A method and device for error prediction and compensation based on pattern recognition technology

Abstract

The present invention discloses an error prediction compensation method and a device for use based on pattern recognition technology, belonging to the field of gas sensing technology; specifically comprising the following steps: first, using light intensity compensation to eliminate the measurement error caused by the initial light intensity change; then adjusting the output of the sensor to a calibration mode through an equivalent coefficient to eliminate the error caused by inaccurate temperature and pressure control; finally, correcting the gas molecule number error based on the deviation between the measured actual gas temperature and pressure and the gas temperature and pressure in the calibration mode. When monitoring the mode of the sensor, the method of the present invention uses an equivalent coefficient to determine the reliability of the sensor output, and the actual pressure of the gas can be predicted through the deviation of the equivalent coefficient, thereby improving the measurement accuracy and realizing a wide range of detection, providing a new method for monitoring and correcting the sensor mode and measurement error in the temperature adaptive TDLAS system.

Description

Technical Field

[0001] The present invention belongs to the technical field of gas sensing, and in particular relates to an error prediction compensation method based on pattern recognition technology and a use device thereof. Background Art

[0002] Gas geochemical exploration methods are effective tools for monitoring and analyzing geological anomalies. During geological activities, geochemical gases containing a large amount of information will be discharged to the surface along the fault zone, resulting in abnormal gas concentrations near the fault zone. Changes in gas concentrations are important early warning signals for geological anomalies. 2 It is an important geochemical gas, closely related to geological formation and the Earth's ecosystem. Due to human activities and seasonal changes, CO in the atmosphere 2 The gas concentration growth rate is 0.7% / year. Therefore, determining the cause of the change in gas concentration is the key to geological analysis, and accurate monitoring of CO 2 The spatial and temporal distribution of earthquakes is of great significance for earthquake early warning and geological analysis.

[0003] Gas sensors based on tunable diode laser absorption spectroscopy have been successfully applied in industrial analysis, environmental monitoring, and medical diagnosis due to their high sensitivity, high stability, and simple structure. Sensors must simultaneously overcome challenges such as low power consumption, high stability, and high precision to meet the requirements of field gas geochemical exploration. At present, research on improving measurement accuracy is divided into two directions: one is to use a system with a special structure to eliminate errors. For example, RK Hanson developed a gas sensor that uses scanning wavelength modulation spectroscopy for 1f phase detection to measure CH 4 / air plane burner exhaust temperature and water mole fraction, which improves the measurement accuracy, but its complex system structure makes its stability in the field easily disturbed. The other is the error correction method, which is usually used to reduce power consumption while ensuring measurement accuracy. For example, Zhu Xiaorui et al. proposed a temperature correction method based on the influence of temperature on spectral line intensity in 2019 and applied it to power plant exhaust gas measurement. At ambient temperature, when CO 2 At a concentration of 10%, the measurement error was reduced from -2.86% to -2.3%. However, these sensors are susceptible to changes in initial light intensity, gas temperature and pressure in the field. At the same time, the reliability of the measurement results is not monitored and further compensated. Therefore, it is necessary to monitor and analyze the sensor mode and further compensate to overcome the problem of inaccurate sensor parameter control.

[0004] In summary, monitoring sensor patterns and prediction errors and compensation are of great significance and are one of the important topics in the field of gas geochemical detection. Summary of the invention

[0005] The purpose of the present invention is to provide an error prediction compensation method and a device based on pattern recognition technology to solve the technical problem in the prior art that when the TDLAS technology detection system is used for field measurement, the sensor is affected by factors such as initial light intensity, gas temperature and pressure, resulting in inaccurate control of multiple parameters.

[0006] To achieve the above purpose, the specific technical scheme of an error prediction compensation method based on pattern recognition technology and a device for using the same is as follows:

[0007] An error prediction compensation method based on pattern recognition technology includes the following steps, and the following steps are performed in sequence:

[0008] Step 1: using the first temperature sensor to measure the ambient temperature, and using the second temperature sensor and the third temperature sensor to measure and set the target temperature of the external temperature control device and the multi-pass pool temperature control device according to the ambient temperature, and setting the target temperature of the external temperature control device to be 2°C higher than the ambient temperature, and the target temperature of the multi-pass pool temperature control device to be 4°C higher than the ambient temperature;

[0009] Step 2: After completing step 1, turn on the external temperature control device and the multi-pass pool temperature control device. After the temperature of the external temperature control device and the multi-pass pool temperature control device is stable, open the inlet proportional valve, the outlet proportional valve and the vacuum pump to charge the CO to be tested. 2 The sample gas was used for 300 s to completely replace the residual gas inside the multi-pass cell, the inlet proportional valve was closed, and the pressure of the multi-pass cell was controlled to the target pressure (40 Torr), and the outlet proportional valve and vacuum pump were closed;

[0010] Step 3: After completing step 2, turn on the CO 2 Laser and multi-channel acquisition card to start measuring CO in the sample gas to be tested 2 Gas concentration and record it;

[0011] Step 4: After completing step 3, use the light intensity compensation formula to eliminate the measurement error caused by the initial light intensity change;

[0012] Step 5: After completing the light intensity compensation in step 4, use the equivalent coefficient Q to monitor the mode of the sensor to complete the equivalent compensation;

[0013] Step 6: Use the temperature characteristic difference of different absorption line intensities to measure the CO to be measured 2 The actual temperature of the gas;

[0014] Step 7: Calculate the actual pressure of the gas based on the actual temperature of the gas and the equivalent coefficient before and after equivalent compensation;

[0015] Step 8: Finally, the gas molecule number error is corrected based on the deviation between the measured actual gas temperature and pressure and the gas temperature and pressure in the calibration mode.

[0016] Furthermore, the error prediction compensation process can be expressed as:

[0017] c o =DQMc (12)

[0018] Among them, c is the original gas concentration data, c o It is the gas concentration data output after compensation.

[0019] Furthermore, in step 4, the measurement error caused by the change in initial light intensity is corrected according to formulas (1), (2) and (3);

[0020] According to the relevant principles of wavelength modulation spectroscopy technology, the amplitude of the second harmonic signal is expressed as:

[0021] 2f A =-nσKLI 0 (1)

[0022] Where n is the number density of gas molecules, σ is the absorption cross section at the center of the absorption line, K is a constant related to the system, L is the effective absorption path length of light, and I 0 is the initial light intensity during calibration.

[0023] Therefore, compared with the initial light intensity I 0 The relevant errors include the output fluctuation of the laser, the change of the system-related constant K, and the initial light intensity error caused by the thermal expansion of the multi-pass cell. The light intensity compensation is used to correct the above errors; the light intensity compensation coefficient D is expressed as:

[0024]

[0025] Among them, K T is a constant related to the system at temperature T, and I is the light intensity during measurement; the output after light intensity compensation is described as:

[0026] 2f a =2f A D (3).

[0027] Further, in step 5, when Q is not equal to Q 0 When the sensor mode is not in the calibration mode, it will cause measurement error. The data corrected in step 4 is corrected using formulas (4), (5) and (6) to correct its influence on the measurement results.

[0028] The amplitude of the second harmonic in calibration mode is expressed as:

[0029] 2fA0 =-nσ 0 K 0 L 0 I 0 (4)

[0030] Among them, σ 0 is the absorption cross section at the center of the absorption line under the gas temperature and gas pressure during calibration, K 0 is a constant related to the system at the gas temperature during calibration, L 0 is the effective absorption path length of light during calibration; the equivalent coefficient in calibration mode is defined as:

[0031]

[0032] When the equivalent coefficient Q is equal to Q 0 When the temperature and pressure combination at this time is defined as the calibration mode of the sensor, that is, the output of the sensor is independent of the temperature and pressure of the gas in the calibration mode, and the error caused by the temperature and pressure of the gas can be corrected according to the principle of equivalent compensation;

[0033] By adjusting the gas pressure to adjust the sensor output until it is in calibration mode, the output after equivalent compensation is specifically expressed as:

[0034]

[0035] Furthermore, in step 6, the deviation between the actual temperature of the gas and the gas temperature during calibration is calculated according to formula (7), and the actual temperature of the gas can be calculated using the temperature deviation and the gas temperature during calibration;

[0036]

[0037] Where k is the Boltzmann constant, h is the Planck constant, c is the speed of light, ΔE is the energy level difference at the ground state, A is the amplitude ratio of the second harmonic signal, and I 1 and I 2 All of them are obtained through the synchronous detection of light intensity, T c is the reference temperature, s 1 and 2 are the intensities of two different absorption lines at the reference temperature.

[0038] Furthermore, in step 7, the pressure change can be calculated according to formulas (8) and (9) by using the response of the sensor at different pressures and amplitudes of the second harmonic signal before and after equivalent compensation; the ratio of the amplitudes of the second harmonic signal before and after equivalent compensation is:

[0039]

[0040] The pressure deviation is expressed as:

[0041] δP=F(A q ) (9)

[0042] Among them, F(A q ) is the sensor response function under different pressure values.

[0043] Further, in step 8, the last step of molecular number compensation is completed according to formulas (10) and (11);

[0044] According to the universal gas law, the molecular number density is affected by the gas temperature and pressure; the molecular number density is corrected by the temperature and pressure deviations and expressed as:

[0045]

[0046] Where R is the universal gas constant, V is the volume of the multi-pass cell, T 0 is the gas temperature during calibration, P 0 is the gas pressure during calibration; the sensor output after molecular number compensation can be expressed as:

[0047] 2f c =2f A DQM (11)

[0048] Based on the above equivalent principle, the mode of the sensor can be monitored and the measurement error can be corrected.

[0049] The present invention also provides a device used in the error prediction compensation method based on pattern recognition technology, comprising an external temperature control device, a multi-pass cell temperature control device is arranged inside the external temperature control device, and a multi-pass cell and a first photoelectric detector are arranged inside the multi-pass cell temperature control device;

[0050] An air inlet proportional valve and a pressure sensor are arranged on the inlet pipeline of the multi-pass pool, and an air outlet proportional valve and a vacuum pump are arranged on the outlet pipeline of the multi-pass pool;

[0051] The outlet of the optical fiber coupler is optically connected to the inlet of the multi-pass cell and one end of the second photodetector respectively, and the outlet of the multi-pass cell is optically connected to one end of the first photodetector;

[0052] The other end of the first photodetector is electrically connected to the phase-locked amplifier and the multi-channel acquisition card respectively, and the multi-channel acquisition card is electrically connected to the phase-locked amplifier at the same time, the other end of the second photodetector is electrically connected to the multi-channel acquisition card, the multi-channel acquisition card is also electrically connected to one end of the microcomputer, and the other end of the microcomputer is electrically connected to one end of the signal generating circuit;

[0053] The other end of the signal generating circuit is electrically connected to the control circuit, and the other end of the control circuit is electrically connected to the CO2 Laser driver and temperature control electrical connection, CO 2 Laser drive and temperature control and CO 2 Laser electrical connection; CO 2 An inlet optical connection of the laser to the fiber coupler;

[0054] A first temperature sensor is arranged on the external temperature control device, and a second temperature sensor and a third temperature sensor are arranged on the multi-pass pool. The first temperature sensor is used to measure the ambient temperature, the second temperature sensor is used to measure and set the target temperature of the external temperature control device, and the third temperature sensor is used to measure and set the target temperature of the multi-pass pool temperature control device.

[0055] The error prediction and compensation method based on pattern recognition technology and the use device of the present invention have the following advantages: when monitoring the mode of the sensor, the equivalent coefficient is used to determine the reliability of the sensor output, and the actual pressure of the gas can be predicted by the deviation of the equivalent coefficient, thereby improving the measurement accuracy and realizing large-range detection, providing a new method for monitoring and correcting the sensor mode and measurement error in the temperature adaptive TDLAS system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The present invention is a structural block diagram of a device using an error prediction and compensation method based on pattern recognition technology.

[0057] Figure 2 The present invention is a detection flow chart of an error prediction and compensation method based on pattern recognition technology.

[0058] Explanation of the markings in the figure: 1. Proportional valve at the air inlet; 2. Proportional valve at the air outlet; 3. Vacuum pump; 4. Pressure sensor; 5. Multi-pass cell; 6. First photodetector; 7. Second photodetector; 8. Lock-in amplifier; 9. Multi-channel acquisition card; 10. Microcomputer; 11. Signal generating circuit; 12. Control circuit; 13. CO 2 Laser drive and temperature control; 14. CO 2 Laser; 15. Fiber optic coupler; 16. Multi-pass pool temperature control device; 17. External temperature control device; 18. First temperature sensor; 19. Second temperature sensor; 20. Third temperature sensor. DETAILED DESCRIPTION

[0059] In order to better understand the purpose, structure and function of the present invention, the following is a further detailed description of an error prediction compensation method based on pattern recognition technology and a device for using the present invention in conjunction with the accompanying drawings.

[0060] The present invention discloses an error prediction compensation method and a device for use based on pattern recognition technology, which specifically include the following steps: first, light intensity compensation is used to eliminate the measurement error caused by the change in initial light intensity; then, the output of the sensor is adjusted to the calibration mode through the equivalent coefficient to eliminate the error caused by inaccurate temperature and pressure control; finally, the gas molecule number error is corrected according to the deviation between the measured actual gas temperature and pressure and the gas temperature and pressure in the calibration mode. When monitoring the mode of the sensor, the method of the present invention uses the equivalent coefficient to determine the reliability of the sensor output, and the deviation of the equivalent coefficient can predict the actual pressure of the gas, improve the measurement accuracy, and achieve large-scale detection, which provides a new method for monitoring and correcting the sensor mode and measurement error in the temperature adaptive TDLAS system.

[0061] An error prediction compensation method based on pattern recognition technology includes the following steps, and the following steps are performed in sequence:

[0062] Step 1: using the first temperature sensor 18 to measure the ambient temperature, and using the second temperature sensor 19 and the third temperature sensor 20 to measure and set the target temperatures of the external temperature control device 17 and the multi-pass pool temperature control device 16 according to the ambient temperature, and setting the target temperature of the external temperature control device 17 to be 2°C higher than the ambient temperature, and the target temperature of the multi-pass pool temperature control device 16 to be 4°C higher than the ambient temperature;

[0063] Step 2: After completing step 1, turn on the external temperature control device 17 and the multi-pass pool temperature control device 16. After the temperatures of the external temperature control device 17 and the multi-pass pool temperature control device 16 are stable, open the inlet proportional valve 1, the outlet proportional valve 2 and the vacuum pump 3 to charge the CO to be measured. 2 The sample gas is used for 300 s to completely replace the residual gas inside the multi-pass cell 5, the inlet proportional valve 1 is closed, and the pressure of the multi-pass cell 5 is controlled to the target pressure (40 Torr), and the outlet proportional valve 2 and the vacuum pump 3 are closed;

[0064] Step 3: After completing step 2, turn on the CO 2 The laser 14 and the multi-channel acquisition card 9 start measuring the CO in the sample gas to be tested. 2 Gas concentration and record it;

[0065] Step 4: After completing step 3, use the light intensity compensation formula to eliminate the measurement error caused by the initial light intensity change;

[0066] Step 5: After completing the light intensity compensation in step 4, use the equivalent coefficient Q to monitor the mode of the sensor to complete the equivalent compensation;

[0067] Step 6: Use the temperature characteristic difference of different absorption line intensities to measure the CO to be measured 2 The actual temperature of the gas;

[0068] Step 7: Calculate the actual pressure of the gas based on the actual temperature of the gas and the equivalent coefficient before and after equivalent compensation;

[0069] Step 8: Finally, the gas molecule number error is corrected based on the deviation between the measured actual gas temperature and pressure and the gas temperature and pressure in the calibration mode.

[0070] Furthermore, the error prediction compensation process can be expressed as:

[0071] c o =DQMc (12)

[0072] Among them, c is the original gas concentration data, c o It is the gas concentration data output after compensation.

[0073] Furthermore, in step 4, the measurement error caused by the change in initial light intensity is corrected according to formulas (1), (2) and (3);

[0074] According to the relevant principles of wavelength modulation spectroscopy technology, the amplitude of the second harmonic signal is expressed as:

[0075] 2f A =-nσKLI 0 (1)

[0076] Where n is the number density of gas molecules, σ is the absorption cross section at the center of the absorption line, K is a constant related to the system, L is the effective absorption path length of light, and I 0 is the initial light intensity during calibration.

[0077] Therefore, compared with the initial light intensity I 0 The relevant errors include the output fluctuation of the laser, the change of the system-related constant K and the initial light intensity error caused by the thermal expansion of the multi-pass cell 5. The light intensity compensation is used to correct the above errors; the light intensity compensation coefficient D is expressed as:

[0078]

[0079] Among them, K T is a constant related to the system at temperature T, and I is the light intensity during measurement; the output after light intensity compensation is described as:

[0080] 2f a =2f A D (3).

[0081] Further, in step 5, when Q is not equal to Q0 When the sensor mode is not in the calibration mode, it will cause measurement error. The data corrected in step 4 is corrected using formulas (4), (5) and (6) to correct its influence on the measurement results.

[0082] The amplitude of the second harmonic in calibration mode is expressed as:

[0083] 2f A0 =-nσ 0 K 0 L 0 I 0 (4)

[0084] Among them, σ 0 is the absorption cross section at the center of the absorption line under the gas temperature and gas pressure during calibration, K 0 is a constant related to the system at the gas temperature during calibration, L 0 is the effective absorption path length of light during calibration; the equivalent coefficient in calibration mode is defined as:

[0085]

[0086] When the equivalent coefficient Q is equal to Q 0 When the temperature and pressure combination at this time is defined as the calibration mode of the sensor, that is, the output of the sensor is independent of the temperature and pressure of the gas in the calibration mode, and the error caused by the temperature and pressure of the gas can be corrected according to the principle of equivalent compensation;

[0087] By adjusting the gas pressure to adjust the sensor output until it is in calibration mode, the output after equivalent compensation is specifically expressed as:

[0088]

[0089] Furthermore, in step 6, the deviation between the actual temperature of the gas and the gas temperature during calibration is calculated according to formula (7), and the actual temperature of the gas can be calculated using the temperature deviation and the gas temperature during calibration;

[0090]

[0091] Where k is the Boltzmann constant, h is the Planck constant, c is the speed of light, ΔE is the energy level difference at the ground state, A is the amplitude ratio of the second harmonic signal, and I 1 and I 2 All of them are obtained through the synchronous detection of light intensity, T c is the reference temperature, s 1 and 2 are the intensities of two different absorption lines at the reference temperature.

[0092] Furthermore, in step 7, the pressure change can be calculated according to formulas (8) and (9) by using the response of the sensor at different pressures and amplitudes of the second harmonic signal before and after equivalent compensation; the ratio of the amplitudes of the second harmonic signal before and after equivalent compensation is:

[0093]

[0094] The pressure deviation is expressed as:

[0095] δP=F(A q ) (9)

[0096] Among them, F(A q ) is the sensor response function under different pressure values.

[0097] Further, in step 8, the last step of molecular number compensation is completed according to formulas (10) and (11);

[0098] According to the universal gas law, the molecular number density is affected by the gas temperature and pressure; the molecular number density is corrected by the temperature and pressure deviations and expressed as:

[0099]

[0100] Where R is the universal gas constant, V is the volume of the multi-pass cell 5, T 0 is the gas temperature during calibration, P 0 is the gas pressure during calibration; the sensor output after molecular number compensation can be expressed as:

[0101] 2f c =2f A DQM (11)

[0102] Based on the above equivalent principle, the mode of the sensor can be monitored and the measurement error can be corrected.

[0103] The present invention also provides a device used in the error prediction compensation method based on pattern recognition technology, comprising a multi-pass cell temperature control device 16, wherein a multi-pass cell 5 and a first photodetector 6 are arranged in the multi-pass cell temperature control device 16;

[0104] An air inlet proportional valve 1 and a pressure sensor 4 are arranged on the inlet pipeline of the multi-pass pool 5, and an air outlet proportional valve 2 and a vacuum pump 3 are arranged on the outlet pipeline of the multi-pass pool 5;

[0105] The outlet of the optical fiber coupler 15 is optically connected to the inlet of the multi-pass cell 5 and one end of the second photodetector 7 respectively, and the outlet of the multi-pass cell 5 is optically connected to one end of the first photodetector 6;

[0106] The other end of the first photodetector 6 is electrically connected to the phase-locked amplifier 8 and the multi-channel acquisition card 9 respectively, and the multi-channel acquisition card 9 is also electrically connected to the phase-locked amplifier 8, the other end of the second photodetector 7 is electrically connected to the multi-channel acquisition card 9, the multi-channel acquisition card 9 is also electrically connected to one end of the microcomputer 10, and the other end of the microcomputer 10 is electrically connected to one end of the signal generating circuit 11;

[0107] The other end of the signal generating circuit 11 is electrically connected to the control circuit 12, and the other end of the control circuit 12 is electrically connected to the CO 2 Laser driver and temperature control 13 electrical connection, CO 2 Laser drive and temperature control 13 and CO 2 Laser 14 is electrically connected to CO 2 The laser 14 is optically connected to the entrance of the fiber coupler 15;

[0108] All the above components are arranged in the external temperature control device 17;

[0109] A first temperature sensor 18 is arranged on the external temperature control device 17, and a second temperature sensor 19 and a third temperature sensor 20 are arranged on the multi-pass pool 5. The first temperature sensor 18 is used to measure the ambient temperature, the second temperature sensor 19 is used to measure and set the target temperature of the external temperature control device 17, and the third temperature sensor 20 is used to measure and set the target temperature of the multi-pass pool temperature control device 16.

[0110] Embodiment 1:

[0111] The proposed multi-component gas detection device was used to detect the standard gas (3000ppmv CO 2 ) is taken as an example to illustrate the implementation process of an error prediction compensation method based on pattern recognition technology proposed by the present invention, which includes the following steps, and the following steps are performed in sequence:

[0112] Step 1: The ambient temperature is measured by the first temperature sensor 18 as 30°C, the target temperature of the external temperature control device 17 is measured and set to 32°C according to the ambient temperature by the second temperature sensor 19, and the target temperature of the multi-pass pool temperature control device 16 is measured and set to 34°C according to the ambient temperature by the third temperature sensor 20;

[0113] Step 2: Turn on the external temperature control device 17 and the multi-pass cell temperature control device 16. After the temperature of the external temperature control device 17 and the multi-pass cell temperature control device 16 is stable, open the inlet proportional valve 1, the outlet proportional valve 2 and the vacuum pump 3 to charge the CO to be measured. 2The sample gas is used for 300 s to completely replace the residual gas inside the multi-pass cell 5, the inlet proportional valve 1 is closed, and the pressure of the multi-pass cell 5 is controlled to the target pressure, and the outlet proportional valve 2 and the vacuum pump 3 are closed;

[0114] Step 3: Turn on CO 2 Laser 14 and multi-channel acquisition card 9 start measuring CO 2 Gas concentration, CO 2 The output concentration is 2824.5ppmv;

[0115] Step 4: Use the light intensity compensation formulas (1), (2) and (3) to eliminate the measurement error caused by the initial light intensity change. The compensated CO 2 The concentration is 2746.5ppmv;

[0116] Step 5: After completing the light intensity compensation, use the equivalent coefficient Q to monitor the sensor mode to complete the equivalent compensation. 0 When the sensor mode is not in calibration mode, it will cause measurement error. The data corrected in step 4 is corrected using formulas (4), (5) and (6). The compensated CO 2 The concentration is 2949.6ppmv;

[0117] Step 6: Calculate the deviation between the actual gas temperature and the calibration temperature according to formula (7), and use the temperature characteristic difference of different absorption line intensities to measure the CO 2 The actual temperature of the gas;

[0118] Step 7: The pressure change can be calculated according to formulas (8) and (9), and the actual pressure of the gas can be calculated using the actual temperature of the gas and the equivalent coefficient before and after equivalent compensation;

[0119] Step 8: Correct the gas molecule number error based on the deviation between the measured gas temperature and pressure and the gas temperature and pressure in the calibration mode. Complete the last step of molecule number compensation according to formulas (10) and (11). The compensated CO 2 The concentration is 2992.5ppmv.

[0120] After compensation CO 2 The accuracy increased from 94.15% to 99.75%. It can be seen that the use of error prediction compensation method can improve measurement accuracy and achieve large-scale detection.

[0121] It is to be understood that the present invention is described by some embodiments, and it is known to those skilled in the art that various changes or equivalent substitutions may be made to these features and embodiments without departing from the spirit and scope of the present invention. In addition, under the teachings of the present invention, these features and embodiments may be modified to adapt to specific circumstances and materials without departing from the spirit and scope of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the scope of protection of the present invention.

Claims

1. An error prediction compensation method based on pattern recognition technology, It is characterized in that The following steps are included: And the following steps are performed in sequence: Step 1: using a first temperature sensor (18) to measure the ambient temperature, and using a second temperature sensor (19) and a third temperature sensor (20) to measure and set the target temperatures of the external temperature control device (17) and the multi-pass pool temperature control device (16) according to the ambient temperature, and setting the target temperature of the external temperature control device (17) to be 2°C higher than the ambient temperature, and the target temperature of the multi-pass pool temperature control device (16) to be 4°C higher than the ambient temperature; Step 2: After completing the above Step 1, turn on the external temperature control device (17) and the multi-pass cell temperature control device (16). After the temperatures of the external temperature control device (17) and the multi-pass cell temperature control device (16) are stabilized, open the inlet proportional valve (1), the outlet proportional valve (2) and the vacuum pump (3), and fill in the CO to be measured. 2 The residual gas inside the multi-pass cell (5) is completely replaced by the sample gas. Close the inlet proportional valve (1), control the pressure of the multi-pass cell (5) to the target pressure, and close the outlet proportional valve (2) and the vacuum pump (3). Step 3: After completing step 2, turn on the CO 2 The laser (14) and the multi-channel acquisition card (9) start measuring the CO in the sample gas to be tested. 2 Gas concentration and record it; Step 4: After completing step 3, use the light intensity compensation formula to eliminate the measurement error caused by the initial light intensity change; In step 4, the measurement error caused by the change in initial light intensity is corrected according to formulas (1), (2) and (3); According to the relevant principles of wavelength modulation spectroscopy technology, the amplitude of the second harmonic signal is expressed as: 2f A =-nσKLI 0 (1) Where n is the number density of gas molecules, σ is the absorption cross section at the center of the absorption line, K is a constant related to the system, L is the effective absorption path length of light, and I 0 is the initial light intensity during calibration; Therefore, compared with the initial light intensity I 0 The relevant errors include the output fluctuation of the laser, the change of the system-related constant K and the initial light intensity error caused by the thermal expansion of the multi-pass cell (5). The light intensity compensation is used to correct the above errors; the light intensity compensation coefficient D is expressed as: Among them, K T is a constant related to the system at temperature T, and I is the light intensity during measurement; the output after light intensity compensation is described as: <h2 style=";text-align:left;direction:ltr">2f<h2 style=";text-align:left;direction:ltr"> a <h2 style=";text-align:left;direction:ltr"> =2f<h2 style=";text-align:left;direction:ltr"> A <h2 style=";text-align:left;direction:ltr"> D (3); Step 5: After completing the light intensity compensation in step 4, use the equivalent coefficient Q to monitor the mode of the sensor to complete the equivalent compensation; In step 5, when Q is not equal to Q 0 When the sensor mode is not in the calibration mode, it will cause measurement error. The data corrected in step 4 is corrected using formulas (4), (5) and (6) to correct its influence on the measurement results. The amplitude of the second harmonic in calibration mode is expressed as: 2f A0 =-nσ 0 K 0 L 0 I 0 (4) Among them, σ 0 is the absorption cross section at the center of the absorption line under the gas temperature and gas pressure during calibration, K 0 is a constant related to the system at the gas temperature during calibration, L 0 is the effective absorption path length of light during calibration; the equivalent coefficient in calibration mode is defined as: When the equivalent coefficient Q is equal to Q 0 When the temperature and pressure combination at this time is defined as the calibration mode of the sensor, that is, the output of the sensor is independent of the temperature and pressure of the gas in the calibration mode, and the error caused by the temperature and pressure of the gas can be corrected according to the principle of equivalent compensation; By adjusting the gas pressure to adjust the sensor output until it is in calibration mode, the output after equivalent compensation is specifically expressed as: Step 6: Use the temperature characteristic difference of different absorption line intensities to measure the CO to be measured 2 The actual temperature of the gas; Step 7: Calculate the actual pressure of the gas based on the actual temperature of the gas and the equivalent coefficient before and after equivalent compensation; In step 7, the change in pressure can be calculated according to formulas (8) and (9) by using the response of the sensor at different pressures and amplitudes of the second harmonic signal before and after equivalent compensation; the ratio of the amplitudes of the second harmonic signal before and after equivalent compensation is: The pressure deviation is expressed as: δP=F(A q ) (9) Among them, F(A q ) is the sensor response function under different pressure values; Step 8: Finally, the gas molecule number error is corrected based on the deviation between the measured actual gas temperature and pressure and the gas temperature and pressure in the calibration mode.

2. The error prediction compensation method based on pattern recognition technology according to claim 1, It is characterized in that The error prediction compensation process can be expressed as: c o =DQMc (12) Among them, c is the original gas concentration data, c o is the gas concentration data output after compensation, and M is the molecular number compensation coefficient.

3. The error prediction compensation method based on pattern recognition technology according to claim 1, It is characterized in that In step 6, the deviation between the actual temperature of the gas and the gas temperature during calibration is calculated according to formula (7), and the actual temperature of the gas can be calculated using the temperature deviation and the gas temperature during calibration; Where k is the Boltzmann constant, h is the Planck constant, c is the speed of light, ΔE is the energy level difference at the ground state, A is the amplitude ratio of the second harmonic signal, and I 1 and I 2 All of them are obtained through the synchronous detection of light intensity, T c is the reference temperature, s 1 and 2 are the intensities of two different absorption lines at the reference temperature.

4. The error prediction compensation method based on pattern recognition technology according to claim 3, It is characterized in that In step 8, the last step of molecular number compensation is completed according to formulas (10) and (11); According to the universal gas law, the molecular number density is affected by the gas temperature and pressure; the molecular number density is corrected by the temperature and pressure deviations and expressed as: Where R is the universal gas constant, V is the volume of the multi-pass cell (5), T 0 is the gas temperature during calibration, P 0 is the gas pressure during calibration; the sensor output after molecular number compensation can be expressed as: 2f c =2f A DQM (11) To monitor sensor modes and correct measurement errors.

Images