TDLAS (Tunable Diode Laser Absorption Spectroscopy)-based mine confined space gas concentration adaptive detection method

Abstract

The invention belongs to the technical field of gas detection, and discloses a TDLAS-based mine confined space gas concentration adaptive detection method. Comprising the following steps: S1, acquiring environmental parameters and approximate background spectral signals of a mine confined space, performing composite analysis, and determining current control parameters of a laser based on output wavelength drift compensation amount; s2, driving a laser to output wavelength by applying the current control parameter, scanning a characteristic absorption spectral line area of the target gas, obtaining an initial absorption signal, and analyzing and determining a gain absorption signal of the target gas; s3, performing secondary harmonic detection and phase-locked amplification treatment on the gain absorption signal to generate a pure absorption spectral line of the target gas, determining an initial concentration value of the target gas based on characteristic parameters of the pure absorption spectral line, and performing further analysis to obtain a real-time concentration value of the target gas; and S4, generating a parameter instruction or a standard exceeding early warning signal based on the real-time concentration value of the target gas, and providing stable signal guarantee for concentration inversion.

Description

Technical Field

[0001] This invention relates to the field of gas detection technology, and more specifically, to an adaptive detection method for gas concentration in confined spaces in mines based on TDLAS. Background Technology

[0002] Tunable semiconductor laser absorption spectroscopy (TDLAS) is a gas detection technique that scans specific absorption lines of gas molecules and inverts concentration based on Beer-Lambert's law. It has been widely used in gas safety detection in confined spaces in mines. However, significant fluctuations in temperature and pressure within confined spaces in mines can cause shifts in the position of gas absorption lines.

[0003] In existing TDLAS detection systems, a mathematical model is typically established to correlate temperature with the position of gas absorption lines. Software algorithms calculate wavelength compensation based on ambient temperature sensor data and adjust laser temperature or current parameters to compensate for temperature-induced wavelength shifts. However, this single-path, single-compensation-source-dependent approach ignores nonlinear interferences such as pressure changes and humidity coupling. When environmental factors interact unexpectedly, the model output deviates significantly, leading to inaccurate measurement standards. For example, when detecting methane concentration in a confined space in a mine, technicians relied solely on a temperature-wavelength model for compensation. On a particular day, while the underground temperature remained stable, the pressure surged, and humidity increased due to water seepage from the tunnel. Because the model failed to consider the coupling effect of pressure and humidity, it mistakenly attributed the absorption peak shift to temperature changes, resulting in a lower displayed methane concentration than the actual value, nearly causing a misjudgment and highlighting the limitations of a single compensation path.

[0004] In view of this, the present invention proposes an adaptive detection method for gas concentration in confined spaces in mines based on TDLAS to solve the above problems. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above objectives, the present invention provides the following technical solution: an adaptive detection method for gas concentration in a mine confined space based on TDLAS, comprising: S1, acquiring environmental parameters and approximate background spectral signals of the mine confined space, performing composite analysis on the environmental parameters and approximate background spectral signals, outputting the wavelength drift compensation amount of the laser, and determining the current control parameters of the laser based on the wavelength drift compensation amount;

[0006] S2. Apply the current control parameters to drive the laser output wavelength, scan the characteristic absorption spectral region of the target gas and obtain the initial absorption signal, analyze and evaluate the initial absorption signal, and determine the gain absorption signal of the target gas.

[0007] S3. Perform second harmonic detection and lock-in amplification on the gain absorption signal to generate a pure absorption spectrum of the target gas. Analyze the characteristic parameters of the pure absorption spectrum to determine the preliminary concentration value of the target gas. Further analyze the preliminary concentration value to obtain the real-time concentration value of the target gas.

[0008] S4. Based on the real-time concentration value of the target gas, determine the status and generate parameter instructions or over-limit warning signals according to the determination results.

[0009] Furthermore, methods for combined analysis of environmental parameters and approximate background spectral signals include:

[0010] After collecting the transmission spectrum signal obtained by laser scanning of the gas in the confined space of the mine, signal points with no absorption characteristics on both sides of the absorption line in the transmission spectrum signal are selected and fitted to obtain an approximate background spectrum signal; temperature, humidity and air pressure values ​​of the confined space of the mine are collected simultaneously to form environmental parameters.

[0011] Obtain a sample dataset of environmental parameters and wavelength drift, perform regression fitting analysis on the sample dataset to determine the coefficients of the linear term and the cross-coupling term of the environmental parameters, and solve the environmental parameters based on the coefficients of the linear term and the cross-coupling term to determine the environmental-induced wavelength drift.

[0012] Mutual information entropy analysis is performed on the reference spectral signal and the approximate background spectral signal to determine the spectral correlation wavelength shift; quality decisions are made on the spectral correlation wavelength shift and the environment-induced wavelength shift to determine the wavelength shift compensation amount of the laser.

[0013] Furthermore, methods for performing mutual information entropy analysis on the reference spectral signal and the approximate background spectral signal include:

[0014] The reference spectral signal and the approximate background spectral signal are processed by wavelength axis alignment and amplitude normalization; the processed approximate background spectral signal is slid on the wavelength axis with a fixed step size, and the shifted approximate background spectral signals corresponding to different sliding offsets are combined to form a displacement signal sequence.

[0015] Each shifted approximate background spectral signal and the processed reference spectral signal in the displacement signal sequence are discretized to form multiple pairs of histograms with the same interval; the mutual information entropy value of each pair of histograms is calculated to obtain the mutual information entropy-offset relationship sequence.

[0016] A curve is formed by smoothing the mutual information entropy-sliding offset relationship sequence, the global maximum point of the curve is located, and the sliding offset corresponding to the global maximum point is used as the spectral correlation wavelength drift.

[0017] Furthermore, methods for determining the current control parameters of the laser based on wavelength drift compensation include:

[0018] Based on the temperature sensitivity coefficient and current sensitivity coefficient calibrated by the laser, the wavelength drift compensation is calculated to determine the temperature setpoint adjustment and drive current adjustment.

[0019] The temperature setpoint adjustment and drive current adjustment are respectively superimposed on the laser's current operating point to determine the laser's current control parameters, which include the current temperature controller setpoint and the current drive current setpoint.

[0020] Furthermore, methods for scanning the characteristic absorption spectral region of the target gas and obtaining the initial absorption signal include:

[0021] A sawtooth wave scanning signal is superimposed on the current driving current setting, and the laser wavelength is driven to scan the characteristic absorption spectral region of the target gas. At the same time, the transmitted light intensity sequence during the scanning process is synchronously acquired based on the data acquisition card.

[0022] The transmitted light intensity sequence is analyzed to determine the precise driving current value corresponding to the absorption peak and valley values. The precise driving current value is set as the current new driving current setting value of the laser, and a high-frequency sinusoidal modulation signal is superimposed on the current new driving current setting value to complete the laser wavelength locking and modulation.

[0023] The locked and modulated laser is introduced into a Heriot-type multipass cell for several reflections, and the received emitted laser is converted into a raw electrical signal. The raw electrical signal is then pre-amplified and filtered to form an initial absorption signal.

[0024] Furthermore, methods for analyzing and evaluating the initial absorption signal include:

[0025] Spectral analysis is performed on the initial absorption signal to extract the peak intensity and signal-to-noise ratio of the initial absorption signal; the peak intensity is compared with a preset intensity threshold; if the peak intensity is lower than the intensity threshold, the mirror spacing of the Heriot-type multi-pass cell is gradually adjusted to re-reflect the locked and modulated laser until the peak intensity of the initial absorption signal is not lower than the intensity threshold.

[0026] The signal-to-noise ratio (SNR) is compared with a preset SNR threshold. If the SNR is less than the preset SNR threshold, the integration time of the data acquisition card is gradually extended for rescanning until the SNR of the initial absorbed signal is not lower than the SNR threshold.

[0027] The initial absorption signal that meets the threshold judgment condition is determined as the gain absorption signal of the target gas.

[0028] Furthermore, methods for generating pure absorption lines of the target gas include:

[0029] The modulation frequency component extracted from the gain absorption signal is frequency multiplied to generate a second harmonic reference signal; the gain absorption signal is filtered based on a bandpass filter with a center frequency of second harmonic to determine the second harmonic signal in the gain absorption signal.

[0030] The second harmonic signal and the second harmonic reference signal are simultaneously input into the lock-in amplifier, and the output is a mixed signal containing DC component and high-frequency noise; the mixed signal is low-pass filtered to obtain the second harmonic amplitude signal.

[0031] The peak position in the second harmonic amplitude signal is located. Based on the center position of the target gas absorption spectrum, the second harmonic amplitude signal is linearly fitted to determine the fitting baseline. The second harmonic amplitude signal is subtracted from the fitting baseline to obtain the pure absorption spectrum of the target gas.

[0032] Furthermore, methods for determining the preliminary concentration value of the target gas include:

[0033] Record the peak heights of each peak on the pure absorption spectrum to determine the absorption peak height parameters; determine the effective optical path value based on the number of reflections and single optical path length of the Heriot-type multipass cell; extract the center wavelength from the pure absorption spectrum and query the preset spectral database to obtain the absorption line intensity coefficient of the target gas.

[0034] Based on Beer-Lambert's law, the absorption peak height parameter, effective optical path value, and absorption line intensity coefficient were analyzed to obtain the preliminary concentration value of the target gas.

[0035] Furthermore, methods for further analyzing the preliminary concentration values ​​and obtaining the real-time concentration values ​​of the target gas include:

[0036] Calculate the area enclosed by the pure absorption line and the fitted baseline to determine the absorption peak area parameters;

[0037] A calibration curve is established based on a standard target gas of known concentration. The calibration curve includes the measured absorption peak area corresponding to each concentration. The absorption peak area parameter is input into the calibration curve to determine the area method concentration value of the target gas. The preliminary concentration value and the area method concentration value are weighted and fused to obtain the real-time concentration value of the target gas.

[0038] Furthermore, methods for generating parameter commands or over-limit warning signals based on the judgment results include:

[0039] A preset safe concentration threshold is set, which includes an upper threshold and a lower threshold. The real-time concentration value of the target gas is compared with the safe concentration threshold. If the real-time concentration value of the target gas is lower than the lower threshold, it is determined that the current state is safe, and a parameter instruction to maintain the current detection frequency is generated.

[0040] If the real-time concentration of the target gas is between the upper and lower thresholds, it is determined that the current state is alert, and a parameter instruction to increase the current detection frequency is generated.

[0041] If the real-time concentration of the target gas exceeds the upper limit threshold, an over-limit warning signal will be generated.

[0042] The technical effects and advantages of the adaptive detection method for gas concentration in confined spaces in mines based on TDLAS of this invention are as follows:

[0043] 1. This invention collects environmental parameters and calculates the environmentally induced wavelength drift based on a pre-calibrated sample dataset. Simultaneously, it uses mutual information entropy analysis to compare the reference spectrum with the measured approximate background spectrum to obtain the spectral-related wavelength drift. Then, it generates the optimal wavelength drift compensation amount through quality decision fusion. This effectively overcomes the interference of drastic temperature and humidity fluctuations and air pressure changes on the laser wavelength in the confined space of a mine, ensuring stable overlap between the scanning spectrum and the gas absorption spectrum. The wavelength drift compensation amount determines the temperature setpoint adjustment and drive current adjustment, dynamically superimposing and updating the current control parameters. Furthermore, it utilizes the optical path enhancement effect of a Heriot-type multi-pass cell and adaptive mirror spacing adjustment to actively optimize signal strength and signal-to-noise ratio, thus providing a stable and high-quality original signal guarantee for concentration inversion in the harsh environment of a mine.

[0044] 2. This invention calculates concentration using the peak height method based on Beer-Lambert's law and the area method based on the calibration curve, respectively. Then, through weighted fusion, it effectively complements the sensitivity of a single algorithm to wavelength jitter or baseline drift. This significantly improves the accuracy and anti-interference capability of real-time concentration values. By presetting a safety threshold range, the system can automatically trigger differentiated instructions such as maintaining the detection frequency, increasing the detection frequency, or issuing over-limit warnings based on the different ranges in which the real-time concentration value is located. This achieves intelligent switching from energy-saving monitoring in a safe state to emergency alarms and linkage control in a dangerous state. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the adaptive gas concentration detection method for confined spaces in mines based on TDLAS according to the present invention.

[0046] Figure 2 This is a flowchart illustrating the process of obtaining the spectral correlation wavelength shift in this invention.

[0047] Figure 3 This is a flowchart for generating parameter instructions for this invention. Detailed Implementation

[0048] 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, and 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.

[0049] Example 1

[0050] Please see Figures 1-3 As shown in the figure, the main design contents of the adaptive detection method for gas concentration in a confined space in a mine based on TDLAS described in this embodiment are as follows:

[0051] In the confined spaces of mines, which are high-risk working environments, gas concentration monitoring is crucial. However, complex factors such as drastic temperature fluctuations (0-50°C), air pressure changes (±10kPa), and humidity as high as 90% often cause relative drift between the wavelength of the TDLAS laser and the gas absorption spectrum, resulting in signal weakening, noise inundation, and concentration inversion deviation. Relying on a single temperature compensation model cannot effectively cope with multi-source nonlinear interference, which can easily lead to inaccurate measurements and create potential safety hazards such as explosions or poisoning.

[0052] Based on this, an adaptive detection method for gas concentration in confined spaces in mines based on TDLAS is designed, including:

[0053] 1. An adaptive detection method for gas concentration in confined spaces in mines based on TDLAS, characterized in that it includes:

[0054] S1. Obtain environmental parameters and approximate background spectral signals of the confined space in the mine, perform composite analysis on the environmental parameters and approximate background spectral signals, output the wavelength drift compensation amount of the laser, and determine the current control parameters of the laser based on the wavelength drift compensation amount;

[0055] Methods for combined analysis of environmental parameters and approximate background spectral signals include:

[0056] Transmission spectral signals obtained by laser scanning of gas in a confined space in a mine are collected. Signal points without absorption characteristics on both sides of the absorption lines in the transmission spectral signals are selected and fitted to obtain approximate background spectral signals.

[0057] The transmitted light intensity during the laser scanning of the target gas's characteristic absorption spectral line region is synchronously acquired using a data acquisition card. This yields a transmission spectral signal data sequence composed of wavelength indices and transmitted light intensity data points. Within this transmission spectral signal data sequence, the center wavelength position of the target gas's characteristic absorption spectral line is located. Using this center wavelength position as a reference, regions with smooth spectral curves, gradual intensity changes, and no obvious absorption dips are searched to the left and right, respectively. These two regions are designated as the left-side no-absorption region and the right-side no-absorption region. The spectral signal of these regions is then used as a background signal that is approximately unaffected by the target gas's absorption.

[0058] Simultaneously collect temperature, humidity, and air pressure values ​​in the enclosed space of the mine to form environmental parameters. Temperature and humidity values ​​are collected using temperature and humidity sensors, while air pressure values ​​are collected using air pressure sensors.

[0059] To obtain an environmental parameter-wavelength drift sample dataset, the laser is placed in an environmental test chamber where temperature, pressure, and humidity can be precisely controlled. Within the laser's expected operating range, multiple different combinations of temperature, humidity, and pressure values ​​are systematically traversed. Under each specific combination of environmental parameters, the actual output wavelength of the laser is measured, and the wavelength drift relative to the baseline environmental parameters is calculated. Each "environmental parameter combination" and its corresponding "wavelength drift" are recorded as a data point. This process is repeated until a sufficient number of sample points are obtained, forming an environmental parameter-wavelength drift sample dataset.

[0060] Regression fitting analysis was performed on the environmental parameter-wavelength drift sample dataset to determine the coefficients of the linear terms and cross-coupling terms of the environmental parameters. Based on the environmental parameter-wavelength drift sample dataset, temperature, humidity, and air pressure were used as independent variables, and wavelength drift was used as the dependent variable. A multiple linear regression algorithm was used to fit the environmental parameter-wavelength drift sample dataset and solve the regression equation. This regression equation includes the coefficients of the linear terms of temperature, humidity, and air pressure, as well as the coefficients of the cross-coupling terms of temperature and humidity, temperature and air pressure, and humidity and air pressure.

[0061] The environmental parameters are solved based on the linear and cross-coupling coefficients to determine the environmental-induced wavelength drift. A set of environmental parameters (temperature, air pressure, humidity) corresponding to the precise locking of the laser wavelength under a reference environment is used as the reference environmental parameters. The environmental parameters are subtracted one by one from the reference environmental parameters to obtain the temperature deviation, air pressure deviation, and humidity deviation, respectively. Each environmental parameter deviation is multiplied by its corresponding linear coefficient, and the sum is obtained to obtain the linear drift component. The product of each environmental parameter deviation is multiplied by its corresponding cross-coupling coefficient, and the sum is obtained to obtain the cross-coupling drift component. All calculated components are algebraically summed to obtain the environmental-induced wavelength drift.

[0062] For example, the linear coefficient of environmental parameters: the coefficient of the temperature linear term. =0.010nm / ℃, humidity linear term coefficient =-0.0002nm / %RH, pressure linearity coefficient =+0.001nm / kPa.

[0063] Cross-coupling coefficients: Temperature-humidity cross-coupling coefficients =+0.00002nm / (℃·%RH)Temperature-pressure cross-coupling term coefficient =-0.0001nm / (℃·kPa), Humidity-Pressure Cross-Coupling Coefficient =+0.00001nm / (%RH·kPa;

[0064] Reference environmental parameters: reference temperature T1 = 25.0℃, reference air pressure P1 = 101.3kPa, reference humidity H1 = 50.0%RH;

[0065] Environmental parameters collected in real time in the confined space of the mine: real-time temperature T=28.0℃, real-time air pressure P=95.0kPa, real-time humidity H=80.0%RH;

[0066] The environmental parameter deviations are: temperature deviation ΔT = T - T1 = +3.0℃, air pressure deviation ΔP = P - P1 = -6.3kPa, humidity deviation ΔH = H - H1 = +30.0%RH;

[0067] Linear drift component = ( ×ΔT)+( ×ΔH)+( ×ΔP)=+0.0177nm;

[0068] Cross-coupling drift component = [ ×(ΔT×ΔH)]+[ ×(ΔT×ΔP)]+[ [×(ΔH×ΔP)]=+0.0018nm;

[0069] Environmentally induced wavelength shift = linear drift component + cross-coupling drift component = +0.0195nm.

[0070] Mutual information entropy analysis is performed on the reference spectral signal and the approximate background spectral signal to determine the spectral-related wavelength drift. A quality decision is made on the spectral-related wavelength drift and the environment-induced wavelength drift to determine the wavelength drift compensation amount for the laser. The spectral-related wavelength drift and the environment-induced wavelength drift are mapped to low, medium, and high fuzzy sets using triangular membership functions to generate their respective fuzzy sets. The activation intensity of each rule is calculated using Mamdani inference rules (e.g., if the spectral-related wavelength drift is medium and the environment-induced wavelength drift is high, then the compensation amount is medium), generating an activation rule set. The activation rule sets are aggregated, and the output fuzzy sets are merged through a max-min synthesis operation to generate a compensation amount fuzzy set. The compensation amount fuzzy set is defuzzified, and the average value is calculated using weighted averages (e.g., centroid method) to generate the wavelength drift compensation amount.

[0071] Methods for mutual information entropy analysis of reference spectral signals and approximate background spectral signals include:

[0072] Wavelength axis alignment and amplitude normalization are performed on the reference spectral signal and the approximate background spectral signal. The reference spectral signal and the approximate background spectral signal are preprocessed, and the approximate background spectral signal is resampled onto the wavelength grid of the reference spectral signal to ensure that the two spectral signals have the same wavelength coordinate axis and data point resolution. The amplitude of the two wavelength-aligned spectral signals is normalized, and their intensity values ​​are uniformly scaled to the [0,1] interval.

[0073] The processed approximate background spectral signal is slid along the wavelength axis with a fixed step size. A displacement signal sequence is formed by combining the shifted approximate background spectral signals corresponding to different sliding offsets. The amplitude-normalized approximate background spectral signal is used as the reference signal. The reference signal is slid left and right along the wavelength axis with a fixed wavelength offset step size. After each slide, the portion of the reference signal that overlaps with the wavelength range of the processed reference spectral signal is extracted to form a new shifted approximate background spectral signal. All shifted approximate background spectral signals generated under different sliding offsets are arranged in order of offset to form a displacement signal sequence.

[0074] It should be explained that the fixed wavelength setting can be based on the wavelength resolution of the reference spectral signal. For example, the step size can be set to 1 to 2 times the minimum wavelength interval between adjacent data points on the wavelength axis.

[0075] Each shifted approximate background spectral signal in the displacement signal sequence is discretized with the processed reference spectral signal (using equal-width intervals, with a certain number of intervals). M represents the total number of spectral data points (ensuring each partitioned interval contains at least 3 data points), forming multiple pairs of histograms with the same interval; calculating the mutual information entropy value of each histogram pair to obtain the mutual information entropy-offset relationship sequence. Each shifted approximate background spectral signal and the amplitude-normalized reference spectral signal in the displacement signal sequence, with their light intensity values ​​at the same wavelength point, are considered a data pair; traversing all wavelength points, the spatial distribution of all data pairs is statistically analyzed to construct a two-dimensional joint probability distribution histogram; based on the two-dimensional joint probability distribution histogram and its corresponding two one-dimensional marginal probability distributions, the mutual information entropy value of this signal pair is calculated; each sliding offset and its corresponding mutual information entropy value are recorded to form the mutual information entropy-offset relationship sequence.

[0076] A smooth curve is formed by fitting the mutual information entropy-sliding offset relationship sequence. The global maximum point of the curve is located, and the sliding offset corresponding to the global maximum point is used as the spectral correlation wavelength drift. Cubic spline interpolation is performed on the data points in the mutual information entropy-sliding offset relationship sequence to generate a smooth fitting curve. The fitting curve is traversed, and the point with the largest mutual information entropy value on the curve is located and recorded as the global maximum point. The x-coordinate value (sliding offset) corresponding to the global maximum point is read, which is the spectral correlation wavelength drift.

[0077] S2. Apply the current control parameters to drive the laser output wavelength, scan the characteristic absorption spectral region of the target gas and obtain the initial absorption signal, analyze and evaluate the initial absorption signal, and determine the gain absorption signal of the target gas.

[0078] Methods for determining the current control parameters of a laser based on wavelength drift compensation include:

[0079] Based on the laser's calibrated temperature and current sensitivity coefficients, the wavelength drift compensation is calculated to determine the temperature setpoint adjustment and drive current adjustment. Both the temperature and current sensitivity coefficients are inherent physical characteristics of the laser. The temperature sensitivity coefficient describes the change in the output laser wavelength when the laser's operating temperature changes by 1°C; the current sensitivity coefficient describes the change in the output laser wavelength when the laser's drive current changes by 1 mA. The laser's temperature controller setpoint and drive current setpoint are read, and the wavelength drift compensation is allocated to the temperature and current adjustments. The quotient obtained by dividing the wavelength drift compensation by the temperature sensitivity coefficient is used as the temperature setpoint adjustment; the quotient obtained by dividing the remaining wavelength drift compensation by the current sensitivity coefficient is used as the drive current adjustment.

[0080] The temperature setpoint adjustment and drive current adjustment are respectively superimposed on the laser's current operating point to determine the laser's current control parameters, which include the current temperature controller setpoint and the current drive current setpoint.

[0081] For example, the laser's current operating point can be read: the temperature controller setpoint is 25.0℃; the drive current setpoint is 80.0mA.

[0082] The temperature sensitivity coefficient is 0.1 nm / ℃, the current sensitivity coefficient is 0.01 nm / mA, the wavelength drift compensation is +0.0265 nm, the wavelength drift compensation for temperature regulation is +0.026 nm, and the wavelength drift compensation for driving current is +0.005 nm.

[0083] Temperature setpoint adjustment amount = +0.026nm ÷ 0.1nm / ℃ = +0.26℃;

[0084] Drive current adjustment = +0.05nm;

[0085] The current temperature controller setpoint is 25.26℃, and the current drive current setpoint is 80.05mA.

[0086] Methods for scanning the characteristic absorption spectral region of a target gas and obtaining the initial absorption signal include:

[0087] A sawtooth wave scanning signal is superimposed on the current driving current setting, and the laser wavelength is driven to scan the characteristic absorption spectral region of the target gas. Simultaneously, the transmitted light intensity voltage signal output by the photodetector during the scanning process is synchronously acquired based on the data acquisition card, forming a transmitted light intensity sequence. The transmitted light intensity sequence is a corresponding data set of "time-transmitted light intensity voltage signal".

[0088] It should be explained that the sawtooth wave scanning signal is a linearly rising / falling electrical signal superimposed on the current drive current setting, used to control the laser wavelength to "continuously and uniformly" cover the characteristic absorption spectral region of the target gas.

[0089] The transmitted light intensity sequence is analyzed to determine the precise driving current value corresponding to the absorption peak and valley values. Digital signal processing is performed on the transmitted light intensity sequence to find the minimum light intensity point, which is taken as the absorption peak and valley value (when the laser wavelength falls at the position of strongest absorption in the target gas, the least amount of light passes through the gas, resulting in the minimum light intensity value). The precise time point corresponding to the absorption peak and valley value in the scanning cycle is recorded, and the voltage value of the sawtooth wave scanning signal at that instant is deduced from this time point. Combined with the current driving current setting value, the precise driving current value that perfectly corresponds to the absorption peak and valley value is calculated.

[0090] The precise drive current value is set to the current new drive current setting value of the laser, and a high-frequency sinusoidal modulation signal is superimposed on the current new drive current setting value to complete the locking and modulation of the laser wavelength. The purpose of superimposing the high-frequency sinusoidal modulation signal is to apply a fast, small-amplitude periodic modulation to the locked laser wavelength so that the gas absorption signal can be extracted through harmonic detection technology.

[0091] The locked and modulated laser is introduced into a Heriot-type multipass cell for several reflections, allowing it to fully interact with the target gas inside the cell and increase the effective optical path of the target gas. The attenuated emitted laser is then received by a photodetector at the output end and converted into a raw electrical signal.

[0092] The original electrical signal is pre-amplified and filtered to form the initial absorption signal. The original electrical signal is input into a pre-amplifier circuit, where it undergoes transimpedance amplification (converting the current signal into a voltage signal) and voltage amplification to enhance the weak signal. The amplified electrical signal is then subjected to bandpass filtering to remove the DC component and high-frequency noise outside the modulation frequency range, forming the initial absorption signal.

[0093] Methods for analyzing and evaluating the initial absorption signal include:

[0094] Spectral analysis is performed on the initial absorbed signal to extract its peak intensity and signal-to-noise ratio.

[0095] The initial absorbed signal is transformed from the time domain to the frequency domain using Fourier transform to obtain its spectrum. The amplitude value at the frequency corresponding to the high-frequency sinusoidal modulation signal is located in the spectrum and recorded as the peak intensity of the initial absorbed signal. A noise band far from the signal frequency components and without significant interference is selected from the spectrum, and the average amplitude within the noise band is calculated. The average value is recorded as the noise floor. The ratio of the peak intensity to the noise floor is determined as the signal-to-noise ratio of the initial absorbed signal.

[0096] The peak intensity is compared with a preset intensity threshold. If the peak intensity is lower than the intensity threshold, a step control command is sent to the piezoelectric ceramic driver controlling the mirror spacing of the Heriot-type multi-pass cell to gradually adjust the mirror spacing of the Heriot-type multi-pass cell. The locked and modulated laser is re-reflected until the peak intensity of the initial absorption signal is not lower than the intensity threshold. After the adjustment is completed, the laser wavelength scanning and signal acquisition process is re-executed to obtain a new initial absorption signal, and the peak intensity evaluation is re-performed. This process is repeated until the peak intensity of the newly acquired initial absorption signal is not lower than the preset intensity threshold.

[0097] It should be explained that the intensity threshold is set by the effective linear operating range of the photodetector and subsequent amplification circuit and the optimal input voltage range of the lock-in amplifier.

[0098] The signal-to-noise ratio (SNR) is compared with a preset SNR threshold. If the SNR is less than the preset SNR threshold, the program is triggered to gradually extend the integration time of the data acquisition card (for example, the extension ratio can be set to 20% of the initial integration time each time) to rescan, re-execute the laser wavelength scanning and signal acquisition process, obtain a new initial absorption signal and calculate the SNR until the SNR of the initial absorption signal is not lower than the SNR threshold.

[0099] It should be explained that the signal-to-noise ratio threshold is set based on the minimum measurement accuracy required by the target gas concentration inversion algorithm.

[0100] The initial absorption signal that meets the threshold judgment condition is determined as the gain absorption signal of the target gas.

[0101] S3. Perform second harmonic detection and lock-in amplification on the gain absorption signal to generate a pure absorption spectrum of the target gas. Analyze the characteristic parameters of the pure absorption spectrum to determine the preliminary concentration value of the target gas. Further analyze the preliminary concentration value to obtain the real-time concentration value of the target gas.

[0102] Methods for generating pure absorption lines of the target gas include:

[0103] The modulation frequency component extracted from the gain absorption signal is frequency-multiplied to generate a second harmonic reference signal. This second harmonic reference signal is used to accurately extract and amplify weak second harmonic signals. A Fourier transform is performed on the gain absorption signal to locate the modulation frequency component with the same modulation frequency as the laser in the resulting spectrum. The phase information of the modulation frequency component is extracted, and based on this phase information, a sine wave sequence with twice the frequency and synchronized phase is synthesized and determined as the second harmonic reference signal.

[0104] The gain absorption signal is filtered using a bandpass filter with a center frequency of twice the harmonic frequency to determine the second harmonic signal within it. A bandpass filter with a center frequency of twice the harmonic frequency of the reference signal is set up. The gain absorption signal is input into the bandpass filter, which filters out noise and interference at non-second harmonic frequencies, thus obtaining a preliminary second harmonic signal.

[0105] The second harmonic signal and the second harmonic reference signal are simultaneously input into a lock-in amplifier, and the output is a mixed signal containing DC components and high-frequency noise. The mixed signal is low-pass filtered to obtain the second harmonic amplitude signal. The second harmonic signal and the second harmonic reference signal are multiplied point by point (phase-sensitive detection) to obtain the mixed signal. The mixed signal is passed through a digital low-pass filter to filter out the high-frequency components generated after multiplication. The cutoff frequency of the low-pass filter is set below the second harmonic. The signal output after low-pass filtering is the second harmonic amplitude signal.

[0106] The peak position in the second harmonic amplitude signal is located. Based on the center position of the acquired target gas absorption spectrum, a linear fit is performed on the second harmonic amplitude signal to determine the fitting baseline. Within the wavelength scanning range corresponding to the second harmonic amplitude signal, the highest point of the signal amplitude is located (the highest point is the peak position of the second harmonic amplitude signal, and its corresponding wavelength is the center position of the target gas absorption spectrum). Background data points located on both sides of the peak position and far from the center in the second harmonic amplitude signal are selected. Least square regression fitting is performed on the selected background data points to generate a fitting baseline that can describe the background trend.

[0107] Subtracting the second harmonic amplitude signal from the fitted baseline yields the pure absorption spectrum of the target gas. Then, subtracting the fitted baseline amplitude at each corresponding wavelength point from the second harmonic amplitude signal covering the entire scanning wavelength range yields the pure absorption spectrum that retains only the characteristic absorption information of the target gas.

[0108] Methods for determining the preliminary concentration value of the target gas include:

[0109] Record the peak heights on the pure absorption spectrum to determine the absorption peak height parameter. On the pure absorption spectrum of the target gas, locate the global maximum point, record the amplitude of the ordinate of this maximum point, and use this amplitude as the absorption peak height parameter.

[0110] The effective optical path value is determined based on the number of reflections and the single optical path length of the Heriot-type multipass cell; when generating the gain absorption signal, the number of reflections and the single optical path length of the Heriot-type multipass cell are read (obtained by measuring the mirror spacing); the effective optical path value is obtained by multiplying the number of reflections by the single optical path length value.

[0111] The center wavelength is extracted from the pure absorption spectrum and queried in a preset spectral database to obtain the absorption line intensity coefficient of the target gas. The wavelength corresponding to the point of maximum amplitude recorded in the pure absorption spectrum is read as the center wavelength. The center wavelength value is used as the query condition to perform a matching query in the preset spectral database. The absorption line intensity coefficient of the target gas at the center wavelength is read.

[0112] It should be explained that the preset spectral database is a subset of the HITRAN database, containing the spectroscopic parameters of various gas molecules at different wavelengths.

[0113] Based on Beer-Lambert's law, the absorption peak height parameter, effective optical path value, and absorption line intensity coefficient were analyzed to obtain the preliminary concentration value of the target gas.

[0114] The initial concentration value of the target gas is obtained as follows:

[0115] ;

[0116] in This represents the initial concentration value of the target gas; The absorption peak height parameter for a pure absorption line; The Doppler half-width of the target gas absorption line (obtained from a spectral database); Indicates the target gas at standard temperature The absorption line intensity coefficient below; This is expressed as the effective optical path value of a Heriot-type multipass cell; The system response coefficient is denoted as .

[0117] Further analysis of the preliminary concentration values ​​to obtain the real-time concentration values ​​of the target gas includes the following methods:

[0118] Calculate the area enclosed by the pure absorption line and the fitted baseline to determine the absorption peak area parameter. Within the wavelength scanning range corresponding to the pure absorption line, plot the wavelength value on the x-axis and the pure absorption line amplitude on the y-axis, and use the trapezoidal integral method to calculate the area of ​​the closed region enclosed between the pure absorption line curve and the x-axis (i.e., the zero amplitude baseline); the calculated area value is the absorption peak area parameter.

[0119] A calibration curve is established based on a standard target gas of known concentration. The calibration curve includes the measured absorption peak area corresponding to each concentration. The absorption peak area parameter is input into the calibration curve to determine the area method concentration value of the target gas. Several standard target gas samples of known concentrations are measured, and the absorption peak area parameter corresponding to each concentration is recorded to form several data points (concentration - absorption peak area parameter). The least squares method is used to fit the data points into a concentration-area relationship straight line (i.e., the calibration curve). The determined absorption peak area parameter is substituted into the mathematical expression of the calibration curve for calculation, and the output concentration value is the area method concentration value of the target gas.

[0120] For example, if the absorption peak area parameter is 15.7 V·nm, and the calibration curve equation is: concentration = 0.5 × absorption peak area parameter + 0.1, then substituting the absorption peak area parameter of 15.7 V·nm into the equation, we obtain the area method concentration value = 0.5 × 15.7 + 0.1 = 7.95 ppm.

[0121] The preliminary concentration value and the area method concentration value are weighted and fused to obtain the real-time concentration value of the target gas. Weighting coefficients are assigned to the preliminary concentration value and the area method concentration value (the sum of the two weighting coefficients is 1); the area method concentration value and the preliminary concentration value are multiplied by their respective weighting coefficients, and the sum is determined as the real-time concentration value of the target gas.

[0122] It should be explained that the weighting coefficients for both the preliminary concentration value and the area method concentration value are dynamically adjusted based on their corresponding uncertainties. The uncertainty of the preliminary concentration value is quantified by calculating the standard deviation of the center wavelength of the target gas absorption spectrum across multiple measurements. The greater the fluctuation in the center wavelength, the higher the uncertainty of the preliminary concentration value, and the corresponding weighting coefficient decreases accordingly. The uncertainty of the area method concentration value is quantified by calculating the standard deviation of the residual between the pure absorption spectrum and the fitted baseline. The larger the residual, the higher the uncertainty of the area method concentration value, and its weight decreases accordingly.

[0123] S4. Based on the real-time concentration value of the target gas, determine the status and generate parameter instructions or over-limit warning signals according to the determination results.

[0124] Methods for generating parameter commands or over-limit warning signals based on judgment results include:

[0125] Preset safe concentration thresholds, including an upper limit threshold and a lower limit threshold. The upper limit threshold is the concentration limit that triggers an exceedance warning; the lower limit threshold is the lowest concentration limit that defines a safe state.

[0126] It should be explained that the specific value of the safe concentration threshold needs to be set based on the occupational exposure limit, lower explosive limit of the target gas, as well as the specific geological conditions, ventilation conditions and historical data of the mine.

[0127] The real-time concentration value of the target gas is compared with a safe concentration threshold. If the real-time concentration value of the target gas is lower than the lower threshold, it is determined that the current state is safe and a parameter instruction to maintain the current detection frequency is generated. The parameter instruction includes an instruction object (e.g., a timing manager for controlling the detection cycle) and instruction content (e.g., "maintain the current detection cycle" or "do not modify the scan interval").

[0128] If the real-time concentration of the target gas is between the upper and lower thresholds, the system is determined to be in an alert state, and a parameter command to increase the current detection frequency is generated. This command includes the target object (e.g., the timing controller) and the command content (e.g., "Change the detection cycle from T3 to T4").

[0129] If the real-time concentration of the target gas exceeds the upper limit threshold, an over-limit warning signal will be generated (for example, simultaneously triggering the on-site audible and visual alarm, updating the alarm information on the local display screen, and sending an emergency alarm signal to the remote monitoring center through the communication interface).

[0130] In this embodiment, environmental parameters are collected and the environmentally induced wavelength drift is calculated based on a pre-calibrated sample dataset. Simultaneously, mutual information entropy analysis is used to compare the reference spectrum with the measured approximate background spectrum to obtain the spectral-related wavelength drift. Then, the optimal wavelength drift compensation is generated through quality decision fusion. This effectively overcomes the interference of drastic temperature and humidity fluctuations and air pressure changes on the laser wavelength in the confined space of the mine, ensuring the stable overlap between the scanning spectrum and the gas absorption spectrum. The wavelength drift compensation determines the temperature setpoint adjustment and the drive current adjustment, dynamically superimposing and updating the current control parameters. Furthermore, the optical path enhancement effect of the Heriot-type multi-pass cell and the adaptive mirror spacing adjustment are used to actively optimize the signal strength and signal-to-noise ratio, thereby providing a stable and high-quality original signal guarantee for concentration inversion in the harsh environment of the mine.

[0131] Concentrations are calculated using the peak height method based on Beer-Lambert's law and the area method based on the calibration curve, respectively. Weighted fusion effectively complements the sensitivity of a single algorithm to wavelength jitter or baseline drift, significantly improving the accuracy and anti-interference capability of real-time concentration values. By pre-setting safety threshold ranges, the system can automatically trigger differentiated commands such as maintaining the detection frequency, increasing the detection frequency, or issuing over-limit warnings based on the different ranges in which the real-time concentration value is located, realizing intelligent switching from energy-saving monitoring in a safe state to emergency alarms and linkage control in a dangerous state.

[0132] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this invention can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0133] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only one method, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0134] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0135] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An adaptive detection method for gas concentration in confined spaces in mines based on TDLAS, characterized in that, The adaptive detection method for gas concentration in confined spaces in mines based on TDLAS includes: S1. Obtain environmental parameters and approximate background spectral signals of the enclosed space in the mine, perform composite analysis on the environmental parameters and approximate background spectral signals, output the wavelength drift compensation amount of the laser, and determine the current control parameters of the laser based on the wavelength drift compensation amount; S2. Apply the current control parameters to drive the laser output wavelength, scan the characteristic absorption spectral region of the target gas and obtain the initial absorption signal, analyze and evaluate the initial absorption signal, and determine the gain absorption signal of the target gas. S3. Perform second harmonic detection and lock-in amplification on the gain absorption signal to generate a pure absorption spectrum of the target gas. Analyze the characteristic parameters of the pure absorption spectrum to determine the preliminary concentration value of the target gas. Further analyze the preliminary concentration value to obtain the real-time concentration value of the target gas. S4. Based on the real-time concentration value of the target gas, determine the status and generate parameter instructions or over-limit warning signals according to the determination results.

2. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 1, characterized in that, The method for combined analysis of environmental parameters and approximate background spectral signals includes: After collecting the transmission spectrum signal obtained by laser scanning of the gas in the confined space of the mine, signal points with no absorption characteristics on both sides of the absorption line in the transmission spectrum signal are selected and fitted to obtain an approximate background spectrum signal; temperature, humidity and air pressure values ​​of the confined space of the mine are collected simultaneously to form environmental parameters. Obtain a sample dataset of environmental parameters and wavelength drift, perform regression fitting analysis on the sample dataset to determine the coefficients of the linear term and the cross-coupling term of the environmental parameters, and solve the environmental parameters based on the coefficients of the linear term and the cross-coupling term to determine the environmental-induced wavelength drift. Mutual information entropy analysis is performed on the reference spectral signal and the approximate background spectral signal to determine the spectral correlation wavelength shift; quality decisions are made on the spectral correlation wavelength shift and the environment-induced wavelength shift to determine the wavelength shift compensation amount of the laser.

3. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 2, characterized in that, The method for performing mutual information entropy analysis on the reference spectral signal and the approximate background spectral signal includes: The reference spectral signal and the approximate background spectral signal are processed by wavelength axis alignment and amplitude normalization; the processed approximate background spectral signal is slid on the wavelength axis with a fixed step size, and the shifted approximate background spectral signals corresponding to different sliding offsets are combined to form a displacement signal sequence. Each shifted approximate background spectral signal and the processed reference spectral signal in the displacement signal sequence are discretized to form multiple pairs of histograms with the same interval; the mutual information entropy value of each pair of histograms is calculated to obtain the mutual information entropy-offset relationship sequence. A curve is formed by smoothing the mutual information entropy-sliding offset relationship sequence, the global maximum point of the curve is located, and the sliding offset corresponding to the global maximum point is used as the spectral correlation wavelength drift.

4. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 3, characterized in that, The method for determining the current control parameters of the laser based on wavelength drift compensation includes: Based on the temperature sensitivity coefficient and current sensitivity coefficient calibrated by the laser, the wavelength drift compensation is calculated to determine the temperature setpoint adjustment and drive current adjustment. The temperature setpoint adjustment and drive current adjustment are respectively superimposed on the laser's current operating point to determine the laser's current control parameters, which include the current temperature controller setpoint and the current drive current setpoint.

5. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 4, characterized in that, The method for scanning the characteristic absorption spectral region of the target gas and acquiring the initial absorption signal includes: A sawtooth wave scanning signal is superimposed on the current driving current setting, and the laser wavelength is driven to scan the characteristic absorption spectral region of the target gas. At the same time, the transmitted light intensity sequence during the scanning process is synchronously acquired based on the data acquisition card. The transmitted light intensity sequence is analyzed to determine the precise driving current value corresponding to the absorption peak and valley values. The precise driving current value is set as the current new driving current setting value of the laser, and a high-frequency sinusoidal modulation signal is superimposed on the current new driving current setting value to complete the laser wavelength locking and modulation. The locked and modulated laser is introduced into a Heriot-type multipass cell for several reflections, and the received emitted laser is converted into a raw electrical signal. The raw electrical signal is then pre-amplified and filtered to form an initial absorption signal.

6. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 5, characterized in that, The method for analyzing and evaluating the initial absorption signal includes: Spectral analysis is performed on the initial absorption signal to extract the peak intensity and signal-to-noise ratio of the initial absorption signal; the peak intensity is compared with a preset intensity threshold; if the peak intensity is lower than the intensity threshold, the mirror spacing of the Heriot-type multi-pass cell is gradually adjusted to re-reflect the locked and modulated laser until the peak intensity of the initial absorption signal is not lower than the intensity threshold. The signal-to-noise ratio (SNR) is compared with a preset SNR threshold. If the SNR is less than the preset SNR threshold, the integration time of the data acquisition card is gradually extended for rescanning until the SNR of the initial absorbed signal is not lower than the SNR threshold. The initial absorption signal that meets the threshold judgment condition is determined as the gain absorption signal of the target gas.

7. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 6, characterized in that, The method for generating pure absorption lines of the target gas includes: The modulation frequency component extracted from the gain absorption signal is frequency multiplied to generate a second harmonic reference signal; the gain absorption signal is filtered based on a bandpass filter with a center frequency of second harmonic to determine the second harmonic signal in the gain absorption signal. The second harmonic signal and the second harmonic reference signal are simultaneously input into the lock-in amplifier, and the output is a mixed signal containing DC component and high-frequency noise; the mixed signal is low-pass filtered to obtain the second harmonic amplitude signal. The peak position in the second harmonic amplitude signal is located. Based on the center position of the target gas absorption spectrum, the second harmonic amplitude signal is linearly fitted to determine the fitting baseline. The second harmonic amplitude signal is subtracted from the fitting baseline to obtain the pure absorption spectrum of the target gas.

8. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 7, characterized in that, The method for determining the preliminary concentration value of the target gas includes: Record the peak heights of each peak on the pure absorption spectrum to determine the absorption peak height parameters; determine the effective optical path value based on the number of reflections and single optical path length of the Heriot-type multipass cell; extract the center wavelength from the pure absorption spectrum and query the preset spectral database to obtain the absorption line intensity coefficient of the target gas. Based on Beer-Lambert's law, the absorption peak height parameter, effective optical path value, and absorption line intensity coefficient were analyzed to obtain the preliminary concentration value of the target gas.

9. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 8, characterized in that, The method for further analyzing the preliminary concentration value to obtain the real-time concentration value of the target gas includes: Calculate the area enclosed by the pure absorption line and the fitted baseline to determine the absorption peak area parameters; A calibration curve is established based on a standard target gas of known concentration. The calibration curve includes the measured absorption peak area corresponding to each concentration. The absorption peak area parameter is input into the calibration curve to determine the area method concentration value of the target gas. The preliminary concentration value and the area method concentration value are weighted and fused to obtain the real-time concentration value of the target gas.

10. The adaptive detection method for gas concentration in a confined space in a mine based on TDLAS according to claim 9, characterized in that, The method for generating parameter instructions or over-limit warning signals based on the judgment result includes: A preset safe concentration threshold is set, which includes an upper threshold and a lower threshold. The real-time concentration value of the target gas is compared with the safe concentration threshold. If the real-time concentration value of the target gas is lower than the lower threshold, it is determined that the current state is safe, and a parameter instruction to maintain the current detection frequency is generated. If the real-time concentration of the target gas is between the upper and lower thresholds, it is determined that the current state is alert, and a parameter instruction to increase the current detection frequency is generated. If the real-time concentration of the target gas exceeds the upper limit threshold, an over-limit warning signal will be generated.

Images