A peroxyl radical source generation and calibration system

Abstract

This invention provides a peroxide free radical source generation and calibration system, relating to the field of pollutant monitoring technology. The system includes a free radical generation module and an online calibration module. The free radical generation module includes a main photolysis cavity, which generates ·H and ·OH by photolysis of H2O using ultraviolet light. The ·H generated in the main photolysis cavity reacts with O2 to generate HO2 free radicals. The online calibration module includes a calibration channel connected to the main photolysis cavity and a separate background channel. NO standard gas is introduced into the calibration channel for calibration; NO standard gas is added to the background channel as background. This invention can accurately acquire NO2 signals under both active and passive scenarios, effectively deducting background interference and achieving quantitative measurement of HO2.

Description

Technical Field

[0001] This invention relates to the field of pollutant monitoring technology, and in particular to a peroxide free radical source generation and calibration system. Background Technology

[0002] O3 is composed of volatile organic compounds (VOCs) and nitrogen oxides (NOx). x Ozone is a secondary pollutant formed through a series of photochemical reactions under light conditions. Among these, total peroxide radicals (RO2*, HO2+RO2) are key intermediates in the generation and accumulation of O3 during VOCs degradation, and their concentration changes are inevitably linked to local O3 concentration changes. Therefore, accurate and highly sensitive measurement of atmospheric RO2* concentration is crucial for a deeper understanding of ozone formation mechanisms, assessing the degree of photochemical pollution, and developing effective ozone pollution control strategies. However, achieving accurate monitoring of atmospheric RO2* requires a standard source capable of generating stable, known concentrations of peroxide radicals for calibrating various field measuring instruments.

[0003] Currently, common methods for generating peroxide free radicals in the laboratory mainly include hydrogen peroxide pyrolysis HO2 free radical source, H2O photolysis, and acetone photolysis. However, these methods generally face core technical challenges. First, the concentration uncertainty is high. Most existing methods rely on indirect measurement and theoretical calculation of parameters such as precursor concentration, light intensity, and quantum yield. These parameters themselves have errors and drift over time, leading to increased uncertainty in the concentration of generated free radicals, making them unsuitable as "standard sources." Second, there is a lack of online calibration capabilities. Traditional offline sampling or indirect estimation methods cannot perform real-time, online verification and absolute measurement of free radical concentration, resulting in low reliability of source strength. In particular, existing generation and calibration systems are usually single generation and calibration structures, making it difficult to accurately acquire NO2 signals under both active and passive scenarios, and difficult to effectively subtract background interference to achieve quantitative measurement of HO2. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides a peroxy free radical source generation and calibration system. This invention can accurately acquire NO2 signals under both active and passive scenarios, effectively deduct background interference, and achieve quantitative measurement of HO2.

[0005] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0006] A peroxide free radical source generation and calibration system includes a free radical generation module and an online calibration module. The free radical generation module includes a main photolysis cavity, which generates ·H and ·OH by photolysis of H2O with ultraviolet light. The ·H generated by the main photolysis cavity reacts with O2 to generate HO2 free radicals. The online calibration module includes a calibration channel connected to the main photolysis cavity and an independent background channel. NO standard gas is introduced into the calibration channel for calibration, and NO standard gas is added into the background channel as background. The online calibration module is used to accurately measure the concentration of generated NO2 molecules. By directly obtaining the net NO2 signal generated by the HO2 reaction, the absolute calibration of the free radical concentration is achieved.

[0007] Existing calibration systems still suffer from problems such as crude stability control. Although existing technologies have proposed controlling the concentration of OH radicals, they mostly use simple PID control or feedforward control, which have limited ability to suppress fluctuations in key parameters such as humidity and light intensity, making it difficult to achieve long-term stability.

[0008] Although existing technologies focus on different aspects of radical generation or detection, none of them solve the aforementioned systemic problems. For example, in existing technologies, a dual-generation system is used for cross-validation to improve accuracy, but the system is complex and does not involve high-precision closed-loop control of the stability of the radical generation source itself.

[0009] Therefore, developing a peroxy radical source system that integrates stable generation and online precise calibration functions, and introduces advanced control algorithms to comprehensively improve the system's stability and accuracy, is of great significance for promoting basic research in atmospheric chemistry and the calibration of related instruments. This is also one of the core problems that this invention aims to solve.

[0010] To solve one or more of the above-mentioned technical problems, the present invention also provides the following preferred technical solutions.

[0011] Preferably, the free radical generation module further includes a secondary reaction chamber connected to the main photolysis chamber. Oxygen-containing gas is introduced into the secondary reaction chamber to react with the ·H generated in the main photolysis chamber to generate HO2 free radicals.

[0012] Preferably, it further includes a stabilization module, which includes a thermostatic reaction tube connected to the outlet of the free radical generation module. The free radicals remain in the thermostatic reaction tube for a sufficient time, so that their concentration distribution reaches dynamic equilibrium and spatial uniformity.

[0013] Preferably, the residence time of the gas containing free radicals is greater than 30 seconds; a constant temperature jacket is wrapped around the outside of the stabilization module to maintain the temperature of the stabilization module at 25.0±0.1℃.

[0014] Preferably, it also includes a fully automatic concentration control module, which includes a high-precision mass flow meter and a pressure controller, as well as a nitrogen standard gas cylinder. The nitrogen standard gas cylinder is connected to a dry gas path and a wet gas path. The wet gas path is connected in series with a nano-distribution pipe. The dry gas path and the wet gas path are independently controlled by the high-precision mass flow meter, and the ratio of dry gas to wet gas flow can be dynamically adjusted.

[0015] Preferably, a model relating humidity and flow rate is established by fitting real-time humidity data using the discrete least squares method, and the ratio of dry to wet air flow rate is dynamically adjusted through a feedback adjustment mechanism to stabilize the relative humidity within a set range; the fitting formula for the discrete least squares dynamic fitting method is as follows:

[0016] ;

[0017] Where a and b are the model parameters to be determined, R is the dry air flow rate / wet air flow rate, and RH is the real-time fitted humidity; during system operation, k historical data points R are recorded. i and RH i , where i=1,2,...,k; find parameters a and b by least squares fitting such that the sum of squared errors of all data points is minimized.

[0018] Preferably, the formula for the sum of squared errors function is as follows:

[0019] ;

[0020] By taking the partial derivative and setting it to zero, we can obtain:

[0021] ;

[0022] ;

[0023] Solving the system of equations yields the current optimal model parameters a and b; where E is the sum of squared errors. Let be the relative humidity at time i. Let be the ratio of dry to wet air flow at time i, and a, b, and k be model parameters. When it is necessary to adjust the humidity to the target value RH... tar At that time, the system uses the newly fitted model to solve for the target ratio R. tar For example:

[0024] ;

[0025] In the formula: The target dry-to-wet air flow ratio, Let a and b be the target humidity, and a and b be the model parameters. Then, the controller determines the humidity based on R. tarBased on the total flow requirement, the set values ​​for dry gas flow rate and wet gas flow rate are calculated, and the high-precision mass flow meter is driven to execute.

[0026] Preferably, the fully automatic concentration control module further includes an ultraviolet light intensity control system, and an ultraviolet irradiance monitoring module that collects irradiance values ​​in real time at a high frequency;

[0027] A model predictive control algorithm is used to dynamically adjust the output power of the programmable power supply based on the light intensity deviation, ensuring a constant photolysis rate. First, a discrete state-space model is established between the ultraviolet lamp irradiance and the power supply.

[0028] ;

[0029] Where k is the current time, k+1 is the next time, A and B are model parameters, I is the UV lamp irradiance, and W is the power supply. Then, rolling optimization is performed. At each time k, the model predictive control algorithm solves an optimization problem, searching for the control sequence {W} for the next N_c time steps. k W k+1 ,...,W k+N_c-1}, so that the predicted output at the next N_p time steps is similar to the target value I. t The deviation is minimized, and the change in the control quantity is not too drastic;

[0030] The cost function is:

[0031] ;

[0032] Where the summation extends from i=1 to N_p, ΔW is the increment of the control quantity, λ is the weighting coefficient, and J is the cost function. It is the irradiance value at time k+i. It is the target irradiance value. It controls the incremental power of the UV lamp; finally, it performs feedback calibration of the UV lamp power, implementing only the first control variable W in the optimization sequence. k At the next moment, use the new measured value I. k+1 The prediction is refreshed, and the above optimization is repeated to form rolling time-domain control, thereby achieving stable control of ultraviolet lamp irradiance.

[0033] Preferably, it also includes a data acquisition and processing module, which includes an industrial control computer and a dedicated software platform. The industrial control computer and the dedicated software platform synchronously acquire all data such as flow rate, temperature, pressure, humidity, light intensity and optical signals, which are used to correct the influence of temperature, pressure and humidity on NO2 concentration measurement and final free radical concentration calculation.

[0034] Preferably, the environmental factor correction function is obtained by fitting the experimental data using a multiple linear regression model, as follows:

[0035] ;

[0036] Where β0, β1, β2...β7 are regression coefficients; T is temperature; P is pressure; H is humidity; Correction functions for temperature, pressure, and humidity factors;

[0037] The calibration formula is as follows:

[0038] ;

[0039] Among them, S sample To calibrate the NO2 signal value of the module, S background σ represents the NO2 signal value of the background module, σ represents the absorption cross-section of NO2 at the CEAS laser wavelength, L represents the effective absorption optical path of the CEAS optical cavity, and CEAS_r represents the calibration factor of the CEAS instrument for NO2.

[0040] The beneficial effects of this invention are as follows:

[0041] Compared with existing technologies, this invention, by setting up a calibration channel connected to the main photolysis cavity and an independent background channel, can accurately acquire NO2 signals under both active and passive scenarios, effectively deduct background interference, realize quantitative measurement of HO2, and use an online calibration module to accurately measure the concentration of generated NO2 molecules. By directly obtaining the net NO2 signal generated by the HO2 reaction, absolute calibration of free radical concentration can be achieved. Attached Figure Description

[0042] Figure 1 This is a system composition diagram of the present invention.

[0043] Figure 2 This is a schematic diagram of the overall structure of the system of the present invention.

[0044] Figure 3 This is a flowchart of the method of the present invention.

[0045] Figure 4 The result diagram is the actual source strength calibration.

[0046] In the diagram: 11. Zero gas generator; 12. Nitrogen standard gas cylinder; 13. Nitric oxide standard gas cylinder; 21. Nanometer; 22. Temperature and humidity sensor; 31-38. High-precision mass flow meter; 4. Main photolysis cavity; 5. Secondary reaction cavity; 6. Constant temperature reaction tube; 71. Calibration channel; 72. Background channel; 81. First NO2 detection module; 82. Second NO2 detection module; 9. Industrial control computer and dedicated software platform. Detailed Implementation

[0047] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0048] A method for generating and calibrating a peroxy free radical source, using a system for generating and calibrating a peroxy free radical source, the system comprising a free radical generation module 100, a stabilization module 200, an online calibration module 300, a fully automatic concentration control module 400, and a data acquisition and processing module 500.

[0049] The free radical generation module 100 includes a main photolysis chamber 4 and an optional secondary reaction chamber 5. The main photolysis chamber uses ultraviolet light (185nm) to photolyze H2O to produce ·H and ·OH, which are the primary sources for generating HO2. The secondary reaction chamber 5 is connected to the main photolysis chamber 4 and is used to introduce oxygen-containing gas (zero air is selected here). The ·H generated by the main photolysis chamber 4 reacts with the O2 therein to generate the corresponding HO2 free radicals.

[0050] The stabilization module 200 is connected to the outlet of the free radical generation module 100 and is a constant temperature reaction tube 6. The interior is inertized with a polytetrafluoroethylene coating to provide sufficient residence time for free radicals, so that their concentration distribution reaches dynamic equilibrium and spatial uniformity.

[0051] The online calibration module 300 employs an absolute measurement principle combining chemical conversion and optical detection. This module quantitatively reacts the system's output free radicals (HO2) with an excess of NO standard gas in a specially designed reaction chamber (HO2 + NO → OH + NO2). Subsequently, the concentration of the generated NO2 molecules is precisely measured using highly sensitive dual-channel cavity enhanced absorption spectroscopy. Through differential measurement (one channel measures the sample gas, and the other measures the background gas), the net NO2 signal generated by the HO2 reaction is directly obtained, thereby achieving absolute calibration of the free radical concentration.

[0052] The fully automatic concentration control module 400 precisely controls the flow and pressure of all gas paths through an integrated high-precision mass flow meter and pressure controller. The system is equipped with multi-parameter monitoring sensors to monitor the temperature, humidity, and pressure of the core area in real time. The ultraviolet light source is driven by a constant-power, programmable power supply and has a built-in ultraviolet irradiance monitor to provide real-time feedback on light intensity changes.

[0053] The fully automatic concentration control module 400 includes a humidity control system 410, which uses high-purity nitrogen from a nitrogen standard gas cylinder 12 as the carrier gas and is divided into a dry gas path and a humid gas path. The humid gas path humidifies through a nano-distribution pipe 21. Dry nitrogen flows inside the nano-distribution pipe 21, while water flows outside the pipe. Because the water vapor partial pressure inside the pipe (nitrogen side) is extremely low, while the water vapor partial pressure outside the pipe (humid gas side) is high, a concentration gradient is formed. Water molecules are released on the inner side of the pipe wall and enter the dry nitrogen flow, thereby achieving nitrogen humidification. The dry gas path uses pure nitrogen.

[0054] By independently controlling the flow rates of dry and wet air using two high-precision mass flow meters, and fitting real-time humidity data using the discrete least squares method to establish a relationship model between humidity and flow rate, a feedback adjustment mechanism is used to dynamically adjust the dry and wet air flow rate ratio, thereby stabilizing the relative humidity within a set range (e.g., 25%~30%RH). This method overcomes the instability of traditional humidification methods and improves the accuracy and response speed of humidity control.

[0055] The discrete least squares dynamic fitting method achieves adaptive adjustment by fitting a linear relationship model between humidity (RH) and the ratio of dry to wet air flow (R = dry air flow / wet air flow) in real time. The fitting formula is as follows:

[0056] (1)

[0057] Where a and b are the model parameters to be determined.

[0058] During system operation, k historical data points (R) will be recorded. i ,RH i ), where i = 1, 2, ..., k.

[0059] Least squares fitting: The goal is to find parameters a and b that minimize the sum of squared errors for all data points.

[0060] The formula for the sum of squared errors (E) function is as follows:

[0061] (2)

[0062] By taking the partial derivatives and setting them to zero, we can obtain the normal equations (3) and (4):

[0063] (3)

[0064] (4)

[0065] Solving this system of equations will yield the current optimal model parameters a and b.

[0066] When it is necessary to adjust the humidity to the target RH value tar At that time, the system uses the newly fitted model to solve for the target ratio R. tar For example, in formula (5):

[0067] (5)

[0068] Then, the controller according to R tar Based on the total flow requirement, the set values ​​for dry gas flow rate and wet gas flow rate are calculated, and the high-precision mass flow meter is driven to execute.

[0069] The fully automatic concentration control module 400 also includes an ultraviolet light intensity control system 420, and an ultraviolet irradiance monitoring module that collects irradiance values ​​in real time at a high frequency (10 times per second). The main control module adopts a model predictive control (MPC) algorithm to dynamically adjust the output power of the programmable power supply according to the light intensity deviation. This algorithm, by predicting system behavior and handling nonlinear characteristics, maintains light intensity stability more effectively than traditional PID control, ensuring a constant photolysis rate.

[0070] The MPC is an advanced control algorithm that uses a dynamic model to predict the system's behavior over a future period. Through rolling optimization, it calculates a series of control variables to make the predicted output as close as possible to the target value. It effectively handles the inertia and nonlinearity of UV lamp systems, exhibiting excellent suppression of sudden changes and slow drift in light intensity, and providing faster response and less overshoot than PID.

[0071] First, a discrete state-space model is established between the ultraviolet lamp irradiance (I) and the power supply (W):

[0072] (6)

[0073] Where k is the current time, k+1 is the next time, and A and B are model parameters (which can be obtained through system identification).

[0074] Then, rolling optimization is performed. At each time step k, the MPC controller solves an optimization problem, finding the control sequence {W} for the next N_c time steps. k W k+1 ,...,W k+N_c-1}, so that the predicted output at the next N_p time steps is similar to the target value I. t The deviation is minimized, and the change in the control quantity is not too drastic.

[0075] The cost function is typically:

[0076] (7)

[0077] The summation extends from i=1 to N_p, where ΔW is the increment of the control quantity and λ is the weighting coefficient.

[0078] Finally, the power of the UV lamp is calibrated using feedback, and only the first control variable W in the optimization sequence is implemented. k At the next moment, use the new measured value I. k+1 The prediction is refreshed, and the above optimization is repeated to form "rolling time domain" control, thereby achieving stable control of ultraviolet lamp irradiance.

[0079] The data acquisition and processing module 500, based on an industrial control computer and a dedicated software platform 9, synchronously acquires all data, including flow rate, temperature, pressure, light intensity, and optical signals. The software incorporates a real-time inversion algorithm, introducing an environmental factor correction function F(T,P,H). This function is obtained by fitting the experimental data using a multiple linear regression model, and is used to correct the effects of temperature (T), pressure (P), and humidity (H) on NO2 concentration measurement and the final free radical concentration calculation. The calibration formula is shown in formula (8):

[0080] (8)

[0081] Among them, S sample To calibrate the NO2 signal value of the module, S background σ represents the NO2 signal value of the background module, σ represents the absorption cross-section of NO2 at the CEAS laser wavelength, L represents the effective absorption optical path of the CEAS optical cavity, and CEAS_r represents the calibration factor of the CEAS instrument for NO2.

[0082] The correction function F(T,P,H) for the environmental factors is obtained by using a multiple linear model to obtain the parameter values ​​of the model, as shown in formula (9):

[0083] (9)

[0084] Where β0, β1, β2...β7 are regression coefficients; T is temperature; P is pressure; H is humidity;

[0085] In the laboratory, a NO2 standard gas was regulated using a dynamic calibrator to fix a known NO2 concentration. Then, the temperature, pressure, and humidity were systematically varied, and the signal value S measured by the CEAS instrument was recorded. m Then, the concentration S of the dynamic calibrator was... B By comparison, using a large amount of data and a multiple linear regression algorithm, S was fitted. B =F(T,P,H)×S m The function F in the equation, i.e., the coefficient β, is obtained. i .

[0086] During real-time calibration, the system synchronously collects the current T, P, and H values, and substitutes them into the pre-trained correction function F(T,P,H) to perform real-time correction on the NO2 signal measured by CEAS. Then, it substitutes the NO2 signal into the concentration calculation formula to obtain the concentration value of free radicals.

[0087] The invention will be further described below with reference to specific implementation examples.

[0088] like Figure 1As shown, the peroxy free radical source generation and calibration system of the present invention mainly includes a free radical generation module 100, a stabilization module 200, an online calibration module 300, a fully automatic concentration control module 400, and a data acquisition and processing module 500.

[0089] Specific implementation methods for each module:

[0090] 1. Free radical generation module 100.

[0091] This module includes a main photolysis chamber 4 and a secondary reaction chamber 5. The main photolysis chamber 4 is made of quartz and contains a 185nm low-pressure mercury lamp. Its ultraviolet irradiance is precisely controlled by the ultraviolet light intensity control system 420 in the fully automatic concentration control module. Humidified gas enters the main photolysis chamber 4 through a high-precision mass flow meter 36 at a flow rate of 150 mL / min, and undergoes a photolysis reaction under 185nm ultraviolet light irradiation.

[0092] H₂O + hν(185nm) → H + OH⁻

[0093] The secondary reaction chamber 5 uses a high-precision mass flow meter 34 to precisely control the zero air flow (500 mL / min), allowing the H atoms generated in the main photolysis chamber to react with O2 to generate HO2 free radicals.

[0094] H+O2+M→HO2+M.

[0095] 2. Stabilization module 200.

[0096] This module includes a thermostatic reaction tube 6. The thermostatic reaction tube 6 has a diameter of 10 mm, and its inner wall is inertized with a Teflon coating. The module is wrapped with a thermostatic jacket, and the temperature is maintained at 25.0±0.1℃ by a heating cable. This design ensures that the gas residence time is greater than 30 seconds, allowing the HO2 concentration to reach dynamic equilibrium and achieve spatial uniformity.

[0097] 3. Online calibration module 300.

[0098] This module includes a calibration channel 71 and a background channel 72. A nanotube is used as the reaction chamber to reduce wall loss of free radicals. In the calibration channel, HO2 reacts quantitatively with excess NO standard gas. In the background channel, NO standard gas is also added and reacts with the gas that does not pass through the free radical generation module to serve as the background.

[0099] HO2 + NO → OH + NO2

[0100] It also includes a first NO2 detection module 81 and a second NO2 detection module 82, which use broadband cavity enhanced absorption spectroscopy technology to measure the signals generated by the calibration channel and the background channel, respectively.

[0101] 4. Fully automatic concentration control module 400.

[0102] This module is the core of achieving system stability and includes two innovative subsystems: the humidity control system 410 and the ultraviolet light intensity control system 420.

[0103] The fully automatic concentration control module 400 includes a humidity control system 410, using a nitrogen standard gas cylinder 12 as the carrier gas, and is divided into a dry gas path and a wet gas path. The wet gas path humidifies through a Nafion tube. Dry nitrogen flows inside the Nafion tube 21, while water flows outside. Due to the extremely low partial pressure of water vapor inside the tube (nitrogen side) and the high partial pressure outside the tube (wet gas side), a concentration gradient is formed. Water molecules are released on the inner wall of the tube and enter the dry nitrogen flow, thus achieving nitrogen humidification. The dry gas path uses pure nitrogen. A high-precision mass flow meter 31 controls the wet gas flow rate, and a high-precision mass flow meter 32 controls the dry gas flow rate. Based on discrete least squares method, real-time humidity data is fitted to establish a model of the relationship between humidity and flow rate ratio. A feedback adjustment mechanism dynamically adjusts the dry and wet gas flow rate ratio to stabilize the relative humidity within a set range (e.g., 25%~30%RH). This method overcomes the instability of traditional humidification methods and improves the accuracy and response speed of humidity control.

[0104] The discrete least squares dynamic fitting method achieves adaptive adjustment by fitting a linear relationship model between humidity (RH) and the ratio of dry to wet air flow (R = dry air flow / wet air flow) in real time. The fitting formula is as follows:

[0105] (1)

[0106] Where a and b are the model parameters to be determined.

[0107] During system operation, k historical data points (R) will be recorded. i ,RH i ), where i = 1, 2, ..., k.

[0108] Least squares fitting: The goal is to find parameters a and b that minimize the sum of squared errors for all data points.

[0109] The formula for the sum of squared errors (E) function is as follows:

[0110] (2)

[0111] By taking the partial derivatives and setting them to zero, we can obtain the normal equations (3) and (4):

[0112] (3)

[0113] (4)

[0114] Solving this system of equations will yield the current optimal model parameters a and b.

[0115] When it is necessary to adjust the humidity to the target RH value tar At that time, the system uses the newly fitted model to solve for the target ratio R. tar For example, in formula (5):

[0116] (5)

[0117] Then, the controller according to R tar Based on the total flow requirements, the set values ​​for dry gas flow rate and wet gas flow rate are calculated, and the high-precision mass flow meters 31 and 32 are driven to perform the operation.

[0118] The fully automatic concentration control module 400 also includes an ultraviolet light intensity control system 420, and an ultraviolet irradiance monitoring module that collects irradiance values ​​in real time at a high frequency (10 times per second). The main control module adopts a model predictive control (MPC) algorithm to dynamically adjust the output power of the programmable power supply according to the light intensity deviation. This algorithm, by predicting system behavior and handling nonlinear characteristics, maintains light intensity stability more effectively than traditional PID control, ensuring a constant photolysis rate.

[0119] The MPC is an advanced control algorithm that uses a dynamic model to predict the system's behavior over a future period. Through rolling optimization, it calculates a series of control variables to make the predicted output as close as possible to the target value. It effectively handles the inertia and nonlinearity of UV lamp systems, exhibiting excellent suppression of sudden changes and slow drift in light intensity, and providing faster response and less overshoot than PID.

[0120] First, a discrete state-space model is established between the ultraviolet lamp irradiance (I) and the power supply (W):

[0121] (6)

[0122] Where k is the current time, k+1 is the next time, and A and B are model parameters (which can be obtained through system identification).

[0123] Then, rolling optimization is performed. At each time step k, the MPC controller solves an optimization problem, finding the control sequence {W} for the next N_c time steps. k W k+1 ,...,W k+N_c-1}, so that the predicted output at the next N_p time steps is similar to the target value I. t The deviation is minimized, and the change in the control quantity is not too drastic.

[0124] The cost function is typically:

[0125] (7)

[0126] The summation extends from i=1 to N_p, where ΔW is the increment of the control quantity and λ is the weighting coefficient.

[0127] Finally, the power of the UV lamp is calibrated using feedback, and only the first control variable W in the optimization sequence is implemented. k At the next moment, use the new measured value I. k+1 The prediction is refreshed, and the above optimization is repeated to form "rolling time domain" control, thereby achieving stable control of ultraviolet lamp irradiance.

[0128] 5. Data acquisition and processing module 500.

[0129] The data acquisition and processing module, based on an industrial control computer and a dedicated software platform 9, synchronously acquires all data, including flow rate, temperature, pressure, light intensity, and optical signals. An environmental factor correction function F(T,P,H) is introduced. This function is obtained by fitting the experimental data using a multiple linear regression model and is used to correct the effects of temperature (T), pressure (P), and humidity (H) on NO2 concentration measurement and the final free radical concentration calculation. The calibration formula is shown in formula (8):

[0130] (8)

[0131] Among them, S sample To calibrate the NO2 signal value of the module, S background σ represents the NO2 signal value of the background module, σ represents the absorption cross-section of NO2 at the CEAS laser wavelength, L represents the effective absorption optical path of the CEAS optical cavity, and CEAS_r represents the calibration factor of the CEAS instrument for NO2.

[0132] The correction function F(T,P,H) for the environmental factors is obtained by using a multiple linear model to obtain the parameter values ​​of the model, as shown in formula (9):

[0133] (9)

[0134] Where β0, β1, β2...β7 are regression coefficients; T is temperature; P is pressure; H is humidity;

[0135] In the laboratory, a NO2 standard gas was regulated using a dynamic calibrator to fix a known NO2 concentration. Then, the temperature, pressure, and humidity were systematically varied, and the signal value S measured by the CEAS instrument was recorded. m Then, the concentration S of the dynamic calibrator was... B By comparison, using a large amount of data and a multiple linear regression algorithm, S was fitted. B =F(T,P,H)×S mThe function F in the equation, i.e., the coefficient β, is obtained. i .

[0136] During real-time calibration, the system synchronously collects the current T, P, and H values ​​and substitutes them into the pre-trained correction function F(T,P,H) to perform real-time correction on the NO2 signal measured by CEAS. Then, it substitutes the NO2 signal into the concentration calculation formula to obtain the concentration value and stability index (relative standard deviation) of free radicals.

[0137] Figure 4 In a case study of actual measurements, the average source intensity was (238.3±15.0) ppt during 5 hours of continuous operation. The relative error of 12 consecutive measurements was within ±10%. The temperature was controlled at (25.0±0.1) ℃, the humidified gas was (25.0%±0.5%) RH, and the ultraviolet light intensity fluctuation was <5%. The system demonstrated excellent stability and repeatability, meeting the accuracy requirements for atmospheric free radical monitoring.

[0138] The present invention has the following advantages:

[0139] (1) Compared with the prior art, the present invention is the first to use dynamic fitting feedback control based on discrete least squares method. This combination realizes rapid, accurate and adaptive adjustment of humidity and reduces the change in free radical generation rate caused by water vapor concentration fluctuation.

[0140] (2) This invention explicitly adopts the model predictive control (MPC) algorithm to replace the conventional PID, which effectively addresses the nonlinearity, aging and environmental interference of the ultraviolet lamp, controls the light intensity fluctuation at an extremely low level, and ensures the stable generation of free radical concentration.

[0141] (3) The present invention introduces an environmental factor correction function, which automatically compensates for the influence of environmental fluctuations on CEAS measurement through an algorithm model, significantly reducing the overall uncertainty of the system and making the calibration results of peroxy free radical concentration more reliable.

[0142] (4) Stable and controllable concentration: Combining the clean reaction path of ultraviolet photolysis of H2O and high-precision flow control, the generated HO2 has high purity and the concentration can be adjusted in the range of (200~550)ppt, and has good stability.

[0143] (5) Online absolute calibration: accurately acquire NO2 signals under both active and passive scenarios, effectively deduct background interference, and realize quantitative measurement of HO2.

[0144] (6) Real-time and fast: The entire occurrence and calibration process is carried out online continuously. The data acquisition and processing system can display the concentration results in real time. The response time is fast and can meet the needs of dynamic chemical reaction research.

[0145] 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 principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A peroxide free radical source generation and calibration system, comprising a free radical generation module (100) and an online calibration module (300), characterized in that: The free radical generation module (100) includes a main photolysis chamber (4), which generates ·H and ·OH by photolysis of H2O with ultraviolet light, and controls the ·H generated by the main photolysis chamber (4) to react with O2 to generate HO2 free radicals; The online calibration module (300) includes a calibration channel (71) connected to the main photolysis cavity (4) and an independent background channel (72). NO standard gas is introduced into the calibration channel (71) for calibration, and HO2 reacts quantitatively with excess NO standard gas here. NO standard gas is added to the background channel (72) to react with gas that does not pass through the free radical generation module as background. The online calibration module (300) is used to accurately measure the concentration of generated NO2 molecules. By directly obtaining the net NO2 signal generated by the HO2 reaction, the absolute calibration of the free radical concentration is achieved. It also includes a fully automatic concentration control module (400), which includes a high-precision mass flow meter and a pressure controller, as well as a nitrogen standard gas cylinder (12). The nitrogen standard gas cylinder (12) is connected to a dry gas path and a wet gas path. The wet gas path is connected in series with a nano-distribution pipe (21). The dry gas path and the wet gas path are independently controlled by the high-precision mass flow meter, and the ratio of dry gas to wet gas flow can be dynamically adjusted.

2. The peroxy free radical source generation and calibration system according to claim 1, characterized in that, The free radical generation module (100) also includes a secondary reaction chamber (5), which is connected to the main photolysis chamber (4). Oxygen-containing gas is introduced into the secondary reaction chamber (5) to react with the ·H generated by the main photolysis chamber (4) to generate HO2 free radicals.

3. The peroxy free radical source generation and calibration system according to claim 1, characterized in that, It also includes a stabilization module (200), which includes a thermostatic reaction tube (6) connected to the outlet of the free radical generation module (100). The free radicals remain in the thermostatic reaction tube (6) to achieve dynamic equilibrium and spatial uniformity in their concentration distribution.

4. The peroxide free radical source generation and calibration system according to claim 3, characterized in that, The residence time of the gas containing free radicals is greater than 30 seconds; the outside of the stabilization module (200) is wrapped with a constant temperature jacket to maintain the temperature of the stabilization module (200) at 25.0±0.1℃.

5. The peroxide free radical source generation and calibration system according to claim 4, characterized in that, Based on the discrete least squares method, a model relating humidity to flow rate is established by fitting real-time humidity data. A feedback adjustment mechanism is then used to dynamically adjust the dry-to-wet air flow rate ratio, stabilizing the relative humidity within a set range. The fitting formula for this dynamic fitting method using discrete least squares is as follows: ； Where a and b are the model parameters to be determined, R is the dry air flow rate / wet air flow rate, and RH is the real-time fitted humidity; During system operation, k historical data points R are recorded. i and RH i where i = 1, 2, ..., k; The parameters a and b are found by fitting using the least squares method, which minimizes the sum of squared errors for all data points.

6. The peroxy free radical source generation and calibration system according to claim 5, characterized in that, The formula for the sum of squared errors function is as follows: ； By taking the partial derivative and setting it to zero, we can obtain: ； ； Solving the system of equations yields the current optimal model parameters a and b. In the formula: E is the sum of squared errors. Let be the relative humidity at time i. Let be the ratio of dry to wet air flow rate at time i, and a, b, and k be model parameters; When it is necessary to adjust the humidity to the target RH value tar At that time, the system uses the newly fitted model to solve for the target ratio R. tar For example: ； In the formula: The target dry-to-wet air flow ratio, Let a be the target humidity, and b be the model parameters. Then, the controller according to R tar Based on the total flow requirement, the set values ​​for dry gas flow rate and wet gas flow rate are calculated, and the high-precision mass flow meter is driven to execute.

7. The peroxy free radical source generation and calibration system according to claim 1, characterized in that, The fully automatic concentration control module (400) also includes an ultraviolet light intensity control system (420), and an ultraviolet irradiance monitoring module that collects irradiance values ​​in real time at high frequency. A model predictive control algorithm is used to dynamically adjust the output power of the programmable power supply according to the light intensity deviation to ensure a constant photolysis rate; First, establish a discrete state-space model between the ultraviolet lamp irradiance and the power supply: ； Where k is the current time, k+1 is the next time, A and B are model parameters, I is the UV lamp irradiance value, and W is the power supply. Then, rolling optimization is performed. At each time k, the Model Predictive Control algorithm solves an optimization problem to find the control sequence {W} for the next N_c time steps. k W k+1 ,...,W k+N_c-1 }, so that the predicted output at the next N_p time steps is similar to the target value I. t The deviation is minimized, and the change in the control quantity is not too drastic; The cost function is: ； Where the summation extends from i=1 to N_p, ΔW is the increment of the control quantity, λ is the weighting coefficient, and J is the cost function. It is the irradiance value at time k+i. It is the target irradiance value. It controls the increment of the ultraviolet lamp power; Finally, the power of the UV lamp is calibrated using feedback, and only the first control variable W in the optimization sequence is implemented. k At the next moment, use the new measured value I. k+1 The prediction is refreshed, and the above optimization is repeated to form rolling time-domain control, thereby achieving stable control of ultraviolet lamp irradiance.

8. The peroxy free radical source generation and calibration system according to claim 1, characterized in that, It also includes a data acquisition and processing module (500), which includes an industrial computer and a dedicated software platform (9). The industrial computer and the dedicated software platform (9) synchronously acquire flow rate, temperature, pressure, humidity, light intensity and optical signals to correct the influence of temperature, pressure and humidity on NO2 concentration measurement and final free radical concentration calculation.

9. The peroxy free radical source generation and calibration system according to claim 8, characterized in that, The environmental factor correction function was obtained by fitting the experimental data using a multiple linear regression model, as follows: ； Where β0, β1, β2...β7 are regression coefficients; T is temperature; P is pressure; H is humidity; Correction functions for temperature, pressure, and humidity factors; The calibration formula is as follows: ； Among them, S sample To calibrate the NO2 signal value of the module, S background σ represents the NO2 signal value of the background module, σ represents the absorption cross-section of NO2 at the CEAS laser wavelength, L represents the effective absorption optical path of the CEAS optical cavity, and CEAS_r represents the calibration factor of the CEAS instrument for NO2.

Images