JPH0257253B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention allows light emitted from an object to be measured to pass through multiple types of filters having different wavelengths at which the transmittance reaches a maximum value (hereinafter referred to as "transmission wavelength"). This relates to a radiation thermometer that determines the temperature of an object to be measured based on transmitted light.

Radiation thermometers have the advantage of being able to measure the temperature of an object to be measured without contact, and are therefore used to measure the temperature of heated steel.

First, the principle of such a radiation thermometer will be explained.

The spectral radiant volatility L〓 _,T emitted by an object depends on the temperature T of the object, the transmission wavelength λ of the filter through which the emitted light from the object is transmitted, and the emissivity ε of the object, and is based on the Wien formula and Planck's It is given by the following formula using the formula:

L〓 _,T = εC ₁ /λ ⁵ e ^-C2/ 〓〓 [W/sr/m ³ ] (1) Here, C ₁ = 1.19096×10 ^-16 [W・m ² ] C ₂ = 0.014388 [m・K] If the transmission wavelengths of the filter through which the emitted light from the object is transmitted are λ ₁ and λ ₂ , and the emissivity of the object at the transmission wavelengths λ ₁ and λ ₂ are ε ₁ and ε ₂ , then each transmission wavelength λ ₁ , The spectral radiances L ₁ and L 2 at λ ₂ and emissivity ε ₁ and ε ₂ are: L ₁ = ε ₁ C ₁ /λ ⁵ / _{1 e -C2/} 〓 ₁ 〓 (2) L ₂ = ε ₃ C ₂ /λ ⁵ / ₂ e _-C2/ 〓 ₂ 〓 (3). Dividing equation (2) by equation (3), L ₁ /L ₂ =ε ₁ /ε ₂ (λ ₂ /λ ₁ ) ⁵ e ^-C2/ 〓 ² 〓 ^(1/ 〓 ^1
-1/ 〓 ²⁾ (4), and when calculating the temperature T from this formula, T=-C ₂ (1/λ ₁ -1/λ ₂ )/ln{L ₁ /L ₂ (λ ₁
/λ ₂ ) ⁵ }+l _o ε ₁ /ε ₂ (5). This formula is the formula for determining the temperature by measuring the spectral radiances L ₁ and L ₂ .

In general, it is difficult to determine the emissivity ε ₁ and ε ₂ because they vary depending on the wavelength and temperature as well as the surface oxides and deposits of the object. From this, the emissivity ε ₁ and ε ₂ are the transmission wavelength λ ₁ ,
It is assumed that ε ₁ = ε ₂ without changing depending on λ ₂ . In this way, equation (5) becomes the following equation.

T=−C ₂ (1/λ ₁ −1/λ ₁ )/ _lo {1/λ ₁ (1
/λ ₁ ) ² }(6) Among radiation thermometers, the principle formula for a two-color thermometer that measures temperature using light of two different wavelengths is
It becomes equation (6).

Next, an example of a conventional configuration of a radiation thermometer to which such a principle is applied is shown in FIG. Here, the case of a two-color thermometer among radiation thermometers is shown.

In FIG. 1, 10 is a light converging section, 20 is a transmission wavelength conversion means such as a filter wheel, 30 is a detection section, and 40 is a calculation section.

In the light converging section 10 , light emitted from the object to be measured is converged by an objective lens 11 .

Filters 21 ₁ and 21 ₂ having different transmission wavelengths λ ₁ and λ ₂ are attached to the filter wheel 20 and rotated by a motor 22 . Depending on the rotational position of the filter wheel 20, the light converged by the objective lens 11 passes through the filter 21 ₁
and 21 ₂ are transmitted alternately.

In the detection unit 30, 31 is a detector that detects the light that has passed through the filter 21, converts it into an electrical signal, and outputs it. 32 is an automatic gain amplifier (hereinafter referred to as
(AGC amplifier), which amplifies the electrical signal from the detector 31. 33 is a signal separator;
A signal corresponding to the light transmitted through the filter 21 ₁ and a signal corresponding to the light transmitted through the filter 21 ₂ are separated into electrical signals V ₁ and V ₂ . These electrical signals V ₁ , V ₂
is sent to the calculation section 40. 34 is a synchronization signal generator, which is arranged close to the filter wheel 20 and transmits light through filters 21 ₁ ,
21 ₂ is sent to the signal separator 33.

In the calculation unit 40, the electric signal V ₁ is sent to the zero clamp 41. Further, the electrical signal V ₂ is sent to a zero clamp 41 and a comparison amplifier 42 . The zero clamp 41 outputs a signal when a signal corresponding to a temperature above a certain temperature is sent. The output signal of the zero clamp 41 is amplified by an amplifier 43 and becomes an output signal of the thermometer. The comparison amplifier 42 is supplied with the electrical signal V ₂ and the output V ₀ of the reference voltage source 44 . Then, the output of the comparison amplifier 42 is
The signal is supplied to the AGC amplifier 32, whereby the gain of the AGC amplifier 42 is adjusted so that the electric signal _V1 becomes a constant value _V0 .

The electrical signals V ₁ and V ₂ of such a radiation thermometer are
It is given by the following formula.

V ₁ = L (λ ₁ , T) Rv ₁ Gμ ₁ (7) V ₂ = L (λ ₂ , T) Rv ₁ Gμ ₂ (8) When formula (7) is divided by formula (8), V ₁ / V ₂ =L(λ ₁ ,T)/L(λ ₂ ,T)Rv ₁ /R
v ₂ μ ₁ / μ ₂ (9) If the output signal of the thermometer is V _x , since V _x is proportional to the electrical signal V ₁ , V _x = KV ₁ (10) where, K: constant becomes. In addition, since the electrical signal V ₂ is controlled to the output V ₀ of the reference voltage source 44 so that it always has a constant value, V ₀ = KV ₂ , and from this formula, K = V ₀ /V ₂ (11) becomes. Substituting equation (11) into equation (10) gives V _x = V ₁ /V ₂ V ₀ , and substituting equation (9) into this equation gives V _x = L(λ ₁ , T)/L(λ ₂ , T) Rv ₁ / Rv ₂ μ ₁
/μ ₂ V ₀ (12). In Equation (12), Rv ₁ /Rv ₂ , μ ₁ /μ ₂ and V ₀ can be regarded as constant values specific to the thermometer. Further, L(λ ₁ , T)/L(λ ₂ , T) corresponds to equation (4).

When an output signal V _x is obtained from such a radiation thermometer, L ₁ /L ₂ in equation (4) can be obtained. However,
Emissivity ε ₁ and ε ₂ are difficult to determine. Therefore, assuming that the emissivities ε ₁ and ε ₂ are equal, the measured temperature of the object to be measured is determined using equation (6) in which ε ₁ and ε ₂ are eliminated. The calculation for determining the measured temperature will be explained below.

As the filters 21 ₁ and 21 ₂ , the transmission wavelength is λ ₁ =
0.75×10 ^-6 [m], λ ₂ = 0.85×10 ^-6 [m], the emissivity of the object to be measured is set to ε ₁ = 0.7, ε ₂ = 0.8, and the temperature of the object to be measured is from 1000 [K] to 1600
When the temperature is changed to [K], the error between the true temperature of the object to be measured and the temperature measured by the radiation temperature becomes as shown in the graph of FIG.

In the graph of FIG. 2, the horizontal axis is the true temperature T, and the vertical axis is the error temperature and error rate. In the graph of FIG. 2, the solid line is the difference between the true temperature and the temperature measured by the radiation thermometer, that is, the error temperature, and the broken line is the ratio of the error temperature to the true temperature (hereinafter referred to as error rate).

As is clear from the graph in Figure 2, the emissivity ε ₁
If there is only a slight difference of 0.1 _between
It is about 140 [℃] (error rate about 9%).

For this reason, there has been a problem in that large errors occur in the measured temperature due to fluctuations in the emissivity of the object to be measured.

The present invention was made in order to eliminate the above-mentioned problems, and an object of the present invention is to provide a radiation thermometer in which the error in the measured temperature is small even if the emissivity of the object to be measured changes. .

FIG. 3 is a diagram showing the configuration of an embodiment of the radiation thermometer according to the present invention, and here the case of a two-color thermometer is shown. In Figure 3, the first
Items that are the same as those in the figure are given the same symbols.

In FIG. 3, 50 is a detection section, and 60 is a calculation means.

The output signal of the detector 31 is amplified by an amplifier 51 and sent to a signal separation means 52. The signal separation means 52 determines the spectral radiances L 1 and L 2 of the light that has passed through the filter 21 ₁ and the light that has passed through the filter _{21 2} , respectively, and calculates these spectral radiances L ₁ and L ₂ .
Separates and outputs the signal according to L ₂ . The synchronizing signal generator 34 is arranged close to the filter wheel 20, and sends signals corresponding to the filters 21 ₁ and 21 ₂ used for transmitting light to the signal separating means 5.
Send to 2.

The calculation means 60 receives signals corresponding to the spectral radiances L ₁ and L ₂ from the signal separation means 52, and performs calculations to determine the measured temperature of the object according to a predetermined calculation method.

Generally, the emissivity ε is determined by the physical properties of the object, the surface condition,
It becomes a function of temperature, wavelength, etc. However, with multicolor radiation thermometers, such as two-color thermometers, the measured temperature is determined using light transmitted through multiple types of filters at almost the same time, using the same measurement system. The surface condition and temperature are considered to be the same, and the dependence of emissivity on the transmitted wavelength is a major error factor in the measured temperature.

For this reason, the emissivity of an object to be measured when temperature is measured using two types of filters with different transmission wavelengths is approximated by a wavelength function having one unknown quantity. Considering ease of calculation, emissivity ε is approximated by the following function.

ε=e〓 ^f( 〓 ⁾ (13) Here, α: unknown number The specific determination of the approximation formula for emissivity will be explained below.

When the temperature measured by the radiation thermometer is T _n , the spectral radiance L ₁ and L ₂ are calculated as follows using equation (13) and Vienna's formula: L ₁ = e〓 ^f( 〓 ¹⁾・C ₁ /λ ⁵ / ₁ e ^-C2/ 〓 ¹ 〓 ^m (14) L ₂ = e〓 ^f( 〓 ²⁾・C ₁ /λ ⁵ / ₂ e ^-C2/ 〓 ² 〓 ^m (15) With α and T _n as unknowns, Solving the simultaneous equations of equations (14) and (15), we get T _n =C ₂ (1/f(λ ₁ )λ ₁ -1/f(λ ₂ )λ ₂ )/1/
f(λ ₂ )l _o L ₂ /λ ⁵ / ₂ -1/f(λ ₁ )l _o L ₁ λ ⁵ / ₁ /C ₁
(16) α=f( _λ1 )/f( _λ2 ) _λ2 −f( _λ1
)λ ₁ (λ ₂ lnL ₂ λ ⁵ / ₂ /C ₁ −λ ₁ l _o L ₁ λ ⁵ / ₁ /C ₁ ) (17
) Here, the error between the true temperature T and the measured temperature T _n is determined using equation (16).

If the true emissivity of the measured object is ε ₁ and ε ₂ , then (2)
Substituting equations and equations (3) into equation (16), T _n =C ₂ T (1/f(λ ₂ )λ ₁ -1/f(λ ₂ )λ ₂ )/
T(1/f(λ ₂ ) _lo ε ₂ −1/f(λ ₁ ) _lo ε ₁ )+C ₂ (
1/f(λ ₁ )λ ₁ −1/f(λ ₂ )λ ₂ ) (18) Here, the error rate is T _n −T/T=ΔT/T (19) However, ΔT=T _n − It becomes T. Substituting equation (18) into equation (19), ΔT/T=-T(1/f(λ ₂ ) lo _{ε 2} -1/f(λ ₁
)l _o ε ₁ )/T(1/f(λ ₂ )l _o ε ₂ −1/f(λ ₁ )l _o
ε ₁ )+C ₂ (1/f(λ ₁ )λ ₁ −1/f(λ ₂ )λ ₂ )(2
0) becomes. Here, A=T1/f(λ ₂ ) _lo ε ₂ −1/f(λ ₁ ) _lo ε ₁ (21) B=C ₂ 1/f(λ ₁ )λ ₁ −1/f(λ ₂ ) λ ₂ (22) Then, equation (20) becomes ΔT/T=-A/A+B (23). In equations (21) and (22), A is ε ₁ ,
It is a function of ε ₂ and is a cause of error in the measured voltage.
Therefore, if B>A, the error will become smaller. For this reason, we will consider A/B.

A/B=T(1/f(λ ₂ ) _lo ε ₂ −1/f(λ ₁
)I _o ε ₁ )/C ₂ (1/f(λ ₁ )λ ₁ −1/f(λ ₂ )λ ₂
) = λ ₁ λ ₂ T/C ₂ f (λ ₁ ) I _o ε ₂ − f (λ ₂ ) l _o ε ₁ /f
(λ ₂ )λ ₂ −f(λ ₁ )λ ₁ (24). In equation (24), consider f(λ) as the following function.

f(λ)=λ ⁿ (25) where n: integer Substituting equation (25) into equation (24), A/B=λ ₁ λ ₂ T/C ₂ λ ₁ ⁿ l _o ε ₂ −λ ₂ ⁿ l _o ε ₁ /λ ₁ ⁿ⁺¹ −
λ ₂ ⁿ⁺¹ (26). Here, by setting λ ₂ = λ ₁ + Δλ, substituting this equation into equation (26), and rearranging by first-order approximation, A/B = λ ₁ λ ₂ /C ₂ Tλ ₁ ⁿ l _o ε ₂ − λ ₁ ⁿ (1+Δ
λ/λ ₁ ) ⁿ _lo ε ₁ /λ ₁ ⁿ⁺¹ −λ ₁ ⁿ⁺¹ (1+Δλ/λ ₁ ) ^n+
1 ≒λ ₁ λ ₂ /C ₂ Tl _o ε ₂ −(1+nΔλ/λ ₁ )l _o ε ₁ /−
(n+1)Δλ(27).

Here, λ ₁ = 0.75×10 ^-6 [m], λ ₂ = 0.85×10 ^-6
[m], T = 1600 [K], C ₂ = 0.014388 [m・K], and for emissivity, ε ₁ = 0.4, ε ₂ = 0.95, ε ₁ =
Consider three cases: 0.95, ε ₂ =0.4 and ε ₁ =ε ₂ =0.4.

If the above values are substituted into equation (27) and graphed, the result will be as shown in FIG. 4.

In the graph of FIG. 4, the vertical axis represents A/B, and the horizontal axis represents the integer n in equation (25). Fourth
In the graphs in the figure, curves a, b, and c are graphs when the emissivities are ε ₁ =0.4 and ε ₂ =0.95, ε ₁ =0.95 and ε ₂ =0.4, and ε ₁ =ε ₂ =0.4. .

From the graph in FIG. 4, as n increases, graphs a and c approach each other, so that A/B is less influenced by fluctuations in the emissivity ratio ε ₁ /ε ₂ . On the other hand, in graph b, as n increases,
As A/B increases, the error in the measured temperature increases. If the error in the measured temperature becomes large in graph b, that is, when the emissivity is ε ₁ =ε ₂ , the advantage of the radiation thermometer will be lost. From such a thing,
n=2 is set so that the error in the measurement signal when the emissivity is ε ₁ =ε ₂ is small, and A/B is less affected by fluctuations in the emissivity ratio ε ₁ /ε ₂ .

Hereinafter, the error in the measured temperature obtained using the approximation formula for emissivity ε will be explained.

Let Lλ ₁ and Lλ ₂ be the black body radiance of light transmitted through filters with transmission wavelengths λ ₁ and λ _2. From Vienna's formula, Lλ ₁ = C ₁ /λ ⁵ / ₁ e ^-C1/ 〓 ¹ 〓 (28) Lλ ₂ = C ₁ / λ ⁵ / ₂ e ^-C1/ 〓 ² 〓 (29) L ₁ = ε ₁ Lλ ₁ (30) L ₂ = ε ₂ Lλ ₂ (31).

Here, from equations (13) and (25) and n=2,
Emissivity ε is approximated by the following function.

ε=e〓〓 ² (32) The relationship between the spectral radiance L ₁ , L ₂ and the measurement temperature T _n is
It will look like this:

L ₁ = e〓〓・C ₁ /λ ⁵ / ₁ e ^-C2/ 〓 ₁ 〓 ^m (33) L ₂ =e〓〓・C ₁ /λ ⁵ / ₂ e ^-C2/ 〓 ¹ 〓 ^m (34) The simultaneous equations of equations (33) and (34) are used to measure the temperature.
Solving for T _n , T _n = C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂ / λ ₁ ) / {λ ² ₂ l _o L ₁ λ ⁵
／ ₁ ／C ₁ −λ ² ₁ l _o L ₂ λ ⁵ ／ ₂ ／C ₁ }
(35) becomes. Substituting equations (35) into equations (30) and (31), T _n = C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂ / λ ₁ )
/{λ ² ₂ l _o λ ⁵ / ₁ /C ₁ ε ₁ Lλ ₁ −λ ² ₁ l _o λ ⁵ / ₂ /C ₁ ε ₂ L
λ ₂ }(36). Substituting equations (28) and (29) into equation (36), T _n = C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂ / λ ₁ ) / {λ
² ₂ I _o (ε ₁ e ^-C2/ 〓 ^1T ) − λ ² ₁ l _o (ε ₂ e ^-C2/ 〓 ^2T )} = C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂ / λ ₁ )/{Tl
_o (ε ₁ λ ² / ₂ / ε ₂ λ ² / ₁ ) + C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂
/λ ₁ )}(37). Here, the error rate ΔT/T is calculated using equation (19) as follows: ΔT/T= _−Tlo (ε ₁ λ ² / ₂ /ε ₂ ² / ₁ )/ _Tlo (ε ₁ λ ²
/ ₂ /ε ₂ λλ ² / ₁ ) + C ₂ (λ ² / ₁ / λ ₂ - λ ² / ₁ / λ ₁ )
(38) becomes. Here, if we replace ε ₁ = γε ₂ , (38)
The formula is ΔT/T=-TI _o γ〓〓ε〓〓〓〓〓／Tl _o γ〓〓〓ε ₂ 〓〓〓
〓+C ₂ (λ ² / ₁ / λ ₂ − λ ² / ₂ / λ ₁ ) (39).

Here, the transmission wavelength of the filter λ ₁ = 0.75×10 ^-6
[m], λ ₂ =0.85×10 ^-6 [m], one emissivity ε ₂ =0.
Four
As, the temperature T of the object to be measured is 1000, 1300 and
Set in 3 ways: 1600 [K]. Under these conditions, the relationship between the error rate ΔT/T in equation (39) and the emissivity ratio γ is graphed as shown in FIG.

In the graph of FIG. 5, the error rate ΔT/T is plotted on the vertical axis, and the emissivity ratio γ is plotted on the horizontal axis. Also,
Curves d, e, and t are graphs when the temperature T is set to 1600, 1300, and 1000 [K].

In the conventional calculation for calculating the measured temperature, that is, the calculation for calculating the measured temperature using equation (6) assuming that the emissivity is ε ₁ = ε ₂ , the relationship between the error rate ΔT/T and the emissivity ratio γ is shown in a graph. and as shown in Figure 6.

In the graph of FIG. 6, the vertical axis, horizontal axis, and transmission wavelengths λ ₁ and λ ₂ are the same as the graph of FIG. 5. In addition, curves g, h, k and l have temperature T
It is a graph when it is set to 1600, 1400, 1200 and 1000 [K].

Based on the spectral radiances L ₁ and L ₂ determined by the detection unit 50 as described above, the calculation means 60 performs the calculation to determine the measured temperature T _n using equation (35).

According to the radiation thermometer having such a configuration, the measured temperature of the object to be measured is determined by approximating the emissivity ε to an appropriate function of the transmission wavelength λ. Therefore, as shown in the comparison between the graph of FIG. 5 and the graph of FIG. 6, even if the emissivity of the object to be measured changes, the error occurring in the measured temperature can be reduced.

In addition, in the example, a case was explained in which two types of filters with different transmission wavelengths were used and the emissivity ε was approximated by a function having one unknown. However, the present invention is not limited to this. The emissivity ε may be approximated by a function number having N-1 unknowns by using a filter having a value of (an integer greater than or equal to).

Further, in the embodiment, a case has been described in which a filter wheel is used as the transmission wavelength conversion means 20, but other than this, for example, filters with different transmission wavelengths are fixedly arranged, and a movable prism is used as the transmission wavelength conversion means 20. Alternatively, the traveling direction of the light may be changed depending on the direction of the filter, and the light may be made to enter each filter alternately.

Further, in the embodiment, a case has been described in which one detector 31 is provided, but a plurality of other detectors 31 may be provided.

Further, in the embodiment, a case where the radiation thermometer is a two-color thermometer has been described, but the radiation thermometer may be a multi-color thermometer that uses light of a plurality of different wavelengths.

As described above, according to the present invention, it is possible to provide a radiation thermometer with small errors in measured temperature even if the emissivity of the object to be measured changes.

[Brief explanation of the drawing]

Fig. 1 is a diagram showing a conventional configuration example of a radiation thermometer, Fig. 2 is a graph showing the temperature characteristics of the error of the radiation thermometer shown in Fig. 1, and Fig. 3 is a diagram showing a radiation thermometer according to the present invention. A diagram showing the configuration of one embodiment, FIG.
The figure is a graph showing the relationship between the error rate of the measured temperature determined by the radiation thermometer in Figure 3 and the ratio of emissivity.
The figure is a graph showing the relationship between the error rate of measured temperature determined by a conventional radiation thermometer and the ratio of emissivity. 21, 21 ₁ , 21 ₂ ... filter, 60 ... calculation means.

Claims (1)

[Scope of Claims] 1. Light emitted from the object to be measured is transmitted through N filters (N is an integer of 2 or more) having different transmission wavelengths, and the light emitted from the object to be measured is determined based on the light that has passed through each filter. In a radiation thermometer that measures the temperature of an object, the emissivity of the object to be measured corresponding to the transmission wavelength of each of the filters is approximated by a function of the transmission wavelength having N-1 unknowns, and these approximated radiation 1. A radiation thermometer comprising a calculation means for determining the temperature of an object to be measured from a predetermined calculation formula using a ratio. 2. The radiation thermometer according to claim 1, wherein the calculation formula uses Wien's formula and Planck's formula.