JP2018169203A - Apparatus and method for gas absorption spectroscopy - Google Patents

Abstract

To provide a method and an apparatus for gas absorption spectroscopy capable of measuring gas concentration or the like with high accuracy in high-speed measurements.SOLUTION: A measurement target gas is irradiated with laser light whose wavelength changes, and a spectral profile representing a change in intensity with respect to the wavelength of the laser light having passed through the measurement target gas is obtained. The spectrum profile is polynomially approximated by an approximate polynomial within a range of a predetermined wavelength modulation width (a) at each point of the wavelength, and based on coefficients of each term of the approximate polynomial at each point, an nth-order differential curve including the zero order of the spectral profile is created. Characteristic values (peak area, height, and so on) of the created nth-order differential curve including the zero order are calculated by a plurality of wavelength modulation widths a1, ..., an, and are regarded as unknown vectors. The unknown vector is compared with a vector of the known gas calculated in advance in the same manner and the physical quantity of the measurement target gas is determined from the nearest known vector.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a gas absorption spectroscopic apparatus and method for measuring the concentration, temperature, pressure, and the like of a gas based on a laser light absorption spectrum of a measurement target gas. This gas absorption spectroscopy device and method can be applied to non-contact, high-speed measurement of gas concentration, temperature and pressure in the automotive industry, and gas measurement in high-temperature and high-pressure environments such as combustion gas in plant furnaces. It can be applied in various fields.

There are the following three methods of gas absorption spectroscopy using a laser.
(1) DLAS (Direct Laser Absorption Spectroscopy)
(2) WMS (Wavelength Modulated Spectroscopy)
(3) CRDS (Cavity Ring Down Spectroscopy)

  Among these, as an industrial gas absorption spectrometer, WMS balanced in sensitivity and robustness (easiness of measurement) is suitable (Patent Document 1). In WMS, the gas concentration is easily calculated from the intensity of the obtained absorption spectrum. In addition, even in an environment where the pressure or temperature cannot be measured directly in an environment where the temperature and pressure constantly change, it can be used for applications that measure gas concentration and temperature by measuring using two wavelengths.

However, WMS has the following problems.
1. In order to perform high-speed measurement, the sweep cycle is shortened and a high wavelength modulation frequency is required. However, when an injection current control type tunable diode laser, which is most widely used as a wavelength tunable laser, is used, if the modulation frequency is increased, the wavelength change rate with respect to the injection current decreases, and a sufficient wavelength modulation width ( modulation depth) cannot be obtained (Non-Patent Document 1).
2. In particular, it is difficult to accurately measure the wavelength modulation width for high-speed modulation exceeding MHz, and an accurate wavelength modulation width cannot be determined in high-speed measurement. Therefore, the uncertainty of information such as gas concentration and temperature calculated from the measurement result is increased.

Due to the above factors, in the high-speed measurement, the conventional WMS has a problem that the difficulty of measuring the gas concentration, temperature, etc. with high accuracy is remarkably increased.
On the other hand, in Patent Document 2,
a) detecting the intensity of the measurement target gas passing light that is light from the light source that has passed through the measurement target gas while continuously changing the wavelength of the light source;
b) approximating a curve (absorption line) of a change in intensity of the light passing through the measurement target gas with a change in the light source wavelength by an approximation polynomial within a predetermined wavelength modulation width a at each wavelength point;
c) creating a k-th order differential curve including the zeroth order of the absorption line based on the coefficient of each term of the approximate polynomial of each point;
d) determining at least one of a temperature, a concentration, and a pressure of the measurement target gas based on a characteristic value (peak area, peak height, etc.) of the k-th order differential curve including the zeroth order. (This is referred to as “Numerical Wavelength Modulated Spectroscopy: NWMS”).

  The “wavelength” uniquely corresponds to the “wave number”, and it is of course possible to assemble a similar configuration using the “wave number”.

  In the above-described operational wavelength modulation spectroscopy, the light irradiated to the gas to be measured (usually, but not necessarily limited to laser light) changes the wavelength (performs a wavelength sweep), similar to DLAS, but is similar to WMS. Don't modulate it. The wavelength sweep may be changed only once (sweep) between a predetermined minimum frequency and a maximum frequency, or the sweep may be repeated a plurality of times.

  This light passes through the measurement target gas and is then received by the photodetector, and the intensity change is detected. Since the wavelength range for performing the wavelength sweep is set in advance to include the absorption wavelength of the measurement target gas, the spectrum profile of the light detected by the photodetector (see above “the measurement target gas passage accompanying the change in the light source wavelength”). In the curve of the change in light intensity (absorption line) "), an absorption peak centered on the wavelength specific to the measurement target gas appears.

  In the calculation wavelength modulation spectroscopy, a mathematical calculation similar to the WMS processing is performed on the spectrum profile including the absorption peak. Specifically, a k-order polynomial approximation is performed on the spectral profile of the section corresponding to the wavelength modulation width of WMS, centering on each wavelength point, and the WMS signal is used using the coefficient of the k-order polynomial based on the principle of Fourier transform. Reproduce the amplitude. The principle is as follows.

  In general, in WMS processing, it is known that the spectrum profile of the k-th harmonic obtained by synchronous detection is a waveform obtained by approximately k-order differentiation of the absorption spectrum (Non-Patent Document 2: Equation 8). Therefore, it is considered that a spectrum corresponding to k-th order synchronous detection can be obtained by k-order differentiation of a spectrum obtained by wavelength sweeping. However, if k-th order differentiation is performed, the influence of measurement data noise is large, which causes a practical problem. Therefore, in the calculation wavelength modulation spectroscopy, k-th order polynomial approximation is performed for a certain range centering on a wavelength for which a harmonic signal is desired to be obtained. The polynomial coefficient obtained is a harmonic signal obtained by WMS processing. At this time, the range in which the polynomial approximation is performed corresponds to the modulation amplitude in the WMS process.

  The higher the order k of the approximate polynomial, the more accurate approximation can be performed, but in general, a first-order or second-order polynomial approximation is sufficient.

  In the calculation wavelength modulation spectroscopy, only the wavelength sweep of several hundred kHz or less is performed in the light source, and therefore the oscillation wavelength with respect to the injection current of the light source is accurately determined. Since WMS processing is performed by mathematical calculation based on the wavelength information, high-order synchronization detection with an accurate wavelength modulation width is possible without being affected by the nonlinearity of the light source driving power source and the light source itself.

  In addition, because of the mathematical processing, it is possible to acquire WMS spectra with multiple modulation widths simultaneously using one light source, and temperature measurement that could not be measured without using two or more light sources with one light source. It is also easy to realize. Furthermore, when measuring temperature, it is necessary to optimally adjust the modulation width in advance in order to minimize the pressure dependence of the measurement temperature. However, in calculation wavelength modulation spectroscopy, the modulation width can be adjusted by post-analysis. it can.

JP 2011-106560 A International Publication WO2014 / 106940 Publication

J.T.C. Liu, "Near-infrared diode laser absorption diagnostics for temperature and species in engines," Ph.D. dissertation, Dept. Mechanical Engineering, Stanford Univ., Stanford, CA, 2004 (Figure 3.12) Reid, J. and Labrie, D., "Second-harmonic detection with tunable diode lasers-comparison of experiment and theory," Appl. Phys. B 26, 203-210 (1981) "Calculation of molecular spectra with the Spectral Calculator", [searched January 7, 2013], Internet <URL: http: //www.spectralcalc.com/info/CalculatingSpectra.pdf> Katsuhiko Fukusato, Yuji Ikeda, Ken Nakajima, "Measurement of CO2 gas using semiconductor laser spectroscopy system (2nd report)", Transactions of the Japan Society of Mechanical Engineers, B, 2002, 68, 2901-2907 G. B. Rieker, J. B. Jeffries, and R. K. Hanson, "Calibration-free wavelength modulation spectroscopy for measurements of gas temperature and concentration in harsh environments," Appl. Opt., Submitted 2009

  In general, when the temperature of the gas to be measured is increased and the pressure is increased, the peak width of the absorption line is widened by collision between gas molecules. Also, the peak of the absorption line is widened when the concentration of the measurement target gas is low. When calculation is performed with a certain wavelength modulation width in the calculation wavelength modulation spectroscopy, the state (temperature, pressure, etc.) of the measurement target gas changes, and the peak width changes. For this reason, there is a case where the optimum value of the wavelength modulation width varies depending on the shape of the absorption line obtained by measurement, and there is a problem that the reliability of the calculation result cannot be ensured.

  The problem to be solved by the present invention is to provide a gas absorption spectroscopic measurement method and apparatus capable of always performing a highly reliable calculation regardless of a change in the state of the measurement target gas.

The gas absorption spectroscopy according to the present invention, which has been made to solve the above problems,
a) For each of a plurality of known absorption lines for gases with different conditions, n (n is an integer of 2 or more) wavelength modulation widths are used to calculate n k (k is 0 or more). An integer) calculating a known vector, which is a vector composed of the second derivative curve feature values;
b) detecting the intensity of light passing through the measurement target gas that is light from the light source that has passed through the measurement target gas while continuously changing the wavelength of the light source;
c) Calculate an unknown vector, which is a vector composed of n k-th order differential curve characteristic values obtained by the same method, based on a curve of a change in intensity of the light passing through the measurement target gas with a change in the light source wavelength. And steps to
d) detecting a known vector closest to the unknown vector among the plurality of known vectors.

In the gas absorption spectroscopy according to the present invention, first, known absorption lines of a plurality of gases having different conditions, that is, different gas types, concentrations, temperatures, or pressures, are respectively calculated by arithmetic wavelength modulation spectroscopy. Such a vector (known vector) is calculated. Note that the gas absorption line that is the basis for calculating each known vector may be an actual measurement of actual gas, or it may be stored in the data of an existing database (for example, HITRAN database = high-resolution transmission molecular absorption database). The absorption line produced | generated based on may be sufficient.
The arithmetic wavelength modulation spectroscopy is a method described in Patent Document 2. Specifically, as mentioned above,
a) detecting the intensity of the measurement target gas passing light that is light from the light source that has passed through the measurement target gas while continuously changing the wavelength of the light source;
b) approximating a curve (absorption line) of a change in intensity of the light passing through the measurement target gas with a change in the light source wavelength by an approximation polynomial within a predetermined wavelength modulation width a at each wavelength point;
c) creating a k-th order differential curve including the zeroth order of the absorption line based on the coefficient of each term of the approximate polynomial of each point;
d) determining at least one of a temperature, a concentration, and a pressure of the measurement target gas based on a characteristic value (peak area, peak height, etc.) of the k-th order differential curve including the zeroth order. It is.

In the gas absorption spectroscopy according to the present invention, for each of a plurality of known absorption lines, a k-th order differential curve is calculated by this operation wavelength modulation spectroscopy to obtain a characteristic value (peak area, peak height, etc.). This characteristic value reflects the type, concentration, temperature, or pressure of the measurement target gas. In the gas absorption spectroscopy according to the present invention, two or more different values of the wavelength modulation width a used in the calculation wavelength modulation spectroscopy (these are assumed to be a ₁ ,..., _An (n ≧ 2)). Then, a k-th order differential curve is calculated for each known absorption line to obtain a feature value. Then, for one absorption line, different wavelength modulation width a _1, ..., wherein values f _ka1 of k order derivative curve calculated by a _n, ..., vector a consisting of _{f kan (f ka1, ...,} f kan) Is obtained. This is the known vector. In the gas absorption spectrometer according to the present invention described later, these known vectors are stored in a known vector storage unit. For a ₁ ,..., _An (n ≧ 2) having different wavelength modulation widths, since the relative sizes of f _ka1 ,..., F _kan are determined by the wavelength modulation widths, the wavelength modulation widths a _{1 ,.} ..., vector element f _ka1 corresponding to _{a n (n ≧ 2),} ..., respectively coefficients alpha ₁ relative to f _kan, ..., may be subjected to weighting by alpha _n. Coefficient alpha _1, ..., for the alpha _n, the ratio with respect to temperature, pressure, all _{f ka (f ka1, ...,} f kan) obtained by changing the density average value or the maximum value of each wavelength modulation width a _n It is good to decide by. Further, the weighting coefficient α _x corresponding to the vector element f _x corresponding to the wavelength modulation width a _x suitable for the measurement target condition may be increased.

Next, for the measurement target gas whose concentration, temperature, or pressure is unknown, the intensity of the measurement target gas passing light, which is the light from the light source that has passed through the measurement target gas, while continuously changing the wavelength of the light source. To detect. Based on the intensity change curve, an unknown vector b (b _ka1 ,..., B _kan ), which is a vector composed of n k-th order differential curve feature values, is calculated by the same method as described above. The unknown vector b, a plurality of known vectors that had been the calculated (these a _1, ..., and a _m) is compared respectively, to detect the closest known vectors ap unknown vector. That two vectors are closest means that the angle between both vectors is the smallest and the value R _ab expressed by the following equation is the largest.
R _ab = ( a・b ) / (| a | ・ | b |)
However, ( a · b ) is an inner product of vectors a and b . Ratio of length of known vector (recently known vector) a _p and unknown vector b detected in this way
Q _b = | a _p | / | b |
Is considered to be equal to the concentration ratio of the gas (type, temperature, pressure) corresponding to the recently known vector and the gas to be measured corresponding to the unknown vector. Therefore, from the gas concentration c _p corresponding to the recently known vector, the concentration of the gas to be measured
c = c _p / (| a _p | / | b |)
Can be calculated. The temperature and pressure of the measurement target gas are considered to be equal to the temperature and pressure of the gas corresponding to the recently known vector.

  Note that the temperature or pressure is measured at the same time that the transmitted light intensity is measured for the gas to be measured, and the range of known vectors to be detected is limited in advance by the actual measurement value. From the point of view is desirable.

  For the light source, it is desirable to measure the intensity change of the light source light simultaneously with the transmitted light intensity measurement and normalize (compensate) the intensity of the measurement target gas passage light.

  In addition, it is also possible to normalize the intensity change of light source light by calculation according to the method described in [0033] and after in Patent Document 2 without actually measuring the intensity of such light source light.

In order to solve the above problems, a gas absorption spectrometer according to the present invention comprises:
a) For each of a plurality of known absorption lines for gases with different conditions, n k (where k is an integer equal to or larger than 2) wavelength modulation widths obtained in the operation wavelength modulation spectroscopy are used. An integer greater than or equal to 0) a known vector storage unit that stores a known vector that is a vector composed of a characteristic value of the second derivative curve;
b) a tunable light source;
c) a light source controller that changes the wavelength of light generated by the light source;
d) a photodetector that detects the intensity of light generated by the light source and passed through the measurement target gas;
e) A vector composed of n k-th order differential curve feature values obtained by the same method based on a curve of a change in light intensity detected by the light detector accompanying a change in wavelength by the light source control unit. An unknown vector generator for generating an unknown vector;
and f) a determination unit that detects a known vector closest to the unknown vector among a plurality of known vectors stored in the known vector storage unit.

  In the gas absorption spectroscopy or gas absorption spectrometer according to the present invention, the concentration, temperature, and pressure of the unknown gas can be determined only by measuring the absorbance at one time and at one absorption wavelength. In addition, by preparing a corresponding known vector in advance, it is possible to deal with a measurement target gas whose temperature and pressure change.

1 is a schematic configuration diagram of an embodiment of a gas absorption spectrometer according to the present invention. The flowchart regarding one Example of the gas absorption spectroscopy method which concerns on this invention. An example of the measurement condition setting screen in a present Example. An example of the drive current supplied to a light source in a present Example. An example of the detection signal of the photodetector in a present Example. An example of the absorption spectrum in a present Example. The figure explaining the polynomial approximation of a spectrum profile. An example of the profile of the 2nd harmonic in a present Example. The figure explaining the structure of the database of the known gas in a present Example. The schematic block diagram of the modification of the gas absorption spectrometer which concerns on this invention.

  Embodiments of a gas absorption spectroscopy apparatus and a gas absorption spectroscopy method according to the present invention will be described below with reference to the drawings.

  As shown in FIG. 1, the gas absorption spectroscopic device 1 of this embodiment is a device that measures the concentration of a measurement target gas flowing in the gas cell 11, and is roughly composed of a measurement unit 10 and a control unit 20. The measurement unit 10 includes a laser light source 12 and an incident side lens 16 on one side with a gas cell 11 interposed therebetween, and an emission side lens 17 and a photodetector 13 on the other side, and supplies a current for driving the laser light source 12. A power supply unit 14, an A / D converter 15 that digitally converts an output signal from the photodetector 13, a beam splitter 18 that extracts a part of light from the laser light source 12 as reference light, and a photodetector 19 for reference light And. The laser light source 12 of the present embodiment is a semiconductor laser diode, and emits light having a wavelength and intensity corresponding to the magnitude of the drive current supplied from the power supply unit 14.

  In addition to the storage unit 21, the control unit 20 includes a light source control unit 22, a spectrum profile acquisition unit 23, an unknown vector generation unit 24, a recently known vector determination unit 25, and a measurement result presentation unit 26 as functional blocks. The entity of the control unit 20 is a personal computer, and each functional block is embodied by executing a gas absorption spectroscopy program by the CPU. An input unit 30 and a display unit 40 are connected to the control unit 20.

  A procedure for measuring the concentration, temperature, pressure and the like of the measurement object gas in this embodiment will be described with reference to the flowchart of FIG.

  When the user instructs the start of measurement, a measurement condition setting screen as shown in FIG. 3 is displayed. On the measurement condition setting screen, there are columns for selecting the type of gas to be measured, the number and order of wavelength modulation widths used to generate an unknown vector, which will be described later, and columns for measuring items such as the concentration, temperature, and pressure of the gas to be measured. It is displayed. The user selects and inputs each item to determine measurement conditions (step S1). The unit of the wavelength modulation width in this embodiment is a wave number, but it is called a wavelength modulation width in view of the fact that it corresponds to the wavelength modulation width in WMS that has been used conventionally. Further, a wavelength modulation width in units of wavelength may be used.

  When the measurement condition is determined by the user, the light source control unit 22 radiates laser light having a predetermined minimum wavelength from the laser light source 12, and sequentially changes the wavelength and sweeps the wavelength to the maximum wavelength T. (Step S2, see FIG. 4). The predetermined minimum wavelength and the maximum wavelength are determined in advance so that the wavelength of light absorbed by the measurement target gas is included between them. The light from the laser light source 12 passes through the measurement target gas in the gas cell 11 and is absorbed there at a wavelength corresponding to the measurement target gas. The intensity of the laser light that has passed through the measurement target gas is detected by the photodetector 13. The detection signal output from the photodetector 13 is digitized by the A / D converter 15 and stored in the storage unit 21 (see FIG. 5).

  The spectrum profile acquisition unit 23 acquires the absorption spectrum (spectrum profile) of the measurement target gas as shown in FIG. 6 from the detection signal and the reference light intensity stored in the storage unit 21 (step S3). The peak depth of the absorption spectrum reflects the concentration of the measurement target gas, and the peak width reflects the pressure and temperature of the measurement target gas.

When the spectrum profile of the measurement target gas is obtained, the unknown vector generation unit 24 uses different wavelength modulation widths a _{1 to} a ₄ (this embodiment) having different numbers of wavelength modulation widths (4 in this embodiment) determined by the user. Then, this spectral profile is approximated by a polynomial using (a ₁ = 0.1 cm ⁻¹ , a ₂ = 0.2 cm ⁻¹ , a ₃ = 0.4 cm ⁻¹ , a ₄ = 0.8 cm ⁻¹ ) (step S4). The number and value of the wavelength modulation width are associated with each other in advance and stored in the storage unit 21. When the user determines the number of wavelength modulation widths in step S1, these values are automatically read out.

  Here, a numerical wavelength modulation spectroscopy (NWMS) by polynomial approximation performed by the unknown gas spectrum creation unit 24 will be described in comparison with a method performed by a conventional WMS.

ただし、νは波数、Aはピーク面積、ν_cはピーク波数、α_Lはローレンツ広がりの半値半幅である。 In general, it is known that an absorption peak is represented by the following Lorentz function in an environment under atmospheric pressure.

Where ν is the wave number, A is the peak area, ν _c is the peak wave number, and α _L is the half-value half-width of the Lorentz broadening.

ν：波数（なお、本願の本文中では、電子出願の制限により、上付線を下線で表現する。）、τ：透過スペクトルのプロファイル、a：波長変調幅 Based on the WMS method, modulation of wavelength modulation width a is applied to the incident laser light, and when synchronous detection with a lock-in amplifier is performed on the light transmitted through the gas having the above absorption profile, the spectrum by the n-th order synchronous detection is It is obtained by the following formula (Non-patent Document 2).

ν : wave number (in the text of the present application, the superscript line is expressed as an underline due to limitations of electronic application), τ: profile of transmission spectrum, a: wavelength modulation width

  Even with equation (2), processing equivalent to WMS can be performed by mathematical calculation. However, the form of the equation is complicated and not practical. Therefore, in NWMS, by performing calculations using polynomials, processing equivalent to WMS high-order (including zero-order) detection processing is performed at high speed and simply, and various physical quantities of the measurement target gas are measured.

で近似する。これを図７に模式的に示す。式(3)のn次微分を求めると

となる。一般に、WMS処理で同期検波して得られるn次の高調波のスペクトルプロファイルは、近似的に次式で示されることが知られている（非特許文献２：Equation 8）。

したがって、式(4)、(5)より

となる。したがって、DLASスペクトルにおいて波数νに対するWMS信号を算出するためには、[ν−a'<ν<ν＋a']の波数の範囲を最小二乗法等によりフィッティングし、多項式の係数b₀、b₁、b₂、b₃…を求める。νを逐次変化させてフィッティングにより求めた係数b₁とb₂のプロファイルが、1f（一次）と2f（二次）のWMSプロファイルに相当するものとなる。なお、フィッティングの範囲を表す波長幅a'は、WMSにおける波長変調幅に相当する。 In this example, the range [ ν− a ′ <ν < ν + a ′] of the width 2a ′ around each point ν of the wavenumber axis of the obtained spectral profile is expressed as

Approximate. This is schematically shown in FIG. Finding the nth derivative of equation (3)

It becomes. In general, it is known that the spectrum profile of the n-th harmonic obtained by synchronous detection by WMS processing is approximately expressed by the following equation (Non-patent Document 2: Equation 8).

Therefore, from equations (4) and (5)

It becomes. Therefore, in order to calculate the WMS signal for the wave number ν in the DLAS spectrum, the wave number range of [ ν− a ′ <ν < ν + a ′] is fitted by the least square method or the like, and the coefficients b ₀ and b _{1 of the} polynomial are obtained. , B ₂ , b ₃ ... The profiles of the coefficients b ₁ and b ₂ obtained by fitting by sequentially changing ν correspond to the WMS profiles of 1f (primary) and 2f (secondary). The wavelength width a ′ representing the fitting range corresponds to the wavelength modulation width in WMS.

In the present embodiment, a plurality of (four in the present embodiment) wavelength modulation widths (a ₁ = 0.1 cm ⁻¹ , a ₂ = 0.2 cm ⁻¹ , a with respect to the spectrum profile obtained by measuring the measurement target gas. ₃ = 0.4 cm ^-1 and a ₄ 0.8 cm ^-1 ) respectively. As a result, four types of polynomials are obtained. FIG. 8 shows spectral profiles of second harmonics obtained for wavelength modulation widths of 0.1 cm ⁻¹ , 0.2 cm ⁻¹ , 0.4 cm ⁻¹ , and 0.8 cm ⁻¹ , respectively. As can be seen from FIG. 8, the shape of the second harmonic profile differs depending on the value of the wavelength modulation width.

Subsequently, the unknown vector generation unit 24 obtains the coefficient of the k-th order term of the polynomial at the peak wavelength ν _c of the absorption spectrum as the feature value, and the k-th order feature value relating to each of the wavelength modulation widths a _{1 to} a ₄ as components. (Unknown vector) is generated (step S5). In the present embodiment, a four-dimensional vector b (b _2a1 ,..., B _2a4 ) is calculated using the coefficients of the second-order terms determined in advance by the user. This vector b (b _2a1 ,..., B _2a4 ) reflects the depth and shape of the absorption peak in the spectrum profile of the measurement target gas. More specifically, the direction of the vector b reflecting the shape of the absorption peak, the magnitude of the vector b reflects the concentration of the measured gas.

  In the storage unit 21, the components of known vectors (values corresponding to the respective wavelength modulation widths) obtained by the same calculation process from spectrum profiles having different concentrations, pressures, and temperatures for the plurality of gases are stored in the concentration, pressure, And stored as a database in association with gas profiles such as temperature (see FIG. 9). The known vector component stored in the database may be obtained by performing the same calculation process as above on the measured data, or the same calculation as described above for the spectrum profile stored in the database such as the HITRAN database. It may be obtained by performing processing.

The recently known vector determination unit 25 reads out the value of the wavelength modulation width corresponding to the unknown vector b (b _2a1 ,..., B _2a4 ) calculated by the unknown vector generation unit, and generates a known vector having these as components. Specifically, the measurement target gas and the same type of gas (gas A), for each of the pressure and temperature combinations 1 to m, the wavelength modulation width is ^{0.1cm -1, 0.2cm -1, 0.4cm -1} , known vectors a ₁ four-dimensional read the value of the component is 0.8 cm ^-1, ..., to generate a _m. Then, these are compared with an unknown vector, and a known vector (most recently known vector) a _p closest to the unknown vector is determined (step S6). Here, “the two vectors are closest” means that the angle between the two vectors is the smallest and the value R _ab expressed by the following equation is the largest.
R _ab = ( a・b ) / (| a | ・ | b |)
However, ( a · b ) is an inner product of vectors a and b .

Once the recently known vector is determined, the ratio of the length of the recently known vector a _p to the unknown vector b
Q _b = | a _p | / | b |
Ask for. This value corresponds to the concentration ratio of the gas (type, temperature, pressure) corresponding to the recently known vector and the measurement target gas corresponding to the unknown vector. Therefore, from the gas concentration c _p corresponding to the recently known vector, the concentration of the gas to be measured
c = c _p / (| a _p | / | b |)
Is required.

When the concentration of the measurement target gas is determined, the measurement result presentation unit 26 displays the concentration on the display unit 40 together with the pressure and temperature associated with the recently known vector ap (step S7).

  As described in Patent Document 1, the NWMS method itself has been conventionally used, but the conventional NWMS method has obtained a polynomial using one wavelength modulation width. The absorption spectrum (spectral profile) of the gas varies depending on the temperature, pressure, and concentration of the gas, and the optimum wavelength modulation width varies depending on the peak shape. Therefore, for example, when the concentration of the measurement target gas contained in the measurement target gas changes with time and the peak width of the absorption spectrum changes, the concentration of the measurement target gas is determined by an arithmetic process using a wavelength modulation width that is not suitable for the shape. In some cases, the correct concentration or the like cannot be obtained. In contrast, in this embodiment, a polynomial is obtained for each of a plurality of wavelength modulation widths, and a multidimensional unknown vector is created using a coefficient of a term of a predetermined order. Further, the concentration or the like is obtained by comparing the unknown vector with a known vector created by a similar method for a gas whose concentration or the like is known. In this manner, since a concentration or the like is determined by using a plurality of wavelength modulation widths and comparing with a known vector, a highly reliable calculation can always be performed regardless of a change in the state of the measurement target gas.

The above-described embodiment is an example, and can be appropriately changed in accordance with the gist of the present invention.
In the above embodiment, the concentration, pressure, and temperature are determined from the result of comparing the unknown vector with the known vector. However, as in the modification shown in FIG. 10 (only the configuration of the measurement unit 10 is shown in FIG. 10), the gas cell 11 A pressure / temperature value is obtained by using a temperature / pressure sensor 50 for measuring the pressure and temperature in the inside, and only those close to those values are extracted from the component data of known vectors stored in the storage unit 21. You may make it compare with an unknown vector. In this case, the number of known vectors to be compared with the unknown vector is reduced, and the load related to the arithmetic processing is reduced, so that high-speed processing is possible. Can be monitored. In addition, since the pressure and temperature are actually measured values, there is no fear of errors.

  In the above embodiment, the coefficient of the quadratic term of the polynomial is used as it is, but a value obtained by normalizing this may be used. Here, the normalization process is a process for reducing the influence of a change in light intensity caused by a change in the optical axis due to contamination of an optical component used in the gas cell 11 or vibration in a poor environment. Since the specific contents are described in Patent Document 1, detailed description is omitted here. For example, the coefficient of the second-order term (value corresponding to 2f signal in WMS) is changed to the coefficient of the first-order term ( Normalization processing can be performed by dividing by a coefficient of a zero-order term) (value corresponding to 1f signal in WMS). This makes it possible to perform robust gas measurement independent of light intensity.

DESCRIPTION OF SYMBOLS 10 ... Measurement part 12 ... Laser light source 13 ... Photodetector 14 ... Power supply part 15 ... A / D converter 16 ... Incident side lens 17 ... Outlet side lens 18 ... Beam splitter 19 ... Optical detector 50 ... Pressure / temperature sensor 20 ... Control unit 21 ... Storage unit 22 ... Light source control unit 23 ... Spectral profile acquisition unit 24 ... Unknown vector generation unit 25 ... Recently known vector determination unit 26 ... Measurement result presentation unit

Claims (6)

a) For each of a plurality of known absorption lines for gases with different conditions, n (n is an integer of 2 or more) wavelength modulation widths are used to calculate n k (k is 0 or more). An integer) calculating a known vector, which is a vector composed of the second derivative curve feature values;
b) detecting the intensity of light passing through the measurement target gas that is light from the light source that has passed through the measurement target gas while continuously changing the wavelength of the light source;
c) Calculate an unknown vector, which is a vector composed of n k-th order differential curve characteristic values obtained by the same method, based on a curve of a change in intensity of the light passing through the measurement target gas with a change in the light source wavelength. And steps to
d) detecting a known vector closest to the unknown vector among the plurality of known vectors.   In the step of detecting the intensity of the measurement target gas passing light, the temperature or pressure of the measurement target gas is measured, and in the step of detecting the known vector, a range of known vectors to be detected in advance based on the measured value of the temperature or pressure. The gas absorption spectroscopy according to claim 1.   2. The intensity of light emitted from a light source is measured in the step of detecting the intensity of the measurement target gas passing light, and the unknown vector is normalized based on the measured value of the light intensity in the step of calculating the unknown vector. Or the gas absorption spectroscopy of 2.   The gas absorption spectroscopy according to any one of claims 1 to 3, wherein the k-th derivative curve characteristic value is a peak area of the k-th derivative curve.   The gas absorption spectroscopy according to any one of claims 1 to 4, wherein k is 2. a) For each of a plurality of known absorption lines for gases with different conditions, n k (where k is an integer equal to or larger than 2) wavelength modulation widths obtained in the operation wavelength modulation spectroscopy are used. An integer greater than or equal to 0) a known vector storage unit that stores a known vector that is a vector composed of a characteristic value of the second derivative curve;
b) a tunable light source;
c) a light source controller that changes the wavelength of light generated by the light source;
d) a photodetector that detects the intensity of light generated by the light source and passed through the measurement target gas;
e) A vector composed of n k-th order differential curve feature values obtained by the same method based on a curve of a change in light intensity detected by the light detector accompanying a change in wavelength by the light source control unit. An unknown vector generator for generating an unknown vector;
f) A gas absorption spectrometer comprising: a determining unit that detects a known vector closest to the unknown vector among a plurality of known vectors stored in the known vector storage unit.

Images