Differential tunable gas detection method and device

Abstract

The invention discloses a differential tunable gas detection method and device, and belongs to the technical field of photoacoustic spectrometry gas detection.The method comprises the steps that standard gas is input into a resonant cavity, a tunable infrared laser is controlled to periodically output infrared laser to the resonant cavity, the lengths of two secondary resonant cavities in the resonant cavity are adjusted respectively, and the resonant cavity is obtained; the strongest resonance signal of the resonant cavity is found by observing the change of the photoacoustic signal measurement value in the adjustment process; after the strongest resonant signal is obtained, the lengths of the two secondary resonant cavities are fixed, the standard gas is discharged, then the detected gas is input into the resonant cavities, the detected gas excites periodic resonant sound waves after absorbing the infrared laser, and the resonant sound waves are detected by using an acoustic-electric conversion device and are converted into electric signals; differential amplification and filtering processing are carried out on the electric signals output by the acoustic-electric conversion device, and the processed electric signals are input into a computer; and calculating the concentration of the detected gas in the computer by using the pre-constructed multi-modal feature fusion and self-adaptive weighted regression model.

Description

Technical Field

[0001] The present invention relates to the technical field of photoacoustic spectroscopy gas detection, and in particular to a differential tunable gas detection method and device. Background Art

[0002] With the continuous advancement of science and technology, people's requirements for the sensitivity, accuracy and speed of gas detection are increasing. Photoacoustic spectroscopy gas detection technology, which uses the photoacoustic effect, provides high sensitivity, fast response and excellent gas selectivity, and is an important technology in the field of modern gas detection. The core of this technology is to use lasers to excite gas molecules and convert the absorbed light energy into heat energy through the photoacoustic effect, thereby generating sound waves. The intensity of these sound waves is proportional to the gas concentration. These sound wave signals are collected and analyzed by highly sensitive sound wave detectors to accurately determine the gas concentration.

[0003] The sensitivity of the photoacoustic spectroscopy system is mainly affected by three core components: the light source, the photoacoustic cell, and the acoustic wave detector. The photoacoustic cell, as the source of the photoacoustic signal, is the core part of the measurement system. According to the working mode, the photoacoustic cell is divided into two types: resonant and non-resonant. Among them, the resonant photoacoustic cell can enhance the photoacoustic signal and improve the sensitivity of gas detection, so it is more popular in practical applications. However, traditional resonant cavity designs, such as dumbbell-shaped resonant cavities, have obvious limitations. For example, in order to reduce the resonant frequency, it is necessary to increase the length of the photoacoustic cell, which not only increases the volume, but also prolongs the response time, and the effect of reducing the resonant frequency is limited. In addition, the fixed length of the resonant cavity limits the continuous adjustability of the frequency, affects the ability to reach the maximum resonant frequency, and thus limits the analytical sensitivity of the system. Traditional photoacoustic spectroscopy gas detection devices usually adopt a fixed structure. Once assembled, their performance is difficult to adjust and can only rely on external electronic systems for performance improvement, which limits the flexibility of system adjustment.

[0004] In order to optimize the gas detection performance, achieving the tunability of the resonant cavity is extremely critical to improving the accuracy of signal measurement. This requires the design of a more flexible and adjustable resonant cavity so that it can be adjusted according to different measurement requirements and improve the sensitivity and response speed of the system.

[0005] The photoacoustic spectroscopy gas detection system used in the prior art includes a photoacoustic resonant cavity, which resonates and enhances the photoacoustic signal generated by the gas to be tested entering the cavity absorbing the incident light, the first reflected light and the second reflected light, thereby obtaining an output photoacoustic signal. It uses a traditional buffer cavity, an absorption cavity and a resonance cavity to be mechanically connected in a fixed manner, which cannot achieve precise adjustment of the length of the resonance cavity, resulting in the inability to continuously adjust the resonance frequency of the resonance cavity. The suppression of background noise of non-detection gases in the mixed gas is limited. In order to improve the measurement accuracy, a complex optical path is required to increase the absorption optical path of the detection gas, and the operation process is cumbersome. Summary of the invention

[0006] In order to solve the problems existing in the prior art, the present invention provides the following technical solutions.

[0007] The first aspect of the present invention provides a differential tunable gas detection method, comprising the following steps:

[0008] The standard gas is input into the resonant cavity, and the tunable infrared laser is controlled to periodically output infrared laser to the resonant cavity. The lengths of the two secondary resonant cavities in the resonant cavity are adjusted respectively, and the strongest resonant signal of the resonant cavity is found by observing the changes in the measured values ​​of the photoacoustic signal during the adjustment process.

[0009] After obtaining the strongest resonance signal, the lengths of the two secondary resonance cavities are fixed, the standard gas is discharged, and the gas to be measured is input into the resonance cavity. The gas to be measured absorbs the infrared laser and excites periodic resonance sound waves. The resonance sound waves are detected by the acoustic-to-electric conversion device and converted into electrical signals.

[0010] Perform differential amplification and filtering on the electrical signal output by the acoustic-electric conversion device, and input the processed electrical signal into a computer;

[0011] The concentration of the measured gas is calculated in the computer using a pre-built multi-modal feature fusion and adaptive weighted regression model.

[0012] Optionally, the processed electrical signal contains characteristic data related to gas concentration, including frequency, amplitude and phase.

[0013] Optionally, the pre-built multimodal feature fusion and adaptive weighted regression model includes:

[0014] The signal preprocessing dynamic equation includes the state transfer equation and the observation equation. The state transfer equation is:

[0015] x k =Φx k-1 +w k

[0016] in, is the state vector at time k, is the state transfer matrix, w k is process noise, with mean 0 and covariance matrix Q k Gaussian distribution of

[0017] The state vector x k for:

[0018]

[0019] In the formula, f inst ,A env ,φ accum , They represent the instantaneous frequency, envelope amplitude, cumulative phase, and frequency change rate, respectively, which are obtained from the processed electrical signal;

[0020] The observation equation is:

[0021] z k =Hx k +v k

[0022] Where H is the observation matrix, v k is the observation noise, with mean 0 and covariance matrix R k Gaussian distribution of

[0023] Multidimensional feature tensor construction, defining the time-frequency-phase joint feature matrix:

[0024]

[0025] Among them, f i is the instantaneous frequency, A i is the Teager energy amplitude, is the phase time differential, the instantaneous frequency f i With the state vector x k The instantaneous frequency f inst Correspondingly, the first column element of X is filled with values ​​from the state variables containing the instantaneous frequency in sequence; the Teager energy amplitude A i The Teager energy operator is used to calculate the energy from the state vector x k The envelope amplitude A env Obtain; the phase time differential By changing the state vector x k The cumulative phase φ in accum and frequency change rate Calculate and obtain;

[0026] Elastic net regularized regression model, the objective function is defined as:

[0027]

[0028] In the formula, is the data fitting term, is the regularization term, β is the parameter vector in the model, λ>0 is the regularization strength parameter, ||β|| 1 represents the L1 norm, represents the square of the L2 norm, α∈[0,1] controls the L1 / L2 regularization mixing ratio, X is the input time-frequency-phase joint feature matrix, and C is the output gas concentration;

[0029] The dynamic environment compensation equation includes the temperature drift compensation function and the residual feedback correction term. The temperature drift compensation function is:

[0030] β i ′(T)=β i exp(-γΔT),γ=0.023℃ -1

[0031] In the formula, β i ′(T) is the parameter value after temperature drift compensation, β i is the original parameter value before compensation, γ is the temperature drift coefficient, and ΔT is the temperature change;

[0032] Residual feedback correction term:

[0033]

[0034] In the formula, is the estimated residual value after correction by the LSTM network at time t+1, ε t-3 ,ε t-2 ,ε t-1 ,ε t are the residuals at time t-3, t-2, t-1, and t respectively.

[0035] A second aspect of the present invention provides a differential tunable gas detection device, based on the differential tunable gas detection method described in the first aspect of the present invention, the device comprises:

[0036] Laser module, control module, adjustable resonant cavity module, gas transmission module and exhaust module;

[0037] The laser module is connected to the control module and is used to input a reference signal into the control module. The laser module is also used to output a periodically modulated infrared laser to the tunable resonant cavity module;

[0038] The gas transmission module and the exhaust module are respectively connected to the adjustable resonant cavity module. The gas transmission module is used to input the measured gas into the adjustable resonant cavity module. The adjustable resonant cavity module includes two adjustable resonant cavities 7 with the same structure. The lengths of the two adjustable resonant cavities 7 can be independently adjusted by corresponding resonant cavity length adjustment knobs 5 to obtain the strongest resonant signal. After the measured gas reacts with the infrared laser in the adjustable resonant cavity module, it is discharged from the adjustable resonant cavity module through the exhaust module.

[0039] The control module is also connected to the adjustable resonant cavity module, and is used to receive the electrical signal output by the adjustable resonant cavity module, perform differential amplification and filtering on the electrical signal, and perform calculation based on the processed electrical signal to obtain the concentration of the measured gas.

[0040] Optionally, the laser module includes a tunable infrared laser 1, a function generator 16 and a laser driver 17, which are connected in sequence, the function generator 16 outputs a modulation signal to the laser driver 17, and the laser driver 17 controls the tunable infrared laser 1 to periodically emit infrared laser.

[0041] Optionally, the adjustable resonant cavity module includes a first infrared window 2, a photoacoustic resonant cavity 3, a reference resonant cavity 22, an air inlet 4, a resonant cavity length adjustment knob 5, a built-in microphone 6, an adjustable resonant cavity 7, a second infrared window 9 and an air outlet 10; wherein the infrared laser output by the laser module is input into the photoacoustic resonant cavity 3 through the first infrared window 2, and is emitted from the second infrared window 9 to the photodetector 14 of the control module;

[0042] The gas to be measured enters the photoacoustic resonant cavity 3 through the gas inlet 4 , and is discharged from the resonant cavity through the gas outlet 10 after the reaction.

[0043] Optionally, the resonant cavity includes two primary resonant cavities of the same structure, namely, the photoacoustic resonant cavity 3 and the reference resonant cavity 22, wherein two identical secondary resonant cavities are installed in the middle position, and the secondary resonant cavity is an adjustable resonant cavity 7, each adjustable resonant cavity 7 corresponds to a resonant cavity length adjustment knob 5, and the lengths of the two adjustable resonant cavities 7 can be independently adjusted by the corresponding resonant cavity length adjustment knob 5, so that the resonant frequency can be continuously adjusted, thereby achieving the best resonant position to obtain the strongest resonant signal;

[0044] The same built-in microphone 6 is provided in each of the two secondary resonant cavities. The two built-in microphones 6 are respectively connected to the two input ends of the differential and filtering amplifier circuit 8 in the control module. The built-in microphone 6 is used to detect the resonant sound wave signal and convert the sound signal into an electrical signal and output it to the differential and filtering amplifier circuit 8.

[0045] Optionally, the control module includes a differential amplifier and filter circuit 8, a phase-locked amplifier 15, a computer 18 and a photodetector 14, which are connected in sequence; wherein the phase-locked amplifier 15 is connected to the function generator 16 of the laser module, and the reference signal of the phase-locked amplifier 15 is provided by the function generator 16; the two input ends of the differential amplifier and filter circuit 8 are respectively connected to the two built-in microphones 6 of the adjustable resonant cavity module, for receiving the electrical signal outputted therefrom, and amplifying and filtering the electrical signal; the electrical signal processed by the differential amplifier and filter circuit 8 is transmitted to the phase-locked amplifier 15, and the phase-locked amplifier 15 further amplifies the processed electrical signal in combination with the reference signal and inputs it into the signal collector; the signal collector collects the signal and inputs it into the computer 18, and the computer 18 further processes these signals to calculate the gas concentration.

[0046] Optionally, the gas delivery module includes a measured gas source 21, a gas valve 20 and a first mass flow meter 19, which are connected in sequence, wherein the first mass flow meter 19 is connected to the gas inlet 4 of the adjustable resonant cavity module.

[0047] Optionally, the exhaust module includes a second mass flow meter 11, a mechanical pump 12 and a gas recovery chamber 13, which are connected in sequence, wherein the second mass flow meter 11 is connected to the gas outlet 10 of the adjustable resonant cavity module.

[0048] 1. In the differential tunable gas detection method provided by the present invention, a multimodal feature fusion and an adaptive weighted regression model are proposed. Through the triple innovation of multidimensional feature fusion, adaptive regularization constraint and dynamic environmental compensation, the nonlinear distortion, feature collinearity and environmental sensitivity problems existing in traditional acoustic gas detection are effectively solved;

[0049] 2. In a differential tunable gas detection method provided by the present invention, the resonant cavity of the photoacoustic spectrometer can be continuously tuned. Compared with the traditional fixed resonant cavity, whose resonant frequency is fixed, this scheme can continuously search for the strongest resonant position, effectively improving the sensitivity of gas measurement;

[0050] 3. In a differential tunable gas detection method provided by the present invention, a differential structural design is adopted for the differential tunable gas detection device. This structural design can effectively reduce interference factors such as environmental noise and airflow noise, and significantly improve the resolution of gas detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A flow chart of a differential tunable gas detection method provided by an embodiment of the present invention;

[0052] Figure 2 A schematic structural diagram of a differential tunable gas detection device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only part of the embodiments of the present invention, not all of them. Based on the spirit of the present invention, other embodiments obtained by ordinary technicians in this field without creative work are all within the scope of protection of the present invention.

[0054] Embodiment 1, the present invention provides a differential tunable gas detection method, such as Figure 1 As shown, the following steps are included:

[0055] Step 1, inputting standard gas into the resonant cavity, controlling the tunable infrared laser to periodically output infrared laser to the resonant cavity, adjusting the lengths of the two secondary resonant cavities in the resonant cavity respectively, and finding the strongest resonant signal of the resonant cavity by observing the change of the measured value of the photoacoustic signal during the adjustment process;

[0056] It is worth noting that the present invention realizes precise adjustment of the length of the resonance cavity by continuously adjusting the resonance cavity length adjustment knob, thereby realizing continuous regulation of the resonance frequency.

[0057] Step 2, after obtaining the strongest resonance signal, the lengths of the two secondary resonance cavities are fixed, the standard gas is discharged, and the gas to be measured is input into the resonance cavity. After the gas to be measured absorbs the infrared laser, it excites periodic resonance sound waves, and the resonance sound waves are detected by the sound-to-electric conversion device and converted into electrical signals;

[0058] In a further preferred but non-limiting embodiment, in step 2, adjusting the length of the resonant cavity further comprises:

[0059] First, a standard gas containing a known concentration of the gas to be tested is given into the photoacoustic spectroscopy system, and then the position of the knob is continuously adjusted to obtain the resonance position of the strongest signal, and then the gas to be tested is measured.

[0060] It should be noted that, although the traditional T-shaped fixed structure has been simulated, it still has a big gap with the actual situation. The present invention directly calibrates the gas with known concentration to eliminate the gap.

[0061] Specifically, in response to the interference and noise problems existing in the prior art, different from the prior art in which background noise is reduced by complex optical path adjustment, the present invention introduces two second gas buffer chambers in the technical solution, one of which has an optical path passing through it, and the other two second gas buffer chambers have no optical path passing through them. As a differential comparison, it can achieve significant suppression of the background noise of non-detected gases in the mixed gas, has higher measurement sensitivity, and is easier to operate.

[0062] In a further preferred but non-limiting embodiment, in step 2, the sound-to-electric conversion device includes a built-in microphone, and two built-in microphones symmetrically arranged in two secondary resonant cavities are used to detect the resonant sound waves and convert them into electrical signals.

[0063] Step 3, differentially amplifying and filtering the electrical signal output by the acoustic-electric conversion device, and inputting the processed electrical signal into a computer;

[0064] In a further preferred but non-limiting embodiment, in step 3, the processed electrical signal contains characteristic data related to gas concentration, including frequency, amplitude, phase, etc.

[0065] In a further preferred but limiting embodiment, in step 3, processing the electrical signal includes:

[0066] The electrical signal is amplified and filtered using a differential amplifier and filter circuit, and then further amplified using a lock-in amplifier.

[0067] In a further preferred but limiting embodiment, the reference signal of the lock-in amplifier is provided by a function signal amplifier.

[0068] In a further preferred but limiting embodiment, in step 3, inputting the processed electrical signal into a computer further comprises:

[0069] The electrical signal processed by the lock-in amplifier is collected by a signal collector and then input into the computer system.

[0070] Step 4: Calculate the concentration of the measured gas in a computer using a pre-built multi-modal feature fusion and adaptive weighted regression model.

[0071] In a further preferred but non-limiting embodiment, in step 4, the multimodal feature fusion and adaptive weighted regression model is a nonlinear mapping relationship between gas concentration and acoustic feature vector, and the mathematical modeling of the model includes:

[0072] (1) Signal preprocessing dynamic equation

[0073] Construct the state space model of the joint filtering algorithm:

[0074] The state transfer equation is:

[0075] x k =Φx k-1 +w k

[0076] in, is the state vector at time k, is the state transfer matrix, w k is process noise, with mean 0 and covariance matrix Q k Gaussian distribution, that is, w k ~N(0,Q k );

[0077] Specifically, the state vector x k It can be expressed as:

[0078]

[0079] In the formula, f inst ,A env ,φ accum , The four-dimensional state quantities representing instantaneous frequency, envelope amplitude, cumulative phase, and frequency change rate are obtained from the processed electrical signal;

[0080] The observation equation is:

[0081] z k =Hx k +v k

[0082] Where H is the observation matrix, v k is the observation noise, with mean 0 and covariance matrix R k Gaussian distribution, that is, v k ~N(0,R k );

[0083] (2) Multidimensional feature tensor construction

[0084] Define the time-frequency-phase joint feature matrix:

[0085]

[0086] Among them, f i is the instantaneous frequency, A i is the Teager energy amplitude, is the phase time differential.

[0087] Specifically, the instantaneous frequency f i With the state vector x k The instantaneous frequency f instCorrespondingly, the first column element of X is filled in order from the state variables containing the instantaneous frequency; the Teager energy amplitude A i The Teager energy operator is used to calculate the energy from the state vector x k The envelope amplitude A env Obtain; the phase time differential By changing the state vector x k The cumulative phase φ in accum and frequency change rate Calculation is obtained. For example, the phase time differential at different times can be calculated by the central difference formula:

[0088]

[0089] Where Δt is the sampling time interval.

[0090] (3) Elastic Net Regularized Regression Model

[0091] The objective function is defined as:

[0092]

[0093] In the formula, is the data fitting term, is the regularization term, β is the parameter vector in the model, λ>0 is the regularization strength parameter (adaptively optimized by the BIC criterion), ||β|| 1 represents the L1 norm, represents the square of the L2 norm, α∈[0,1] controls the L1 / L2 regularization mixing ratio, X is the input time-frequency-phase joint feature matrix, and C is the output gas concentration.

[0094] (4) Dynamic environmental compensation equation

[0095] Temperature drift compensation function:

[0096] β i ′(T)=β i exp(-γΔT),γ=0.023℃ -1

[0097] In the formula, β i ′(T) is the parameter value after temperature drift compensation, β i is the original parameter value before compensation, γ is the temperature drift coefficient, and ΔT is the temperature change;

[0098] Residual feedback correction term:

[0099]

[0100] In the formula, is the residual estimation value corrected by the LSTM network at time t + 1, ε t-3 , ε t-2 , ε t-1 , ε t are the residuals at times t - 3, t - 2, t - 1, and t respectively, which are used as the inputs of the LSTM network for predicting and correcting the residuals at the next time.

[0101] It should be noted that the multi-modal feature fusion and adaptive weighted regression model proposed by the present invention effectively solves the problems of non-linear distortion, feature collinearity, and environmental sensitivity existing in traditional acoustic gas detection through triple innovations of multi-dimensional feature fusion, adaptive regularization constraint, and dynamic environment compensation.

[0102] In a further preferred but non-limiting embodiment, the gas to be measured includes SF 6 gas decomposition characteristic products, such as SO 2 F 2 , SO 2 F 4 , SOF 2 , SO 2 and HF, etc.

[0103] In a further preferred but non-limiting embodiment, in the method, the structure of the resonant cavity is specifically:

[0104] The resonant cavity includes two first-level resonant cavities with the same structure, which are used as the photoacoustic resonant cavity and the reference resonant cavity respectively. Two identical second-level resonant cavities are installed at the middle position, and the lengths of these second-level resonant cavities can be independently adjusted through their respective adjustment knobs, so that the resonant frequency of the resonant cavity is continuously adjustable, and the best resonant position is achieved to obtain the strongest resonant signal.

[0105] Embodiment 2, the present invention provides a differential tunable gas detection device for implementing the differential tunable gas detection method described in Embodiment 1, including:

[0106] a laser module, a control module, a tunable resonant cavity module, a gas input module, and an exhaust module;

[0107] The laser module is connected to the control module for inputting a reference signal into the control module, and the laser module is also used for outputting periodically modulated infrared laser to the tunable resonant cavity module;

[0108] The gas input module and the exhaust module are respectively connected to the tunable resonant cavity module. The gas input module is used for inputting the gas to be measured into the tunable resonant cavity module. After the gas to be measured reacts with the infrared laser in the tunable resonant cavity module, it is discharged from the tunable resonant cavity module through the exhaust module;

[0109] The control module is also connected to the tunable resonator module, and is configured to receive and process the electrical signals output by the tunable resonator module to obtain the concentration of the gas to be measured.

[0110] In a further preferred but non-limiting embodiment, the laser module includes a tunable infrared laser 1, a function generator 16, and a laser driver 17, which are connected in sequence. The function generator 16 outputs a modulation signal to the laser driver 17, and the laser driver 17 controls the tunable infrared laser 1 to emit infrared laser periodically.

[0111] In a further preferred but non-limiting embodiment, the tunable resonator module includes a first infrared window 2, a photoacoustic resonator 3, a reference resonator 22, an air inlet 4, a resonator cavity length adjustment knob 5, a built-in microphone 6, an adjustable resonator 7, a second infrared window 9, and an air outlet 10. The infrared laser output by the laser module enters the photoacoustic resonator 3 through the first infrared window 2 and exits through the second infrared window 9 to the photodetector 14 of the control module.

[0112] The air inlet 4 is connected to the first mass flowmeter 19 of the gas delivery module. The gas to be measured enters the photoacoustic resonator 3 through the air inlet 4. The air outlet 10 is connected to the second mass flowmeter 11 of the exhaust module. After the reaction of the gas to be measured, it is discharged from the air outlet 10 out of the resonator.

[0113] The resonator includes two first-stage resonators with the same structure, namely the photoacoustic resonator 3 and the reference resonator 22. Two identical second-stage resonators are installed in the middle position. The second-stage resonator is the adjustable resonator 7. Each adjustable resonator 7 corresponds to a resonator cavity length adjustment knob 5. The lengths of the two adjustable resonators 7 can be independently adjusted through the corresponding resonator cavity length adjustment knobs 5, so that the resonance frequency is continuously adjustable, and then the best resonance position is achieved to obtain the strongest resonance signal.

[0114] Specifically, the present invention introduces two completely identical second buffer cavities, one of which has an optical path passing through and the other has no optical path passing through. As a comparison of background noise, the differential signal is measured to reduce the background noise. Two completely identical first-stage resonators are installed with a second-stage resonator with adjustable cavity length that is completely consistent in the middle. The telescopic adjustment of the resonator cavity length is achieved through the second-stage resonator cavity length adjustment knob.

[0115] Identical built-in microphones 6 are provided in both second-stage resonators. The two built-in microphones 6 are respectively connected to the two input ends of the differential and filter amplification circuit 8 in the control module. The built-in microphone 6 is used to detect the resonance acoustic wave signal and convert the acoustic signal into an electrical signal and output it to the differential and filter amplification circuit 8.

[0116] In a further preferred but non-limiting embodiment, the control module includes a differential amplifier and filter circuit 8, a phase-locked amplifier 15, a computer 18 and a photodetector 14, which are connected in sequence; wherein the photodetector 14 is used to monitor the remaining intensity of the laser after being absorbed, so as to monitor the stability of the optical path system; the phase-locked amplifier 15 is connected to the function generator 16 of the laser module, and the reference signal of the phase-locked amplifier 15 is provided by the function generator 16; the two input ends of the differential amplifier and filter circuit 8 are respectively connected to the two built-in microphones 6 of the adjustable resonant cavity module, so as to receive the electrical signal outputted therefrom, and amplify and filter the electrical signal; the electrical signal processed by the differential amplifier and filter circuit 8 is transmitted to the phase-locked amplifier 15, and the phase-locked amplifier 15 further amplifies the processed electrical signal in combination with the reference signal and inputs it into the signal collector; the signal collector collects the signal and inputs it into the computer 18, and the computer 18 further processes these signals to calculate the gas concentration.

[0117] In a further preferred but non-limiting embodiment, the gas delivery module includes a measured gas source 21, a gas valve 20 and a first mass flow meter 19, which are connected in sequence, wherein the first mass flow meter 19 is connected to the gas inlet 4 of the adjustable resonant cavity module.

[0118] In a further preferred but non-limiting embodiment, the exhaust module includes a second mass flow meter 11, a mechanical pump 12 and a gas recovery chamber 13, which are connected in sequence, wherein the second mass flow meter 11 is connected to the gas outlet 10 of the adjustable resonant cavity module.

[0119] The main features of the present invention are as follows: 1: tunable infrared laser; 2: first infrared window; 3: photoacoustic resonant cavity; 4: air inlet; 5: resonant cavity length adjustment knob; 6: built-in microphone; 7: adjustable resonant cavity; 8: differential amplifier and filter circuit; 9: second infrared window; 10: air outlet; 11: second mass flowmeter; 12: mechanical pump; 13: gas recovery chamber; 14: photoelectric detector; 15: phase-locked amplifier; 16: function generator; 17: laser driver; 18: computer; 19: first mass flowmeter; 20: gas valve; 21: measured gas source; 22: reference resonant cavity.

[0120] The innovative configuration of this differentially tunable photoacoustic spectrometer not only overcomes the limitations of the traditional fixed resonant cavity in continuously adjusting the resonance signal, but also significantly reduces the influence of external interference factors such as environmental noise and airflow noise, significantly improving the accuracy and resolution of gas detection.

Claims

1. A differential tunable gas detection method, characterized in that: The steps include: The standard gas is input into the resonant cavity, and the tunable infrared laser is controlled to periodically output infrared laser to the resonant cavity. The lengths of the two secondary resonant cavities in the resonant cavity are adjusted respectively, and the strongest resonant signal of the resonant cavity is found by observing the changes in the measured values ​​of the photoacoustic signal during the adjustment process. After obtaining the strongest resonance signal, the lengths of the two secondary resonance cavities are fixed, the standard gas is discharged, and the gas to be measured is input into the resonance cavity. The gas to be measured absorbs the infrared laser and excites periodic resonance sound waves. The resonance sound waves are detected by the acoustic-to-electric conversion device and converted into electrical signals. Perform differential amplification and filtering on the electrical signal output by the acoustic-electric conversion device, and input the processed electrical signal into a computer; The concentration of the measured gas is calculated in the computer using a pre-built multi-modal feature fusion and adaptive weighted regression model.

2. A differential tunable gas detection method according to claim 1, characterized in that: The processed electrical signal contains characteristic data related to gas concentration, including frequency, amplitude and phase.

3. A differential tunable gas detection method according to claim 2, characterized in that: The pre-built multi-modal feature fusion and adaptive weighted regression model includes a signal pre-processing dynamic equation; The signal preprocessing dynamic equation includes a state transfer equation and an observation equation. The state transfer equation is: x k =Φx k-1 +w k in, is the state vector at time k, is the state transfer matrix, w k is process noise, with mean 0 and covariance matrix Q k Gaussian distribution of The state vector x k for: In the formula, f inst ,A env ,φ accum , They represent the instantaneous frequency, envelope amplitude, cumulative phase, and frequency change rate, respectively, which are obtained from the processed electrical signal; The observation equation is: z k =Hx k +v k Where H is the observation matrix, v k is the observation noise, with mean 0 and covariance matrix R k Gaussian distribution.

4. A differential tunable gas detection method according to claim 3, characterized in that: The multimodal feature fusion and adaptive weighted regression model includes a multidimensional feature tensor construction, defining a time-frequency-phase joint feature matrix: Among them, f i is the instantaneous frequency, A i is the Teager energy amplitude, is the time differential of the phase, the instantaneous frequency f i With the state vector x k The instantaneous frequency f inst Correspondingly, the first column element of X is filled with values ​​from the state variables containing the instantaneous frequency in sequence; the Teager energy amplitude A i The Teager energy operator is used to calculate the energy from the state vector x k The envelope amplitude A env Get; the phase time differential By changing the state vector x k The cumulative phase φ in accum and frequency change rate Calculate and obtain; Elastic net regularized regression model, the objective function is defined as: In the formula, is the data fitting term, is the regularization term, β is the parameter vector in the model, λ>0 is the regularization strength parameter, ||β||1 represents the L1 norm, represents the square of the L2 norm, α∈[0,1] controls the L1 / L2 regularization mixing ratio, X is the input time-frequency-phase joint feature matrix, and C is the output gas concentration.

5. A differential tunable gas detection method according to claim 4, characterized in that: The multimodal feature fusion and adaptive weighted regression model includes a dynamic environment compensation equation; The dynamic environment compensation equation includes a temperature drift compensation function and a residual feedback correction term. The temperature drift compensation function is: b i ′(T)=β i ·exp(-γΔT),γ=0.023℃ -1 In the formula, β i ′(T) is the parameter value after temperature drift compensation, β i is the original parameter value before compensation, γ is the temperature drift coefficient, and ΔT is the temperature change; The residual feedback correction term is: In the formula, is the estimated residual value after correction by the LSTM network at time t+1, ε t-3 ,ε t-2 ,ε t-1 ,ε t are the residuals at time t-3, t-2, t-1, and t respectively.

6. A differential tunable gas detection device, used to perform a differential tunable gas detection method according to any one of claims 1 to 5, characterized in that: The device includes: Laser module, control module, adjustable resonant cavity module, gas transmission module and exhaust module; The laser module is connected to the control module and is used to input a reference signal into the control module. The laser module is also used to output a periodically modulated infrared laser to the tunable resonant cavity module; The gas transmission module and the exhaust module are respectively connected to the adjustable resonant cavity module. The gas transmission module is used to input the measured gas into the adjustable resonant cavity module. The adjustable resonant cavity module comprises two adjustable resonant cavities (7) with the same structure. The lengths of the two adjustable resonant cavities (7) can be independently adjusted by corresponding resonant cavity length adjustment knobs (5) to obtain the strongest resonant signal. After the measured gas reacts with the infrared laser in the adjustable resonant cavity module, it is discharged from the adjustable resonant cavity module through the exhaust module. The control module is also connected to the adjustable resonant cavity module, and is used to receive the electrical signal output by the adjustable resonant cavity module, perform differential amplification and filtering on the electrical signal, and perform calculation based on the processed electrical signal to obtain the concentration of the measured gas.

7. A differential tunable gas detection device according to claim 6, characterized in that: The laser module comprises a tunable infrared laser (1), a function generator (16) and a laser driver (17), which are connected in sequence. The function generator (16) outputs a modulation signal to the laser driver (17), and the laser driver (17) controls the tunable infrared laser (1) to periodically emit infrared laser light.

8. A differential tunable gas detection device according to claim 7, characterized in that: The adjustable resonant cavity module comprises a first infrared window (2), a photoacoustic resonant cavity (3), a reference resonant cavity (22), an air inlet (4), a resonant cavity length adjustment knob (5), a built-in microphone (6), an adjustable resonant cavity (7), a second infrared window (9) and an air outlet (10); wherein the infrared laser output by the laser module is input into the photoacoustic resonant cavity (3) through the first infrared window (2), and is emitted from the second infrared window (9) to the photoelectric detector (14) of the control module; The gas to be measured enters the photoacoustic resonant cavity (3) through the gas inlet (4), and after the gas to be measured reacts, it is discharged from the resonant cavity through the gas outlet (10); The resonant cavity comprises two primary resonant cavities of identical structure, namely a photoacoustic resonant cavity (3) and a reference resonant cavity (22), wherein two identical secondary resonant cavities are arranged in the middle, wherein the secondary resonant cavity is an adjustable resonant cavity (7), each adjustable resonant cavity (7) corresponds to a resonant cavity length adjustment knob (5), and the lengths of the two adjustable resonant cavities (7) can be independently adjusted by the corresponding resonant cavity length adjustment knobs (5), so that the resonant frequency can be continuously adjusted, thereby achieving an optimal resonant position to obtain the strongest resonant signal; The same built-in microphone (6) is arranged in each of the two secondary resonant cavities. The two built-in microphones (6) are respectively connected to two input ends of a differential and filter amplifier circuit (8) in the control module. The built-in microphone (6) is used to detect the resonant sound wave signal and convert the sound signal into an electrical signal and output it to the differential and filter amplifier circuit (8).

9. A differential tunable gas detection device according to claim 8, characterized in that: The control module comprises a differential amplifier and filter circuit (8), a phase-locked amplifier (15), a computer (18) and a photodetector (14), which are connected in sequence; wherein the phase-locked amplifier (15) is connected to a function generator (16) of a laser module, and a reference signal of the phase-locked amplifier (15) is provided by the function generator (16); two input ends of the differential amplifier and filter circuit (8) are respectively connected to two built-in microphones (6) of the tunable resonant cavity module, for receiving the electrical signals outputted therefrom, and amplifying and filtering the electrical signals; The electrical signal processed by the differential amplification and filtering circuit (8) is transmitted to the phase-locked amplifier (15), which further amplifies the processed electrical signal in combination with the reference signal and inputs it into the signal collector; the signal collector collects the signal and inputs it into the computer (18), and the computer (18) further processes the signal to calculate the gas concentration.

10. A differential tunable gas detection device according to claim 9, characterized in that: The gas transmission module comprises a measured gas source (21), a gas valve (20) and a first mass flow meter (19), which are connected in sequence, wherein the first mass flow meter (19) is connected to the gas inlet (4) of the adjustable resonant cavity module; The exhaust module comprises a second mass flow meter (11), a mechanical pump (12) and a gas recovery chamber (13), which are connected in sequence, wherein the second mass flow meter (11) is connected to the gas outlet (10) of the adjustable resonant cavity module.

Images