JPWO2019150053A5 - - Google Patents

Description

The technical field of the invention is optical methods for the analysis of gases using a black-body or gray-body light source and measuring the absorption of the light waves emitted by the light source.

Optical methods are often used to analyze gases. The sensor makes it possible to determine the composition of the gas based on the fact that the constituent species of the gas have different absorption spectral characteristics. Therefore, if the absorption spectral band of a gas species is known, its concentration can be determined by calculating the absorption of light through the gas using the Beer-Lambert law. This principle makes it possible to calculate the concentration of gas species present in the gas.

In the most common method, the gas to be analyzed is between a light source and a photodetector, called a measuring photodetector, which is intended to measure the light wave transmitted by the gas to be analyzed, and this light wave is partially absorbed by the photodetector. The light source is usually a light source that emits in the infrared and the method used is usually called NDIR detection, the acronym NDIR meaning non-dispersive infrared. Such principles are frequently used and disclosed, for example, in US Pat. No. 5,026,992 and WO 2007/064370.

Conventional methods generally involve measuring a light wave emitted from a light source, called a reference light wave, which is not or negligibly absorbed by the analyte gas. Measurement of the reference light wave makes it possible to estimate the intensity of the light wave emitted by the light source or to estimate the light wave that would be detected at the measurement photodetector in the absence of absorption by the analyte gas. This technique is called "dual beam". A comparison of the light wave in the presence of gas and the light wave in the absence of gas makes it possible to characterize the gas absorption. For example, a technique called "absorption NDIR" can be used to determine the amount of gas species in a gas.

A reference light wave is measured by a reference light detector. This can be a reference light detector, different from the measurement light detector, arranged opposite the light source, which reference light detector is used together with the reference optical filter. This reference optical filter defines a reference spectral band in which the analyte gas exhibits no significant absorption.

One approach, disclosed in U.S. Patent Application Publication No. 2011/0042570, uses a measurement photodetector and a reference photodetector, where the two photodetectors are in the same spectral band, In this case, light waves in the absorption spectral band of _CO2 are detected. The reference photodetector is placed closer to the light source than the measurement photodetector. A comparison of the signals respectively measured by the measurement and reference photodetectors can obviate the knowledge of the intensity of the light waves emitted by the light source.

FR 3000548 discloses a CO ₂ sensor comprising a measurement channel in the infrared spectral band and a reference channel (0.4 μm to 0.8 μm) in the visible spectral band. . This reference channel is assumed to be insensitive to the _CO2 concentration in the measured gas. In order to account for variations in the emission spectrum of the light source, this document mentions the use of a function F representing the aging of the light source in the visible and infrared spectral bands, respectively. Since this function F is approximated by an identity function, the temporal change of the light source in the infrared is considered to be equal to the temporal change of the light source in the visible.

U.S. Pat. No. 5,026,992 WO2007/064370 U.S. Patent Application Publication No. 2011/0042570 French Patent No. 3000548

The inventors have found that relying on a reference light wave has several drawbacks. The inventor proposes a method to overcome these drawbacks and improve the measurement accuracy.

A first subject of the present invention is
A method for measuring the amount of gas species present in a gas and capable of absorbing light within an absorption spectral band, comprising:
a) placing said gas between a light source and a measurement photodetector, wherein said light source is capable of emitting an incident light wave, said incident light wave propagating through said gas to said measurement photodetector; ) and
b) illuminating said gas with said light source;
step c) of measuring, with said measuring photodetector, the intensity, called measuring intensity, of the light wave transmitted by said gas in a measuring spectral band comprising an absorption spectral band;
d) using a reference light detector to measure the intensity, called the reference intensity, of the reference light wave emitted by said light source in a reference spectral band;
steps b)-d) are performed at a plurality of measurement times;
At each measurement time,
e) calculating the absorption of the incident light wave by the gas based on the reference intensity measured at the reference photodetector and the measured intensity measured at the measurement photodetector;
f) calculating the amount of said gas species based on the amount of absorption calculated in step e),
In step e), the method factors in a correction function representing the temporal variation of the intensity of the incident light wave in the measurement spectral band with respect to the intensity of the incident light wave in the reference spectral band.

The light source may include a filament raised to a temperature that allows emission of light within the illumination spectral band.

The correction function represents a comparison between the intensity of the incident light wave in the measured spectral band and the intensity of the incident light wave in the reference spectral band, and the comparison can take different values from measurement to measurement. .

Said comparison can be expressed in the form of a ratio or a difference.

Said correction function is preferably:
step cal-i) placing a test light source opposite a measurement test light detector and opposite a reference test light detector (wherein said test light source, said measurement test light detector and said reference test light detectors respectively represent the light source, the measurement light detector, and the reference light detector);
step cal-ii) illuminating said measurement test photodetector and said reference test photodetector with said test light source for calibration times within a calibration period;
step cal-iii of comparing the intensity over time detected at the measurement test photodetector in the measurement spectral band with the intensity over time detected at the reference test photodetector in the reference spectral band. ) is preset in the calibration stage.

The test light source can be pulsed and each pulse can correspond to one calibration time. The calibration period can include at least 1000 calibration times.

The correction function, at various calibration times, is:
an intensity detected by the measurement test photodetector normalized with the initial intensity detected by the measurement test photodetector;
It can be calculated based on a comparison between the intensity detected by the reference test photodetector and normalized with the initial intensity detected by the reference test photodetector.

By initial intensity is meant the intensity measured at the beginning of the calibration period.

Step e) is based on the reference intensity and the correction function measured at the measurement time, and in the absence of gas at the measurement time, detected by the measuring photodetector in the measured spectral band. It may also include the step of estimating the intensity that will be Step e) may comprise correcting the measured intensity based on the reference intensity measured at the measurement time and the correction function, wherein the corrected measured intensity is free of aging of the light source. corresponds to the measured intensity of the case.

A second subject of the present invention is
A device for measuring the amount of gas species in a gas, comprising:
a light source configured to emit an incident light wave that is within the absorption spectral band of the gas species and that propagates through the gas;
a measurement photodetector configured to detect the light wave transmitted by the gas in a measurement spectral band and to measure the intensity of the light wave, referred to as the measurement intensity, at various measurement times;
a reference light detector configured to measure the intensity, called the reference intensity, of a reference light wave emitted from said light source in a reference spectral band at different times of measurement;
a processor for performing steps e) and f) of the method of the first subject of the invention on the basis of said reference intensity and said measured intensity.

Other advantages and features will become more apparent from the following description of specific embodiments of the invention. These embodiments are given as non-limiting examples and illustrated in the figures listed below.

1 illustrates an example of an apparatus in which the invention may be implemented; FIG. Fig. 2 schematically shows the emission spectrum of a black body type light source;

Fig. 3 shows that a decrease in the light intensity emitted by the light source is observed in two different spectral bands;

FIG. 4 shows the loss of emissivity of the light source in the measured spectral band as a function of the loss of emissivity of the light source in the reference spectral band;

FIG. 4 shows the loss of emissivity of the light source in the measured spectral band as a function of the loss of emissivity of the light source in the reference spectral band for three different light source supply voltages.

Fig. 3 shows the reference intensity that is taken into account to compensate for the loss of emissivity of the light source in the conventional method and the method of the invention, respectively;

1 shows the main steps of a method embodying the invention; FIG.

FIG. 1A is an example of an apparatus 1 for analyzing gases. The device comprises a chamber 10 defining an interior space, the interior space containing:
a light source 11 capable of emitting a light wave 12, called the incident light wave, so as to illuminate the gas G present in the interior space, where the incident light wave 12 is within the illumination spectral band Δ ₁₂ ;
A photodetector 20, called a measurement photodetector, arranged to detect a light wave 14 transmitted by the gas G under the influence of illumination of the gas G by an incident light wave 12 (here, the light wave 14 is called a measurement light wave). is detected in the measurement spectral band Δ ₂₀ by the measurement photodetector 20), and
A reference photodetector 20 _ref configured to detect the _light wave 12 _ref , called the reference light wave, in a reference spectral band Δ _ref , where the absorption of the light wave 12 by the gas G is considered negligible. are the spectral bands where the

The reference spectral band Δ _ref is different from the measured spectral band Δ ₂₀ . The measured spectral band _Δ20 may be significantly broader than the reference spectral band _Δref . The measured spectral band Δ ₂₀ can include the reference spectral band Δ _ref .

The gas G contains a gas species G _x whose quantity (c _x (k)), eg its concentration, is sought to be determined at some measurement time k. This gas species absorbs a measurable percentage of light within the absorption spectral band _Δx .

_The light source 11 may emit an incident light wave 12 in an illumination spectral band Δ ₁₂ , which is between the near ultraviolet and mid-infrared, e.g. between 200 nm and 10 μm, often 1 μm. It can extend between 10 μm. The absorption spectral band _Δx of the gas species _Gx to be analyzed is contained in the illumination spectral band _Δ12 . The light source 11 may in particular be pulsed, in which case the pulse duration of the incident light wave 12 is typically between 100 ms and 1 s. The light source 11 can in particular be a suspended filament light source heated to a temperature of 400-800° C., whose emission spectrum corresponds to that of a black body in the emission spectral band _Δ12 .

The measurement photodetector 20 is preferably used with an optical filter 18 that defines a measurement spectral band Δ ₂₀ that encompasses all or part of the absorption spectral band _Δx of the gaseous species.

In this example, the measuring photodetector 20 is a thermopile, capable of sending a signal dependent on the intensity of the detected light waves. Alternatively, the measurement photodetector may be a photodiode or another type of photodetector.

The reference photodetector 20 _ref is arranged next to the measurement photodetector 20 and is of the same type as the measurement photodetector 20 . It is used with an optical filter called reference optical filter 18 _ref . The reference optical filter 18 _ref defines a reference spectral band Δ _ref corresponding to the range of wavelengths not absorbed by the gas species of interest. The reference spectral band Δ _ref is centered, for example, at a wavelength of 3.91 μm.

ここで、
μ（ｃ_ｘ（ｋ））は、回数ｋにおける量ｃ_ｘ（ｋ）に依存する吸収係数であり、
ｌは、チャンバ１０内で光波によって通過されるガスの厚さであり、
Io（ｋ）は、チャンバー内に吸収性ガスが存在しない状態で測定光検出器２０に到達する光波の、測定スペクトルバンドΔ_２０における強度に対応する、回数ｋにおける入射光波の強度である。 The intensity I(k), called the measured intensity, of the light wave 14 detected by the measurement photodetector 20 at a measurement epoch k depends, according to the Beer-Lambert equation, on the quantity c _x (k) at a measurement epoch.

here,
μ(c _x (k)) is the absorption coefficient dependent on the quantity c _x (k) at times k,
l is the thickness of the gas passed by the light wave in the chamber 10;
Io(k) is the intensity of the incident light wave at times k corresponding to the intensity in the measured spectral band Δ20 of the light wave reaching the measurement photodetector ₂₀ in the absence of absorbing gas in the chamber.

Comparing I(k) and I ₀ (k) in the form of the ratio I(k)/I ₀ (k) to determine the absorption abs(k) caused by the gas species of interest at times k can be done.

Therefore, if μ(c _x (k)) can be determined during each pulse of light source 11 and the relationship between c _x (k) and μ(c _x (k)) is known, This makes it possible to calculate c _x (k).

Equation (1) presupposes control of the intensity Io(k) of the incident light wave 12 at measurement times k.

ここで、
Ｌ（λ，Ｔ）は、黒体の波長λと表面温度Ｔに依存する放射輝度であり、
ｈは、プランク定数であり、
ｋは、ボルツマン定数であり、
ｃは、空気中の光の速度である。 FIG. 1B schematically shows the emission spectrum of a black body type light source 11 according to Planck's law.

here,
L(λ, T) is the radiance that depends on the wavelength λ of the black body and the surface temperature T,
h is Planck's constant,
k is the Boltzmann constant,
c is the speed of light in air.

The emission spectrum S of the light source 11 corresponds to the change in radiance L(λ,T) as a function of λ as the source increases to a temperature T. Generally, the temperature T is between 400 and 800°C.

FIG. 1B shows an illumination spectral band Δ ₁₂ extending between 1 μm and 10 μm of light source 11 . The reference spectral band Δ _ref is also indicated by a dashed line.

This type of light source is particularly advantageous because the illumination spectrum S can be modulated simply by modulating the temperature T of the light source. Thus, for each temperature T, one illumination spectrum S is associated.

を算定することができる。参照強度はまた、光源１１の経時変化を考慮するように測定強度Ｉ（ｋ）を補正することを可能にする。 It is known that the emissivity of blackbody or graybody type light sources varies over time and decreases significantly as a result of aging of the light source. This time variation in the emission of the light source 11 is taken into account by the reference photodetector 20 _ref . The reference light detector 20 _ref is arranged to detect a reference light wave 12 _ref representative of the incident light wave 12 emitted by the light source 11 . This reference light wave 12 _ref reaches the reference light detector 20 _ref without interacting with the gas G or without interacting with the gas G significantly. The intensity of the reference light wave 12 _ref detected by the reference photodetector 20 _ref at measurement time k is denoted by the term reference intensity I _ref (k). If the emission spectrum of the light source 11 is known, from I _ref (k), the intensity of the light wave that would reach the measuring photodetector 20 in the absence of gas G

can be calculated. The reference intensity also allows the measured intensity I(k) to be corrected to account for changes in the light source 11 over time.

The apparatus includes a microprocessor 30 connected to a memory 32 containing instructions enabling the steps of the method of the invention (described below) to be performed.

を算定する。 According to a first embodiment, the microprocessor 30 is arranged to receive a signal representative of the intensity I _ref (k) of the reference lightwave 12 _ref measured by the reference photodetector 20 _ref at each measurement time k. . This microprocessor 30 converts I _ref (k) into intensity

to calculate

式（１）を用いて、μ（ｃ_ｘ（ｋ））、次いでｃ_ｘ（ｋ）が得られる。 Based on I(k), the absorption of the incident light wave can be calculated using the following equation.

Using equation (1), μ(c _x (k)) and then c _x (k) are obtained.

ここで、
Ｉ₀（ｋ＝０）は、チャンバ内に吸収性ガスが存在しない状態で、初期測定回ｋ＝０において、すなわち、光源１１が新品であると考えられるときに、測定光検出器に入射する光波を表す。
式（１）を用いて、μ（ｃ_ｘ（ｋ））、次いでｃ_ｘ（ｋ）が得られる。 According to a second embodiment, the microprocessor 30 is arranged to receive a signal representative of the reference intensity I _ref (k) and then perform a correction of the measured intensity I(k). Denote the corrected intensity by I ^* (k). This corrected intensity corresponds to the intensity that would be measured by the measuring photodetector in the absence of aging of the light source. The absorption amount abs(k) of the incident light wave can be obtained by the following equation.

here,
I ₀ (k=0) is incident on the measurement photodetector at the initial measurement time k=0, i.e. when the light source 11 is considered new, with no absorbing gas present in the chamber. Represents light waves.
Using equation (1), μ(c _x (k)) and then c _x (k) are obtained.

のような式をを用いて、
強度

は、簡単にＩ_ｒｅｆ（ｋ）から算定される；
第２実施形態が実施されるとき、補正関数を適用して、Ｉ_ｒｅｆ（ｋ）から補正強度Ｉ^＊（ｋ）
が得られる。

The ratio of the emissivity of the light source 11 in the reference spectral band Δ _ref and the measured spectral band Δ ₂₀ is generally considered to decrease similarly under the following assumptions.
When the first embodiment is practiced, based on knowledge of the theoretical emission spectrum of the light source, or

Using an expression like
Strength

is easily calculated from I _ref (k);
When the second embodiment is implemented, a correction function is applied to obtain the corrected intensity I ^* (k) from I _ref (k)
is obtained.

However, the inventors have found that aging of the light source 11 affects the reference spectral band Δ _ref and the measured spectral band Δ ₂₀ differently. In contrast to what is suggested in FR 3 000 548, it cannot be assumed that the time course in the measured spectral band is similar to the time course in the reference spectral band _Δref . To establish this, the inventor performed an experimental calibration described below with reference to FIGS. 2A-2C. He used a measurement test sensor 20' and a reference test sensor 20' _ref similar to the measurement and reference sensors, respectively, described with reference to FIG. 1A. During calibration, the analyzed gas was a known gas, in this example _CO2 with a concentration of 400 ppm. The experimental parameters were as follows.
Measurement filter 18: Heimann F4.26-180 filter, center wavelength 4.26 μm.
Reference filter 18 _ref : Heimann F3.91-90 filter, center wavelength 3.91 μm.
Measurement and reference photodetectors 20, 20 _ref : Heimann HCM Cx2 Fx thermopiles.

In this test, the measurement filter 18 deliberately defined a narrow measurement spectral band Δ ₂₀ so as to be able to clearly show the aging observed by the inventors. It will be appreciated that the present invention is also applicable to other measured spectral bands Δ ₂₀ , especially wider measured spectral bands Δ ₂₀ than the reference spectral band Δ _ref .

A test light source 11', similar to the light source described with reference to FIG. 1A, was pulsed at various times k between an initial calibration time k=0 and a final calibration time k=K. The duration of each pulse was 300 ms and the time interval between subsequent pulses was 300 ms. Approximately 26 million pulses were generated. FIG. 2A shows the following time variations.
Measured intensity I′(k) in measurement spectral band Δ ₂₀ , measured by measurement test photodetector 20′; and Reference intensity I′ in reference spectral band _Δref , measured by reference test photodetector 20′ _ref . _ref (k).

These variations were normalized by the measured and reference intensities at the initial calibration round (k=0), respectively.

The notation I'(k) and I' _ref (k) indicates that these intensities are measured in the calibration stage using test sensors, test light sources, and known gases. During the calibration phase it was possible to measure the aging of the light source 11 of the same nature as the test light source 11'.

The observed variations in each curve correspond to temporal and intentional variations in _CO2 concentration.

It will be seen that during calibration, the measured intensity I'(k) and the reference intensity I' _ref (k) decreased over time, as expected. This corresponds to aging of the light source 11 . It will also be seen that the reduction in the measured spectral band _Δ20 and the reference spectral band _Δref are different. This means that the aging of the light source 11 in the measured spectral band _Δ20 is different than the aging of the light source 11 in the reference spectral band _Δref . Therefore, the ratio I'(k)/I' _ref (k) varies as a function of the number k. This means that the emission spectrum slightly changes as the light source 11 changes over time.

FIG. 2B shows the emissivity loss EL ₂₀ (y-axis) in the measured spectral band as a function of the emissivity loss EL _ref (x-axis) in the reference spectral band. The emissivity loss (%) in each spectral band is obtained for each number k using the following equation.

The variations in the curve of Figure 2B correspond to temporal variations in _CO2 concentration, ie variations as described with reference to Figure 2A. It will be seen that EL ₂₀ varies linearly as a function of EL _ref with a slope A greater than one. In FIG. 2B the curve EL ₂₀ =EL _ref is drawn with a dashed line.

A similar test was conducted while changing the supply potential V of the light source 11 . Three identical photodetectors were used, each facing one of the three light sources of the same type. The potentials of the three light sources were V=1.48 V (which corresponds to the elevated potential of the light source in the tests reported in FIG. 2B), V=1.28 V, and V=1.18 V, respectively. there were. FIG. 2C shows the curve EL ₂₀ as a function of EL _ref for each potential V, respectively. As the potential decreases, the loss of emissivity is lower than at high potentials because the aging of the light source 11 is less noticeable. However, it will be seen that the three curves overlap. Therefore, the effects of time-varying emission in the measured spectral band Δ ₂₀ and the reference spectral band Δ _ref are believed to be independent of the increase in the potential of the light source 11 .

The test described with reference to FIGS. 2A-2C can be considered a calibration test that measured the difference in aging of the light source 11 in the measured spectral band Δ ₂₀ relative to the reference spectral band Δ _ref . This test was performed with a test sensor having components similar to those provided by sensors for analyzing unknown gases.

Such a calibration test makes it possible to determine a correction function δ characterizing the relative emissivity change in the two spectral bands Δ ₂₀ and Δ _ref . The correction function δ involves comparing the intensity of the measured spectral band with the reference intensity at each iteration k.

In a first approach, the correction function .delta.

A is the slope of the line obtained by applying linear regression to the data shown in FIGS. 2B and 2C. A is a scalar value representing the difference in illuminant aging in each spectral band.

本発明の実施に対応する第２の補正（曲線２）。この補正は、次式のように、時間（time、回数）の関数として、Ｉ_ｒｅｆ（ｋ）に対して変化する補正関数δを適用して得られる。

FIG. 3 shows an example of the time change of the measured intensity I(k). This variation corresponds to the variation measured in FIG. 2A. The following two corrections to the measured signal I(k) are shown in FIG.
A first correction (curve 1) corresponding to the prior art. This correction is obtained by applying a constant correction factor, as follows:

A second correction (curve 2) corresponding to the practice of the invention. This correction is obtained by applying a correction function δ that varies with I _ref (k) as a function of time, as follows:

During tests performed at V=1.48 V, the drift ε(K) affecting the correction of I(k) at times K was calculated and shown in FIG. Drift ε(K) is expressed in % according to the following equation.

You will find the following.
When applying the first correction according to equation (6), the ε(K) value is 1.34%;
When applying the second correction according to equation (10), the ε(K) value is 0.08%.

FIG. 4 shows the main steps of a measurement method embodying the invention.

Step 100: At some number k, illuminate the gas.

Step 110: Measure the reference intensity I _ref (k) in the reference spectral band Δ _ref using the reference photodetector 20 _ref .

Step 120: Using the measurement photodetector 20, measure the intensity I(k) of the radiation 14 transmitted by the gas in the measurement spectral band _Δ20 .

を算定する。この算定は、補正関数δ（ｋ）を考慮し、次式を適用して実行される。

は、ｋ＝０における強度

の算定に対応する。 Step 130: The intensity that would be detected in the measured spectral band Δ ₂₀ by the measuring photodetector 20 in the absence of gas in the chamber.

to calculate This calculation is performed by taking into account the correction function .delta.(k) and applying the following equation.

is the intensity at k=0

corresponds to the calculation of

を算定する。 Step 140: Absorption in the measured spectral band _Δ20

to calculate

Step 150: Based on the absorption, apply equation (1) to calculate the quantity c _x (k) of the gas species G _x from the ratio.

Step 160: Increase the number of measurements k and repeat steps 100-150 or terminate the algorithm.

This embodiment follows the calibration only with the value of the time-varying difference A, in order to be able to apply the correction function δ for each measurement of the reference intensity I _ref (k) measured at each measurement time k. , which is advantageous.

It is also possible to consider other expressions for the correction function δ. For example, the correction function δ may be the following equation.

の算定値が、次式によって得られる。

Next, in step 130,

is given by the following equation:

補正関数δはまた、式（１４）で表すことができ、この場合、

である。 According to a variant, in step 130 the intensity values I(k) measured by the measurement sensor are corrected using a correction function δ. A corrected intensity I ^* (k) is obtained. The correction function δ can be expressed as in Equation (9).

The correction function δ can also be expressed in equation (14), where

is.

According to this variant, in step 140 the absorption is obtained using the equation:

の算定;
または、光源の経時変化がない場合に測定光検出器によって測定されるであろう補正強度Ｉ^＊（ｋ）。 Thus, in general, the calibration stage makes it possible to assess the relative decrease over time of the intensity of the illumination radiation produced by the light source in the reference spectral band and the measurement spectral band. A correction function includes a comparison of the reduction in each spectral band. The use of the correction function δ(k) takes into account variations in the intensity reduction of the illumination radiation 12 in the two spectral bands, making it possible to obtain the following.
Intensity that would be measured by the measuring photodetector in the absence of gas

calculation of;
Or the corrected intensity I ^* (k) that would be measured by the measuring photodetector in the absence of light source aging.

The present invention can be used to detect the amount of gas species G _x whose absorption spectrum Δ ₂₀ falls within the measured spectral band Δ ₂₀ . The measured spectral band Δ ₂₀ may be narrow, as in the experimental example above. The measured spectral band Δ ₂₀ may also be broad, eg, to include absorption spectral bands Δ _x of a plurality of different gas species.

Claims (7)

A method of measuring the amount (c _x ) of gas species (G _x ) present in a gas and configured to absorb light within an absorption spectral band (Δ _x ), comprising:
a) placing said gas between a light source (11) arranged to emit an incident light wave (12) and a measuring light detector (20);
b) illuminating said gas (G) with said light source (11), whereby said incident light wave propagates through said gas to said measurement light detector (20);
using said measuring photodetector (20) to measure the intensity (I(k)) of the light wave (14) transmitted by said gas in a measuring spectral band (Δ ₂₀ ) including an absorption spectral band (Δ _x ); step c) of measuring;
measuring a reference intensity (I _ref (k)) of a reference light wave (12 _ref ) emitted by said light source (11) in a reference spectral band (Δ _ref ) using a reference photodetector (20 _ref ); d) and
steps b)-d) are performed at a plurality of measurement times (1, k, K),
At each measurement time,
using the reference intensity (I _ref (k)) measured with the reference photodetector and the measured intensity (I(k)) measured with the measurement photodetector, the incident light wave (12) by the gas; step e) of calculating the absorption (abs(k)) of
and step f) of calculating the amount (c _x (k)) of the gas species (G _x ) based on the amount of absorption calculated in step e),
In step e) a correction function representing the temporal variation of the intensity of the incident light wave (12) in the measured spectral band (Δ ₂₀ ) with respect to the intensity of the incident light wave (12) in the reference spectral band (Δ _{ref )} . Taking into account (δ),
The correction function (δ) is
step cal-i) of placing a test light source (11') opposite a measurement test light detector (20') and opposite a reference test light detector ( _20'ref ) (wherein said test light source , said measurement test photodetector and said reference test photodetector are respectively similar to said light source (11), said measurement photodetector (20) and said reference photodetector ( _20ref ). ,
step cal-ii) illuminating said measurement test photodetector and said reference test photodetector with said test light source for calibration times within a calibration period;
time variation of intensity (I'(k)) detected at the measurement test photodetector in the measurement spectral band (Δ ₂₀ ) and at the reference test photodetector in the reference spectral band (Δ _ref ) a step cal-iii) comparing the detected intensity (I' _ref (k)) with the time variation. The correction function (δ(k)) is between the intensity of the incident light wave (12) in the measured spectral band (Δ ₂₀ ) and the intensity of the incident light wave (12) in the reference spectral band (Δ _ref ). 2. The method of claim 1, wherein said comparison takes a different value for each measurement. Method according to claim 1 or 2, wherein said test light source (11') is pulsed, each pulse corresponding to one calibration time, said calibration period comprising at least ¹⁰³ calibration times. . The correction function (δ) is, at various calibration times,
In an initial calibration round, the intensity (I' (k)) and
In an initial calibration round, the intensity (I _{' ref} (k ))). Step e) comprises, at said measurement time (k), based on said reference intensity (I _ref (k)) measured at said measurement time (k) and said correction function (δ), in the absence of gas: , calculating the intensity that will be detected in the measurement spectral band (Δ ₂₀ ) by the measurement photodetector (14). Step e) includes correcting the measured intensity based on the reference intensity (I _ref (k)) measured at the measurement times (k) and the correction function (δ), A method according to any one of the preceding claims, wherein the measured intensity (I ^* (k)) corresponds to the measured intensity in the absence of aging of the light source. A device (1) for measuring the amount (c _x (k)) of a gas species (G _x ) in a gas,
a light source (11) configured to emit an incident light wave (12) lying within an absorption spectral band (Δ _x ) of said gas species (G _x ) and propagating in said gas (G);
a measurement light detector adapted to detect the light wave (14) transmitted by said gas in a measurement spectral band at different number of measurements (k) and to measure the measured intensity (I(k)) of said light wave. a vessel (20);
Reference light detection arranged to measure a reference intensity (I _ref (k)) of a reference light wave (12 _ref ) emitted from said light source (11) in a reference spectral band at different number of measurements (k). a device (20 _ref );
for performing steps e) and f) of the method according to any one of claims 1 to 6 on the basis of said reference intensity (I _ref (k)) and said measured intensity (I(k)) and a processor (30) of