JPS63305228A - Radiation thermometer - Google Patents

Abstract

PURPOSE:To compute temperature at small errors even if actual emissivity spectrums are complicated, by inputting the values of measured reflection data into the emissivity. CONSTITUTION:Two or more reference beams having the different wavelengths are projected on a material to be measured TG from a light source L through a chopper CP. The reference beams are measured with photodetectors S1-S3. The reflected beam from the material to be measured TG are measured with photodetectors S1'-S3'. The outputs of the photodetectors S1-S3 and S1'-S3' are inputted into a reflectivity operating means RC and a temperature operating means TC. In the temperature operating means TC, the radiances from the material to be measured TG are measured for every wavelength. The temperature of the material to be measured TG is computed by using the measured reflection data from the reflectivity operating means RC.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to a radiation thermometer that measures the temperature of an object using light as a medium.

BACKGROUND OF THE INVENTION In such a radiation thermometer, it is necessary to know the emissivity of the object to be measured, but such emissivity is often unknown. Therefore, calculations are performed assuming this emissivity appropriately, but in JP-A-56-130622, two colors (three wavelengths) are used as light, and the emissivity ε is calculated as ε(λ,)=
The calculation is performed on the assumption that ε(λ2), that is, the wavelengths λ1 and λ2 are equal. Also, JP-A-56-13062
In No. 3, three colors (three wavelengths) are used as light, and the emissivity is assumed to be ε(λ) =exp(a6+a+λ). Furthermore, in JP-A No. 57-7529, three colors (three wavelengths) are used, and the temperature T is assumed to be emissivity ε(λ, )=ε(λ2).
1□, Tt3 and T31 are similarly calculated, and T={TI2+TZ3+T:l1)/3 is taken as the true temperature.

On the other hand, JP-A No. 61-30727 uses two colors to measure the reflected light L (λ1) and L (λ2), respectively, and the reflection expressed by the emissivity ε (λ1) and ε (λ2). Compared to the intensity, λ1)/L (λri1(1-ε(λ,))/(1-
The temperature is calculated based on ε(λ2)). That is, if the measured radiance is D(λ) and the measured blackbody radiance at temperature T is Do(λ, T), ε(λ)=D(λ)/Do(
λ, T) and Do(λ, T) can be calculated using Blank's formula and device constants, and in the end L(λ, )/L (λ2)
Since the only unknown quantity in the equation is T, the temperature can be found.

Problems to be Solved by the Invention However, since the above-mentioned JP-A-56-130622 and JP-A-57-7529 assume that ε(λI) = ε(λ2), when the emissivity spectrum is constant, Only in this way can the true temperature be calculated. That is, if the emissivity spectrum is not constant, the assumption of ε(λI)=ε(λ2) itself is incorrect, and the correct temperature cannot be calculated. In any case, since these JP-A-56-130622 and JP-A-57-7529, including JP-A-56-130623, do not include information from reflected light in the emissivity assumption,
It is impossible to determine the true temperature except in certain cases.

On the other hand, in the calculation described in JP-A-61-30727, two different temperatures may be obtained.

This is R(λ1)/used in this calculation.
R(λg)={1−ε (λ, ))/(1−ε (λ2
)) (where R is the reflectance) is true only when the reflected light from the object to be measured is collected at the front or when the object to be measured is a completely diffusing surface.

An object of the present invention is to provide a new and effective radiation thermometer that solves the above problems.

Means for Solving the Problems: An illumination means for irradiating two or more reference lights of different wavelengths toward an object to be measured; a reference light photometer for measuring the reference lights of each of the wavelengths; reflected light photometry means for measuring the reflected light of the reference light of each wavelength; radiance photometry means for measuring the radiance from the object to be measured for each wavelength; =L-k(λ)×L(
λ) (k(λ) is a function of wavelength λ including unknowns, L(λ)
a calculation means for calculating the temperature of the object to be measured from photometric values obtained by each of the photometric means for each of the wavelengths as a measured reflection information value];

In this configuration, the estimated emissivity includes measured reflection information that is closely related to the emissivity, so even if the actual emissivity spectrum is complex, it is possible to calculate a temperature with a small error. Moreover, at that time, the emissivity ε is simply changed to ε=1−.
If we simply assume that L(λ) (k is a constant and L(λ) is the measured reflection information value), the temperature error will be relatively large if k varies depending on the wavelength, but in this configuration, ε= 1-
Since it is assumed that k(λ)×L(λ) and k is a function of wavelength λ, the correct temperature can be calculated.

Embodiment An embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a schematic configuration diagram of a radiation thermometer according to an embodiment of the present invention. In this embodiment, it is assumed that light of three different wavelengths λ1, λ2, and λ is measured. The illumination means includes a light source that generates reference light including light of three different wavelengths λ3, λ2, and λ3, and a chopper CP that intermittents the reference light from the light source, and irradiates the object to be measured TO with intermittent light. do. The chopper CP is realized by rotating a rotating body having a light-transmitting part and a light-blocking part. LS is a lens that condenses the light from the object to be measured TG, and the condensed light is branched into three by the fiber FB. Fiber FBI, FB2, FB
The light transmitted through 3 is received by photodetectors S'l, 312, and S'3, respectively, through optical filters F'+ and F'Z% F'3 that transmit light of wavelengths λ1, λ2, and λ3. be done. Photodetector 511, S'2.51. Then, the radiance B(λ, T) of the synchrotron radiation emitted by the object to be measured TG itself is
and the intensity signal of the reflected light from the light source reflected by the object to be measured TG are superimposed and detected. Photodetector S”.

, S"l, s'2, the radiance B(λ, T) of the synchrotron radiation emitted by the object to be measured TG itself is constant, so it is detected as a DC signal, and the detection signal is D.C.
Amplified by amplifiers 1゛, 2゛, and 3'' respectively, and wavelength λ3
, λ2, λ, the radiance measurements D(λ1), D
(λ2) and D(λ3). These DC amplifiers 1'', 2'', 3'' and the optical filters F'+, F
's, F"x % and photodetector s11, Sl, %
S'3 constitutes the above-mentioned radiation 1 degree photometry means for three wavelength light.

Among the output signals of the photodetectors S''+, S'z, and S'3, the intensity signal of the reflected light when the light from the light source is reflected by the object to be measured TG is the intensity signal of the reflected light that is irradiated from the light rAL to the object to be measured TG. Since the light is modulated into intermittent light by the chopper CP, it is detected as an AC signal via the DC component cutting capacitor, and the detection signal is amplified by AC amplifiers 1'', 2'', and 3'', respectively, and the wavelength Regarding λ6, λ2, and λ, reflected light intensity signals R(λ, ), R(λ2), and R(λ,) are output.

These AC amplifiers 1", 2", 3", the optical filters F+, , p+, , F"1, and the photodetector S',
, S'' gate, and Sl constitute the above-mentioned reflected light photometry means.

By the way, in order to calculate the reflectance, wavelengths λ1, λ8
, λ is required, so photodetectors St , Sz , Ss are provided. photodetector S,
, S,, S, are directed toward the light source via optical filters F,, F2, F, which transmit only light corresponding to wavelengths λ3, λ2, λ3.

The light source contains light of various wavelengths, and the photodetector S1 detects the light intensity of the component of wavelength λ necessary for calculating the reflectance. Further, the photodetectors Sz and Ss similarly detect the light intensity of the components of wavelengths λ2 and λ3. The detection outputs of the photodetectors s, , s, , Sl are amplified by DC amplifiers 1 and 2.3 for light source correction, respectively, and the reference intensity signals P(λI ), P
(λt) and P(λ,) are obtained. AC amplifier 1”, 2
", 3" output signals R(λ1), R(λ2), R(λ,
) and the output signal of the DC amplifier 1.2.3 P(λ, ), P
(λ2), P(λ,) are input to the reflectance calculating means, and the measured reflectance L(λ,) of the measured object TG at the wavelength λ
1)=R(λ1)/P (λ1) and measured reflectance L of the measured object TG at wavelengths λ2 and λ3 (λz>=RCλ2)
/P(λ2), L(1m)=R(λ3)/P(λ, ) is calculated as the calculation output of the 0 reflectance calculation means RC (λ1)
, L(λt), L(λ3) are DC amplifiers 1゛, 2゛, 3
Radiance measurement values D(λ1), D(λ2) obtained from “
Together with D(λ,), it is input to the temperature calculation means TC, and the true temperature is calculated. Hereinafter, a method of calculating the true temperature from these data will be explained.

When the object to be measured is a black body, the energy of light of wavelength λ emitted from the object per unit area, that is, the radiance E(λ,
T) is given by Blank's formula. Now, measure the radiance of the object to be measured at the wavelength λ, Do(λ, T)
If we obtain the measured value of Do(λ, T)=crXE (λ, T) −−−−
--・・・・−(1) Here, α is a constant determined by the measuring device, specifically, a constant determined by calibration measurement. Therefore, the temperature T is calculated by solving the above equation (1). Although the object to be measured is treated as a black body in the above, the actual object to be measured is not a black body, so it is necessary to detect the emissivity of the object and correct it to the temperature required for a black body. If the emissivity of an object that is not a black body is ε(λ, T), then the radiance B(λ, T) of light with wavelength λ at the temperature T of the object is: B(λ, T)=ε(λ, T )xE (λ, T) The radiance measurement value D(λ) of any object is D (λ)−α×ε (
λ, T)xE (λ, T)-----1・----
--.= (2) α is a constant determined by the measuring device.

Next, as shown in Fig. 2, a beam of light 4 with a wavelength λ having a light intensity of 1 inch (λ) is incident on the object to be measured TG, and the reflected light 5 of the light is
is reflected with a light intensity of 1 out (θ, λ) in the direction of various angles θ, then its hemispherical reflectance γ (λ, T
) is expressed by equation (3) shown in FIG.

This hemispherical reflectance T(λ, T) and emissivity ε(λ.

For non-transparent objects, the relationship of T) is ε(λ, T) +r (λ, T)=1・−1−−−
---...-(4). Also, on the one hand, the hemispherical reflectance γ (λ).

Between T) and the measured reflection λ), L(λ)=α'xk'(λ)×γ(λ,T)−...
−・−−m−−−−・−・・−・−(5) There is the following relationship. From equations (4) and (5) above, equation (6) shown in Figure 3 is obtained.
The formula is obtained. Here, α' is a constant determined by the measuring device, similar to α described above, and k'(λ) is a coefficient related to the angular characteristics of the reflected light determined by the surface condition of the object to be measured. Specifically, if a reflected light measuring device with a photometric viewing angle △θ reflects the light at the positions θ=θm and ψ=ψm, then k'
(λ) becomes as shown in equation (7) in FIG. In other words,
k'(λ) is the ratio of the actually measured reflected light to the total reflected light. Therefore, the emissivity ε(λ, T) is expressed as follows.

ε(λ, T)=1−k(λ)×L(λ)−・−(8)
Here, k(λ) is (n-
1) It is a function of λ that includes no more than 1) unknowns.

Two calculation methods for determining temperature based on such assumptions will be described below.

<First calculation method> When k(λ) is considered as a function including (n-1) unknowns, there are a total of n unknowns, including (n-1) unknowns and the temperature T. Therefore, since wavelengths of n colors are used, the number of unknowns and conditional expressions is reduced, and a solution is found. As a specific example, calculation will be shown using three wavelengths and assuming that k(λ) is as follows.

k (λ)=a1+a1 ・λ・−・−−−−−・
・−・−・−(9) From equations (1) and (2), for λ1, ε(λ+, T) D o (λ,)=D(λ1)
From equations (8) and (9), (t-(ao +a, λ1)L(λ1))XDo(
λ1)=D(λ,) a1+a1-λ1=Fl However, Fl = (LD(λ,)/Do(λ,)l/
Similarly for L (λ1)λ2, λ, a6+a+ ・λ2=F2 a1+a1 ・λ,=F.

Eliminating aOlal, (pg-p1) / (λ2-λI) = {F:I -p
g) / (λ, −λ2)・−・−・−・−・−・・
-・-------00) This equation 00) cannot be solved analytically for the temperature T. However, for a certain temperature T, E(λ, T) can be calculated using Blank's formula. Then, Do(λ, T) can be found from equation (1). Since F can be calculated from the measured values D(λ) and L(λ), the left and right sides of equation 00 can be found.

By repeating this calculation for several T's prepared in advance, it is possible to find a T that satisfies the αω expression.

The temperature T at that time can be considered to be the temperature of the object to be measured. Moreover, if the three wavelengths are selected so that λ2={λ1 +λ, )/2 at that time, the 0ω equation becomes the following equation.

2 Fg = Ft +F3・・−・・−−−−−・・
・−・・−・−(11) This 00 formula has a simple form and can be expected to make calculations easier.

Second Calculation Method> The second calculation method is a method in which the temperature is calculated taking into account the error from the beginning. ! II. The accuracy of the calculated temperature is determined by how k(λ) is assumed in equation (8), but no matter what k(λ) is introduced, it will contain errors. there is a possibility. Furthermore, in practice, measured values almost always include measurement errors.

Therefore, temperature calculations are performed using the following concept.

Estimating the function h(T) the radiance ε(λ)XDo(λ.

T) (= (1-k <λ)×L (λ))XD
o(λ.

T)) and the radiance measurement value D(λ) is defined as a function that evaluates the magnitude of the error.

As a specific example, the background radiance Db(λ) is measured using wavelengths of n colors (n≧3), or if Db(λ) is known, the background radiance Db(λ) input from the background radiance input means DI is measured.
λ) to take into account the removal of background radiation effects, and h
(T) is the estimated radiance [ε(λ,T)XDo(λ,T
) + (1-ε(λ,T)) Db(λ)] and the measured radiance value D(λ), defined as the sum of the squared differences for each wavelength, and k
The temperature is calculated assuming that (λ) is the equation (6). This h (
The equations of T) are shown in FIG. 3 as equations 021 and 02.

Since the equation 0 is a quadratic equation for the unknowns a0 and at,
ao(
” a ,, win) and a + (= a tIIlin
) are atl and al when h (T) takes an extreme value. Therefore, 2 h (T) / 2 as = o and 2 h
It can be determined from (T)/2a,=O.

These a, min, a, and win are shown in a vertical matrix as 0 in Figure 3.
Shown as a formula. a expressed by this formula 04). Substitute rRin, a, and min into 02 to find h(T).

From the standpoint of error theory, in order to evaluate the error regardless of the number of wavelengths n, we further calculate the square root least squares error h'' (T) as shown in equation 09 in Figure 3 from h (T) obtained above. Calculate.

The above calculation is repeated by changing T to find T that minimizes h'' (T).Then, the temperature that minimizes this h'' (T) is output as the temperature of the object to be measured.

In reality, (i) various constraints [ε(λ.

(T) ≦ 1, etc.) and within the temperature range determined by the measurement situation, (ii) search for the characteristic h" (T), (iii) determine a certain small number △, The temperature of the object to be measured may be determined by searching for T for which d h'' (T) /d T is smaller than Δ.

FIGS. 4 and 5 show flowcharts of a calculation program for determining the correct answer within the calculation means TC by the above-mentioned second calculation method.

First, in the example shown in FIG. 4, in (11), from the measured values, the measured radiance value D (λi), the measured reflectance L (λi),
Background radiance measurement value Db(λi) and optical constant αi
Load.

Next, in (112), the calculation start temperature of the temperature T to be substituted into equation 0 is set to To, and the number of calculations from the To to the measurement temperature range is set to injection and end, respectively.

For example, if the number of calculations is 500 in the temperature range from 1"C to 500°C, 1@C,
Set 500 to m-end. In the above (#2), a numerical value is also set for h"min, but this numerical value is compared with h" (Tm) obtained as described later, and h"
If (Tm) is smaller, h"(Tw) will take its place, so to speak, by sequentially comparing the data for each calculation result performed by changing the temperature, the number one is finally determined. Since this is the initial value in the process of processing so that a value remains, any relatively large value can be chosen.Next, the energy E (λi, Tm ) to find DIll(λi) (I3). Then, calculate Dffl(λi) and D(λi), L(λi), Db( read in above (#1)).
λi) into the formula shown in (#4), allm and al
Find m. Note that this equation corresponds to the equation 04) in FIG. 3 mentioned above. Next, in the formula shown in (#5), the above a e
l! l, a 1lllls D (λi), Dn+(λ
1), L(λi), and Db(λi) to obtain h(T
Find m). Furthermore, from this h(Ts1) and n, h”
(Tm) is calculated (116), (117), h
Compare the size of ``(Tm) and h''min, and find h''(Tm
) is smaller, h” sin = (#10)
Set h" (Tw) and set the temperature Tenp to Tm, then proceed to (#8). If h" (Tm) is larger, proceed directly to (I8) and check whether m is set to H-end. If they do not match, increment m by 1 (Il1) and return to (!13). If they match, Temp is the desired temperature, so change this to (19)
Output with and end the calculation flow. In this way, the true temperature is obtained.

Next, in the embodiment shown in FIG. 5, a small darkness value △ is set in advance, and h'' (Tm) is determined by two subsequent calculations.
The true temperature is the temperature when the difference between the two becomes less than the threshold value for the first time. Therefore, it is not necessary to perform calculations for all the numbers as in the fourth case, so the calculation time can be shortened. You can expect it. Note that the temperature Tm at which the difference between the two successive h''(Tm) is the smallest can be the true temperature because h(T) is the two of the unknowns a0 and a, as shown in the measurement above. This can be seen from the fact that the following function produces a downwardly convex graph (the white area is the minimum value).

In FIG. 5, first, in (#1), the measured radiance value D (λi), the measured reflectance L (λi), the background radiance value D
b(λi) and optical constant αi.

Subsequently, in (#2), the calculation start temperature is set to To and the darkness value is set to Δ. After that, in (!3), the optical constant αi of each wavelength and the energy E(
Find Da(λi) from λi,'r1) and calculate this Dm(λ
i) and previously read L(λi), D(λi),
Aom and a- are determined by substituting Db(λi) into the equation shown in (114). And in (#5), this a0...
aIII+1 arrangement L(λi), Dm(λi), D(λi
), h(Tm) is calculated from Db(λi). h”(T
m) can be easily calculated from the h(Tm) and n in (#6).

In (#7), it is determined whether the difference between h" (Tm) and h" (T-1) calculated at the previous temperature is smaller than the dark value Δ, and if it is larger, m is incremented ( #9) Then return to (l13) and repeat the flow after (#3). If it is smaller than Δ, then the temperature Tm
Outputs as the true temperature. In addition, instead of Tm, T, -
1 may be output as the true temperature.

Although the present invention has been described above with reference to examples, the present invention is not limited to these examples. It is possible to modify the estimated radiance and the measured radiance D(λ), for example, by changing the evaluation function h (T) to the equation of 0.
h (T) may be any function as long as it evaluates the error between the estimated radiance and the measured radiance value. It may also be the sum of the absolute values of the differences for each wavelength.

In addition, in that case, instead of the 0!9 formula, h"(T)=h(
The error is evaluated as T)/n.

In addition, in the above embodiment, the background radiance Db(λ) was measured or inputted in order to remove the influence of background radiation, but instead, the background radiance is measured or inputted using the representative ambient temperature Ts and the radiation from an object at the temperature Ts for the entire background radiance. contribution rate β(
Db(λ)=β(λ)
It may be calculated as Do(λ, Ts).

Although the embodiment shown in FIG. 1 uses three wavelengths, four or more wavelengths may be used. In that case, only the number of detectors, amplification sections, etc. will change.

Further, in FIG. 1, an optical fiber FB is used to branch the light into three, but a half mirror may be used or three filters may be mechanically inserted and exited alternately. Regarding the light source, it is also possible to eliminate chopper by using pulsed light like a strobe.

Effects of the Invention In the present invention, the measured reflection information value L(λ), which is closely related to the emissivity, is entered into the emissivity, and the emissivity ε is calculated as 1−k by setting its coefficient as a function k(λ) of the wavelength λ including the unknown. (λ)×L
(λ), the effect is that the temperature of the object to be measured can be determined with high accuracy even if the emissivity spectrum of the object to be measured is complex.

[Brief explanation of drawings]

The figures are all related to the radiation thermometer of the present invention, and Figure 1 is a block diagram showing the overall schematic configuration, Figure 2 is an explanatory diagram of reflectance, and Figure 3 is the principle of the present invention. FIG. 4 is a flowchart for arithmetic operations, and FIG. 5 is a flowchart showing different examples of arithmetic operations. L...Light source, TG...Object to be measured, CP...
・Chopper, S Is St, S2, S
'l % S'g, S゛, ... photodetector, ■
, 2.3, l", 2", 3"...AC amplifier, 1
°, 2°, 3゛...DC amplifier, Fl, F2, F
3, Fl, ,F'! ,p+,...optical filter,
TC...Calculation means, DM...Background radiance measurement means, DI...Background radiance input means.

Claims (4)

[Claims] (1) Illumination means for irradiating three or more reference lights with different wavelengths toward the object to be measured, a reference light metering means for measuring the reference lights of each of the wavelengths, and each of the wavelengths reflected by the object to be measured. a reflected light photometer for measuring the reflected light of the reference light; a radiance photometer for measuring the radiance from the object to be measured for each wavelength; (λ
)×L(λ) [k(λ) is a function of wavelength λ including unknowns,
L(λ) is a measured reflection information value]; computing means for calculating the temperature of the object to be measured from the photometric values obtained by the photometric means for each of the wavelengths; (2) When the reference light has three wavelengths, the calculation means a_
0, a_1 is a constant, λ is the wavelength, D(λ) is the radiance measurement value, Do(λ) is the blackbody radiance measurement value found from the blackbody radiance and the measurement device constant, then the above k(λ) k(
Assuming that λ)=a_1+a_1λ, (F_2-F_1)
/(λ_2-λ_1)=(F_3-F_2)/(λ_3
-λ_2) [However, F={1-D(λ)/Do(λ)}/
Claim 1, characterized in that the temperature that satisfies [L(λ)] is calculated by using several prepared temperature values.
The radiation thermometer described in section. (3) The calculation means calculates the temperature that minimizes the difference between the estimated radiance obtained using the emissivity and the measured radiance obtained by the radiance photometry means as the true temperature. A radiation thermometer according to claim 1. (4) The calculation means calculates a temperature that minimizes the difference between the estimated radiance plus the background radiance and the measured radiance as the true temperature. The radiation thermometer described in Section.