JPS60169015A - Burning condition diagnosing method - Google Patents

Abstract

PURPOSE:To monitor and diagnose continuously the burning condition by a method wherein at least two wave length emission quantities or luminance distributions are detected from a flame under a boiler operation condition, then the correlation between said two factors is obtained. CONSTITUTION:An emission of a flame 3 detected by an optical fiber 5 which is protected by a cooling device 4 is distributed to plural passages by utilizing a half miller 6. Each specific wave length light is extracted from the distributed light by utilizing band-pass-filter 7, 7' having respectively different permeable wave length. The specific wave length lights are converted to electric signals with light receiving elements 8, 8', then inputted to integrator 9, 9'. Thus, different wave length light volumes Wlambda1, Wlambda2 are obtained. A correlation value P1 obtained from Wlambda1, Wlambda2 with a calculating device 11 is outputted to a comparator 12, then compared and verified with the burning condition previously stored in a memory device. Thereby, the burning condition is diagnosed precisely, accordingly, the accomodate process can be indicated.

Description

Detailed Description of the Invention [Field of Application of the Invention] The present invention detects at least two emission spectra during combustion, and determines the combustion state from the correlation value between the detected spectra of an index representing the amount of the spectra. The present invention relates to a method for diagnosing a combustion state.

[Background of the invention]

Conventionally, methods for determining the combustion state from the flame in a boiler furnace include monitoring using an ITV camera through a viewing window installed on the wall facing the burner nozzle, or visual observation through a viewing window installed on the furnace wall. There are methods such as inspection and temperature control using heat transfer tubes. However, these methods require skill and are a burden to the operator, as the operator visually observes and makes decisions based on experience and intuition. The only automatic monitoring device used was a flame detector to determine burner ignition and extinguishment for furnace protection.

Regarding the monitoring of the combustion state during boiler operation, there has been no quantified reference value in the past, so there has been a problem in that the decision-making becomes a burden on the operators and prompt action cannot be taken.

In addition, when monitoring the combustion status using an IT camera, an ITV camera attached to a viewing window on the furnace wall is used to monitor the burner nozzle on the opposite wall, so when the rated operation is approached, the flame becomes swirly. However, it was extremely difficult to accurately judge the combustion state, and the operator had no choice but to rely on the experience and intuition of experienced operators.

On the other hand, a typical method for extracting an object image is to use several types of filters for different wavelength regions of light and take photos for each wavelength region.
There is a method of sampling and quantizing the film, inputting it into a computer, and extracting and recognizing a specific object using the characteristics of the spectral composition of the specific object.

However, such conventional methods have the following drawbacks.

Since a specific object is recognized only by the shape of its spectral distribution, other noise objects with the same spectral distribution are also extracted. As a method to eliminate such drawbacks, a specific object extraction device has recently been proposed that extracts a specific object while minimizing the inclusion of noise objects by performing analog calculations between several spectral components ( Special Publication No. 58-39102).

Such methods and devices can be very effective when extracting the shape of an object by exposing it to light or other radiation.

However, when extracting the shape and features of a burning flame,
The wavelength band that is radiated is not only the wavelength of light. Flame has a thermal emission spectral distribution and a spectral distribution caused by the emission of chemical components contained in fuel or produced during combustion.

In the thermal spectral distribution, the absolute amount of light emitted tends to increase as the wavelength increases from the ultraviolet to the visible to infrared regions. Furthermore, the emission spectrum distribution of chemical components appears independently of the thermal emission spectrum distribution, and the emission spectra of chemical components having the same spectral distribution cannot be noise. This is a phenomenon peculiar to flame during combustion.

The present invention focuses on physical phenomena unique to combustion (for example, the emission spectrum distribution of chemical components in the near-ultraviolet region), and proposes a method for well and accurately understanding the combustion state.

[Purpose of the invention]

The object of the present invention is to prevent at least 2
The object of the present invention is to provide a method for continuously monitoring and diagnosing the combustion state by detecting the luminescence amount or luminance distribution of two wavelengths and from the correlation thereof.

[Summary of the invention]

The present invention detects the amount of luminescence or luminance distribution detected in at least two wavelength bands from a flame during boiler operation, and focuses on the fact that the difference in the correlation between the detected values is related to the difference in the combustion state. The present invention is characterized in that the combustion state corresponding to the correlation value is stored in advance, and the combustion state corresponding to the actually measured correlation value is diagnosed as the combustion state of the flame.

[Embodiments of the invention]

An embodiment of the present invention is shown in FIG. In FIG. 1(a), 1 indicates a boiler furnace, 2 indicates a burner port, and 3 indicates a flame. The light emitted from the flame 3 captured by the optical fiber 5 protected by the cooling device 4 is branched into a plurality of light beams (in this embodiment, two beams) using a half mirror 6, and the branched light is divided into bandpasses with different passing wavelengths. A filter 7.7' is used to extract light of a specific wavelength. Light of a specific wavelength is transmitted to light receiving elements 8, 8.
' is converted into an electrical signal and input to an integrator 9.9'. 10 is an integrator 9.9 which transmits the integration timing signal 15.
is a timing signal generator that provides

The obtained light amount of different wavelengths "i + WJ□" is given here by t: time, i: combustion condition, and λ: wavelength, and the arithmetic unit 11 calculates the correlation value PL of these light amounts, that is, the correlation value at the combustion condition cover. For example, it can be expressed by equation (2).

PI=Wλ1,1/WJ2-...Old...(2)W
The correlation 1tP+ determined by the arithmetic unit 11 from JI,+, Wλ2,1 is output to the comparator 12. Comparison device 12
Then, the combustion state is compared with the combustion state previously stored in the storage device. The storage device 13 stores diagnostic condition setting values W for the combustion state.
It stores LM t+W+, ut2+ PLMr, and diagnoses the combustion state by comparing and checking as shown in Table 1, for example.

Here, WJI, l is the band/bass ratio in combustion state 1.
The time integral value of the signal ux(t) passed through the filter (wavelength λI), Wλ3. , is the band in combustion state i.
The signal L'5(
t) time 41゛1 minute value, both times 0 to 1. It shows the integral value of . WLMTI is the diagnostic reference value for λl, % WLMτ2 is the diagnostic reference value for λ2,
PLMT indicates the diagnostic standard values for the ratio of WJI, turtle and Wλ8,1, respectively. and 0≦Wλ(
(During WLMTI, the diagnosis result is A in a normal combustion state. The same applies to WJ2.P below, and when the relationship shown in Table 1 is satisfied, it is determined to be A or B, respectively.

Table 2 shows examples of combustion state determination based on the determination results for Wλi and Wλ2°P obtained in Table 1, and examples of corresponding operations.

Note that the combustion condition 1 mentioned above is a combustion condition determined by, for example, the size of the load or the type of fuel. 1st
Diagnosis using the table is performed by the comparative diagnostic device 12, and diagnosis using the @2 table is performed by the diagnostic device 14.

22 is a storage device that stores the state determination and corresponding operations in Table 2.

Figure 1(b) shows u h (t), (C) shows uz(t
) (7) Give an example t, -tc-t, -, le.

(d) A signal 15 which is a timing signal from the H timing signal generator 10 and determines the integration period of, for example, oxtl.
It shows. The diagnostic device 14 then instructs the operator to perform corresponding operation instructions 16 based on the diagnostic results.

FIG. 1(e) shows an example of the characteristics of filters 7 and 7', both of which have transmittances of 46 to 47%. As the wavelength becomes shorter, the transmittance decreases. Figure 1(e) shows carbon radicals (C2 = emission wavelength 516.5■), hydrocarbon radicals (
This is an example of CH: emission wavelength 430.9 m). The emission of light from C2 radicals and CH radicals has wavelengths whose generation largely depends on the state of combustion.

In the example in Figure 1, us(t), uz(t) o o
In this example, integral values of ~t and K are used, but their average values may also be used. It is also possible to calculate the correlation without using the integral value of the detected signal.

In this case, the integrator 9 and timing signal generator 1° become unnecessary.

FIG. 2 shows another embodiment of the present invention. Using band pass filters 7 and 7' with a plurality of passing wavelengths (two in this embodiment), images of specific wavelengths are transferred to the image fiber 2.
3, passed through half mirror 6 and filter 7, received by ITV camera 21°21', and sent the image signal to A/
It is input to the image storage device 17 via the D converter 19°19'.

20 is an A/D conversion timing signal.

The image data for each wavelength (average image or instantaneous image over a certain period of time) taken into the image storage device 17 is inputted into the electronic computer 18, and the set value for the luminescence brightness (or temperature) stored in advance in the image data is inputted into the computer 18. TLM? Processing to clear the following to 0 (hereinafter referred to as clipping processing)
Where is the tl. The concept of clipping processing is shown in Figure 3 (a
) (b). FIG. 3(a) is a flame image captured at wavelength λ1. Observe the luminance distribution of the t-t' cross section of this image (Fig. 3(b)), and find the set value TLM? Perform clipping processing with .

Figure 4 (a) and (b) show the TL extracted by clipping processing.
M? The above portion is shown (the broken line is the outline portion before clipping processing).

The obtained wavelength λ! , the area of the extracted part at λ2 AJ s
m AJ2 , for example, using equation (3), the correlation value P
Calculate +.

P+=Aλ1 / Aλ2 ・Old...Ichikawa...(3)
The determined AJI*AJ*#P amount is shown in Table 1 of the above example.

It is used for diagnosing combustion conditions in the same manner as in Table 2. However, in Table 1, Wλ1 is AJI, WJ! is AI
used as

Although the embodiment shown in FIG. 1 or FIG. 2 has been described in the case of two wavelengths, it is also possible to use a method of correlating other wavelengths or a plurality of other wavelengths that are related to the combustion state.

Furthermore, as an example of correlation, the example of the ratio of time integral values of signals of different wavelengths is shown in the embodiment of FIG. 1, but the index representing correlation is not limited to the ratio.

For example, a difference signal may be taken.

In addition, when determining the correlation between three wavelengths and four wavelengths, the signal ratio of each wavelength may be taken, or, for example, the ratio of the signal of each wavelength to the total signal of three wavelengths or four wavelengths may be taken.

〔Effect of the invention〕

By implementing the present invention, it is possible to properly diagnose the combustion state during boiler operation without relying on the operator's experience or intuition, and to instruct countermeasures.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing one embodiment of the present invention, Fig. 2 is a diagram showing another embodiment of the invention, Fig. 3 is a diagram showing the concept of clipping processing, and Fig. 4 is a diagram showing portions extracted by wavelength. FIG. DESCRIPTION OF SYMBOLS 1... Boiler furnace, 2... Burner mouth, 3... Flame, 4... Cooling pipe, 5... Optical fiber (or image fiber), 6... Half mirror, 7...・Band bus filter, 8... Light receiving device, 9... Integrator, 10... Timing signal generator, 11... Correlation calculation device, 12... Comparison device, 13... Comparison Condition storage device, 14... Judgment device, 15... Integral timing signal, 16... Judgment result output signal, 17... Image storage device, 18... Electronic word calculator, 19... A
/D converter, 20...A/D conversion timing signal, 2
1...-Imaging device, 22... Judgment condition storage device. Agent: Akio Takahashi, Patent Attorney Figure 1, Figure 2, Figure 3 (Phantom brightness -

Claims (1)

[Claims] 1. In a method for diagnosing the combustion state in a furnace, the amount of emission spectrum at a plurality of wavelengths generated during combustion is detected, and the detected amount of emission spectrum is combined with a predetermined amount for each of the plurality of wavelengths. A method for diagnosing a combustion state, comprising diagnosing a combustion state of the furnace based on a magnitude relationship with a reference value for diagnosing a combustion state. 2. In claim 1, an index representing the correlation between the detected emission spectrum amounts for each of the plurality of wavelengths is calculated, and a predetermined combustion state diagnosis reference value for the index representing the correlation is calculated. A method for diagnosing a combustion state, comprising diagnosing a combustion state of the furnace based on a magnitude relationship between . 3. In the scope of claim 1, the amount of emission spectrum detected for each of the plurality of wavelengths and the index representing the relative value of the amount of emission spectrum detected between the plurality of wavelengths are each predetermined. A method for diagnosing a combustion state, comprising diagnosing a combustion state of the furnace based on a combination of magnitude relationships with a combustion state diagnosis reference value. 4. In claim 1, a predetermined time integral value of the detected emission spectrum amount is calculated, and the above-described value is determined based on the magnitude relationship between the calculated integral value and a combustion state diagnosis reference value. A combustion state diagnosis method characterized by diagnosing the combustion state of a furnace. 5. The method for diagnosing a combustion state according to claim 1, wherein the plurality of wavelengths are a carbon radical wavelength and a hydrocarbon radical wavelength.