CN101907492A - Object emission spectrum-based temperature measurement method - Google Patents

Abstract

The invention discloses an object emission spectrum-based temperature measurement method, which comprises the following steps of: (1) acquiring continuous spectrums emitted by an object to be measured at a certain temperature by using a spectrograph and filtering and smoothing the continuous spectrums; and (2) acquiring an ln (E(lambda, T)lambda5) related variation curve according to the spectrum data in a spectrum width range of 50 to 150 nanometers, solving the slope A of the variation curve by adopting a linear fitting method, and acquiring the temperature T of the object to be measured according to an equation, wherein E(lambda, T) is the spectral intensity, lambda is the radiation wavelength, T is the temperature of the object, and c2=1.43879*10<-2>m/K is the second radiation constant. The method improves the accuracy and universality of temperature measurement.

Description

A kind of thermometry based on object emission spectrum

(1) technical field

The invention belongs to the temperature measurement technology field, relate to a kind of thermometry of object, be particularly useful for the measurement of high temp objects temperature in laser manufacturing and the manufacture process again.

(2) background technology

Temperature is as laser manufacturing and important parameter in the manufacture process again thereof, in the molten bath, all there is important role the control of crystal grain and the aspects such as mechanical property of material.At present the metering system of temperature is divided into two kinds of contact temperature-measuring and contactless temperature-measurings in the laser fabrication process.Thermopair is as the main mode of present contact temperature-measuring, but it has following several shortcoming: at first, the method has certain hysteresis quality, can only realize the discontinuous measurement of temperature; Secondly, the position and the degree of depth that thermopair inserts are judged that by experience therefore certain randomness arranged; Moreover, when a metal material is uninterruptedly repeated thermal treatment, can't in time change thermopair; Last thermopair costs an arm and a leg, and consumes rapidly.

Non-contact temperature measuring is then based on radiation temperature measurement.Present radiation temperature measurement mainly contains brightness method, radiation method, colourimetry, multi-wavelength method etc., introduce respectively below the whole bag of tricks and with difference of the present invention.

The eyes of 1, brightness method and radiation method personnel selection or photomultiplier are for accepting device, whether the standard of comparison spectrum line equates with tested spectrum line or is proportional under specific wavelength, thereby judges temperature (for example people such as people such as Chinese Academy of Sciences's environmental optics and technology key lab Liu Zhiming " based on the heat radiator remote temperature sensing Study of New Method of Fourier transform infrared spectroscopy " and the Wang Feng of Institutes Of Technology Of Hebei " application of infrared temperature-test technology in high-temp liquid continuous temperature measurement system " etc.).But the as easy as rolling off a log environmental impact that is subjected to of this method, error is very big.This method computation process is very complicated simultaneously, can't further promote.

2, colourimetry (duochrome method) is by selected two special modalities, thereby asks the ratio of its light intensity to reach the purpose of eliminating environmental impact and emissivity.This method research at present is very ripe, be mainly used in the temperature of measuring boiler furnace flame (for example: people's such as the few brightness of Lee of Beijing Institute of Technology " the image colourimetry is in the real-time Study of Monitoring in furnace flame temperature field ", people's such as the people's such as Jiang Shujuan of Shenyang robotization research institute of the Chinese Academy of Sciences " utilizing the image colourimetry to carry out the research that the laser molten pool temperature field is detected in real time " and the Jiang Zhi of the Central China University of Science and Technology are big " based on the flame temperature and the radiance image detection method of Flame Image Process " etc.).But this method still has certain defective, is mainly derived from: (1) colourimetry its selection principle when selecting two wavelength comparing conflicts mutually with the data processing principle; (2) condition set up of this method is that the emissivity of object under test does not change (that is: object is absolute grey body or is approximately grey body) with wavelength change, so certain use restriction is arranged in actual conditions; When (3) to one in the dual wavelength strong absorption being arranged as if the medium between measuring instrument and testee, then the method can't be used.

3, multi-wavelength (multispectral) thermometry is comparatively a kind of method of main flow of present conceptual phase.This method is by supposition emissivity model, be taken in the relation of spectral intensity and temperature, utilize the spectral signal of the object under test of multi-channel spectrometer based collection then, progressively return at last and obtain emissivity and temperature (for example Harbin Institute of Technology wears the scape people, Sun Xiao just waited the people's " measurement of multi-wavelength method metal solar heat protection watt surface temperature and emissivity ", " research of multispectral thermometry modeling method ", Northwestern Polytechnical University once learnt military affairs etc. people " the multispectral survey technical research of material surface temperature " and people's such as the Cheng Xiao of Chinese science and technology technology university boat " based on the structure research of the temperature survey equation of radiation ").The method equally also has certain error, is mainly derived from: the emissivity model of (1) supposition just could be similar within the specific limits and meets truth, and this model is not suitable all things, and certain limitation is arranged; (2) the method also is to accept absolute light intensity, thus be subjected to environmental impact easily, and after changing the work-yard, need again it to be debugged.

(3) summary of the invention

In order to solve above-mentioned technical matters, the purpose of this invention is to provide a kind of thermometry based on object emission spectrum, rationally select for use light intensity, wavelength and temperature relation as the data processing foundation, reduce of the influence of the emissivity of object to measurement result, gather more DATA REASONING dot information to improve antijamming capability, be not subjected to the influence of testee material, thereby improve thermometric accuracy and universality.

To achieve the above object, the present invention has adopted following technical scheme:

A kind of thermometry based on object emission spectrum comprises the steps:

(1) utilizes spectrometer to gather continuous spectrum that the object under test under certain temperature sends and this continuous spectrum is carried out filtering and smoothing processing;

(2) obtain ln (E (λ, T) λ according to the spectroscopic data in 50～150nm spectral width scope ⁵) about

Change curve and adopt linear fit method to obtain the slope A of this change curve, again according to equation

$A=-\frac{c_2}{T}$

Obtain the object under test temperature T; Wherein, (λ T) is spectral intensity to E, and λ is a radiation wavelength, and T is an object temperature, c ₂=1.43879 * 10 ^-2M/K is a second radiation constant.Described spectral intensity and radiation wavelength are the result who records in the spectrometer.

The ultimate principle of heat radiation thermometric is based on blackbody radiation law, and promptly temperature is that spectrum and the T that the black matrix of T sends has certain relation, and its spectral intensity is provided by planck formula:

E(λ，T)＝2hc ²λ ^-5(e ^hc/λkT-1) ^-1 (1)

For general object, following formula then is converted to as follows:

E(λ，T)＝ε(λ，T)2hc ²λ ^-5(e ^hc/λKT-1) ^-1 (2)

(λ T) is the spectral emittance of object to ε in the formula, and its value is the function of object temperature and radiation spectrum wavelength.λ is a radiation wavelength, and T is an object temperature, and c is the light velocity in the vacuum, h=6.6261 * 10 ^-34Js is a Planck's constant, k=1.3807 * 10 ^-23JK ^-1Be Boltzmann constant.

In actual measurement, when (λ T)＜＜1 of object, promptly available Wien formula replaces planck formula, and is as follows:

$E(\mathrm{\&lambda;},T)=\mathrm{\&epsiv;}(\mathrm{\&lambda;},T)c_1{\mathrm{\&lambda;}}^{-5}{\left(e^{-{\mathrm{<figure-callout\; id="2"\; label="c"\; filenames="HDA0000023406770000011.png,HDA0000023406770000032.png"\; state="\{\{state\}\}">c</figure-callout>}}_2/\mathrm{\&lambda;T}}\right)}^{-1}---\left(3\right)$

C in the formula ₁=3.7415 * 10 ^-16Wm ²Be first radiation constant, c ₂=1.43879 * 10 ^-2M/K is a second radiation constant.Because temperature is staggered in spectral emittance in the following formula, therefore can't accurately know under the situation of emissivity the very big error of the certain existence of temperature survey.To following formula transposition and to get with e be can get behind the end logarithm:

$\ln (E(\mathrm{\&lambda;},t)\&CenterDot;{\mathrm{\&lambda;}}^5)=-\frac{c_2}{T}\&CenterDot;\frac{1}{\mathrm{\&lambda;}}+{\ln c}_1+\ln \mathrm{\&epsiv;}(\mathrm{\&lambda;},T)---\left(4\right)$

In the formula (4), comprise the afterbody that the emissivity item is moved to equation separately, separate, reduced thermometric error with wavelength; And in 50～150nm wavelength width scope, (λ is T) with ln (E (λ, T) λ to comprise emissivity item ln ε ⁵) value compares very for a short time, can ignore; Therefore, by formula (4) equation left side ln (E (λ, T) λ as can be seen ⁵) with

Linear, and its slope is the c of inverse temperature ₂Doubly.Like this, only need to record the wavelength and the light intensity value thereof of a plurality of measurement points in certain wavelength width scope, just can obtain the equation slope by spectrometer

Can try to achieve object temperature T.

Further, the technical program is particularly suitable for the thermometric in laser manufacturing and the manufacture process again, and the object under test temperature range is at 1400～1900 ℃.For lower temperature, can record by the spectrometer that adopts the different-waveband measurement range.

The present invention need carry out filtering and smoothing processing by the continuous spectrum that spectrometer collects, and the Filtering Processing that the present invention adopts is by using MATLAB that filtering parameter is set, the high frequency noise in the original spectrum being removed.The filtering method that other are conventional can also be obtained effect.Smoothing processing of the present invention is to adopt least square method that the spectral signal that filters out behind the high frequency noise is carried out match, the result that different approximating methods obtains as long as and the original spectrum shape coincide substantially and can meet the demands.

Further, in order to guarantee to measure accurately above-mentioned steps 2) preferably obtain ln (E (λ, T) λ according to the spectroscopic data in 80～120nm (more preferably 100nm) spectral width scope ⁵) about

Change curve and adopt linear fit method to obtain the slope A of this change curve, and then obtain object temperature T.

Step of the present invention (2) does not have special requirement for the selection of concrete spectrum section on continuous spectrum, but should avoid selecting the spectroscopic data at two ends, the left and right sides as far as possible after definite spectral width scope.

The present invention has reasonably selected spectral intensity and the temperature relation processing foundation as data for use in order to obtain the temperature of object more simply; Consider of the influence of the emissivity of object in the measurement to measurement result, this equation has been carried out the mathematics manipulation that logarithm is asked in transposition, emissivity moves on to the afterbody about the linear equation of spectral intensity and wavelength the most at last, so just can make emissivity that the influence of equation slope is reduced (slope of equation is the inverse of temperature) greatly; The present invention simultaneously has stronger antijamming capability so absorb the measurement point that distorts for one or several selection owing to medium, thereby improves thermometric accuracy and universality owing to collected more DATA REASONING dot information.Comparatively speaking:

A) than brightness method and radiation method thermometric, the present invention is owing to adopt spectrometer as receiving element, and according to the actual conditions of this method application, the physical relation that adopts spectral intensity and temperature is calculated and is found the solution recording data, so have higher measuring accuracy and measuring process simple than traditional brightness method and radiation method.

B) than the colourimetry thermometric of present maturation, the present invention has better universality and accuracy.At first, because the present invention has reduced the influence of emissivity variation to measurement result to a great extent, thereby need not to consider in the colourimetry, when selecting two wave bands comparing, the selection of its wave band and the afoul problem of data processing principle need not also to consider whether object under test is grey body or is approximately grey body; Secondly because the selected measurement point distribution range of the present invention is compared extensively, number is many, so some there is very strong anti-interference owing to being subjected to the measurement point that the environmental selection absorption distorts.

C) with respect to the multi-wavelength thermometry, the present invention transplants and the processing of taking the logarithm by the physical relation to spectral intensity and temperature relation, greatly reduced of the influence of emissivity uncertainty to the linear equation slope, also need not optionally to consider material, and then improved thermometric accuracy and universality.

(4) description of drawings

Fig. 1 is the experimental provision structural representation of embodiment 1;

Fig. 2 is the original spectrum distribution plan of embodiment 1;

Fig. 3 is sample spectrum design sketch after ripple and the smoothing processing after filtration among the embodiment 1;

Fig. 4 is whole sample spectrum design sketch after ripple and the smoothing processing after filtration among the embodiment 1;

Fig. 5 (a), Fig. 5 (b), Fig. 5 (c) are respectively ln (E (λ, T) λ of three sample spectrum ⁵) with Linear relationship chart, the object sample of 1630 ℃ of 5 (a) expressions wherein, the object sample that 5 (b) expression is 1700 ℃, the object sample that 5 (c) expression is 1870 ℃.

(5) embodiment

Below in conjunction with accompanying drawing the specific embodiment of the present invention is made a detailed explanation, but protection scope of the present invention is not limited thereto.

Embodiment 1:

The experimental provision that present embodiment adopts as shown in Figure 1, wherein, heating part adopts the high-power CO2 laser instrument of 7KW 3 as heating source, with common 45# steel as measured material.In order to obtain temperature rising curve stably, it is too fast to prevent that temperature from rising in the measuring process, and the output power of laser instrument is set at 750W, and the prolonged exposure same point.Adopt fiber spectrometer 1 to gather spectrum, it gathers wave band is 400nm～950nm.For preventing that flue gas is to the influence of spectral intensity, in the other smoking system 5 that settled of measured material in the heating process.Adopt thermopair 7 to combine in the experiment with infrared thermography instrument 4, utilize thermopair that the infrared thermography instrument is replenished the checking way and obtain the degree of accuracy of true temperature with check spectrometer measurement system: promptly the opposite side at spectrometer assembles a support, a German Metis infrared thermography instrument 4 is fixed on it, the measurement point of infrared thermography instrument 4 is overlapped with the measurement point of the fiber-optical probe 2 of LASER HEATING point (molten bath 6) and spectrometer, meanwhile, go into a thermopair 7 in that the back side of LASER HEATING point identical point is buried, in conjunction with the electric thermo-couple temperature upcurve, from the infrared thermography instrument curve of cyclical fluctuations, draw the true temperature of LASER HEATING point when measuring beginning.This experiment adopts computing machine 8 to carry out data processing.

Experimentation is as follows:

Choose 11 measured temperatures according to the reading on the infrared thermography instrument, record spectral intensity and temperature separately obtained 11 sample spectrum so altogether respectively, and range of temperature is from 1530 ℃ to 1900 ℃.The wavelength coverage of spectrum is from 400 to 950nm, and corresponding original spectrum as shown in Figure 2; Owing to be subjected to the influence of external environment, each spectral distribution curve is mixing ground unrest and high frequency noise among Fig. 2, therefore measure in order to obtain more exactly, must carry out the denoising smoothing processing to it, present embodiment adopts the MATLAB filter function that the spectrum samples data are carried out filtering, uses the high-order Gaussian curve to its match after the filtering again, effect as shown in Figure 3, adopt this kind method smoothing processing, can obtain the good treatment effect, Fig. 4 is 11 design sketchs after the spectrum samples data smoothing.In order to find out ln (E (λ, T) λ from equation (4) two sides more intuitively ⁵) with

Linear relationship, use the visuality of MATLAB, to equation (4) mapping, obtain Fig. 5; By Fig. 5 can visually see ln (E (λ, T) λ ⁵) with Between exist linear relationship, and its slope is a negative value, in fact also be, (600～700nm) linearly dependent coefficient values are all more than 90% in the radiation wavelength scope of choosing, so good linear dependence is arranged, the cftool tool box is to ln (E (λ, T) λ of 11 spectrum samples among the employing MATLAB ⁵) about

Change curve (the radiation wavelength scope of choosing is 600～700nm) to carry out linear fit (least square fitting) and try to achieve fit slope, can obtain object temperature by this slope.Measurement result and measuring error are as shown in table 1 below:

Table 1

Sample

1

2

3

4

5

6

7

8

9

10

11

True temperature	1530	1550	1630	1650	1700	1720	1750	1800	1830	1870	1900
												Measure temperature	1510	1567	1641	1678	1669	1681	1772	1834	1863	1873	1958
Measuring error	20	17	11	28	31	39	22	34	33	3	58

Last table has reflected thermometric error in the test, as can be seen experiment error range between-40-60 ℃, number percent is less than 3%, and the requirement to the error of temperature is in 5% in the actual production practice, this shows that this method can satisfy actual needs.

Claims (4)

1. the thermometry based on object emission spectrum comprises the steps:

(1) utilizes spectrometer to gather continuous spectrum that the object under test under certain temperature sends and this continuous spectrum is carried out filtering and smoothing processing;

(2) obtain ln (E (λ, T) λ according to the spectroscopic data in 50～150nm spectral width scope ⁵) about

Change curve and adopt linear fit method to obtain the slope A of this change curve, again according to equation

$A=-\frac{c_2}{T}$

Obtain the object under test temperature T; Wherein, (λ T) is spectral intensity to E, and λ is a radiation wavelength, and T is the object under test temperature, c ₂=1.43879 * 10 ^-2M/K is a second radiation constant.

2. the thermometry based on object emission spectrum according to claim 1 is characterized in that: described step 2), obtain ln (E (λ, T) λ according to the spectroscopic data in 80～120nm spectral width scope ⁵) about

Change curve and adopt linear fit method to obtain the slope A of this change curve, and then obtain the object under test temperature T.

3. the thermometry based on object emission spectrum according to claim 1 is characterized in that: described step 2), obtain ln (E (λ, T) λ according to the spectroscopic data in the 100nm spectral width scope ⁵) about

Change curve and adopt linear fit method to obtain the slope A of this change curve, and then obtain the object under test temperature T.

4. the thermometry based on object emission spectrum according to claim 1 is characterized in that: the temperature of described object under test is at 1400 ℃～1900 ℃.

Images