratio of the flame in the previous m seconds D s The calculation formula is as follows: （10） In the formula, is the front m Average area ratio of the flame at each time point S v the mean value; Next, the variances of the average gray values of the high-temperature flames of the images of the first camera and the second camera are calculated. First, the average gray values of the high-temperature flames of the extracted frames in the first m seconds are calculated The judgment basis of the high-temperature flame ≥ 0, and the formula is as follows: （11） wherein is a judging function, when ≥ 0, = 1, when < 0, = 0. Then the variance of the average gray value of the flame in the previous m seconds is calculated D g The formula is as follows: （12） In the formula, the average gray value of the flame in the screenshot image at each time point in the previous m seconds the average value of the average gray value of the flame in the screenshot image at each time point in the previous m seconds Finally, the combustion fluctuation index is calculated D b The formula is as follows: （13） Combustion fluctuation indicator D b The smaller, the more stable the combustion process is; D b The larger, the greater the combustion state fluctuation is.

Images