A gas detection method and system based on photoacoustic cell acoustic pulse excitation

Abstract

This invention discloses a gas detection method and system based on photoacoustic cell acoustic pulse excitation, relating to the field of transformer oil gas detection technology. The method includes the following steps: constructing a gas detection system and a data acquisition system; initializing the photoacoustic cell, light source, and acoustic sensor; performing preliminary calibration of the gas detection system and recording the initial calibration parameters; and using a temperature sensor and a gas analysis sensor to detect changes in the temperature and composition of gases in the transformer oil. This invention utilizes the photoacoustic effect to convert the light energy absorbed by gas molecules into a detectable acoustic signal. Combined with signal preprocessing and feature extraction techniques, it extracts the intensity, frequency, and phase characteristics of the photoacoustic signal, significantly improving the sensitivity of gas detection. Simultaneously, by dynamically adjusting the resonant frequency and modulation frequency of the photoacoustic cell, the signal-to-noise ratio of the photoacoustic signal is optimized, further ensuring the accuracy of gas detection.

Description

Technical Field

[0001] This invention relates to the field of transformer oil gas detection technology, specifically to a gas detection method and system based on photoacoustic cell acoustic pulse excitation. Background Technology

[0002] Transformer oil is an important insulating and cooling medium in transformer operation, and its quality directly affects the operational safety of the transformer. The composition and content of dissolved gases in transformer oil are important bases for judging internal faults in transformers. Traditional transformer oil detection methods have problems such as low detection accuracy, large maintenance requirements, and inability to reflect the true health status of transformers in a timely manner. With the continuous advancement of laser technology, especially the emergence of semiconductor lasers, photoacoustic gas detection technology has been greatly developed.

[0003] In existing technologies, during transformer oil testing, the resonant frequency of the photoacoustic cell changes with variations in gas temperature and composition, leading to significant variations in the photoacoustic signal and affecting detection accuracy. Furthermore, under high-frequency modulation, when the modulation frequency approaches the resonant frequency of the photoacoustic cell, the absorption of light by the gas within the cell weakens, resulting in optical power attenuation and consequently affecting the intensity of the photoacoustic signal, leading to detection distortion. Therefore, determining and analyzing the resonant frequency and modulation frequency of the photoacoustic cell to assess the deviation range during gas detection and dynamically adjust the detection parameters is a problem we need to solve. To this end, we propose a gas detection method and system based on acoustic pulse excitation using a photoacoustic cell. Summary of the Invention

[0004] The purpose of this invention is to provide a gas detection method and system based on photoacoustic cell acoustic pulse excitation to solve the problems mentioned in the background art.

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

[0006] In a first aspect, a gas detection method based on photoacoustic cell acoustic pulse excitation includes the following steps:

[0007] Step 1: Construct the gas detection system and data acquisition system, initialize the photoacoustic cell, light source, and micro-sound sensor, perform preliminary calibration of the gas detection system, ensure that the resonant frequency and modulation frequency of the photoacoustic cell are known and accurate under standard conditions, and record the initial calibration parameters.

[0008] Step 2: Use temperature sensors and gas analysis sensors to detect changes in the temperature and composition of gas in transformer oil, and construct a resonant frequency prediction model based on the detected gas temperature and composition data to analyze the trend of resonant frequency changes.

[0009] Step 3: Based on the resonant frequency prediction model and the analysis results of the resonant frequency, dynamically adjust the modulation frequency of the light source to make it as close as possible to but not exceed the current resonant frequency of the photoacoustic cell, observe the intensity change of the photoacoustic signal, analyze the amplitude of the photoacoustic signal under different modulation frequencies, and establish the relationship between the modulation frequency and the resonant frequency.

[0010] Step 4: Use a modulation system to generate an acoustic pulse, which propagates through a photoacoustic cell. The acoustic pulse causes a change in gas pressure within the photoacoustic cell, exciting a photoacoustic signal. Use a high-sensitivity micro-sound sensor to collect the photoacoustic signal, preprocess the collected photoacoustic signal, analyze the intensity, frequency, and phase changes of the photoacoustic signal, and predict the gas concentration.

[0011] Step 5: Real-time detection of gas temperature, pressure and composition; dynamic adjustment of excitation parameters (resonance frequency and modulation frequency) of photoacoustic cell to adapt to changes in gas temperature and composition; and establishment of deviation prediction model based on preprocessed photoacoustic signal and preset detection standard to evaluate the deviation range of gas detection and correct the predicted gas detection deviation range.

[0012] A further improvement to the technical solution of the present invention is that the preliminary calibration process of the gas detection system in step 1 is as follows:

[0013] Step 101: Construct a gas detection system and a data acquisition system, including a photoacoustic cell, a light source, and a micro-sound sensor. The photoacoustic cell is constructed and its inner and outer surfaces are cleaned with ethanol solvent.

[0014] Step 102: Select a laser as the light source according to the absorption characteristics of the gas to be tested, match the wavelength of the light emitted by the light source with the absorption peak of the gas, check the output power of the light source, and adjust the driving current to make the output light intensity reach the predetermined value and remain stable.

[0015] Step 103: Install the microphone sensor outside the photoacoustic cell and add a filter circuit to the signal output terminal of the microphone sensor to further reduce noise interference;

[0016] Step 104: Set the modulation method of the light source and calibrate the resonant frequency and modulation frequency. The laser pulse method is used to excite the sound wave, and the resonant frequency is measured using a micro-sound sensor. A calibration experiment is performed using a standard gas of known concentration. The modulation frequency of the light source is adjusted to match the resonant frequency of the photoacoustic cell. The current modulation frequency of the light source and the photoacoustic signal intensity are recorded as the reference for subsequent measurements.

[0017] Step 105: By changing the concentration of the standard gas, measure the intensity change of the photoacoustic signal at different concentrations, plot the concentration-signal intensity curve, calculate the sensitivity of the microphone sensor, and record all relevant calibration parameters, including the modulation frequency of the light source, the resonant frequency of the photoacoustic cell, and the sensitivity of the microphone sensor.

[0018] A further improvement to the technical solution of this invention lies in the following: In step 2, the process of constructing the resonant frequency prediction model is as follows:

[0019] Step 201: Deploy temperature sensors and gas analysis sensors to detect changes in the temperature of transformer oil and the composition of dissolved gases in transformer oil, and collect data on the temperature and gas composition in transformer oil. Perform preprocessing operations on the collected data, including data cleaning and standardization steps.

[0020] Step 202: Perform feature analysis on the preprocessed data to extract features related to the resonance frequencies, namely temperature features, gas concentration features, and pressure features. Combine the extracted resonance frequency related features to obtain a feature dataset, and divide the feature dataset into a training set and a test set.

[0021] Step 203: Construct a resonant frequency prediction model using training set data and support vector machine algorithm, and use test set data to verify the accuracy and generalization ability of the model in order to adjust the model parameters and predict the resonant frequency under different temperatures and gas compositions.

[0022] Step 204: Combine the associated data of the feature dataset and the output of the resonance frequency prediction model to obtain the resonance frequency evaluation index and analyze the trend of resonance frequency with temperature and gas composition.

[0023] Step 205: Set an adjustment threshold. When the output value of the resonance frequency evaluation index exceeds the adjustment threshold, adjust the modulation frequency of the light source to correct the impact of resonance frequency changes on gas detection accuracy.

[0024] A further improvement to the technical solution of the present invention is that the calculation expression for the resonance frequency evaluation index is:

[0025]

[0026] Δf res =f res -f res，base ;

[0027] Among them, F res The resonant frequency evaluation index is given by T, where T is temperature and C is temperature. i f is the concentration of the i-th gas. res f is the resonant frequency. res，base The reference resonant frequency is the resonant frequency measured under standard conditions, Δf. res The change in resonant frequency is denoted by n, where n is the number of gas types, and w is the number of gas types. i The weight of the i-th gas reflects the importance of that gas to the resonant frequency. baseThe reference temperature is the temperature corresponding to the resonant frequency measured under standard conditions, and k is the temperature coefficient.

[0028] A further improvement to the technical solution of this invention lies in the following: In step 3, the process of establishing the relationship between the modulation frequency and the resonant frequency is as follows:

[0029] Step 301: Using the established resonant frequency prediction model, predict the resonant frequency of the photoacoustic cell based on the current state of the transformer oil (temperature and gas composition), and set the initial modulation frequency of the light source.

[0030] Step 302: Set the modulation frequency of the light source to the starting value of the search range, gradually increase the modulation frequency of the light source in small steps (a few Hz or tens of Hz), and observe the intensity change of the photoacoustic signal. At each modulation frequency point, after stabilizing for a period of time, record the amplitude of the photoacoustic signal.

[0031] Step 303: Use the data acquisition system to acquire photoacoustic signals and analyze the signal amplitude. If the signal amplitude begins to increase significantly, it indicates that the modulation frequency is approaching the resonant frequency. At this time, gradually change the modulation frequency, reduce the step size, and adjust the modulation frequency more finely. Record the intensity of the photoacoustic signal after each adjustment, find the modulation frequency that produces the strongest photoacoustic signal, and take it as the optimal modulation frequency. The optimal modulation frequency is close to the resonant frequency of the photoacoustic cell.

[0032] Step 304: Record the photoacoustic signal amplitude corresponding to each modulation frequency point, draw a graph showing the relationship between modulation frequency and photoacoustic signal amplitude, observe and confirm the position of the resonance peak, analyze the relationship between modulation frequency and photoacoustic signal amplitude, and confirm the actual value of the resonance frequency.

[0033] A further improvement to the technical solution of the present invention is that: in step 4, the gas concentration prediction process is as follows:

[0034] Step 401: Configure the modulation system, use the modulation system to generate an electrical signal, convert the electrical signal into an acoustic pulse through a piezoelectric transducer, release the acoustic pulse into the photoacoustic cell, propagate in the photoacoustic cell, interact with the gas in the photoacoustic cell, causing a change in gas pressure, and at the same time as the acoustic pulse propagates, use a laser to irradiate the gas in the photoacoustic cell. When the gas pressure change caused by the acoustic pulse occurs at the same time as the light source irradiation, the gas molecules absorb light energy and convert it into heat energy, thereby generating local temperature and pressure changes. The local temperature and pressure changes propagate in the form of sound waves, forming a photoacoustic signal.

[0035] Step 402: Use a micro-sound sensor to collect photoacoustic signals generated by pressure changes and convert them into electrical signals for subsequent processing and analysis. Perform preprocessing on the electrical signals converted from the photoacoustic signals collected by the micro-sound sensor.

[0036] Step 403: Analyze the preprocessed photoacoustic signal data to analyze the intensity, frequency, and phase changes of the photoacoustic signal;

[0037] Step 404: Based on the calibration curve established by measuring the photoacoustic signal intensity at a known concentration, and combining the intensity, frequency, and phase characteristics of the photoacoustic signal, calculate the gas concentration using the calibration curve, and record the photoacoustic signal characteristics and the predicted gas concentration for each measurement.

[0038] A further improvement to the technical solution of the present invention is that the calculation expression for the gas concentration is:

[0039]

[0040] Where C is the predicted gas concentration, representing the target calculated value, I is the intensity of the photoacoustic signal, f is the intensity of the photoacoustic signal, φ is the phase of the photoacoustic signal, C0 is the standard gas concentration at a known concentration, I0 is the photoacoustic signal intensity corresponding to C0, a, b, and β are coefficients obtained from fitting the calibration curve, m is the number of measurements, and I... j Let fj be the photoacoustic signal intensity measured in the j-th measurement, f0 be the photoacoustic signal frequency under standard conditions, and φ be the frequency of the photoacoustic signal. j Let be the phase of the photoacoustic signal measured in the j-th measurement.

[0041] A further improvement to the technical solution of this invention lies in the following: In step 5, the process of establishing the deviation prediction model is as follows:

[0042] Step 501: Use a temperature sensor, a pressure sensor and a gas analyzer to detect the temperature, pressure and composition of the gas, use a micro-sound sensor to continuously collect photoacoustic signals generated by the photoacoustic effect, and convert the photoacoustic signals into electrical signals for preprocessing;

[0043] Step 502: Based on the detected gas temperature, pressure and composition information, analyze the influence of gas parameters on photoacoustic signal characteristics. Based on the physical properties of the gas, dynamically adjust the resonant frequency of the photoacoustic cell by changing the frequency of the excitation signal to match the current acoustic characteristics of the gas. According to the dynamic response of the gas and the detection requirements, adjust the frequency of the modulation signal to optimize the intensity and signal-to-noise ratio of the photoacoustic signal.

[0044] Step 503: Filter, amplify and digitize the acquired photoacoustic signal to remove noise and interference components, and extract intensity, frequency and phase features from the preprocessed photoacoustic signal;

[0045] Step 504: Train the deviation prediction model using historical data and known deviation conditions. Take the detected gas parameters and photoacoustic signal characteristics as the input of the model, output the predicted gas detection deviation range, and calculate the deviation evaluation coefficient based on the results of the deviation prediction model and the preset detection standard to analyze the gas detection deviation trend and quantify the degree of gas detection deviation.

[0046] Step 505: Set different deviation levels according to the deviation evaluation coefficient, namely slight deviation level, moderate deviation level and severe deviation level, and match the corresponding deviation threshold for each deviation level.

[0047] Step 506: Based on the predicted deviation range and deviation level, take corresponding correction measures, and feed the correction results and new detection data back to the deviation prediction model to further optimize the monitoring accuracy of transformer oil gas. Regularly generate detailed monitoring reports to summarize gas detection results, deviation assessment, correction measures and their implementation effects.

[0048] A further improvement to the technical solution of this invention is that the calculation expression for the deviation evaluation coefficient is:

[0049]

[0050] Where PP is the bias evaluation coefficient, S is the set of detected gas parameters, such as temperature, pressure, and composition, F is the set of photoacoustic signal features, such as intensity, frequency, and phase, N is the sample size, μ is the sample mean, σ is the sample standard deviation, and C... r Let C be the predicted concentration of the r-th sample. ac，r S represents the actual concentration of the r-th sample. base Here, λ is the baseline value for the gas parameters, and λ is the adjustment parameter used to adjust the impact of gas parameters deviating from the baseline value on the deviation evaluation coefficient. r PP is the sum of squares of all features in the feature set of the photoacoustic signal of the r-th sample. The value of PP ranges from 0 to 1, where 0 represents no bias and 1 represents complete deviation.

[0051] The multiple deviation levels correspond to multiple deviation thresholds, wherein the deviation thresholds include an upper limit threshold and a lower limit threshold;

[0052] The plurality of deviation levels and the plurality of deviation thresholds satisfy the following relationship:

[0053] Slight deviation level 0 <PP≤PP M ;

[0054] Medium deviation grade PP M <PP≤PP S ;

[0055] Severe deviation level PPS <PP<1;

[0056] Where PP is the deviation evaluation coefficient, PP M PP represents the lower threshold for moderate deviation and the upper threshold for slight deviation. S PP represents the lower threshold for severe deviation and the upper threshold for moderate deviation. M =0.1, PP S =0.5.

[0057] Secondly, a gas detection system based on photoacoustic cell acoustic pulse excitation is provided to implement a gas detection method based on photoacoustic cell acoustic pulse excitation. The system includes a detection center, which is communicatively connected to a gas parameter detection module, a photoacoustic signal generation module, a photoacoustic signal acquisition module, a signal preprocessing module, a feature extraction module, an excitation parameter adjustment module, a deviation prediction and evaluation module, and a correction and optimization module. The modules are electrically connected to each other.

[0058] The gas parameter detection module uses temperature sensors, pressure sensors, and a gas analyzer to detect the temperature, pressure, and composition information of the gas, providing the system with comprehensive gas state data as the basis for subsequent processing and adjustment.

[0059] The photoacoustic signal generating module is used to generate an acoustic pulse excitation signal, which excites gas molecules in the photoacoustic cell to generate a photoacoustic signal through the photoacoustic effect. The photoacoustic effect is used to convert the light energy absorbed by the gas into a detectable acoustic signal, providing a direct basis for gas detection.

[0060] The photoacoustic signal acquisition module uses a micro-sound sensor to continuously acquire photoacoustic signals generated by the photoacoustic effect and converts them into electrical signals, thus converting weak acoustic signals into electrical signals for easier subsequent signal processing and analysis.

[0061] The signal preprocessing module is used to filter, amplify, and digitize the acquired photoacoustic signals to remove noise and interference components, improve the signal-to-noise ratio, and ensure the accuracy of subsequent feature extraction and data analysis.

[0062] The feature extraction module is used to extract signal intensity, frequency and phase features from the preprocessed photoacoustic signal, providing useful input data for the deviation prediction model and helping the model to better understand the relationship between the gas state and the photoacoustic signal.

[0063] The excitation parameter adjustment module dynamically adjusts the resonant frequency and modulation frequency of the photoacoustic cell based on the detected gas parameters and photoacoustic signal characteristics, thereby optimizing the intensity and signal-to-noise ratio of the photoacoustic signal and improving the sensitivity and accuracy of gas detection.

[0064] The deviation prediction and evaluation module uses historical data and known deviation situations to train a deviation prediction model, predict the deviation range of gas detection, calculate the deviation evaluation coefficient, analyze the deviation trend of gas detection, discover potential detection deviations in advance, quantify the degree and trend of deviations, and provide a scientific basis for correction measures.

[0065] The correction and optimization module takes corresponding correction measures based on the predicted deviation range and deviation level, and feeds back the correction results and new detection data to the system to correct the detection deviation.

[0066] Due to the adoption of the above technical solution, the technical progress achieved by this invention compared to the prior art is as follows:

[0067] 1. This invention provides a gas detection method and system based on photoacoustic cell acoustic pulse excitation. By utilizing the photoacoustic effect, the light energy absorbed by gas molecules is converted into a detectable acoustic signal. Combined with signal preprocessing and feature extraction techniques, the intensity, frequency, and phase features of the photoacoustic signal are extracted, which greatly improves the sensitivity of gas detection. At the same time, by dynamically adjusting the resonant frequency and modulation frequency of the photoacoustic cell, the signal-to-noise ratio of the photoacoustic signal is optimized, further ensuring the accuracy of gas detection.

[0068] 2. This invention provides a gas detection method and system based on photoacoustic cell acoustic pulse excitation. According to the detected gas parameters and photoacoustic signal characteristics, the excitation parameters of the photoacoustic cell are dynamically adjusted to adapt to the needs of different gases and different detection conditions, ensuring that the system can maintain stable detection performance under different environments and operating conditions. At the same time, by using a deviation prediction and evaluation module, potential detection deviations can be detected in advance, and corresponding correction measures can be taken, further enhancing the stability and reliability of the system and reducing false alarms and missed alarms caused by detection errors. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0070] Figure 1 This is a flowchart of the method of the present invention;

[0071] Figure 2 This is a flowchart illustrating the construction of the resonance frequency prediction model of the present invention;

[0072] Figure 3 This is a flowchart illustrating the gas concentration prediction process of this invention.

[0073] Figure 4This is a flowchart illustrating the establishment of the deviation prediction model of the present invention;

[0074] Figure 5 This is a diagram showing the module configuration of the present invention. Detailed Implementation

[0075] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] Example 1, such as Figures 1-3 As shown, this invention provides a gas detection method based on photoacoustic cell acoustic pulse excitation, comprising the following steps:

[0077] Step 1: Construct the gas detection system and data acquisition system. Initialize the photoacoustic cell, light source, and microphone sensor. Perform preliminary calibration of the gas detection system to ensure that the resonant frequency and modulation frequency of the photoacoustic cell are known and accurate under standard conditions, and record the initial calibration parameters. The gas detection system and data acquisition system include the photoacoustic cell, light source, and microphone sensor. The photoacoustic cell is constructed and its inner and outer surfaces are cleaned with ethanol solvent to remove impurities and oil, ensuring stable installation, good sealing of connections, and no gas leakage. A laser is selected as the light source based on the absorption characteristics of the gas to be measured, matching the wavelength of the emitted light to the gas absorption peak. The output power of the light source is checked, and the drive current is adjusted to achieve and stabilize the output light intensity at a predetermined value. The optical path is optimized, and optical components such as mirrors and lenses are used to adjust the beam direction and focus point, ensuring that the beam accurately and without damage illuminates the photoacoustic cell, absorbing the microphone. The sensor is installed outside the photoacoustic cell. A filter circuit is added to the signal output terminal of the micro-sound sensor to further reduce noise interference, so as to receive and detect the weak photoacoustic signal generated inside the photoacoustic cell. The modulation mode of the light source is set, and the resonant frequency and modulation frequency are calibrated. The laser pulse method is used to excite the sound wave, and the resonant frequency is measured by the micro-sound sensor. A calibration experiment is conducted using a standard gas of known concentration. The modulation frequency of the light source is adjusted to match the resonant frequency of the photoacoustic cell. The current modulation frequency of the light source and the photoacoustic signal intensity are recorded as the reference for subsequent measurements. By changing the concentration of the standard gas, the intensity change of the photoacoustic signal at different concentrations is measured, and a concentration-signal intensity curve is plotted. The sensitivity of the micro-sound sensor (i.e., the ratio of the signal intensity change to the concentration change) is calculated to ensure that it is within an acceptable range. All relevant calibration parameters are recorded, including the modulation frequency of the light source, the resonant frequency of the photoacoustic cell, and the sensitivity of the micro-sound sensor.

[0078] Step 2: Use temperature sensors and gas analysis sensors to detect changes in the temperature and composition of gases in the transformer oil. Based on the detected gas temperature and composition data, construct a resonant frequency prediction model to analyze the trend of resonant frequency changes. Deploy temperature sensors and gas analysis sensors to detect changes in the temperature and dissolved gas composition of the transformer oil, and collect data on temperature and gas composition in the transformer oil. Perform preprocessing operations on the collected data, including data cleaning and standardization. Perform feature analysis on the preprocessed data to extract features associated with the resonant frequency, namely temperature features, gas concentration features, and pressure features. The extracted resonant frequency is then analyzed. The correlation characteristics of the frequency are used to obtain a feature dataset, which is then divided into a training set and a test set. The training set data is used in conjunction with the support vector machine algorithm to construct a resonant frequency prediction model, and the test set data is used to verify the accuracy and generalization ability of the model in order to adjust the model parameters and predict the resonant frequency under different temperatures and gas compositions. The resonant frequency evaluation index is obtained by combining the correlation data of the feature dataset and the output of the resonant frequency prediction model. The trend of resonant frequency changes with temperature and gas composition is analyzed, and an adjustment threshold is set. When the output value of the resonant frequency evaluation index exceeds the adjustment threshold, the modulation frequency of the light source is adjusted to correct the impact of resonant frequency changes on the gas detection accuracy.

[0079] Furthermore, the expression for calculating the resonant frequency evaluation index is as follows:

[0080]

[0081] Δf res =f res -f res，base ;

[0082] Among them, F res The resonant frequency evaluation index is given by T, where T is temperature and C is temperature. i f is the concentration of the i-th gas. res f is the resonant frequency. res，base The reference resonant frequency is the resonant frequency measured under standard conditions, Δf. res The change in resonant frequency is denoted by n, where n is the number of gas types, and w is the number of gas types. i The weight of the i-th gas reflects the importance of that gas to the resonant frequency. base The reference temperature is the temperature at which the resonant frequency is measured under standard conditions. k is a temperature coefficient that reflects the degree to which temperature changes affect the resonant frequency. The combined effects of different gas concentrations on the resonant frequency were analyzed, as well as the influence of temperature changes on the resonant frequency. Reflecting the change in resonant frequency relative to a reference value, F is a key indicator for evaluating the stability of the resonant frequency. When the temperature and gas composition are close to the reference conditions, F... res A value close to 1 indicates system stability. When the temperature or gas composition deviates from the baseline conditions, F... res Significant changes may occur, and the indicator system may need to be adjusted;

[0083] Step 3: Based on the resonant frequency prediction model and the analysis results of the resonant frequency, dynamically adjust the modulation frequency of the light source to make it as close as possible to but not exceed the current resonant frequency of the photoacoustic cell. Observe the intensity change of the photoacoustic signal and analyze the amplitude of the photoacoustic signal at different modulation frequencies. Establish the relationship between the modulation frequency and the resonant frequency. Using the established resonant frequency prediction model, predict the resonant frequency of the photoacoustic cell based on the current state of the transformer oil (temperature and gas composition), and set the initial modulation frequency of the light source. The initial modulation frequency is a value close to but slightly lower than the predicted resonant frequency to avoid generating an excessively large photoacoustic signal at the beginning. The initial modulation frequency is also slightly wider than the predicted resonant frequency to ensure that the optimal modulation frequency can be found. Set the modulation frequency of the light source as the starting value of the search range, with small steps (a few Hz or tens of Hz). z) Gradually increase the modulation frequency of the light source and observe the intensity change of the photoacoustic signal. At each modulation frequency point, after stabilizing for a period of time, record the amplitude of the photoacoustic signal. Use a data acquisition system to collect the photoacoustic signal and analyze the signal amplitude. If the signal amplitude begins to increase significantly, it indicates that the modulation frequency is approaching the resonant frequency. At this time, gradually change the modulation frequency, reduce the step size, and adjust the modulation frequency more finely. Record the intensity of the photoacoustic signal after each adjustment, find the modulation frequency that produces the strongest photoacoustic signal, and take it as the optimal modulation frequency. The optimal modulation frequency is close to the resonant frequency of the photoacoustic cell. Record the photoacoustic signal amplitude corresponding to each modulation frequency point, draw a graph of the relationship between the modulation frequency and the photoacoustic signal amplitude, observe and confirm the position of the resonance peak, analyze the relationship between the modulation frequency and the photoacoustic signal amplitude, and confirm the actual value of the resonant frequency.

[0084] Step 4: A modulation system generates acoustic pulses, which propagate through a photoacoustic cell, causing gas pressure changes and exciting photoacoustic signals. A high-sensitivity micro-sound sensor collects these signals. The collected photoacoustic signals are preprocessed, and their intensity, frequency, and phase changes are analyzed to predict gas concentration. The modulation system is then configured to generate electrical signals, which are converted into acoustic pulses by a piezoelectric transducer. These acoustic pulses are released into the photoacoustic cell and propagate there, interacting with the gas and causing pressure changes. Simultaneously, a laser irradiates the gas in the photoacoustic cell. When the pressure change caused by the acoustic pulse occurs simultaneously with the light source irradiation, gas molecules absorb light energy and convert it into heat energy, resulting in localized temperature and pressure changes. These localized temperature and pressure changes propagate as sound waves, forming photoacoustic signals. A micro-sound sensor collects these pressure-induced photoacoustic signals and converts them into electrical signals for subsequent processing. The process involves preprocessing the electrical signals converted from the photoacoustic signals acquired by the photoacoustic sensor, applying filters to remove noise and interference components, improving the signal-to-noise ratio, and amplifying the signals appropriately for clearer observation and analysis. The electrical signals are then converted into digital signals for further processing. The preprocessed photoacoustic signal data is analyzed, including its intensity, frequency, and phase changes. Specifically, the intensity of the photoacoustic signals is analyzed, comparing signal intensities under different conditions to determine gas concentration changes. The frequency characteristics of the photoacoustic signals are analyzed to verify the effectiveness of the acoustic pulses and the performance of the photoacoustic cell. The phase changes of the photoacoustic signals provide additional information about the energy conversion and transfer process after gas molecules absorb light energy. Based on a calibration curve established by measuring the photoacoustic signal intensity at known concentrations, and combining the intensity, frequency, and phase characteristics of the photoacoustic signals, the gas concentration is calculated using the calibration curve. The characteristics of the photoacoustic signals measured each time and the predicted gas concentration are recorded.

[0085] Furthermore, the expression for calculating gas concentration is as follows:

[0086]

[0087] Where C is the predicted gas concentration, representing the target calculated value, I is the intensity of the photoacoustic signal, f is the intensity of the photoacoustic signal, φ is the phase of the photoacoustic signal, C0 is the standard gas concentration at a known concentration, I0 is the photoacoustic signal intensity corresponding to C0, a, b, and β are coefficients obtained from fitting the calibration curve, m is the number of measurements, and I... j Let fj be the photoacoustic signal intensity measured in the j-th measurement, f0 be the photoacoustic signal frequency under standard conditions, and φ be the frequency of the photoacoustic signal. j Let j be the phase of the photoacoustic signal measured in the j-th measurement. The model's sensitivity to changes in signal intensity was enhanced by analyzing the average photoacoustic signal intensity from multiple measurements and by power-lawing and averaging the intensity. By analyzing the deviation of the photoacoustic signal frequency from the standard frequency, the coefficient β can be adjusted to control the degree to which frequency changes affect concentration calculation. Analyzing the phase changes of the photoacoustic signal, and by taking the square root of the sum of the squares of the phase, we can provide the comprehensive impact of phase changes on concentration calculation. When I j Approaching I0, f approaches f0, and φ j When the change is small, C is close to C0, indicating that the gas concentration is stable. When these parameters deviate significantly from the standard value, C will change accordingly, indicating the change in gas concentration.

[0088] Step 5: Real-time detection of gas temperature, pressure and composition; dynamic adjustment of excitation parameters (resonance frequency and modulation frequency) of photoacoustic cell to adapt to changes in gas temperature and composition; and establishment of deviation prediction model based on preprocessed photoacoustic signal and preset detection standard to evaluate the deviation range of gas detection and correct the predicted gas detection deviation range.

[0089] Example 2, as Figure 4 As shown, based on Embodiment 1, the present invention provides a technical solution: Preferably, in step 5, the process of establishing the deviation prediction model is as follows:

[0090] Temperature, pressure, and composition of the gas are detected using temperature sensors, pressure sensors, and a gas analyzer. A micro-sound sensor continuously acquires photoacoustic signals generated by the photoacoustic effect, which are then converted into electrical signals for preprocessing. Based on the detected gas temperature, pressure, and composition information, the influence of gas parameters on the photoacoustic signal characteristics is analyzed. Based on the physical properties of the gas, the resonant frequency of the photoacoustic cell is dynamically adjusted by changing the frequency of the excitation signal to match the current acoustic characteristics of the gas. Furthermore, the frequency of the modulation signal is adjusted according to the dynamic response of the gas and detection requirements to optimize the intensity and signal-to-noise ratio of the photoacoustic signal. The acquired photoacoustic signal is filtered, amplified, and digitized to remove noise and interference components. Intensity, frequency, and phase features are extracted from the preprocessed photoacoustic signal, using historical data and known deviations. The system trains a deviation prediction model, taking detected gas parameters and photoacoustic signal characteristics as inputs and outputting the predicted gas detection deviation range. Based on the results of the deviation prediction model and preset detection standards, it calculates deviation evaluation coefficients, analyzes gas detection deviation trends, quantifies the degree of gas detection deviation, sets different deviation levels according to the deviation evaluation coefficients (minor, moderate, and severe deviation levels), and matches corresponding deviation thresholds to each deviation level. Based on the predicted deviation range and deviation level, it takes corresponding correction measures and feeds the correction results and new detection data back to the deviation prediction model to further optimize the monitoring accuracy of transformer oil gas. It also generates detailed monitoring reports regularly, summarizing gas detection results, deviation evaluation, correction measures, and their implementation effects.

[0091] Furthermore, the formula for calculating the deviation evaluation coefficient is as follows:

[0092]

[0093] Where PP is the bias evaluation coefficient, S is the set of detected gas parameters, such as temperature, pressure, and composition, F is the set of photoacoustic signal features, such as intensity, frequency, and phase, N is the sample size, μ is the sample mean, σ is the sample standard deviation, and C... r Let C be the predicted concentration of the r-th sample. ac，r S represents the actual concentration of the r-th sample. base Here, λ is the baseline value for the gas parameters, and λ is the adjustment parameter used to adjust the impact of gas parameters deviating from the baseline value on the deviation evaluation coefficient. r Let be the sum of squares of all features in the feature set of the photoacoustic signal of the r-th sample. The average relative deviation of all samples is calculated, reflecting the overall degree of deviation between the model predictions and the actual values. The gas parameter S was analyzed and its reference value S was compared. base The deviation between them is represented by an exponential function to achieve the non-linear effect of the deviation. Represents the limit mean of the comprehensive intensity of the photoacoustic signal feature set F, reflecting the trend of signal features with the increase in the number of samples. The value range of PP is from 0 to 1. 0 indicates no deviation, and 1 indicates complete deviation. When the predicted concentration is close to the actual concentration and the gas parameters are close to the reference value, PP approaches 0. When the prediction deviation or gas parameters deviate significantly from the reference value, PP increases, indicating a decrease in the reliability of the prediction;

[0094] Multiple deviation levels correspond to multiple deviation thresholds. Among them, the deviation thresholds include an upper threshold and a lower threshold;

[0095] The multiple deviation levels and multiple deviation thresholds satisfy the following relationship:

[0096] Minor deviation level 0 < PP ≤ PP M ; It means that the predicted value is very close to the actual value, the deviation can be ignored, and no immediate action is required, but regular reviews are needed;

[0097] Medium deviation level PP M < PP ≤ PP S ; It means that there is a certain degree of deviation, and further analysis or recalibration is required;

[0098] Severe deviation level PP S < PP < 1; It means that there is a significant deviation between the predicted value and the actual value, and immediate corrective measures are required;

[0099] Among them, PP is the deviation evaluation coefficient, and PP M is the lower threshold corresponding to the medium deviation level and the upper threshold corresponding to the minor deviation level, and PP S is the lower threshold corresponding to the severe deviation level and the upper threshold corresponding to the medium deviation level, and PP M = 0.1, and PP S = 0.5.

[0100] Example 3, as Figure 5 shown, based on Examples 1 - 2, the present invention further provides a gas detection system based on photoacoustic cell acoustic pulse excitation for implementing the gas detection method based on photoacoustic cell acoustic pulse excitation, including a detection center. The detection center is communicatively connected to a gas parameter detection module, a photoacoustic signal generation module, a photoacoustic signal acquisition module, a signal preprocessing module, a feature extraction module, an excitation parameter adjustment module, a deviation prediction and evaluation module, and a correction and optimization module. Among them, the modules are electrically connected to each other;

[0101] The gas parameter detection module uses a temperature sensor, a pressure sensor, and a gas analyzer to detect the temperature, pressure, and component information of the gas, providing comprehensive gas state data for the system as the basis for subsequent processing and adjustment;

[0102] The photoacoustic signal generation module is used to generate acoustic pulse excitation signals. Through the photoacoustic effect, it excites gas molecules in the photoacoustic cell to generate photoacoustic signals. It uses the photoacoustic effect to convert the light energy absorbed by the gas into detectable acoustic signals, providing direct evidence for gas detection.

[0103] The photoacoustic signal acquisition module uses a micro-sound sensor to continuously acquire photoacoustic signals generated by the photoacoustic effect and convert them into electrical signals, thus converting weak acoustic signals into electrical signals to facilitate subsequent signal processing and analysis.

[0104] The signal preprocessing module is used to filter, amplify, and digitize the acquired photoacoustic signals to remove noise and interference components, improve the signal-to-noise ratio, and ensure the accuracy of subsequent feature extraction and data analysis.

[0105] The feature extraction module is used to extract signal intensity, frequency, and phase features from the preprocessed photoacoustic signal, providing useful input data for the deviation prediction model and helping the model better understand the relationship between the gas state and the photoacoustic signal.

[0106] The excitation parameter adjustment module dynamically adjusts the resonant frequency and modulation frequency of the photoacoustic cell based on the detected gas parameters and photoacoustic signal characteristics, thereby optimizing the intensity and signal-to-noise ratio of the photoacoustic signal and improving the sensitivity and accuracy of gas detection.

[0107] The deviation prediction and evaluation module uses historical data and known deviation situations to train a deviation prediction model, predict the deviation range of gas detection, calculate the deviation evaluation coefficient, analyze the deviation trend of gas detection, discover potential detection deviations in advance, quantify the degree and trend of deviations, and provide a scientific basis for correction measures.

[0108] The calibration and optimization module takes corresponding calibration measures based on the predicted deviation range and deviation level, and feeds back the calibration results and new detection data to the system to correct detection deviations and improve the stability and reliability of the system.

[0109] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A gas detection method based on photoacoustic cell acoustic pulse excitation, characterized in that: Includes the following steps: Step 1: Construct the gas detection system and data acquisition system, initialize the photoacoustic cell, light source, and microphone, perform preliminary calibration of the gas detection system, and record the initial calibration parameters; Step 2: Use temperature sensors and gas analysis sensors to detect changes in the temperature and composition of gases in the transformer oil. Based on the detected gas temperature and composition data, construct a resonant frequency prediction model to analyze the trend of resonant frequency changes. The construction process of the resonant frequency prediction model is as follows: Step 201: Deploy temperature sensors and gas analysis sensors to detect changes in the temperature of transformer oil and the composition of dissolved gases in transformer oil, and collect data on the temperature and gas composition in transformer oil, and perform preprocessing operations on the collected data. Step 202: Perform feature analysis on the preprocessed data to extract features related to the resonance frequencies, namely temperature features, gas concentration features, and pressure features. Combine the extracted resonance frequency related features to obtain a feature dataset, and divide the feature dataset into a training set and a test set. Step 203: Construct a resonant frequency prediction model using training set data and support vector machine algorithm, and verify the model using test set data to predict the resonant frequency under different temperatures and gas compositions. Step 204: Combine the associated data of the feature dataset and the output of the resonance frequency prediction model to obtain the resonance frequency evaluation index and analyze the trend of resonance frequency with temperature and gas composition. Step 205: Set an adjustment threshold. When the output value of the resonance frequency evaluation index exceeds the adjustment threshold, adjust the modulation frequency of the light source to correct the impact of resonance frequency changes on gas detection accuracy. The formula for calculating the resonance frequency evaluation index is as follows: ； ； in, The index for evaluating resonant frequency. For temperature, For the first The concentration of the gas, The resonant frequency, The reference resonant frequency is the resonant frequency measured under standard conditions. The change in resonant frequency. The number of gas types, For the first The weight of each gas The reference temperature is the temperature corresponding to the resonant frequency measured under standard conditions. Temperature coefficient; Step 3: Based on the resonant frequency prediction model and the analysis results of the resonant frequency, dynamically adjust the modulation frequency of the light source, observe the intensity change of the photoacoustic signal, and analyze the amplitude of the photoacoustic signal at different modulation frequencies to establish the relationship between the modulation frequency and the resonant frequency. The process of establishing the relationship between the modulation frequency and the resonant frequency is as follows: Step 301: Using the established resonant frequency prediction model, predict the resonant frequency of the photoacoustic cell based on the current state of the transformer oil, and set the initial modulation frequency of the light source. Step 302: Set the modulation frequency of the light source to the starting value of the search range, gradually increase the modulation frequency of the light source in small steps, observe the intensity change of the photoacoustic signal, and record the amplitude of the photoacoustic signal at each modulation frequency point. Step 303: Use a data acquisition system to acquire photoacoustic signals and analyze the amplitude of the signals. Gradually change the modulation frequency and reduce the step size. Record the intensity of the photoacoustic signal after each adjustment. Find the modulation frequency that produces the strongest photoacoustic signal as the optimal modulation frequency. The optimal modulation frequency is close to the resonant frequency of the photoacoustic cell. Step 304: Record the photoacoustic signal amplitude corresponding to each modulation frequency point, draw a graph showing the relationship between modulation frequency and photoacoustic signal amplitude, observe and confirm the position of the resonance peak, analyze the relationship between modulation frequency and photoacoustic signal amplitude, and confirm the actual value of the resonance frequency. Step 4: Use a modulation system to generate acoustic pulses, causing gas pressure changes within the photoacoustic cell and exciting photoacoustic signals. Use a micro-sound sensor to collect the photoacoustic signals, preprocess the collected signals, analyze the intensity, frequency, and phase changes of the photoacoustic signals, and predict the gas concentration. The gas concentration prediction process is as follows: Step 401: Configure the modulation system, use the modulation system to generate an electrical signal, convert the electrical signal into an acoustic pulse through a piezoelectric transducer, release the acoustic pulse into the photoacoustic cell, propagate in the photoacoustic cell, interact with the gas in the photoacoustic cell, causing a change in gas pressure, and at the same time as the acoustic pulse propagates, use a laser to irradiate the gas in the photoacoustic cell. When the gas pressure change caused by the acoustic pulse occurs at the same time as the light source irradiation, the gas molecules absorb light energy and convert it into heat energy, thereby generating local temperature and pressure changes. The local temperature and pressure changes propagate in the form of sound waves, forming a photoacoustic signal. Step 402: Use a micro-sound sensor to collect photoacoustic signals generated by pressure changes and convert them into electrical signals. Perform preprocessing on the electrical signals converted from the photoacoustic signals collected by the micro-sound sensor. Step 403: Analyze the preprocessed photoacoustic signal data to analyze the intensity, frequency, and phase changes of the photoacoustic signal; Step 404: Based on the calibration curve established by measuring the photoacoustic signal intensity at a known concentration, and combining the intensity, frequency, and phase characteristics of the photoacoustic signal, calculate the gas concentration using the calibration curve, and record the photoacoustic signal characteristics of each measurement and the calculated predicted gas concentration. The formula for calculating the gas concentration is: ； in, The predicted gas concentration represents the calculated target value. The intensity of the photoacoustic signal. The intensity of the photoacoustic signal. The phase of the photoacoustic signal. The standard gas concentration is given at a known concentration. To and The corresponding photoacoustic signal intensity, , , The coefficients are obtained by fitting the calibration curve. To measure the number of times, For the first The photoacoustic signal intensity measured in this second measurement The photoacoustic signal frequency under standard conditions. For the first The phase of the photoacoustic signal measured in the second measurement; Step 5: Real-time detection of gas temperature, pressure, and composition; dynamic adjustment of the excitation parameters of the photoacoustic cell to adapt to changes in gas temperature and composition; and establishment of a deviation prediction model based on the preprocessed photoacoustic signal and preset detection standards to evaluate the deviation range of gas detection. The predicted gas detection deviation range is then corrected. The process of establishing the deviation prediction model is as follows: Step 501: Use a temperature sensor, a pressure sensor and a gas analyzer to detect the temperature, pressure and composition of the gas, use a micro-sound sensor to continuously collect photoacoustic signals generated by the photoacoustic effect, and convert the photoacoustic signals into electrical signals for preprocessing; Step 502: Based on the detected gas temperature, pressure and composition information, analyze the influence of gas parameters on photoacoustic signal characteristics. Based on the physical properties of the gas, dynamically adjust the resonant frequency of the photoacoustic cell by changing the frequency of the excitation signal to match the current acoustic characteristics of the gas. According to the dynamic response of the gas and the detection requirements, adjust the frequency of the modulation signal to optimize the intensity and signal-to-noise ratio of the photoacoustic signal. Step 503: Filter, amplify and digitize the acquired photoacoustic signal to remove noise and interference components, and extract intensity, frequency and phase features from the preprocessed photoacoustic signal; Step 504: Train the deviation prediction model using historical data and known deviation conditions. Take the detected gas parameters and photoacoustic signal characteristics as the input of the model, output the predicted gas detection deviation range, and calculate the deviation evaluation coefficient based on the results of the deviation prediction model and the preset detection standard to analyze the gas detection deviation trend and quantify the degree of gas detection deviation. Step 505: Set different deviation levels according to the deviation evaluation coefficient, namely slight deviation level, moderate deviation level and severe deviation level, and match the corresponding deviation threshold for each deviation level. Step 506: Based on the predicted deviation range and deviation level, take corresponding correction measures, and feed the correction results and new detection data back to the deviation prediction model to further optimize the monitoring accuracy of transformer oil gas, generate monitoring reports regularly, and summarize the gas detection results, deviation assessment, correction measures and their implementation effects. The formula for calculating the deviation evaluation coefficient is as follows: ； in, This is the deviation evaluation coefficient. The set of detected gas parameters, It is a set of photoacoustic signal features. For the sample size, The sample mean. The standard deviation of the sample is 1. For the first The predicted concentration for each sample For the first The actual concentration of each sample This serves as the baseline value for the gas parameters. To adjust the parameters, For the first The sum of squares of all features in the feature set of a sample photoacoustic signal. The value ranges from 0 to 1, where 0 represents no deviation and 1 represents complete deviation.

2. The gas detection method based on photoacoustic cell acoustic pulse excitation according to claim 1, characterized in that: In step 1, the process of performing preliminary calibration of the gas detection system is as follows: Step 101: Construct a gas detection system and a data acquisition system, including a photoacoustic cell, a light source, and a micro-sound sensor. The photoacoustic cell is constructed and its inner and outer surfaces are cleaned with ethanol solvent. Step 102: Select a laser as the light source according to the absorption characteristics of the gas to be tested, match the wavelength of the light emitted by the light source with the absorption peak of the gas, check the output power of the light source, and adjust the driving current to make the output light intensity reach the predetermined value and remain stable. Step 103: Install the microphone sensor outside the photoacoustic cell and add a filter circuit to the signal output terminal of the microphone sensor to further reduce noise interference; Step 104: Set the modulation method of the light source and calibrate the resonant frequency and modulation frequency. The laser pulse method is used to excite the sound wave and the resonant frequency is measured using a micro-sound sensor. A calibration experiment is performed using a standard gas of known concentration. The modulation frequency of the light source is adjusted to match the resonant frequency of the photoacoustic cell. The current light source modulation frequency and photoacoustic signal intensity are recorded. Step 105: By changing the concentration of the standard gas, measure the intensity change of the photoacoustic signal at different concentrations, plot the concentration-signal intensity curve, calculate the sensitivity of the microphone sensor, and record all relevant calibration parameters, including the modulation frequency of the light source, the resonant frequency of the photoacoustic cell, and the sensitivity of the microphone sensor.

3. The gas detection method based on photoacoustic cell acoustic pulse excitation according to claim 2, characterized in that: The multiple deviation levels correspond to multiple deviation thresholds, wherein the deviation thresholds include an upper limit threshold and a lower limit threshold; The plurality of deviation levels and the plurality of deviation thresholds satisfy the following relationship: Slight deviation level ; Medium deviation level ; Severe Deviation Level ; in, This is the deviation evaluation coefficient. These are the lower threshold for the moderate deviation level and the upper threshold for the slight deviation level. These are the lower threshold for the severe deviation level and the upper threshold for the moderate deviation level. , .

4. A gas detection system based on photoacoustic cell acoustic pulse excitation, used to implement the gas detection method based on photoacoustic cell acoustic pulse excitation as described in claim 3, comprising a detection center, characterized in that: The detection center is connected to a gas parameter detection module, a photoacoustic signal generation module, a photoacoustic signal acquisition module, a signal preprocessing module, a feature extraction module, an excitation parameter adjustment module, a deviation prediction and evaluation module, and a correction and optimization module. The modules are electrically connected to each other. The gas parameter detection module uses a temperature sensor, a pressure sensor, and a gas analyzer to detect the temperature, pressure, and composition information of the gas. The photoacoustic signal generating module is used to generate an acoustic pulse excitation signal, which excites gas molecules in the photoacoustic cell to generate a photoacoustic signal through the photoacoustic effect, and converts the light energy absorbed by the gas into a detectable acoustic signal using the photoacoustic effect. The photoacoustic signal acquisition module uses a micro-sound sensor to continuously acquire photoacoustic signals generated by the photoacoustic effect and converts them into electrical signals. The signal preprocessing module is used to filter, amplify, and digitize the acquired photoacoustic signals; The feature extraction module is used to extract signal intensity, frequency, and phase features from the preprocessed photoacoustic signal. The excitation parameter adjustment module dynamically adjusts the resonant frequency and modulation frequency of the photoacoustic cell based on the detected gas parameters and photoacoustic signal characteristics. The deviation prediction and evaluation module uses historical data and known deviation conditions to train a deviation prediction model, predict the deviation range of gas detection, calculate the deviation evaluation coefficient, and analyze the deviation trend of gas detection. The correction and optimization module takes corresponding correction measures based on the predicted deviation range and deviation level, and feeds back the correction results and new detection data to the system to correct the detection deviation.

Images