A method for simultaneous rapid quantitative analysis of multiple gaseous pollutants in flue gas

Abstract

The present invention relates to a method for rapid and quantitative analysis of multiple gaseous pollutants in flue gas at the same time, and the method specifically includes the following steps: S1: data preprocessing: obtaining the absorption spectrum of the gas to be tested in the ultraviolet band, performing band interception according to the characteristic absorption spectrum of the target gas, performing differential processing on the intercepted spectrum, and then performing Fourier transform; S2: modeling and prediction: if the intercepted characteristic band is a pure characteristic spectral band of the target gas (that is, the absorption of other gases in this band or the absorption after data preprocessing can be ignored except for the target gas), then a univariate correction regression model is selected for modeling or prediction; otherwise, a multivariate correction regression model is selected for modeling or prediction. The method for rapid and quantitative analysis of multiple gaseous pollutants in flue gas proposed in the present application can effectively solve the problems of cross-interference and spectral overlap between multiple gases, and realize simultaneous, rapid and accurate detection of the concentrations of different gaseous pollutants in flue gas.

Description

Technical Field

[0001] The invention relates to the technical field of gas concentration detection, and in particular to a method for simultaneously and rapidly quantitatively analyzing multiple gaseous pollutants in flue gas. Background Art

[0002] Along with the rapid development of my country's national economy, some air pollution phenomena have also emerged. Air pollution threatens people's life, health and safety, and has caused great damage to the living environment. The source of air pollutants is mainly caused by human pollution. People will produce air pollutants in economic production activities and life, such as waste gas (such as carbon dioxide, sulfur dioxide, nitrogen oxides, etc.) in industrial production processes (such as thermal power plants, steel mills, etc.); for example, exhaust gas emitted by people's vehicles, toxic and harmful gas pollutants produced by burning garbage in life, etc. Atmospheric pollutants will have long-term adverse effects on the human body and the environment. Therefore, monitoring and governance of them are urgent. In recent years, with the continuous progress of the country's ecological civilization construction, people's environmental awareness has gradually strengthened, and they pay more attention to the sustainable development of the economy and the quality of life. In the process of optimizing and upgrading the industrial structure, the country has also issued some relevant environmental protection standards, such as "HJ75-2017 Stationary Pollution Source Flue Gas (SO 2 、NO X "HJ 76-2017 Technical Specification for Continuous Monitoring of Flue Gas (SO 2 、NO X , Particulate matter) emissions continuous monitoring system", "HJ 1045-2019 Technical requirements and test methods for portable ultraviolet absorption measurement instruments for flue gas (sulfur dioxide and nitrogen oxides) from stationary pollution sources", "HJ 1131-2020 Portable ultraviolet absorption method for the determination of sulfur dioxide in waste gas from stationary pollution sources", "HJ 1132-2020 Portable ultraviolet absorption method for the determination of nitrogen oxides in waste gas from stationary pollution sources", etc.

[0003] Pollutant gas detection is widely used, but due to the diversity of its gas components, how to solve the cross-interference between the gas components becomes a problem. The patent application with application number CN201410578821 and patent name "Detection device and detection method for ammonia and sulfur dioxide gas concentration" discloses: it includes two gas chambers, namely a nitrogen reference gas chamber and an ammonia reference gas chamber. 2 When the sample spectrum is obtained, the light passes through the ammonia reference gas chamber, that is, the final spectrum contains NH 3 The saturation absorption spectrum (in other words, the SO 2 and saturated NH 3 After this treatment, regardless of SO 2NH in test gas 3 The content of NH 3 Test spectrum after saturation absorption spectrum, thus eliminating NH 3 About SO 2 Therefore, for SO 2 The gas test can be carried out by performing least squares regression (univariate linear regression, x and y used for regression are both vector data types) on the characteristic peak value and concentration value after differential and Fourier transformation to achieve concentration inversion. The concentration inversion is to take the characteristic peak value (the intensity value at the peak top) and the concentration data for least squares regression, and then use the regression model to invert the concentration. This process only takes the data at one wavelength point for regression, and the data of most wavelength points are abandoned. The abandoned wavelength point data also contains useful information related to the measured gas concentration value data. The present application aims to solve the cross-interference problem in the simultaneous detection of multiple gas concentrations by introducing multivariate correction modeling technology without adding special structural components and without adopting special experimental designs. Summary of the invention

[0004] In order to overcome the above-mentioned deficiencies of the prior art, the present invention proposes a method for simultaneous rapid quantitative analysis of multiple gas pollutants in flue gas, which solves the existing technical problem that the concentration of each gas cannot be detected simultaneously, rapidly and accurately when multiple gases cross-interfere.

[0005] The present invention is achieved through the following technical solutions:

[0006] A method for simultaneously and rapidly quantitatively analyzing multiple gaseous pollutants in flue gas comprises the following steps:

[0007] S1: Data preprocessing: Obtain the absorption spectrum of the gas to be tested in the ultraviolet band, perform band interception according to the characteristic absorption spectrum of the target gas, perform differential processing on the intercepted spectrum, and then perform Fourier transform;

[0008] S2: Modeling and prediction: If only the target gas has a strong characteristic absorption peak in this band in the intercepted characteristic band, select the univariate correction regression model for modeling or prediction; otherwise, select the multivariate correction regression model for modeling or prediction.

[0009] Furthermore, the differential processing and Fourier transform process in S1 includes: performing differential processing on the intercepted spectrum, analyzing and calculating the differential absorption spectrum and the differential absorption coefficient; numerically filtering the obtained differential absorption spectrum, using the periodicity of the differential absorption coefficient, taking some of the coefficients therein, and finally obtaining the corresponding relationship between the gas concentration and the Fourier eigenvalue of the spectrum.

[0010] Further, when there are characteristic absorption peaks of other gases in addition to the target gas in the intercepted characteristic wavelength band in S2, a multivariate calibration regression model is adopted to model or predict the response intensity data matrix at multiple wavelengths and the concentration.

[0011] Further, the adopted modeling model is the partial least squares model, and the final prediction model is: Y = <Beta, (X - Xmeancal)> + Ymeancal, where X represents the spectral response data after preprocessing, Y represents the gas concentration vector data, Beta is the regression coefficient, and Xmeancal and Ymeancal represent the data normalization parameters, and <Beta, (X - Xmeancal)> represents the inner product of the two vectors.

[0012] Further, the target gases in S1 are NO, NO 2 , SO 2 ;

[0013] In S2, the modeled and predicted concentrations of NO, NO 2 , SO 2 and their respective actual concentrations all satisfy the linear relationship, and the goodness of fit is above 0.999.

[0014] An apparatus using the ultraviolet absorption spectrometry for simultaneously and rapidly quantifying multiple gas pollutants.

[0015] The principle of differential absorption spectrometry is to utilize the spectral absorption difference between the position with a characteristic absorption peak in the absorption spectrum of the gas to be measured and the position with similar wavelength but without spectral absorption (or with small spectral absorption), to deduce the concentration of the gas to be measured, which can effectively eliminate the influence of Rayleigh scattering and Mie scattering, improve the measurement accuracy, and measure their absolute concentrations without being interfered by the chemical behavior of the measured object. By analyzing the overlapping absorption spectra of several gases in the same wavelength band, the concentrations of several gases can be measured simultaneously;

[0016] Fourier transform filtering is a numerical filtering process, which essentially decomposes the signal waveform into the superposition of many sine waves, so that it can be analyzed from the time domain to the frequency domain. At this time, the frequency domain energy may be concentrated on a few frequencies. After processing, it can not only effectively reduce the noise influence, but also effectively suppress the influence of other interfering gases with different frequency domain characteristics from the measured gas;

[0017] The purpose of multivariate calibration is to establish the correlation between the substance concentration or other chemical and physical property quantities and the response of the analytical instrument. In this application, partial least squares is adopted to deal with the linear problem between the concentration and the spectral response.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0019] The present invention proposes a method for simultaneous rapid quantitative analysis of multiple gas pollutants in flue gas. Without adding special structural components and special experimental design, the method performs multivariate correction regression modeling of concentration and spectral response matrix after characteristic band interception, differentiation and Fourier transformation to solve the cross interference between gases and realize simultaneous, rapid and accurate measurement of the concentration of each component gas. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic flow chart of a method for simultaneously and rapidly quantitatively analyzing multiple gaseous pollutants in flue gas according to an embodiment of the present invention;

[0021] Figure 2 The original ultraviolet absorption spectra of the target gases and interfering gases described in the embodiments of the present invention;

[0022] Figure 3 A linear fitting relationship diagram of the predicted value and the actual value of the concentration of the target gas NO according to an embodiment of the present invention;

[0023] Figure 4 The target gas NO in the embodiment of the present invention 2 Linear fitting relationship diagram between predicted concentration value and actual value;

[0024] Figure 5 The target gas SO in the embodiment of the present invention is 2 Linear fitting relationship between the predicted concentration and the actual value. DETAILED DESCRIPTION

[0025] Examples are shown to specifically illustrate certain embodiments of the present invention, and should not be construed as limiting the scope of the present invention. The disclosed content of the present invention can be improved from materials, methods and reaction conditions at the same time, and all these improvements should fall within the spirit and scope of the present invention.

[0026] like Figure 1 As shown, a method for simultaneously and rapidly quantitatively analyzing multiple gaseous pollutants in flue gas specifically comprises the following steps:

[0027] S1: Data preprocessing: Obtain the absorption spectrum of the gas to be tested in the ultraviolet band, perform band interception according to the characteristic absorption spectrum of the target gas, perform differential processing on the intercepted spectrum, analyze and calculate to obtain the differential absorption spectrum and differential absorption coefficient; perform numerical filtering on the obtained differential absorption spectrum, use the periodicity of the differential absorption coefficient, take some of the coefficients, and finally obtain the corresponding relationship between the gas concentration and the Fourier eigenvalue of the spectrum;

[0028] S2: Modeling and prediction: If the intercepted characteristic band is a pure characteristic spectral band, a univariate correction regression model is selected for modeling or prediction; if there are characteristic absorption peaks of other gases in addition to the target gas in the intercepted characteristic band, a multivariate correction regression model is used to model or predict the response intensity data matrix and concentration at multiple wavelengths. The modeling model used is the partial least squares model, and the final prediction model is: Y = <Beta, (X-Xmeancal) > +Ymeancal, where X represents the spectral response data after preprocessing, Y represents the gas concentration vector data, Beta is the regression coefficient, Xmeancal and Ymeancal represent data standardization parameters,<Beta,(X-Xmeancal)> Represents the inner product of two vectors.

[0029] The target gas is NO, NO 2 、SO 2 , the interfering gas is NH 3 ,CO,CO 2 Taking HCl as an example, the concentration ranges of various gases are shown in Table 1, and the original UV absorption spectra of various target gases and interfering gases are shown in Table 1. Figure 2 shown.

[0030] Table 1 Gas concentration list

[0031] Gas Name Concentration range Gas Type NO <![CDATA[0~500mg / m 3 ]]> Target gas <![CDATA[NO 2 ]]> <![CDATA[0~500mg / m 3 ]]> Target gas <![CDATA[SO 2 ]]> <![CDATA[0~500mg / m 3 ]]> Target gas <![CDATA[NH 3 ]]> <![CDATA[20mg / m 3 ]]> Interfering gas HCl <![CDATA[200mg / m 3 ]]> Interfering gas CO <![CDATA[20000mg / m 3 ]]> Interfering gas <![CDATA[CO 2 ]]> 20％ Interfering gas

[0032] The measuring equipment using the above detection method shall be verified for performance in accordance with "HJ 1045-2019 Technical Requirements and Detection Methods for Portable Ultraviolet Absorption Measurement Instruments for Stationary Pollution Source Flue Gas (Sulfur Dioxide and Nitrogen Oxides)":

[0033] Test conditions: Measurement unit: ppm; Ambient temperature: 28.7℃; Atmospheric pressure: 101.7kpa; Humidity: 38%rh;

[0034] The final prediction models for the three target gases are as follows:

[0035] The final prediction model of NO is: Y =<Beta,(X-Xmeancal)> +48.9662, the parameters of Xmeancal and Beta are shown in Table 11 below:

[0036] Table 11 Parameters of Xmeancal and Beta in NO model

[0037]

[0038]

[0039]

[0040] NO2 The final prediction model is: Y =<Beta,(X-Xmeancal)> +29.5155, the parameters of Xmeancal and Beta are shown in Table 12 below:

[0041] Table 12 NO 2 Parameters of Xmeancal and Beta in the model

[0042]

[0043]

[0044] SO 2 The final prediction model is: Y =<Beta,(X-Xmeancal)> +19.6608, the parameters of Xmeancal and Beta are shown in Table 13 below:

[0045] Table 13 SO 2 Parameters of Xmeancal and Beta in the model

[0046]

[0047]

[0048]

[0049] The performance indicators calculated with reference to the standard are as follows:

[0050] 1) The measured concentration value of each target gas meets the linear index with the actual value, such as Figure 3-5 As shown:

[0051] The predicted value of NO and the actual value satisfy the linear relationship: y = 1.0024x + 6.2805, the fitting degree R 2 =0.9986;

[0052] NO 2 The predicted value and the actual value satisfy the linear relationship: y = 1.0604x + 2.8034, the fitting degree R 2 =0.9991;

[0053] SO 2 The predicted value and the actual value satisfy the linear relationship: y = 0.9741x + 0.7016, the fitting degree R 2 =0.9995.

[0054] 2) Indication error detection, see Table 2 below:

[0055] Table 2 Instrument indication error detection record

[0056]

[0057] 3) Repeatability test:

[0058] Table 3 Instrument repeatability test records

[0059]

[0060]

[0061] 4) Response time detection, see Table 4 below:

[0062] Table 4 Instrument response time (unit: seconds) test record

[0063]

[0064] 5) Stability test, see Table 5 below:

[0065] Table 5 Instrument stability test record

[0066]

[0067] 6) The results of the ten-minute fluctuation test are shown in Table 6 below:

[0068] Table 6 Instrument ten-minute fluctuation test result record

[0069]

[0070] 7) Minimum detection limit, see Table 7:

[0071] Table 7 Instrument minimum detection limit test record

[0072]

[0073] 8) The results of zero drift and quantity drift detection for 6 consecutive hours are shown in Tables 81-83:

[0074] Table 81 6 hours of continuous NO zero drift and quantity drift detection records

[0075]

[0076] Table 82 Continuous 6h NO 2 Zero drift and quantity drift detection records

[0077]

[0078] Table 83 SO for 6 hours 2 Zero drift and quantity drift detection records

[0079]

[0080]

[0081] 9) Detection of the influence of interfering components, as shown in Table 9 below:

[0082] Table 9 Test records of the influence of instrument interference components

[0083]

[0084] 10) Detection of the impact of ambient temperature changes, as shown in Table 10 below:

[0085] Table 10 Test records of the impact of instrument ambient temperature changes

[0086]

[0087]

[0088] It can be seen from the above table that the various performance index tests of the ultraviolet absorption measuring instrument using the detection method of the present application are in compliance with national regulations; in summary, the present application provides a method for rapid and quantitative analysis of multiple gaseous pollutants in flue gas, which solves the problem of cross-interference between the measurement signals of multiple gaseous pollutants, and realizes simultaneous, rapid and accurate measurement of the concentration of each component gas, providing an effective detection method for simultaneous quantitative analysis of multiple gaseous pollutants using ultraviolet absorption spectroscopy technology.

[0089] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A method for simultaneous rapid quantitative analysis of multiple gaseous pollutants in flue gas, characterized in that: The specific steps include: S1: Data preprocessing: Obtain the absorption spectrum of the gas to be tested in the ultraviolet band, perform band interception according to the characteristic absorption spectrum of the target gas, perform differential processing on the intercepted spectrum, and then perform Fourier transform; S2: Modeling and prediction: If only the target gas has a strong characteristic absorption peak in the intercepted characteristic band, a univariate correction regression model is selected for modeling or prediction; Otherwise, select the multivariate adjusted regression model for modeling or prediction; The process of differential processing and Fourier transformation in S1 includes: differential processing of the intercepted spectrum, analysis and calculation to obtain the differential absorption spectrum and differential absorption coefficient; The obtained differential absorption spectrum is numerically filtered, and part of the coefficients are selected to perform filtering using the periodicity of the differential absorption coefficient, and finally the corresponding relationship between the gas concentration and the Fourier eigenvalue of the spectrum is obtained; In S2, when other gases besides the target gas also have characteristic absorption peaks in the intercepted characteristic band, a multivariate correction regression model is used to model or predict the response intensity data matrix and concentration at multiple wavelengths; The modeling model used is the partial least squares model, and the final prediction model is: Y =<Beta,(X-Xmeancal)> +Ymeancal, where X represents the spectral response data matrix after preprocessing, Y represents the gas concentration vector data, Beta is the regression coefficient, Xmeancal and Ymeancal represent the data standardization parameters,<Beta,(X-Xmeancal)> Represents the inner product of two vectors.

2. The method for rapid and quantitative analysis of multiple gaseous pollutants in flue gas according to claim 1, characterized in that: The target gases in S1 are NO, NO2, and SO2; The modeling predictions of NO, NO2, and SO2 concentrations in S2 all satisfy a linear relationship with their actual concentrations, with a goodness of fit above 0.999.

Images