Method and system for measuring and calculating volatile organic compounds in exhaust gas discharged by stationary pollution source

Abstract

The invention discloses a method and system for measuring and calculating volatile organic compounds in exhaust gas discharged by a stationary pollution source, and the method comprises the steps: determining a sampling position and a sampling point of the exhaust gas discharged by the stationary pollution source, collecting the main product yield, main raw material or fuel consumption of a preset enterprise, and building a database; the method comprises the following steps: collecting waste gas through a waste gas sampling device, acquiring a waste gas chromatogram by adopting a mass spectrum detector, acquiring low-molecular-weight components by adopting a flame ionization detector FID, and acquiring volatile organic components; dividing the molecular weight low components into different concentration series, performing gas chromatographic analysis, drawing a working curve, and calculating correlation coefficients of the molecular weight low components of different concentration series and response areas; analyzing other compounds by adopting a mass spectrometric detector, calculating correlation coefficients of other compounds and response areas, calculating the concentration of the volatile organic compounds, and establishing pollution source data according to the database; and carrying out regression analysis by utilizing the database and the volatile organic compound concentration, and outputting a result.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of measurement, in particular to a method and system for measuring volatile organic compounds in exhaust gas of fixed pollution sources. BACKGROUND

[0002] Volatile organic compounds (VOCs) have different definitions, one of which refers to organic chemical substances with a vapor pressure above 13.3 Pa and a boiling point below 260℃ under normal conditions (20℃, 101.3 kPa). Structurally, volatile organic compounds include short-chain hydrocarbons such as alkanes, alkenes, aromatic hydrocarbons, alkynes, and oxygen-containing organic compounds such as aldehydes and ketones. These compounds not only participate in photochemical reactions and contribute significantly to the formation of ozone, but also have significant carcinogenic effects, so the state has developed a series of policies and standards to strictly control volatile organic compound pollution. Volatile organic compounds (VOCs) in the environment have a wide range of sources, but one of the important sources is the exhaust gas of VOCs-related key industries, i.e., the exhaust gas discharged into the environment through exhaust pipes, including chemical raw materials and chemical products manufacturing, pharmaceutical manufacturing, printing, petroleum, rubber, and other industries.

[0003] In recent years, through the atmospheric pollution control special action, comprehensive management of VOCs-related key industries has been carried out, and air quality has been continuously improved. The testing technology and method of VOCs is an important technical support for environmental comprehensive management, and the testing of VOCs in organized exhaust gas of fixed pollution sources mainly includes sample collection and sample analysis. Sample collection requires on-site investigation at the sampling port and carrying sampling instruments to the sampling port, the sampling port is usually not on the ground but at a certain height from the ground, the sampler needs to be carried to the sampling port, and the sampling personnel need to be on duty during sampling. After sampling, the sampling instrument needs to be moved back to the original place, and the collected samples need to be transported to the designated laboratory for analysis. After receiving the samples, the analyst needs to debug the analysis instrument, analyze the samples according to the established method, and process the data according to the data processing software in the instrument to obtain the required pollutant concentration, all analysis is completed in the laboratory. The whole process requires a lot of manpower and material resources, the efficiency is low, and the cost is high.

[0004] Therefore, there is an urgent need for a new method for rapidly measuring volatile organic compounds in exhaust gas of fixed sources, which covers more pollutants, including more than 100 kinds of alkanes, alkenes, aromatic hydrocarbons, alkynes, and oxygen-containing organic compounds such as aldehydes and ketones, and is suitable for more industries and enterprises of fixed pollution sources; using artificial intelligence technology, real-time remote control of on-site sampling and laboratory analysis, saving a lot of manpower and material resources; making full use of data and mining the value of data. SUMMARY

[0005] The application aims to provide a method for calculating volatile organic compounds in exhaust gas of stationary pollution sources.

[0006] To achieve the above-mentioned purpose, the application is implemented according to the following technical scheme:

[0007] The application comprises the following steps:

[0008] The sampling position and sampling point of the exhaust gas of the stationary pollution source are determined, the main product output, the main raw material or fuel consumption of the preset enterprise are collected, and a database is established;

[0009] The exhaust gas is collected by an exhaust gas sampling device, the composition of the exhaust gas is analyzed by using a mass spectrometer detector to obtain an exhaust gas chromatogram, coarse mass spectrum data are obtained, the molecular weight low component of the exhaust gas is analyzed by using a hydrogen flame ionization detector (FID) to obtain fine molecular data, and the volatile organic compound component is obtained according to the coarse mass spectrum data and the fine molecular data;

[0010] The molecular weight low component is allocated into different concentration series, gas chromatography analysis is performed, a working curve is drawn, and the correlation coefficient of the molecular weight low component in different concentration series and the response area is calculated;

[0011] Other compounds are analyzed by using a mass spectrometer detector, the correlation coefficient of the other compounds and the response area is calculated, the concentration of the volatile organic compound is calculated, and the pollution source data is established according to the database;

[0012] Regression analysis is performed on the database and the concentration of the volatile organic compound, and the calculation result is output.

[0013] Further, the method for obtaining the volatile organic compound component according to the coarse mass spectrum data and the fine molecular data comprises:

[0014] The coarse mass spectrum data and the fine molecular data are subjected to component hierarchical analysis, the first layer is to lock C2-C5 hydrocarbons, to quickly match by using a retention time index library, to quantitatively analyze by using an FID, to introduce a retention time prediction model, and to correct the aging offset of the chromatographic column;

[0015] The second layer is to separate the co-elution peaks by using mass spectrum deconvolution based on machine learning for the components not covered by the FID, to verify the compound structure by combining the fragment particle matching;

[0016] The third layer is to cluster the fragment modes of similar mass spectrum peaks for the unmatched mass spectrum peaks, to infer the unlabeled compounds according to the enterprise raw material database;

[0017] The dynamic time warping algorithm is used to automatically correct the retention time deviation of FID and mass spectrum. If both FID and mass spectrum detect signals, but the mass spectrum confidence is greater than 0.9, the mass spectrum result is used as the reference. For FID sensitive components, mass spectrum is used for verification, and volatile organic component is output.

[0018] Further, the method for distributing the low molecular weight components into different concentration series and performing gas chromatography analysis comprises:

[0019] According to the actual emission concentration range, multiple concentration points are configured to cover the detection linear interval, pure nitrogen is used as the zero concentration point, and double samples are prepared for each concentration point, and the relative deviation is less than or equal to 5%;

[0020] The instrument parameters are set and the sampling mode is selected, three times of sampling are repeated for each concentration point, the peak value is taken, the compound is locked by retention time, the temperature rising program is set according to the separation requirement of the low molecular weight component, the initial temperature is 40 for 3 minutes; the temperature is raised to 120 degrees for the first time, the rate is 10 degrees per second for 8 minutes, and the low boiling point compound is separated quickly; the temperature is raised to 200 degrees for the second time, the rate is 5 degrees per minute for 16 minutes, and the separation degree of the medium and high boiling point compound is improved; finally, 200 degrees is kept for 5 minutes until all the target components flow out;

[0021] Helium and nitrogen are independently controlled, the flow rate is given, the relative response factor of common compounds in external standard method is obtained, the internal standard is selected, and the peak area of the internal standard is obtained, and the compound concentration is calculated:

[0022]

[0023] Wherein the concentration of the i th compound is C i , the chromatographic peak area of the i th compound is A i , the relative response factor of the i th compound is RRF i , and the chromatographic peak area of the internal standard is A os ;

[0024] After analyzing light hydrocarbon, the heavy component is backflushed to shorten the cycle, the center cutting technology is used to separate the co-elution peak, the convolution neural network model is introduced, the convolution neural network model is trained to identify the overlapping peak, and the gas chromatography analysis result is output.

[0025] Further, the method for drawing the working curve comprises:

[0026] Three times of sampling are repeated for each concentration point, and the peak area average value is taken, wherein the peak area is locked by retention time to target compound;

[0027] The target compound is input into the linear regression model, and the expression is:

[0028] A = h × C + z

[0029] wherein the peak area of the target compound is A, the concentration of the target compound is C, the intercept of the target compound is z, and the slope of the target compound is h;

[0030] When the correlation coefficient of the target compound is greater than 0.995, and the intercept relative response value is greater than 0.05, the corresponding working curve is drawn.

[0031] Further, the method for calculating the correlation coefficient of the low molecular weight component and the response area in different concentration series comprises:

[0032] Obtaining the peak area of the concentration level and the corresponding low molecular weight component, wherein the product of the peak area and the concentration is the response area, and calculating the correlation coefficient of the low molecular weight component and the response area:

[0033]

[0034] wherein the correlation coefficient is R, the number of concentration levels is n, the concentration of the i th compound is C i , and the peak area of the i th compound is A i ;

[0035] If the correlation coefficient is greater than or equal to 0.995, the linear relationship between the low molecular weight component and the response area is established; otherwise, the linear relationship between the low molecular weight component and the response area is not established, and it is checked whether there is a standard sample configuration error, an instrument response abnormality, or a chromatographic peak integration abnormality, and if there is, the error is corrected and the correlation coefficient is recalculated, otherwise the result is output.

[0036] Further, the method for calculating the concentration of the volatile organic compound comprises:

[0037] For the chromatographic peak area of the target compound obtained by the FID device, the concentration of a certain type of volatile organic compound is calculated:

[0038]

[0039] wherein the concentration of the c th organic compound is the chromatographic peak area of the c th organic compound is A c , the intercept of the working curve is z, and the slope of the working curve is z;

[0040] For the detector with a known relative response factor value, the concentration of a certain type of volatile organic compound is calculated:

[0041]

[0042] wherein the relative response factor of the c th organic compound is RRF c , the concentration of the internal standard is the chromatographic peak area of the internal standard is Aos ;

[0043] The concentration of volatile organic compounds is obtained by adding the concentration of some volatile organic compounds.

[0044] Further, the method for regression analysis of the database and the concentration of volatile organic compounds comprises:

[0045] The data in the enterprise database, the main product yield, the main raw material or fuel consumption data and the concentration of volatile organic compounds in the exhaust gas are subjected to regression analysis to establish a regression analysis model of the concentration of volatile organic compounds in the exhaust gas and the main product yield, the main raw material or fuel consumption, and the expression is:

[0046] y = β0 + β1x1 + β2x2 + β3x3 + S

[0047] wherein the emission concentration of volatile organic compounds in the exhaust gas is y, the main product yield is x1, the main raw material consumption is x2, the fuel consumption is x3, β0, β1, β3 and S are constants;

[0048] When the model explanation of the regression analysis model is greater than 0.7 and the variable significance is less than 0.05, the regression analysis result is predicted.

[0049] In the second aspect, a system for measuring and calculating volatile organic compounds in exhaust gas from stationary sources comprises:

[0050] An exhaust gas sampling module is used to determine the sampling position and sampling point of the exhaust gas from stationary sources, collect the main product yield, the main raw material or fuel consumption of the preset enterprise and establish a database;

[0051] A component analysis module is used to collect exhaust gas by an exhaust gas sampling device, analyze the components of the exhaust gas by a mass spectrometer detector to obtain exhaust gas chromatography, obtain coarse mass spectrum data, analyze the molecular weight low components of the exhaust gas by a hydrogen flame ionization detector to obtain fine molecular data, and obtain volatile organic compound components according to the coarse mass spectrum data and the fine molecular data;

[0052] A chromatography correlation module is used to allocate the molecular weight low components into different concentration series, perform gas chromatography analysis, draw a working curve, and calculate the correlation coefficient of the molecular weight low components in different concentration series and the response area;

[0053] A concentration calculation module is used to analyze other compounds by a mass spectrometer detector, calculate the correlation coefficient of other compounds and the response area, calculate the concentration of the volatile organic compounds, and establish a pollution source data according to the database;

[0054] A modeling and calculating module is configured to perform regression analysis on the volatile organic compound concentration based on the database and output a calculation result.

[0055] The present application has the following advantages:

[0056] Compared with the prior art, the present application has the following technical effects:

[0057] The present application can improve the accuracy of volatile organic compound calculation in fixed pollution source emission waste gas, thereby improving the precision of volatile organic compound calculation in fixed pollution source emission waste gas, optimizing volatile organic compound calculation in fixed pollution source emission waste gas, greatly saving resources and improving work efficiency, realizing intelligent calculation of volatile organic compounds in fixed pollution source emission waste gas, determining volatile organic compound composition and regression analysis in real time, and having important significance for volatile organic compound calculation in fixed pollution source emission waste gas, and being suitable for different standards of volatile organic compound calculation in fixed pollution source emission waste gas and different requirements of volatile organic compound calculation in fixed pollution source emission waste gas, and having certain universality. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 The present application is a method for calculating volatile organic compounds in fixed pollution source emission waste gas.

[0059] Figure 2 The present application is a method for calculating volatile organic compounds in fixed pollution source emission waste gas. DETAILED DESCRIPTION

[0060] The present application is a method for calculating volatile organic compounds in fixed pollution source emission waste gas.

[0061] The present application is a method for calculating volatile organic compounds in fixed pollution source emission waste gas.

[0062] As shown in the present embodiment, the following steps are included: Figure 1

[0063] Determine the sampling position and sampling point of the fixed pollution source waste gas emission, collect the main product output, main raw material or fuel consumption of the preset enterprise and establish a database;

[0064] ​In the actual evaluation, through field investigation, according to the requirements of the state for waste gas emission port; determine the sampling position and sampling point of the waste gas emission of the stationary source of pollution; collect the main product output, main raw material or fuel consumption of the enterprise, establish a database; under the condition of verifying the normal operation of the production equipment and pollution control facilities, sample the waste gas;

[0065] According to the technical requirements, connect the waste gas sampling device, turn on the power, open the heating button, set the required heating temperature, start sampling for 10s, connect the vacuum bottle, the waste gas collection device is a 1L inner wall deactivation treated brown glass vacuum bottle with a volume of 1L, use a sampling gun with heating function to extract gas, prepare a 1L / h inert flow limiter, and sample into the vacuum sampling bottle; the sampling time is 1h, after the sampling is completed, push the fast insertion cathode head, remove the glass sampling bottle, and transport the collected sample bottle back to the laboratory at low temperature;

[0066] Collect waste gas through the waste gas sampling device, analyze the composition of the waste gas using a mass spectrometer detector to obtain a waste gas chromatogram, obtain coarse mass spectrum data, analyze the low molecular weight components of the waste gas using a hydrogen flame ionization detector FID to obtain fine molecular data, and obtain volatile organic components according to the coarse mass spectrum data and the fine molecular data;

[0067] In the actual evaluation, the low molecular weight components are propane, ethylene, acetylene, ethane and propylene; the chromatographic column 1 is a capillary column 60m*0.25mm*1μm); the chromatographic column 2 is a capillary column 30m*0.32mm*20μm; and the chromatographic column 3 is a damping column 0.67m*0.1mm;

[0068] The low molecular weight components propane, ethylene, acetylene, ethane and propylene are analyzed by a hydrogen flame ionization detector FID, and a series of propane, ethylene, acetylene, ethane and propylene concentrations of 0.2, 0.5, 1, 2, 4, 6, 8 and 10ppb are prepared for gas chromatography / FID analysis; according to the experimental data, a standard curve is drawn, the concentration ppb of the standard series is taken as the abscissa, and the corresponding chromatographic peak area is taken as the ordinate, and a working curve is established;

[0069] The low molecular weight components are prepared into different concentration series and subjected to gas chromatography analysis and working curve drawing, and the correlation coefficients of the low molecular weight components in different concentration series and response area are calculated;

[0070] In the actual evaluation, the correlation coefficients R of the concentrations of propane, ethylene, acetylene, ethane and propylene and the response area are calculated, and the results are all above 0.9998, and the linear relationship is strong;

[0071] Other compounds are analyzed by a mass spectrometer, a correlation coefficient of other compounds and response area is calculated, a concentration of the volatile organic compounds is calculated, and pollution source data is established according to the database;

[0072] Regression analysis is performed on the database and the concentration of the volatile organic compounds, and a calculation result is output.

[0073] In the embodiment, a method for obtaining volatile organic compound components according to the coarse mass spectrum data and the fine molecular data comprises the following steps:

[0074] The coarse mass spectrum data and the fine molecular data are analyzed in layers, the first layer: C2-C5 hydrocarbons are locked, FID quantitative analysis is performed, a retention time prediction model is introduced, and column aging offset is corrected;

[0075] The second layer: components not covered by FID are separated by mass spectrum deconvolution based on machine learning, and the structure of the compound is verified by matching fragment particles;

[0076] The third layer: for unmatched mass spectrum peaks, similar mass spectrum peak fragment patterns are clustered, and unlabeled compounds are inferred according to the enterprise raw material database.

[0077] The dynamic time warping algorithm is used to automatically correct the retention time deviation of FID and mass spectrometry. If both FID and mass spectrometry detect signals, but the mass spectrometry confidence is greater than 0.9, the mass spectrometry result is used as the reference. For FID sensitive components, mass spectrometry is used for verification, and volatile organic compound components are output.

[0078] In the embodiment, a method for analyzing different concentration series of low molecular weight components by gas chromatography comprises the following steps:

[0079] According to the actual emission concentration range, multiple concentration points are configured to cover the linear interval, pure nitrogen is used as the zero concentration point, and each concentration point is prepared with double samples, and the relative deviation is less than or equal to 5%.

[0080] The instrument parameters are set and the sampling mode is selected, each concentration point is sampled three times, the peak value is averaged, the compound is locked by retention time, the temperature program is set according to the separation requirements of the low molecular weight components, the initial temperature is 40 degrees for 3 minutes, the temperature is raised to 120 degrees for the first time, the rate is 10 degrees per second for 8 minutes, and the low boiling point compounds are separated quickly; the temperature is raised to 200 degrees for the second time, the rate is 5 degrees per minute for 16 minutes, and the separation degree of the medium and high boiling point compounds is improved; finally, the temperature is kept at 200 degrees for 5 minutes until all the target components flow out;

[0081] Helium and nitrogen are independently controlled, the flow rate is given, the relative response factor of common compounds in external standard method is obtained, the internal standard is selected, the peak area of the internal standard is obtained, and the compound concentration is calculated:

[0082]

[0083] wherein the concentration of the i-th compound is C i the chromatographic peak area of the i-th compound is A i the relative response factor of the i-th compound is RRF i the chromatographic peak area of the internal standard is A os ;

[0084] After analyzing light hydrocarbons, heavy components are back-flushed, the cycle is shortened, the center cutting technology is used to separate co-elution peaks, the convolution neural network model is introduced, the convolution neural network model is trained to identify overlapping peaks, and the gas chromatography analysis result is output.

[0085] In the embodiment, the method for drawing the working curve comprises:

[0086] Each concentration point is repeatedly sampled 3 times, and the peak area average value is taken, wherein the peak area is locked by the retention time of the target compound;

[0087] The target compound is input into a linear regression model, and the expression is:

[0088] A = h x C + z

[0089] wherein the chromatographic peak area of the target compound is A, the concentration of the target compound is C, the intercept of the target compound is z, and the slope of the target compound is h;

[0090] When the correlation coefficient of the target compound is greater than 0.995, and the intercept relative response value is greater than 0.05, the corresponding working curve is drawn.

[0091] In the embodiment, the method for calculating the correlation coefficient of the molecular weight low component and the response area of different concentration series comprises:

[0092] The concentration levels and the peak areas of the corresponding molecular weight low components are obtained, wherein the product of the peak area and the concentration is the response area, and the correlation coefficient of the molecular weight low component and the response area is calculated:

[0093]

[0094] wherein the correlation coefficient is R, the number of concentration levels is n, the concentration of the i-th compound is C i , and the peak area of the i-th compound is A i ;

[0095] If the correlation coefficient is greater than or equal to 0.995, the linear relationship between the low molecular weight component and the response area is established; otherwise, the linear relationship between the low molecular weight component and the response area is not established, and whether the sample configuration error, instrument response abnormality, or chromatographic peak integration abnormality occurs is checked, if exists, the error is corrected and the correlation coefficient is recalculated, otherwise the result is output.

[0096] In the embodiment, the method for calculating the concentration of the volatile organic matter comprises:

[0097] For the chromatographic peak area of the target compound obtained by the FID device, the concentration of a certain type of volatile organic matter is calculated:

[0098]

[0099] Wherein the concentration of the cth organic matter is The chromatographic peak area of the cth organic matter is A c , the intercept of the working curve is z, and the slope of the working curve is z;

[0100] For the detector with known relative response factor value, the concentration of a certain type of volatile organic matter is calculated:

[0101]

[0102] Wherein the relative response factor of the cth organic matter is RRF c The concentration of the internal standard is The chromatographic peak area of the internal standard is A os ;

[0103] The concentration of a certain type of volatile organic matter is summed to obtain the concentration of the volatile organic matter.

[0104] In the embodiment, the method for performing regression analysis on the database and the concentration of the volatile organic matter comprises:

[0105] Using the data in the enterprise database, the regression analysis is performed on the concentration of the volatile organic matter in the waste gas and the main product yield, the main raw material or fuel consumption data, the regression analysis model of the concentration of the volatile organic matter in the waste gas and the main product yield, the main raw material or fuel consumption is established, and the expression is:

[0106] y = β0 + β1x1 + β2x2 + β3x3 + S

[0107] Wherein the emission concentration of the volatile organic matter in the waste gas is y, the main product yield is x1, the main raw material consumption is x2, the fuel consumption is x3, β0, β1, β3, and S are all constants;

[0108] When the model explanatory power of the regression analysis model is greater than 0.7 and the variable significance is less than 0.05, the prediction output regression analysis result is obtained.

[0109] In a second aspect, a system for estimating volatile organic compounds in exhaust gas from a stationary pollution source comprises:

[0110] An exhaust gas sampling module is configured to determine a sampling position and a sampling point of exhaust gas from the stationary pollution source, collect main product output, main raw material or fuel consumption of a preset enterprise, and establish a database.

[0111] A component analysis module is configured to collect exhaust gas by an exhaust gas sampling device, analyze components of the exhaust gas by a mass spectrometer detector to obtain an exhaust gas chromatogram, obtain coarse mass spectrum data, analyze molecular weight low components of the exhaust gas by a hydrogen flame ionization detector to obtain fine molecular data, and obtain volatile organic compound components according to the coarse mass spectrum data and the fine molecular data.

[0112] A chromatogram correlation module is configured to divide the molecular weight low components into different concentration series, perform gas chromatography analysis, draw a working curve, and calculate a correlation coefficient between the molecular weight low components of the different concentration series and a response area.

[0113] A concentration calculation module is configured to analyze other compounds by a mass spectrometer detector, calculate a correlation coefficient between the other compounds and a response area, calculate a concentration of the volatile organic compounds, and establish pollution source data according to the database.

[0114] A modeling estimation module is configured to perform regression analysis on the database and the concentration of the volatile organic compounds, and output an estimation result.

[0115] The above only describes preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for measuring volatile organic compounds in waste gas emitted from stationary pollution sources, characterized in that: The following steps are involved: Determine the sampling locations and points for waste gas emissions from stationary pollution sources, collect data on the output of major products, and the consumption of major raw materials or fuels of pre-set enterprises, and establish a database; Collecting waste gas through a waste gas sampling device, performing component analysis on the waste gas using a mass spectrometer to obtain a waste gas chromatogram, obtaining crude mass spectrum data, analyzing low molecular weight components of the waste gas using a hydrogen flame ionization detector (FID) to obtain fine molecular data, and obtaining volatile organic compound components based on the crude mass spectrum data and the fine molecular data; Dividing the low molecular weight component into different concentration series and performing gas chromatography analysis and drawing a working curve, and calculating the correlation coefficient between the low molecular weight component in different concentration series and the response area; Analyzing other compounds using a mass spectrometer, calculating correlation coefficients between other compounds and response areas, calculating concentrations of the volatile organic compounds, and establishing pollution source data based on the database; A regression analysis is performed using the database and the volatile organic compound concentration, and a measurement result is output.

2. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method for obtaining volatile organic compound components according to the crude mass spectrum data and the fine molecular data comprises: The crude mass spectrometry data and fine molecular data were analyzed in layers. The first layer: C2-C5 hydrocarbons were identified, quickly matched using a retention time index library, and quantified using FID. A retention time prediction model was introduced to correct for column aging drift. The second layer: For components not covered by FID, we use machine learning-based mass spectrometry deconvolution to separate co-eluting peaks, combined with fragment particle matching to verify the compound structure; The third layer: for unmatched mass spectral peaks, the fragmentation patterns of similar mass spectral peaks are clustered and unlabeled compounds are inferred based on the company's raw material database; A dynamic time warping algorithm is used to automatically correct the time deviations caused by different chromatographic analysis processes in the FID and mass spectrometry analysis systems, build a timing matching model for the FID detection signal and the mass spectrometry particle flow signal, and dynamically correct the asynchronous retention time offsets of the FID detection signal and the mass spectrometry particle flow signal during the mass spectrometry peak process. If both the FID and the mass spectrometry detect signals, but the mass spectrometry confidence level is greater than 0.9, the mass spectrometry result shall prevail. For FID-sensitive components, mass spectrometry is used for verification and the volatile organic compound components are output.

3. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method of dividing the low molecular weight component into different concentration series and performing gas chromatography analysis comprises: According to the actual emission concentration range, multiple concentration points are configured to cover the detection linear range. Pure nitrogen is used as the zero concentration point. Duplicate samples are prepared for each concentration point, and the relative deviation is less than or equal to 5%; Set the instrument parameters and select the injection method. Repeat the sampling three times at each concentration point, take the peak value average, lock the compound by retention time, and set the heating program according to the separation requirements of low molecular weight components. The initial temperature is 40 for 3 minutes; the first temperature is raised to 120 degrees, and the rate is 10 degrees per second for 8 minutes to quickly separate low-boiling point compounds; the second temperature is raised to 200 degrees, and the rate is maintained at 5 degrees per minute for 16 minutes to improve the separation of medium and high boiling point compounds; the final temperature is maintained at 200 degrees for 5 minutes until all target components are separated; Independently control helium and nitrogen, set the flow rate, obtain the relative response factors of common compounds using the external standard method, select the internal standard and obtain the peak area of ​​the internal standard, and calculate the compound concentration: The concentration of the i-th compound is C i , the chromatographic peak area of ​​the i-th compound is A i , the relative response factor of the i-th compound is RRF i The chromatographic peak area of ​​the internal standard is A os ; After analyzing light hydrocarbons, heavy components are backflushed immediately to shorten the cycle. The center-cutting technology is used to separate co-peaks. A convolutional neural network model is introduced and trained to identify overlapping peaks and output the gas chromatography analysis results.

4. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method for drawing a working curve comprises: Each concentration point was sampled three times, and the peak area was averaged. The peak area was locked to the target compound by retention time. The target compound is input into the linear regression model, which is expressed as: A=h×C+z Wherein the chromatographic peak area of ​​the target compound is A, the concentration of the target compound is C, the intercept of the target compound is z, and the slope of the target compound is h; When the correlation coefficient of the target compound is greater than 0.995 and the intercept relative response value is greater than 0.05, the corresponding working curve is drawn.

5. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method for calculating the correlation coefficient between the low molecular weight component and the response area in different concentration series comprises: Obtain the concentration level and the peak area of ​​the corresponding low molecular weight component, where the product of the peak area and the concentration is the response area. Calculate the correlation coefficient between the low molecular weight component and the response area: The correlation coefficient is R, the number of concentration levels is n, and the concentration of the i-th compound is C i , the peak area of ​​the i-th compound is A i ; If the correlation coefficient is greater than or equal to 0.995, the linear relationship between the low molecular weight component and the response area is established; otherwise, the linear relationship between the low molecular weight component and the response area is not established, and check whether there is a standard sample configuration error, instrument response abnormality, or chromatographic peak integration abnormality. If so, correct the error and recalculate the correlation coefficient. Otherwise, output the result.

6. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method for calculating the concentration of volatile organic compounds comprises: Based on the chromatographic peak area of ​​the target compound obtained by the FID device, the concentration of a certain type of volatile organic compound is calculated: The concentration of the cth organic matter is The chromatographic peak area of ​​the cth organic compound is A c , the intercept of the working curve is z, and the slope of the working curve is z; Calculate the concentration of a certain type of volatile organic compound for a detector with a known relative response factor value: The relative response factor of the cth organic compound is RRF c The concentration of the internal standard is The chromatographic peak area of ​​the internal standard is A os ; The concentration of volatile organic compounds is obtained by adding the concentrations of certain types of volatile organic compounds.

7. The method for calculating volatile organic compounds in waste gas emitted from a stationary pollution source according to claim 1, characterized in that: The method for performing regression analysis using the database and the volatile organic compound concentration includes: Using the data in the enterprise database, the main product output, main raw material or fuel consumption data and the volatile organic compound concentration in the exhaust gas were regression analyzed to establish a regression analysis model of the volatile organic compound concentration in the exhaust gas and the main product output, main raw material or fuel consumption. The expression is: y=β0+β1x1+β2x2+β3x3+S The emission concentration of volatile organic compounds in the exhaust gas is y, the output of the main product is x1, the consumption of the main raw materials is x2, the fuel consumption is x3, and β0, β1, β3, and S are all constants; When the explanatory power of the regression analysis model is greater than 0.7 and the variable significance is less than 0.05, the regression analysis results are predicted and output.

8. A system for measuring volatile organic compounds in waste gas emitted from a stationary pollution source, for executing the method according to any one of claims 1 to 7, characterized in that: include: Exhaust gas sampling module: used to determine the sampling location and sampling points of exhaust gas emissions from fixed pollution sources, collect the output of main products, main raw materials or fuel consumption of preset enterprises and establish a database; Component analysis module: used to collect exhaust gas through an exhaust gas sampling device, perform component analysis on the exhaust gas using a mass spectrometer to obtain an exhaust gas chromatogram, obtain crude mass spectrum data, analyze the low molecular weight components of the exhaust gas using a hydrogen flame ionization detector to obtain fine molecular data, and obtain volatile organic compound components based on the crude mass spectrum data and the fine molecular data; Chromatography correlation module: used for dividing the low molecular weight component into different concentration series and performing gas chromatography analysis and drawing a working curve, and calculating the correlation coefficient between the low molecular weight component in different concentration series and the response area; Concentration calculation module: used for analyzing other compounds using a mass spectrometer detector, calculating the correlation coefficient between other compounds and the response area, calculating the concentration of the volatile organic compound, and establishing pollution source data according to the database; Modeling and calculation module: used to perform regression analysis using the database and the volatile organic compound concentration, and output the calculation results.

Images