JPH076838B2 - Flame detection method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a flame detection method for detecting a flame of a burner used in thermal power generation equipment or industrial furnace equipment.

[Prior Art] A large number of burners are provided in a furnace in a thermal power generation facility, etc., and the driver visually confirms the ignition of the burners. However, when there are a large number of burners, even when looking through a small window in the furnace wall, the visual flicker discrimination limit is 16 Hz, and there is no ability to discriminate the characteristics of the flame.
It is often indistinguishable whether it is the self-flame of the burner being observed, the opposite flame, or the adjacent flame. Therefore, each burner is visually inspected for flame by ignition / extinguishing operation, and if there is no problem, all burners are ignited at once, and visual confirmation is not performed at that time. However, such a method not only requires a great deal of labor, time and useless fuel, but also cannot confirm the presence or absence of the flame in the final state.

Therefore, conventionally, there has been proposed a flame detector that measures the presence / absence of flame by measuring visible / near infrared light from the flame. In this visible / near-infrared flame detector, flicker high-frequency components in the primary combustion zone of the flame are extracted, and the high-frequency components in the primary combustion zone are used as an index for determining the presence or absence of the flame.

[Problems to be solved by the invention] However, when a large number of burners are present, such as in a boiler, when a flame detector is installed and the inside of the furnace is viewed, only the arrangement of viewing the flame tip from the burner is allowed. Often
In the field of view, the black skirt (unburnt area from the burner to the primary combustion zone) of the self-flame of the burner, the opposing flame, and the adjacent flame also enter. However, if the light from the opposing flame enters or there is a black skirt, the detection signal amount of the flicker in the primary combustion zone, which is an index, changes. Moreover, the influences of the opposing flame and the adjacent flame vary from moment to moment depending on the load condition of the boiler, and the condition of occurrence of the black skirt also changes periodically by cleaning the burner performed several times a month. In this way, since the discrimination index is unstable, reliable flame detection is difficult, and the method of adopting the determination standard must be adjusted each time, which is troublesome. Further, if the adjacent flame flickers in the visual field, the statistical data of the high frequency component for determining the determination standard will remarkably vary, so that the determination accuracy will be deteriorated.

The object of the present invention is to eliminate the above-mentioned problems of the prior art, eliminate the influence of the opposing flame, the adjacent flame, the black skirt of the self-flame, and the size of the self-flame, which causes the false detection, and determine whether or not there is a flame. It is to provide a flame detection method capable of surely detecting a fire.

[Means for Solving Problems] The flame detection method of the present invention detects high-frequency flicker components in the primary combustion zone of a flame, and detects low-frequency flicker components in the tip of the flame to detect these frequencies. The presence or absence of flame is determined based on the ratio of the components.

The basic configuration of an example of an apparatus for carrying out the present invention is shown by the solid line in FIG.

The self-flame 2 of the burner 1 is composed of a primary combustion zone 3 and a tip portion 4, and a condenser 5 of the flame detector is located near the burner 1 with its optical axis 6 extending through the primary combustion zone 3 to the tip portion 4. Installed as you see. The light condensed by the condenser 5 is guided to the photoelectric converter 8 through the optical fiber 7 and converted into the photocurrent 9. The photocurrent 9 is guided to a high pass filter 10 and a low pass filter 12. The high-frequency component 11 of the photocurrent 9 that has passed through the high-pass filter 10 and the low-frequency component 13 of the photocurrent 9 that has passed through the low-pass filter 12 are divided by the divider 14, and the ratio 15 of the high / low frequency components is calculated by the microcomputer.
Entered in 16. The microcomputer 16 determines the presence or absence of the self-flame 2 based on this ratio 15.

[Operation] As shown in FIG. 2, in front of the self-flame 2 of the burner 1, there is an opposite flame 24 of the opposite burner 23, and on the upper and lower sides or left and right of the self-flame 2, there are adjacent flames 26 of the adjacent burners 25. . Therefore, the photocurrent 9
Is a combination of all the lights from the self-flame 2, the opposing flame 24, and the adjacent flame 26 that are incident on the condenser 5.

The characteristics of the self-flame and the counter-flame are shown in FIG. The horizontal axis represents frequency and the vertical axis represents flicker rate. The flicker rate is the flicker intensity at each frequency divided by the average brightness. The line 29 shows the characteristics of the primary combustion zone, and the flicker rate is large in the high frequency region, and for example, even at 150 Hz, there is a flicker of about 1/4 of that at 0 Hz. As described above, the flicker rate is high even at high frequencies because the combustion material has a molecular level in the primary combustion zone, and the production and disappearance of the material are changed drastically while maintaining a delicate balance.

On the other hand, the line 30 is a characteristic of the flame tip, and the flicker rate rapidly decreases from about 25 Hz, and becomes almost 0 at 150 Hz. This is because the combustion substance at the tip of the flame is carbon particles formed by reducing carbon dioxide gas, has a large mass, and does not cause a rapid state change.

Line 31 is the characteristic of the opposing flame, and is mostly limited to the low frequency range. This is because the field of view of the condenser 5 is widened on the side of the opposing flame far from the condenser, and local changes cannot be detected, so that the average brightness is detected and chemical reaction is also unsatisfactory. This is to see the tip of the vigorous counterflame.

The line 32 is the total characteristic of the self-flame and the counter-flame combined, and changes sharply in the vicinity of the vertical axis, but shows a constant tendency a little away from it. The abrupt change part near the vertical axis is the average brightness, which changes variously depending on the size of the self-flame, the presence or absence of the opposing flame, the state of the black skirt, etc., and cannot be used as a judgment index for flame detection. However, if the ratio of the two frequency components is taken from the fixed tendency portion, that is, the region excluding the DC signal of the photocurrent, the detection interference factor can be eliminated as described later.

FIG. 4 shows a frame of the light source 33 including self-flame in front of the condenser 5 and a schematic view of the situation of light emitted from the frame. The light source 33 is divided into four layers or regions of an opposing flame region 37, a self-flame tip region 34, a primary combustion zone 35, and a black skirt region 36, and the emission of the opposing flame region 37 is L _C , the self-flame tip region. The emission in the region 34 is the DC component L _O , the flicker amount L _L at the tip, and the primary combustion zone
Emission of primary combustion zone Flicker amount L _H , black skirt region
Let τ be the transmittance of 36. The amount of light entering the condenser 5 is composed of a condensing amount 40 of the opposing flame, a condensing amount 38 of the tip portion, and a condensing amount 39 of the primary combustion zone, and the total L of these condensing amounts is converted into the photocurrent 9.

When the photoelectric converter 8 that is proportional to the total amount of collected light is used and its photocurrent I9 is extracted by the high-pass filter 10, the primary combustion zone flicker amount I _H1 11 of the high frequency component is I _H1 = τL _H ( 2) Similarly, the flame tip flicker amount I _L1 13 of the low frequency component extracted by the low-pass filter 12 is I _L1 = τL _L (3) Both I _H1 and I _L1 are affected by the black skirt τ
And the influence of the size of the flame L _H , L _L , but not the influence of the counter-flame with little flicker. The ratio (Bricker ratio) R15 is Therefore, there is a similar relationship in that the effect τ of the black skirt disappears and that L _H and L _L also increase and decrease according to the size of the flame. Therefore, by taking the proportional L _H / L _L , the flame size The effect of disappears. Therefore, the influence of the black skirt, the size of the flame, and the opposing flame is removed from the flicker ratio R.

Since L _H and L _L are much smaller than L _O , if the photoelectric converter 8 has a logarithmic characteristic, the photocurrent I9 of the following equation (5) can be obtained approximately.

Therefore, the primary combustion zone flicker amount I _H2 11 is And the flicker amount I 13 of the flame tip is Becomes

Both I _H2 and I _L2 are affected by the opposing flame L _C , but since they are divided by L _O , the influence of the flame size is small. In this case as well, the flicker ratio R15 is the same as equation (4),
The influence of the oncoming flame L _C disappears.

In the conventional method in which the signal processing is performed by one bandpass filter matched to the frequency of the primary combustion zone, the processing up to the equation (2) or the equation (6) is the limit, and in the equation (2), the black skirt It can be seen that the influence τ and the influence L _{C of the} opposing flame remain in Eq. (6).

Next, a method of determining the presence / absence of flame from the flicker ratio R15 in the microcomputer 16 will be described with reference to FIG.
FIG. 6 shows a momentary change in the distribution of the value of the flicker ratio R.

Distribution 46 is the one without auto-flame and distribution 48,4
9 is when there is self-flame. The solid line 50 shows the average value of the distribution, and the dotted line 47 is the average value of the distribution 46 when there is no auto-flame.

When the self-flame is ignited at time 51, distribution 46 to distribution 48
The distribution of the flicker ratio R clearly changes. Even after that, the distribution of the flicker ratio R continued to change to the distribution 49 due to the fluctuation of the load of the furnace and the ignition / extinction of the opposing flame.
Probability calculation is performed as to whether the new sample value of the flicker ratio R is close to the distribution 46 or close to the distribution 49, and it is determined whether or not there is a flame at present.

Embodiments Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows an example of an apparatus for carrying out the present invention. In the figure,
In the basic configuration shown by the solid line, as described above, the flicker of light emission of the primary combustion zone 3 is detected by the high-pass filter 10 set to, for example, 150 Hz bandpass, and the flicker of light emission of the flame tip 4 is changed to, for example, 25 Hz bandpass. Set low-pass filter 12
The flicker ratio R15, which is less affected by opposing flames and black skirts, is obtained by detecting the ratio of both flicker, and the accuracy of judgment to detect the presence or absence of flame from the transition of statistical distribution as shown in Fig. 6. It uses a high detection method.
Hereinafter, an auxiliary configuration (shown by a broken line) for performing more reliable flame detection will be described.

In combustion in which air is excessively supplied, a black skirt does not occur, and a highly transparent flame is formed. Therefore, only the opposing flame is a detection obstacle. Such flames include those of an ignition tonner for igniting the main burner. On the other hand, in the combustion in which air is supplied inadequately, the flame greatly expands and the transparency is poor, and although the opposing flame is not a problem, a black skirt is likely to occur. Such a flame is the flame of the main burner.

As shown in FIG. 1, the flicker amount 11 in the primary combustion zone is sent to the microcomputer 16 via the branch path 22, and the photocurrent 9 is sent to the branch path.
It is input to the microcomputer 16 via 17 respectively. Microcomputer
In 16, the I _{H1 of the} equation (2) and the I _H2 of the equation (6) are calculated from these inputs and used as an auxiliary discrimination index for the presence or absence of the flame. That is,
A more accurate judgment can be made by discriminating and analyzing the flame with three indexes in addition to the flicker ratio R of the expression (4). For example, in combustion with excess air, the emission L _{C of the} opposing flame
Becomes a problem and I _H2 is easily affected, and in the combustion with insufficient air, the transmittance τ of the black skirt becomes a problem and I _H1 is easily affected.

When the adjacent flame 26 enters the field of view of the condenser 5, the photocurrent 9
Is disturbed. When there is self-flame 2, this disturbance is small,
It is large when there is no self-flame 2. FIG. 5 shows a change in the brightness L when there is no self-flame. The change in the long-term average value of the brightness L is as shown by a line 19, and the instantaneous value of the brightness L is as shown by a line 17. Hatching when the instantaneous value falls below the average value
Shown at 43. The characteristic of the adjacent flame is that the period indicated by the hatching 43 is about 50% of the total period, and by detecting this, it can be determined that the self-flame 2 is not present.

In FIG. 1, the smoothing filter 18 obtains the long-time average value 19 of the photocurrent 9, and the comparator 20 compares the instantaneous value of the photocurrent 9 with the long-time average value 19. When the value falls below 19, the signal (interruption signal 21) is input to the microcomputer 16. The microcomputer 16 calculates the ratio of the generation period of the interruption signal 21 to the total measurement time and compares it with a predetermined threshold value. Although it is possible to confirm that there is no self-flame, it is not possible to confirm that there is self-flame, and it is necessary to make a determination based on the flicker ratio R described above.

The distributions 46, 48, 49 in FIG. 6 are statistics at every moment and are also criteria for judgment. Therefore, the data used for statistics must be of good quality. When the photocurrent 9 obtained from the path 17 is abnormally decreased or the adjacent flame is confirmed by the interruption signal 21, it is clear that there is no self-flame, and the sampling data in such a case has a large variation. , When included in the statistics, the accuracy of the criteria for making delicate judgments decreases. Therefore, the sampling data in such cases should not be included in the statistics.

[Effects of the Invention] The present invention has the following effects.

(1) Since the presence / absence of a flame is determined based on the ratio of the high-frequency flicker component of the primary combustion zone and the low-frequency flicker component of the flame tip, the black skirt of the opposing flame, adjacent flame, and self-flame It is possible to eliminate the influence of erroneous detection such as the size of flame, and it is possible to reliably detect the presence or absence of flame. Therefore, it is possible to fully automate the confirmation of ignition, solve the problem caused by visual confirmation of ignition by the driver, and improve the reliability of combustion monitoring for safe operation in a boiler or an industrial furnace.

(2) Further, since the defect of the conventional flame detector is eliminated by improving the signal processing of the light collected by the collector, the device for carrying out the method of the present invention is easy to manufacture, and the existing device is easy to manufacture. The invention can be realized with only minor modifications to the flame detector.

[Brief description of drawings]

FIG. 1 is a configuration diagram showing an example of an apparatus for carrying out the flame detection method of the present invention, and FIG. 2 is a diagram showing the state of flames visible in the field of view of a condenser, FIG. 3, FIG. 4, FIG. FIG. 5 is a diagram showing the optical characteristics of the flame, and FIG. 6 is an explanatory diagram for explaining the method for determining the presence or absence of the flame. In the figure, 1 is a burner, 2 is a self-flame, 3 is a primary combustion zone, 4 is a tip portion, 5 is a condenser, 6 is a photoelectric converter, 10 is a high-pass filter, 12 is a low-pass filter, and 14 is division. Vessel, 16 is a microcomputer, 24
Is an opposing flame and 26 is an adjacent flame.

Claims (3)

[Claims] 1. A high-frequency flicker component of the primary combustion zone of a flame is detected, a low-frequency flicker component of the tip of the flame is detected, and the presence or absence of the flame is determined based on the ratio of these frequency components. A flame detection method characterized in that 2. A high frequency of a signal proportional to the logarithm of the light quantity from the primary combustion zone and the tip of the flame as an auxiliary discrimination index of the presence or absence of the flame when determining the presence or absence of the flame based on the ratio of the frequency components. The flame detection method according to claim 1, wherein a component and a high frequency component of the signal proportional to the light amount are used. 3. When determining the presence or absence of a flame based on the ratio of the frequency components, the instantaneous value of the amount of light from the primary combustion zone and the tip of the flame is a long-term change of the amount of light as an auxiliary determination index for the presence or absence of flame. The flame detection method according to claim 1 or 2, wherein a ratio of time below a value is used.