A method for measuring NO2 concentration based on DOAS

Abstract

The present invention discloses a NO2 concentration measurement method based on DOAS, which includes the following steps: S1, obtaining original spectral data of different power intensities for preprocessing; S2, calculating the spectral correction coefficient and verifying it; S3, collecting the transmission spectra of NO2 of different concentrations for processing to obtain NO2 spectral data; S4, calculating the relationship between each spectral point and the concentration, calculating the optimal differential absorbance position data set and the optimal differential absorption cross section; S5, establishing an equation through the differential absorbance of the actually measured gas, the optimal differential absorbance position data set and the optimal differential absorption cross section, and obtaining the NO2 concentration value; S6, using the least squares method to correct the NO2 concentration value. The spectral data is corrected by the exponential spectral correction method to overcome the defects of poor stability and repeatability of the xenon lamp light source, thereby greatly improving the accuracy of the spectral data; improving the NO2 measurement accuracy and meeting the actual application needs.

Description

Technical Field

[0001] The present invention relates to the technical field of gas concentration measurement, and in particular to a NO2 concentration measurement method based on DOAS. Background Art

[0002] Based on differential optical absorption spectroscopy (DOAS), SO2, NO x Multi-component gas detection with low concentration and high sensitivity and little cross-interference from moisture and other gases has been increasingly studied at home and abroad in recent years, and has been successfully used in places where pollution sources are concentrated, such as coal-fired power plants, paper mills and cement industries.

[0003] As the country's requirements for environmental protection become increasingly stringent, the emission concentrations of particulate matter, sulfur dioxide, and nitrogen oxides in power plant flue gas emissions must be stably maintained at 10 mg / m3, 35 mg / m3, and 50 mg / m3, respectively. Therefore, how to effectively solve the problem of SO2 and NO x The issue of gas emission concentration measurement accuracy is currently an urgent need in the industrial field.

[0004] DOAS-based analyzers mostly use pulsed xenon lamps as the light source, because of their long life, high power and no need for preheating. However, the poor stability and repeatability of xenon lamps have always been a difficult problem to overcome in DOAS technology. In addition, the absorption cross section of NO2 in the ultraviolet band is much smaller than that of NO and SO2, which has also caused nearly half of the analyzers on the market to use NO x Converter, convert NO2 into NO for measurement, using NO x On the one hand, the converter makes the instrument structure complex and costly; on the other hand, it causes NO2 loss during the conversion process, making it impossible to accurately measure the actual NO2 emission concentration. Summary of the invention

[0005] In view of the problems in the related art, the present invention proposes a NO2 concentration measurement method based on DOAS to overcome the above-mentioned technical problems existing in the existing related technology.

[0006] To this end, the specific technical solution adopted by the present invention is as follows:

[0007] A method for measuring NO2 concentration based on DOAS, the method comprising the following steps:

[0008] S1, obtaining raw spectral data of different power intensities for preprocessing;

[0009] S2, calculate the spectral correction coefficient and verify it;

[0010] S3, collecting and processing transmission spectra of NO2 at different concentrations to obtain NO2 spectrum data;

[0011] S4, calculating the optimal differential absorbance position data set and the optimal differential absorption cross section by calculating the relationship between each spectral point and the concentration;

[0012] S5, establishing an equation by using the differential absorbance of the actually measured gas, the optimal differential absorbance position data set and the optimal differential absorption cross section to obtain the NO2 concentration value;

[0013] S6. Correct the NO2 concentration value using the least square method.

[0014] Furthermore, the acquisition of raw spectral data of different power intensities for preprocessing comprises the following steps:

[0015] S11, selecting a wavelet basis function and a decomposition layer number N, and performing N-layer multi-scale decomposition on the signal;

[0016] S12, reconstructing the low-frequency coefficients of the Nth layer to obtain a reconstructed spectrum;

[0017] S13, adjusting the power of the xenon lamp light source, introducing pure nitrogen into the gas chamber, and collecting multiple groups of original spectral data with different power intensities;

[0018] S14, filtering each piece of original spectrum data using wavelet transform;

[0019] S15, calculating the average value of the spectral data of different intensities after filtering to obtain the preprocessed spectral data.

[0020] Furthermore, the wavelet transform is a signal f(t)∈L 2 The integral transformation of (R), the operation process and expression include:

[0021]

[0022] Where a and b represent the scaling and translation factors of Ψ(t), respectively;

[0023] T represents the signal time, t∈(-∞,+∞);

[0024] It is the result of translation and scaling of Ψ(t);

[0025] Yes ab The complex conjugate of (t);

[0026] The inverse transformation process expression is:

[0027]

[0028] in, To allow conditions.

[0029] Further, the calculation of the spectral correction coefficient and verification include the following steps:

[0030] S21. Select at least one band in the bands that SO2, NO and NO2 do not absorb;

[0031] S22, the spectral data corresponding to the pre-processing of the light source power of 800, 775 and 750 are Lamp_800, Lamp_775 and Lamp_750 respectively, and Lamp_750 is used as the reference spectrum to calculate the correction coefficient, the expression is:

[0032] R1=log(Lamp_750 / Lamp_775)

[0033]

[0034] Among them, R1 is the logarithm of the ratio of two power spectrum data, which represents the change of other power spectrum data relative to the fixed power spectrum data;

[0035] K represents the correction factor;

[0036] Lamp_775 represents the spectrum data when the light source power is 775;

[0037] Lamp_750 represents the spectrum data when the light source power is 750;

[0038] 253.6~256 represents the band that SO2, NO and NO2 do not absorb;

[0039] S23. Calibrate the spectral data using the correction coefficient, the expression is:

[0040] R2=log(Lamp_750 / Lamp_800)

[0041] NEW_Lamp_800=exp(K*mean(R2(253.6~256)))*Lamp_800

[0042] Wherein, R2 represents the change of other power spectrum data relative to the fixed power spectrum data;

[0043] Lamp_800 represents the spectrum data when the light source power is 800;

[0044] New_Lamp_800 represents the spectral data after Lamp_800 correction.

[0045] Furthermore, the collection and processing of transmission spectra of NO2 at different concentrations to obtain NO2 spectrum data includes the following steps:

[0046] S31, keeping the light source power unchanged, respectively collecting multiple groups of NO2 spectrum data with different concentrations;

[0047] S32, filtering the spectral data of each concentration using wavelet transform;

[0048] S33, obtaining the average value of the multiple sets of spectral data after filtering;

[0049] S34, using the correction coefficient to calibrate the average value to obtain calibrated NO2 spectrum data.

[0050] Furthermore, the method of calculating the optimal differential absorbance position data set and the optimal differential absorption cross section by calculating the relationship between each spectral point and the concentration comprises the following steps:

[0051] S41, according to the range of spectrum collected by the spectrometer used and the absorption cross sections of SO2, NO and NO2, the spectrum within the range of 390nm to 415nm is selected as the band for calculating the NO2 concentration;

[0052] S42. According to the differential absorption spectroscopy technique, the differential absorbance and differential absorption cross section of NO2 at concentrations of [0, 20, 40, 60, 80, 100] ppm are obtained;

[0053] S43, calculating the correlation coefficient between the concentration value [0, 20, 40, 60, 80, 100] ppm and the differential absorbance at each wavelength point in the range of 390 nm to 415 nm, and selecting the wavelength points corresponding to the correlation coefficient greater than 0.99 to form the optimal differential absorbance position data set;

[0054] S44, selecting the position where the differential absorption cross section is located corresponding to the optimal differential absorption position data set to form an optimal differential absorption cross section.

[0055] Furthermore, the operation expression of the differential absorption spectroscopy technique includes:

[0056]

[0057]

[0058] σ i (λ)=σ i,slow (λ)+σ i,rapid (λ)

[0059] Where I0(λ) represents the light source intensity of the incident light at wavelength λ;

[0060] I(λ) represents the light source intensity of the emitted light at wavelength λ;

[0061] L represents the optical path;

[0062] A(λ) represents the transfer function of the system;

[0063] c i represents the concentration of the i-th gas;

[0064] σ i (λ) represents the absorption cross section of the i-th gas;

[0065] ε R (λ) and ε M (λ) represents the extinction coefficient of Rayleigh scattering and Mie scattering, respectively;

[0066] D(λ) represents the optical thickness of the material;

[0067] σ i,slow (λ) represents the part that changes slowly with wavelength;

[0068] σ i,rapid (λ) represents the portion that changes rapidly with wavelength, i.e., the differential absorption cross section of the gas.

[0069] Furthermore, the equation is established by actually measuring the differential absorbance of the gas, the optimal differential absorbance position data set and the optimal differential absorption cross section, and the expression is:

[0070] C=(P′·P) -1 ·P′·D

[0071] Wherein, C represents the NO2 concentration value;

[0072] D represents the gas differential absorbance after selection using the optimal differential absorbance position data set;

[0073] P represents the optimal differential absorption cross section, and P′ represents the transpose of P.

[0074] Furthermore, the use of the least square method to correct the NO2 concentration value includes the following steps:

[0075] S61, using the least square method to fit a cubic polynomial to obtain the relationship between the NO2 concentration value and the actual concentration value, and obtaining a cubic polynomial relationship;

[0076] S62, substituting the NO2 concentration value into the cubic polynomial relationship to obtain a corrected concentration value.

[0077] Furthermore, the expression of the cubic polynomial is:

[0078] Fit_c = polyfit(C,y,3)

[0079] con=polyval(Fit_c,C)

[0080] Among them, polyfit() is the polynomial fitting function in Matlab;

[0081] C represents the NO2 concentration value;

[0082] y represents the corresponding actual concentration value;

[0083] 3 is the number of polynomial fits;

[0084] Fit_c represents the coefficients of the cubic polynomial obtained by fitting;

[0085] polyval() represents a polynomial evaluation function;

[0086] con represents the corrected concentration value.

[0087] The beneficial effects of the present invention are as follows: the spectral data is corrected by the exponential spectral correction method to overcome the defects of poor stability and repeatability of the xenon lamp light source, thereby greatly improving the accuracy of the spectral data; and then the correlation coefficient method is used to obtain the correlation coefficient between concentration and differential absorbance, and then the optimal differential absorbance position data set and the optimal differential absorption cross section are constructed, thereby improving the NO2 measurement accuracy and meeting the needs of practical applications.

[0088] In addition, in order to meet practical applications, reduce chip calculations, and improve response time, wavelet transform is used to remove high-frequency noise in spectral data and obtain the slowly changing part, and only the corresponding low-frequency part is reconstructed according to the spectral absorption frequency range. BRIEF DESCRIPTION OF THE DRAWINGS

[0089] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0090] Figure 1 is a flow chart of a method for measuring NO2 concentration based on DOAS according to an embodiment of the present invention;

[0091] Figure 2 is a curve diagram of spectrum preprocessing data in a NO2 concentration measurement method based on DOAS according to an embodiment of the present invention;

[0092] Figure 3is a comparison diagram of two different spectral calibration methods in a NO2 concentration measurement method based on DOAS according to an embodiment of the present invention;

[0093] Figure 4 It is a curve diagram of obtaining a differential absorption cross section in a NO2 concentration measurement method based on DOAS according to an embodiment of the present invention;

[0094] Figure 5 4 is a correlation coefficient curve diagram between differential absorbance and concentration in a NO2 concentration measurement method based on DOAS according to an embodiment of the present invention. DETAILED DESCRIPTION

[0095] According to an embodiment of the present invention, a method for measuring NO2 concentration based on DOAS is provided.

[0096] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to the NO2 concentration measurement method based on DOAS according to an embodiment of the present invention, the method comprises the following steps:

[0097] S1. Obtaining raw spectral data of different power intensities for preprocessing, including the following steps:

[0098] S11, select a wavelet basis function and a decomposition layer number N, and perform N-layer multi-scale decomposition on the signal (in the present invention, the basis function is selected as "bior2.8" and the decomposition layer number is 2 after testing);

[0099] S12, reconstructing the low-frequency coefficients of the Nth layer (it has been verified that the high-frequency part basically does not contain useful information, so in order to reduce the amount of calculation, the high-frequency coefficients are not subjected to threshold quantization processing) to obtain a reconstructed spectrum;

[0100] S13, adjusting the power of the xenon lamp light source, introducing pure nitrogen into the gas chamber, and collecting multiple groups of original spectral data with different power intensities;

[0101] In the present invention, the power of the xenon lamp light source is adjusted to 800, 775 and 750 respectively, and pure nitrogen is introduced into the gas chamber. Under the same temperature, pressure and other conditions, 10 spectral data are collected for each power. After each spectrum is processed by wavelet transform in steps (a) and (b), the average value of the 10 spectra is calculated to obtain the pre-processed spectral data, such as Figure 2 As shown;

[0102] S14, filtering each piece of original spectrum data using wavelet transform;

[0103] S15, calculating the average value of the spectral data of different intensities after filtering to obtain the preprocessed spectral data.

[0104] At present, the commonly used methods for spectral data denoising include principal component analysis, low-pass filtering, Kalman filtering, Fourier transform, etc. However, the principal component analysis is too computationally intensive, and the Fourier transform lacks the ability to analyze local signals. The absorption band in the spectral data is only a part of the entire acquisition range. Finally, after evaluation, wavelet transform was selected to denoise the spectral data, and Matlab was used for simulation.

[0105] The wavelet transform is a signal f(t)∈L 2 The integral transformation of (R), the operation process and expression include:

[0106]

[0107] Where a and b represent the scaling and translation factors of Ψ(t), respectively;

[0108] t represents the signal time, t∈(-∞,+∞);

[0109] It is the result of translation and scaling of Ψ(t);

[0110] Yes ab The complex conjugate of (t);

[0111] The inverse transformation process expression is:

[0112]

[0113] in, To allow conditions.

[0114] In practical applications, ab The continuous variables a and b in (t) are taken as integer discrete forms, and Ψ ab (t) is expressed as:

[0115]

[0116] The corresponding wavelet transform is expressed as discrete wavelet transform:

[0117] W f (j,k)=(f(t),Ψ j,k (t))

[0118] S2. Calculate the spectral correction coefficient and verify it, including the following steps:

[0119] S21, select at least one band in the band that SO2, NO and NO2 do not absorb (the present invention selects the 253.3-256 nm band);

[0120] S22, the spectral data corresponding to the pre-processing of the light source power of 800, 775 and 750 are Lamp_800, Lamp_775 and Lamp_750 respectively, and Lamp_750 is used as the reference spectrum to calculate the correction coefficient, the expression is:

[0121] R1=log(Lamp_750 / Lamp_775)

[0122]

[0123] Among them, R1 is the logarithm of the ratio of two power spectrum data, which represents the change of other power spectrum data relative to the fixed power spectrum data (other power spectrum data can also be selected);

[0124] K represents the correction factor;

[0125] Lamp_775 represents the spectrum data when the light source power is 775;

[0126] Lamp_750 represents the spectrum data when the light source power is 750;

[0127] 253.6~256 represents the band that SO2, NO and NO2 do not absorb;

[0128] S23. Calibrate the spectral data using the correction coefficient, the expression is:

[0129] R2=log(Lamp_750 / Lamp_800)

[0130] NEW_Lamp_800=exp(K*mean(R2(253.6~256)))*Lamp_800

[0131] Wherein, R2 represents the change of other power spectrum data relative to the fixed power spectrum data;

[0132] Lamp_800 represents the spectrum data when the light source power is 800;

[0133] New_Lamp_800 indicates the spectral data after Lamp_800 correction, that is, the spectral data to be calibrated.

[0134] like Figure 3 As shown, the left figure uses the traditional method, that is, using the non-absorption band to obtain a proportional coefficient, and then directly multiplying it by the spectrum to be calibrated. The right figure is the exponential calibration method proposed in the present invention. It can be clearly seen from the figure that the exponential calibration method is better, because as the power of the light source changes, the light intensity changes in different bands are not absolutely linear.

[0135] S3, collecting transmission spectra of NO2 of different concentrations for processing to obtain NO2 spectrum data, including the following steps:

[0136] S31, keeping the light source power unchanged, respectively collecting multiple groups of NO2 spectrum data with different concentrations;

[0137] In the embodiment of the present invention, the xenon lamp light source power is maintained at 750, and the NO2 spectral data with concentrations of [0, 20, 40, 60, 80, 100] ppm are collected respectively.

[0138] S32, filtering the spectral data of each concentration using wavelet transform;

[0139] S33, obtaining the average value of the multiple sets of spectral data after filtering;

[0140] S34, using the correction coefficient to calibrate the average value to obtain calibrated NO2 spectrum data.

[0141] S4, calculating the optimal differential absorbance position data set and the optimal differential absorption cross section by calculating the relationship between each spectral point and the concentration, comprising the following steps:

[0142] S41, according to the range of spectrum collected by the spectrometer used and the absorption cross sections of SO2, NO and NO2 (provided by the HITRAN database), the spectrum within the range of 390nm to 415nm is selected as the band for calculating the NO2 concentration;

[0143] S42. According to the differential absorption spectroscopy (DOAS) technique, the differential absorbance and differential absorption cross section of NO2 at concentrations of [0, 20, 40, 60, 80, 100] ppm are obtained;

[0144] S43, calculating the correlation coefficient between the concentration value [0, 20, 40, 60, 80, 100] and the differential absorbance at each wavelength point in the range of 390nm to 415nm, and selecting the wavelength points corresponding to the correlation coefficient greater than 0.99 to form the optimal differential absorbance position data set;

[0145] S44, selecting the position where the differential absorption cross section is located corresponding to the optimal differential absorption position data set to form an optimal differential absorption cross section.

[0146] The differential absorption spectroscopy (DOAS) technique is based on the detection of narrow-band absorption characteristics of trace gas molecules. The absorbed light intensity complies with the Lambert Beer absorption law. When Rayleigh scattering, Mie scattering and other molecular absorption are considered, the calculation expression includes:

[0147]

[0148] Spectral detection technology uses this law to measure the average concentration of trace gases along the optical path. is the optical thickness of the material, expressed as D(λ), and its expression is:

[0149]

[0150] In order to eliminate the influence of Rayleigh scattering and Mie scattering, filtering technology is usually used in mathematics to separate the spectral changes caused by molecular absorption contained in the atmospheric absorption spectrum. The absorption cross section of the gas can usually be regarded as consisting of two parts, expressed as:

[0151] σ i (λ)=σ i,slow (λ)+σ i,rapid (λ)

[0152] Where I0(λ) represents the light intensity of the incident light at wavelength λ;

[0153] I(λ) represents the light intensity of the outgoing light at wavelength λ;

[0154] L represents the optical path;

[0155] A(λ) represents the transfer function of the system;

[0156] c i represents the concentration of the i-th gas;

[0157] σ i (λ) represents the absorption cross section of the i-th gas;

[0158] ε R (λ) and ε M (λ) represents the extinction coefficient of Rayleigh scattering and Mie scattering, respectively;

[0159] D(λ) represents the optical thickness of the material;

[0160] σ i,slow (λ) represents the part that changes slowly with wavelength;

[0161] σ i,rapid (λ) represents the portion that changes rapidly with wavelength, i.e., the differential absorption cross section of the gas.

[0162] The total absorption cross section σ i (λ) minus the calculated slow variation σ i,slow (λ) is the differential absorption cross section σ of the gas i,rapid (λ), where the slow-varying σ i,slow(λ) can be obtained by reconstructing the low-frequency part of the wavelet multi-scale decomposition. For example, the slow-changing part of SO2 can be obtained by reconstructing the low-frequency part after the wavelet 5-layer decomposition, such as Figure 4 shown.

[0163] S5, establishing an equation by using the differential absorbance of the actually measured gas, the optimal differential absorbance position data set and the optimal differential absorption cross section to obtain the NO2 concentration value;

[0164] The equation is established by actually measuring the differential absorbance of the gas, the optimal gas differential absorbance position data set and the optimal differential absorption cross section, and the expression is:

[0165] C=(P′·P) -1 ·P′·D

[0166] Wherein, C represents the NO2 concentration value;

[0167] D represents the gas differential absorbance after selection using the optimal differential absorbance position data set;

[0168] P represents the optimal differential absorption cross section, and P′ is the transpose of P.

[0169] S6, using the least square method to correct the NO2 concentration value, including the following steps:

[0170] S61, using the least square method to fit a cubic polynomial to obtain the relationship between the NO2 concentration value and the actual concentration value, and obtaining a cubic polynomial relationship;

[0171] S62, substituting the NO2 concentration value into the cubic polynomial relationship to obtain a corrected concentration value.

[0172] Wherein, the expression of the cubic polynomial is:

[0173] Fit_c = polyfit(C,y,3)

[0174] con=polyval(Fit_c,C)

[0175] Among them, polyfit() is the polynomial fitting function in Matlab;

[0176] C represents the NO2 concentration value;

[0177] y represents the corresponding actual concentration value;

[0178] 3 is the number of polynomial fits;

[0179] Fit_c represents the coefficients of the cubic polynomial obtained by fitting;

[0180] polyval() represents a polynomial evaluation function;

[0181] con represents the corrected concentration value.

[0182] In summary, with the help of the above technical scheme of the present invention, the spectral data is corrected by the exponential spectrum correction method to overcome the defects of poor stability and repeatability of the xenon lamp light source, thereby greatly improving the accuracy of the spectral data; and then the correlation coefficient method is used to obtain the correlation coefficient between concentration and differential absorbance, and then the optimal differential absorbance position data set and the optimal differential absorption cross section are constructed to improve the NO2 measurement accuracy and meet the needs of practical applications.

[0183] In addition, in order to meet practical applications, reduce chip calculations, and improve response time, wavelet transform is used to remove high-frequency noise in spectral data and obtain the slowly changing part, and only the corresponding low-frequency part is reconstructed according to the spectral absorption frequency range.

[0184] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for measuring NO2 concentration based on DOAS, characterized in that: The method comprises the following steps: S1, obtaining raw spectral data of different power intensities for preprocessing; S2, calculate the spectral correction coefficient and verify it; S3, collecting and processing transmission spectra of NO2 at different concentrations to obtain NO2 spectrum data; S4, calculating the optimal differential absorbance position data set and the optimal differential absorption cross section by calculating the relationship between each spectral point and the concentration; S5. Establish an equation by using the differential absorbance of the actually measured gas, the optimal differential absorbance position data set and the optimal differential absorption cross section to obtain the NO2 concentration value; S6. Correcting the NO2 concentration value using the least square method; The calculation of the spectral correction coefficient and verification thereof include the following steps: S21. Select at least one band in the bands that SO2, NO and NO2 do not absorb; S22, the spectral data corresponding to the pre-processing of the light source power of 800, 775 and 750 are Lamp_800, Lamp_775 and Lamp_750 respectively, and Lamp_750 is used as the reference spectrum to calculate the correction coefficient, the expression is: R1=log(Lamp_750 / Lamp_775) Among them, R1 is the logarithm of the ratio of two power spectrum data, which represents the change of other power spectrum data relative to the fixed power spectrum data; K represents the correction factor; Lamp_775 represents the spectrum data when the light source power is 775; Lamp_750 represents the spectrum data when the light source power is 750; 253.6~256 represents the band that SO2, NO and NO2 do not absorb; S23. Calibrate the spectral data using the correction coefficient, the expression is: R2=log(Lamp_750 / Lamp_800) NEW_Lamp_800=exp(K*mean(R2(253.6~256)))*Lamp_800, where R2 represents the change of other power spectrum data relative to the fixed power spectrum data; Lamp_800 represents the spectrum data when the light source power is 800; New_Lamp_800 represents the spectral data after Lamp_800 correction; The method of calculating the optimal differential absorbance position data set and the optimal differential absorption cross section by calculating the relationship between each spectral point and the concentration comprises the following steps: S41, according to the range of spectrum collected by the spectrometer used and the absorption cross sections of SO2, NO and NO2, the spectrum within the range of 390nm to 415nm is selected as the band for calculating the NO2 concentration; S42. According to the differential absorption spectroscopy technique, the differential absorbance and differential absorption cross section of NO2 at concentrations of [0, 20, 40, 60, 80, 100] ppm are obtained; S43, calculating the correlation coefficient between the concentration value [0, 20, 40, 60, 80, 100] ppm and the differential absorbance at each wavelength point in the range of 390 nm to 415 nm, and selecting the wavelength points corresponding to the correlation coefficient greater than 0.99 to form the optimal differential absorbance position data set; S44, selecting the position where the differential absorption cross section is located corresponding to the optimal differential absorption position data set to form an optimal differential absorption cross section.

2. A method for measuring NO2 concentration based on DOAS according to claim 1, characterized in that: The method of obtaining raw spectral data of different power intensities for preprocessing comprises the following steps: S11, selecting a wavelet basis function and a decomposition layer number N, and performing N-layer multi-scale decomposition on the signal; S12, reconstructing the low-frequency coefficients of the Nth layer to obtain a reconstructed spectrum; S13, adjusting the power of the xenon lamp light source, introducing pure nitrogen into the gas chamber, and collecting multiple groups of original spectral data with different power intensities; S14, filtering each piece of original spectrum data using wavelet transform; S15, calculating the average value of the spectral data of different intensities after filtering to obtain the preprocessed spectral data.

3. A method for measuring NO2 concentration based on DOAS according to claim 2, characterized in that: The wavelet transform is a signal f(t)∈L 2 The integral transformation of (R), the operation process and expression include: Where a and b represent the scaling and translation factors of Ψ(t), respectively; t represents the signal time, t∈(-∞,+∞); It is the result of translation and scaling of Ψ(t); Yes ab The complex conjugate of (t); The inverse transformation process expression is: in, To allow conditions.

4. The method for measuring NO2 concentration based on DOAS according to claim 1, characterized in that: The collecting and processing of transmission spectra of NO2 at different concentrations to obtain NO2 spectrum data comprises the following steps: S31, keeping the light source power unchanged, respectively collecting multiple groups of NO2 spectrum data with different concentrations; S32, filtering the spectral data of each concentration using wavelet transform; S33, obtaining the average value of the multiple sets of spectral data after filtering; S34, using the correction coefficient to calibrate the average value to obtain calibrated NO2 spectrum data.

5. A method for measuring NO2 concentration based on DOAS according to claim 4, characterized in that: The operation expression of the differential absorption spectroscopy technique includes: σi(λ)=σi,slow(λ)+σi,rapid(λ) Where I0(λ) represents the light intensity of the incident light at wavelength λ; I(λ) represents the intensity of the transmitted light at wavelength λ; L represents the optical path; A(λ) represents the transfer function of the system; c i represents the concentration of the i-th gas; σ i (λ) represents the absorption cross section of the i-th gas; ε R (λ) and ε M (λ) represents the extinction coefficient of Rayleigh scattering and Mie scattering, respectively; D(λ) represents the optical thickness of the material; σi,slow(λ) represents the part that changes slowly with wavelength; σ i,rapid (λ) represents the portion that changes rapidly with wavelength, i.e., the differential absorption cross section of the gas.

6. The method for measuring NO2 concentration based on DOAS according to claim 1, characterized in that: The equation is established by actually measuring the differential absorbance of the gas, the optimal differential absorbance position data set and the optimal differential absorption cross section, and the expression is: C=(P′·P)-1·P′·D Wherein, C represents the NO2 concentration value; D represents the gas differential absorbance after selection using the optimal differential absorbance position data set; P represents the optimal differential absorption cross section, and P′ represents the transpose of P.

7. The method for measuring NO2 concentration based on DOAS according to claim 1, characterized in that: The method of correcting the NO2 concentration value by using the least square method comprises the following steps: S61, using the least square method to fit a cubic polynomial to obtain the relationship between the NO2 concentration value and the actual concentration value, and obtaining a cubic polynomial relationship; S62, substituting the NO2 concentration value into the cubic polynomial relationship to obtain a corrected concentration value.

8. A method for measuring NO2 concentration based on DOAS according to claim 7, characterized in that: The expression of the cubic polynomial is: Fit_c = polyfit(C,y,3) con=polyval(Fit_c,C) Among them, polyfit() is the polynomial fitting function in Matlab; C represents the NO2 concentration value; y represents the corresponding actual concentration value; 3 is the number of polynomial fits; Fit_c represents the coefficients of the cubic polynomial obtained by fitting; polyval() represents a polynomial evaluation function; con represents the corrected concentration value.

Images