JPH0637974B2 - Combustion diagnostic device - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a combustion diagnosis device for a pulverized coal combustion process,
In particular, the present invention relates to a combustion diagnosis device that extracts a feature amount in a flame structure, calculates a flame feature parameter effective for flame combustion diagnosis, and diagnoses / monitors a combustion state.

[Background of the Invention]

As a means for monitoring the combustion state of a burner in a furnace, for example, as shown in Japanese Patent Laid-Open No. 56-23630, a flame is captured by an infrared television camera, and the flame in a standard state is detected based on the temperature distribution of the flame. A method of diagnosing the combustion state as compared with a pattern is known.

However, since the flame shape changes irregularly due to the turbulent diffusion of the combustion gas in the furnace, there is a problem that the combustion state cannot be accurately diagnosed from the shape of the temperature distribution of the flame.

[Object of the Invention]

An object of the present invention is to extract a feature amount in a flame structure with a flame image processing device from a flame image in a coal dust combustion captured by an imaging camera, calculate a feature parameter from the extracted feature amount,
(EN) A combustion diagnosis device capable of quantitatively grasping a stable combustion state of a flame, diagnosing and monitoring the combustion state of the flame, and performing stable combustion.

[Outline of Invention]

The present inventors have paid attention to the flame structure in the pulverized coal combustion system aiming at low NOX and high efficiency combustion. That is, the pulverized coal combustion is separated into carbonization gas (volatile matter) and coke (solid carbon) and ignited and burned, and the flame structure is shown in Fig. 2 (a). The pulverized coal combustion process will be described with reference to FIG.

The pulverized coal mixed with primary air and ejected from the burner into the furnace is rapidly heated by the radiant heat from the high-temperature furnace wall and flame, and a portion of the pulverized coal particles cracks and collapses into finer particles. Becomes At the same time, the volatile components are carbonized, and the carbonized gas of about 500 times the volume of the fine coal particles is rapidly released.

The dry-distilled gas reacts with the air present between the particles surrounding the remaining coke-like sludge that released the dry-distilled gas, and diffuses and burns (forming a volatile matter combustion region). The threshed coke grains become pumice-like. As buoyancy increases, solid carbon burns on the surface. (Formation of reduction / oxidation region) The flame structure is such that, in the oxidation region 1 and the oxidation region 1 where volatile components and solid components are burned in the outer flame in the state of excess oxygen, the combustion of volatile components with a high burning rate is near the burner. A volatile matter combustion region 2 is formed, and the internal flame is constituted by a reduction region 3 formed by combustion of solid matter in a state of lacking oxygen.

FIG. 2 (b) shows the luminance distribution with respect to the distance from the burner in the reducing region 3 and the oxidizing region 1 in the flame, and the luminance distribution (BB ') in the reducing region 3 shows the burning luminance of the solid content. On the other hand, the brightness distribution (A-A ') of the oxidized region 1 shows a brightness distribution obtained by adding the combustion brightness 4 of the solid content and the combustion brightness 5 of the volatile content. The combustion area of volatile matter is the distance from the burner X /
The fact that it is formed in the section of D = 0 to 1 is shown in FIG.
It is also verified from (c).

FIG. 2 (c) shows the distribution of generated gas in the oxidation region,
Carbon dioxide (CO ₂ ) sharply increases and oxygen (O ₂ ) sharply decreases in the section of the distance X / D = 0 to 1 from the burner. This indicates that the volatile matter is being burned in this section because the burning rate of the volatile matter is higher than the burning rate of the solid matter.

From the above, the luminance distribution of the flame is observed, the observed luminance distribution is separated into the burning luminance of the solid content and the burning luminance of the volatile content, and the volatile component combustion region is extracted. This processing is performed by the flame image processing device. The combustion diagnosis device of the present invention quantifies the extracted volatile matter combustion region by the method described below to diagnose and monitor stable combustion. The reason for paying attention to the volatile matter combustion region is as follows.

Since the combustion region is formed near the burner, the influence of the gas in the furnace is small and the shape is stable. In addition, it has high sensitivity to changes in burner characteristics and process amount.

In combustion, it is an important factor in the generation of NOX and unburned components.

The combustion diagnosis device calculates the characteristic parameters from the volatile matter combustion region extracted by the flame image processing device, and stores the characteristic parameters of the limit points required for stable combustion in the storage unit in advance, and the comparison unit stabilizes them. When the characteristic parameter of the combustion limit point and the characteristic parameter of the measured flame are compared, and when the characteristic parameter of the measured flame exceeds the stable combustion limit point, the abnormality factor estimation unit refers to the abnormality factor data and estimates the combustion system abnormality factor. Instructing the combustion control device to check for abnormal factors,
Display the cause of anomalies and warning messages on the CRT display.

Example of Invention

FIG. 3 shows an embodiment of the present invention. 6 is a boiler furnace for pulverized coal combustion, 7 is a burner for pulverized coal combustion, 8 is an image fiber, 9 is an imaging camera, 10 is a flame image input device, 1
1 is an A / D converter, 12 is a memory, 13 is a flame image processing device, 14 is a combustion diagnosis device, 15 is a CRT display device, 16 is a CRT graphic display device,
Reference numeral 17 is a combustion control device. Boiler furnace 6 for pulverized coal combustion
The flame when the pulverized coal is burned by the pulverized coal burning burner 7 is captured by the image fiber 8 installed in the furnace with fire resistance, and converted into an electric signal by the imaging camera 9. The flame image data converted into an electric signal is taken into the image input device 10, A / D converted by the A / D converter 11, and then stored in the memory 12. The flame image data stored in the memory is taken in by the flame image processing device 13, and the volatile matter combustion region in the flame structure is extracted by image processing.

The flame image data processed by the flame image processing device 13 is sent to the combustion diagnosis device 14, and the stable combustion of the flame is diagnosed and monitored by the processing described below. The processing result is CR
It is displayed graphically on the T-graphic display device 16. When the abnormal combustion of the flame occurs, the abnormal factor is estimated, the combustion control device 17 is instructed to check the abnormal factor, and the CRT display device 15 displays the abnormal factor and the alarm message.

FIG. 4 shows a configuration diagram of the combustion diagnosis device 14. The flame image processing device 13 extracts a volatile combustion region as a feature amount from the flame image data, and the image data of the extraction result is the combustion diagnosis device 1.
4 is sent to the characteristic parameter calculator 18. The characteristic parameter calculator 18 calculates a characteristic parameter (characteristic parameter) in order to quantitatively express the characteristic amount in the image data. In the present embodiment, the distance between the centers of gravity, the distance between the centers of gravity and the burners, the thickness, the average brightness, and the fluctuation are calculated as the characteristic parameters. The processing procedure is first
Shown in the figure. First, the barycentric position and average brightness of the volatile matter combustion region are calculated 23. The position of the center of gravity and the average brightness are Here, R (x, y): brightness at x, y coordinates δ: when R (x, y)> 0 δ = 1, when R (x, y) ≦ 0 δ = 0 x, y: x, y-coordinate In this calculation, the image data is divided into two in the Y-axis direction along the burner center line, and the barycentric positions G ₁ and G ₂ and the average luminances R ₁ and R ₂ are calculated for each of the divided image data. As a result, the center-of-gravity distance L _G , the center-of-gravity / burner distance L _B , and the average brightness of the feature parameters are Here, x ₁ , y ₁ : x, y coordinates of the center of gravity G ₁ x ₂ , y ₂ : x, y coordinates of the center of gravity G ₂ x ₀ , y ₀ : x, y coordinates x _B , y _{B of the} center of gravity center : X and y coordinates of burner position R ₁ , R ₂ : average brightness of the bisected area S ₁ , S ₂ : area of the bisected area Next, in order to facilitate the subsequent processing, the image data is binarized. Chemical treatment 24 is performed. Image data IP
Let (x, y) be the brightness of each pixel with respect to the threshold t, And The threshold is the average brightness for the next thickness calculation.
Set to R ₁ and R ₂ .

The thickness of the characteristic parameter is calculated by the following equation (thickness calculation 25) for a region whose brightness is equal to or higher than the average brightness R ₁ , R _{2 with} respect to the volatile matter combustion region.

Where D ₁ and D ₂ are the thicknesses in the bisected region: Average thicknesses S ₁ ′ and S ₂ ′ are the areas of the target region l _X1 , l _X2 : The maximum length of the target region in the X-axis direction Next, the contour extraction processing 26 and the skeleton extraction processing 27 are processing for feature-extracting and displaying the contour of the volatile matter combustion region when the image data is graphically displayed on the CRT graphic display device 16. The contour extraction process is the binarization process 2
The edge processing is used to detect the contour of the image data that has been processed. Further, the skeleton extraction processing 25 performs skeleton extraction in order to emphasize the contour extracted by the contour extraction processing 24. For edge processing and skeleton extraction (skeleton) processing, general-purpose image processing software packages are commercially available and can be applied.

Finally, the fluctuation of the characteristic parameter is calculated (not shown). Fluctuation is a parameter indicating the degree of flame fluctuation, and various defining equations can be considered. The following equation is shown as an example.

Here, f: fluctuation P: measurement parameter P ₀ : previous measurement parameter As the measurement parameter, the distance between the center of gravity of the characteristic parameter,
The fluctuation f is calculated for each of the distance between the center of gravity and the burner, the thickness, and the average brightness, and the maximum value of the fluctuations is taken as the fluctuation.

In FIG. 4, the flame characteristic parameter calculated by the characteristic parameter calculation unit 18 is compared by the comparison unit 19 with the characteristic parameter of the stable combustion limit point stored in the storage unit 20. Also, a CRT graphic display device 16
In addition, the volatile matter combustion region feature-extracted from the measured flame and its center of gravity position are graphically displayed together with the stable combustion limit line, and the combustion state is visually evaluated. As a result of comparison by the comparison unit 19, if the characteristic parameter of the measured flame exceeds the stable combustion limit point, it is diagnosed as an abnormal flame and the abnormality factor estimation unit 21
In reference to the abnormal factor data stored in the storage unit 22, the combustion system abnormal factor is estimated, the combustion control device 17 is instructed to check the abnormal factor, and the CRT display device 15 displays the abnormal factor and the alarm message. .

Next, the abnormality factor estimation method will be described. First, the flame feature parameters are manipulated variables (pulverized coal flow rate, primary air flow rate,
A relational expression expressing the relationship with the secondary air flow rate and the tertiary air flow rate) is created.

The production method is as follows: (1) One of the manipulated variables (for example, pulverized coal flow rate) is changed stepwise at regular intervals from the plant equilibrium state. (Fig. 5
(Refer to (a)) The time width shall be such that the flame is sufficiently stable. The flame at that time is measured, and the characteristic parameters are calculated as shown in FIG. 5 (b). From the test result of FIG. 5 (b), the relationship between the change width of the manipulated variable and the change width of the characteristic parameter is approximated by a linear or non-linear function.

y ′ ₁ = a ₁₁ x ′ ₁ or y ′ ₁ = f (x ′ ₁ ) (14) where x ′ ₁ is the change amount of the manipulated variable y ′ ₁ is the change amount of the characteristic parameter a ₁₁ : Proportional coefficient f: Non-linear function The above operation is performed for all operation amounts.

(2) From the relationship (Equation (14)) obtained in (1), the relationship between the characteristic parameter and the manipulated variable can be expressed as a matrix: Y = A · X or Y = f (X) (15) where X: manipulated variable vector (X = (x ₁ , X ₂ , ...)) Y: feature parameter vector (Y = (y ₁ , y ₂ , ...)) A: coefficient matrix f: nonlinear function matrix The above procedure for obtaining the relational expression with the manipulated variable is performed in each required load band during the plant test operation. After that, the relational expressions are corrected at the time of periodic inspection of the plant.

The coefficient data A (combustor structure data) in the relational expression thus obtained is stored in the abnormality factor data storage unit (see FIG.
Store in 2).

Next, the operation of the abnormality factor estimation unit (FIG. 4) will be described.

(1) When an abnormal flame occurs, the operation amount (factor item) is calculated based on the abnormal factor data (stored in the abnormal factor data storage unit 22) obtained above from the characteristic parameter (abnormal item) at that time. presume. In other words, from equation (15) (for linear case) (Non-linear or when A is not normal) age, Convergence calculation so that Ask for. here, : Estimated manipulated variable Y _E : Characteristic parameter of abnormal fire K: Acceleration coefficient (2) Estimated manipulated variable And the measured manipulated variable X _R are compared.

The factor estimation method is divided into the following two depending on the magnitude of the deviation E.

When | E | <ε (ε: allowable deviation vector) The abnormality factor has a problem in the operation amount instructed by the combustion control device. The operation amount X _R is checked, and the operation amount deviating greatly from the equilibrium point is estimated to be the main factor.

When | E |> ε First, check the actual manipulated variable and confirm that it does not deviate significantly from the equilibrium point. If there is a deviation, it is one of the factors.

Actual operation amount X _R and estimated operation amount The fact that there is a difference between the two is presumed to be abnormal in the combustion device (burner, piping, measurement system, etc.). Therefore, the cause of abnormality is estimated by the following procedure.

1) A factor item for (1) and an evaluation function (range: 0 to 1) representing the occurrence probability of the factor are set in advance. (Fifth
(See Figures (c) and (d)) This evaluation function is set based on the know-how of the designer / operator, and is gradually corrected throughout the operation to complete the function.

2) Calculate the evaluation function of each factor item based on the deviation E.
An example is shown in Table 1.

It is estimated that the main factor is the factor with which the evaluation function takes the maximum value. In the example of Table 1, it is estimated that the clogging of the pulverized coal piping system is the main factor.

Thus, the inference in two stages is executed by the abnormality factor estimation unit 21.

According to the embodiment of the present invention, from the volatile matter combustion region in the flame extracted by the image processing device, the characteristic parameter is calculated, and the flame is detected by quantitatively monitoring whether or not it is located within the stable combustion limit line. It is possible to detect abnormal combustion in the early stage and take countermeasures.

〔The invention's effect〕

According to the present invention, a stable combustion state during pulverized coal combustion can be quantitatively grasped, diagnosed, and monitored, so that abnormal combustion can be detected early.
Measures can be taken, and safe and stable combustion can be realized.

Further, according to the present invention, since the combustion state of pulverized coal can be accurately diagnosed, low NOX and high efficiency combustion becomes possible.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a flow chart of a characteristic parameter calculation process according to an embodiment of the present invention, FIG. 2 is a structural diagram of a pulverized coal combustion flame to which the present invention is applied, and FIG. FIG. 4 is a system configuration diagram of an embodiment of the present invention, FIG. 4 is a configuration diagram of a combustion diagnosis device, and FIG. 5 is an explanatory diagram of an abnormality factor estimating method. 13 ... Flame image processing device, 18 ... Characteristic parameter calculation unit, 19 ... Comparison unit.

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Yoshihiro Shimada 52-1 Omika-cho, Hitachi City, Ibaraki Prefecture Hitachi Ltd. Omika Plant (72) Inventor Nobuo Kurihara 3-chome, Hitachi City, Ibaraki Prefecture No. 1 In Hitachi Research Laboratory, Hitachi Ltd. (72) Inventor Mitsuyo Nishikawa 3-1-1, Saiwaicho, Hitachi City, Ibaraki In Hitachi Research Laboratory, Hitachi Ltd. (72) Inventor Yoshio Sato Hitachi City, Ibaraki Prefecture 3-1-1, Machi, Hitachi Research Laboratory, Hitachi, Ltd. (72) Inventor, Atsumi Watanabe 3-1-1, Saiwaicho, Hitachi City, Ibaraki Hitachi Research Laboratory, Hitachi, Ltd. (56) References 56-23630 (JP, A) JP 59-69623 (JP, A) JP 55-128724 (JP, A) JP 57-117083 (JP, A)

Claims (5)

[Claims] 1. A combustion diagnosis device for diagnosing and monitoring the combustion state of pulverized coal, comprising: input means for inputting image data indicating a volatile matter combustion region in a flame image captured by an imaging camera; A characteristic parameter calculation unit for calculating a characteristic parameter quantitatively representing a volatile matter combustion region, and a characteristic parameter storage unit for storing the characteristic parameter of a stable combustion limit point indicating a limit point necessary for stable combustion in advance, A means for comparing the characteristic parameter of the stable combustion limit point with the characteristic parameter of the flame image; and when the characteristic parameter of the flame image exceeds the characteristic parameter of the stable combustion limit point, an abnormality factor estimation unit The abnormal factor data storage unit that stores the abnormal factor data is referred to, and the abnormal factor estimation unit that estimates the abnormal factor is provided. A combustion diagnosis device, characterized in that it outputs data indicating the above-mentioned abnormality factor to a display means. 2. The combustion diagnosis device according to claim 1, wherein the distance between the centers of gravity of the volatile matter combustion regions is used as the characteristic parameter. 3. The combustion diagnostic device according to claim 1, wherein a barycenter-burner distance indicating a distance between a barycenter of the volatile matter combustion region and a burner tip is used as the characteristic parameter. 4. The combustion diagnostic device according to claim 1, wherein the characteristic parameter is a thickness of the volatile matter combustion region. 5. The combustion diagnosis device according to claim 1, wherein an average brightness of the volatile matter combustion region is used as the characteristic parameter.