Ultra-long distance laser multi-gas component non-contact detection system and method

Abstract

The invention relates to the technical field of laser gas detection, and discloses an ultra-long-distance laser multi-gas component non-contact detection system and method, which are used for solving the problem that the central wavelength set by a laser is not matched with the actually emitted central wavelength during gas concentration detection. The method comprises the following steps: acquiring a central wavelength influenced parameter of a laser, evaluating to obtain a central wavelength influenced index, judging whether central wavelength correction is needed or not, and if the central wavelength correction is needed, correcting the central wavelength set by the laser according to the central wavelength influenced index, constructing a detection path and obtaining a concentration estimated value. The absorption coefficient of unit concentration gas is obtained through the TDLAS mathematical inversion model, the absorption coefficient is dynamically corrected, the corrected absorption coefficient is obtained, the accuracy of gas concentration detection is effectively improved, the gas signal recognition misalignment rate is reduced, the method can be used for ultra-long distance and tunnel dust scenes, and interference of distance and dust on detection is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of laser gas detection, and more particularly to an ultra-long-distance laser multi-gas component non-contact detection system and method. Background Art

[0002] With the widespread application of laser spectroscopy technology in environmental protection, rail transit, industrial safety, and other scenarios, multi-gas telemetry systems based on tunable diode laser absorption spectroscopy (TDLAS) are becoming an important means of remote gas leak monitoring. These systems typically use a laser beam of a specific wavelength for non-contact scanning and detection of target gases. Combined with highly sensitive spectral demodulation algorithms, they enable rapid identification and concentration estimation of typical hazardous gases such as methane (CH4), carbon monoxide (CO), and carbon dioxide (CO2). This is particularly true in long-distance tunnels, underground pipeline corridors, and other structural scenarios. By constructing ultra-long-range absorption paths through multiple reflections of the optical path and combining them with multi-component inversion models, they can effectively support leak location tasks in complex spaces.

[0003] To meet the detection needs of complex environments, existing technologies typically employ three-component or multi-component parallel laser detection mechanisms, relying on multiple central wavelengths to match different gas absorption lines, and improving detection accuracy through time-division multiplexing and harmonic extraction. System architectures often incorporate strategies such as diffuse reflection to construct long paths, high-frequency modulation to extract harmonic signals, and Bayesian algorithms for spectral decoupling, resulting in continuous improvements in overall detection accuracy and anti-interference capabilities.

[0004] However, the above technology has at least the following technical problems: Existing TDLAS laser emission modules are generally based on multiple DFB lasers that independently emit laser beams of specific central wavelengths, corresponding to the characteristic absorption peaks of different gases. However, under long-term operation or complex environmental interference, the central wavelength of the actual laser output will cause a slight shift, which reduces the degree of matching between the laser beam and the absorption peak of the target gas, thereby reducing the system sensitivity and concentration inversion accuracy. Currently, most systems only complete wavelength calibration before leaving the factory, and lack adaptive central wavelength offset perception and real-time correction mechanisms during operation. This leads to problems such as inaccurate gas signal recognition, amplified component concentration decoupling errors, and increased leak location deviations in long-distance detection scenarios. The impact is particularly prominent in the detection of low-concentration gases or multi-component overlapping spectra. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an ultra-long-distance laser multi-gas component non-contact detection system and method to solve the problems existing in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions: A non-contact detection system for multiple gas components using ultra-long-distance lasers, comprising: a center wavelength offset judgment module for obtaining parameters affecting the center wavelength of each laser, the center wavelength affecting parameters including current drive fluctuation data, temperature change data, and output power fluctuation data; an evaluation of a center wavelength affected index based on the center wavelength affected parameters; and a determination of whether center wavelength correction is required based on the center wavelength affected index; a TDLAS laser emission module for obtaining the center wavelength currently set for the laser if it is determined that center wavelength correction is required, correcting the center wavelength set for the laser based on the center wavelength affected index, obtaining the center wavelength to be set for each laser after correction, setting the center wavelength value for each laser according to the center wavelength to be set after correction, and sequentially emitting center wavelength laser beams; and an ultra-long-distance optical path construction module for constructing a detection path with an equivalent absorption optical path length exceeding one kilometer by multiple diffuse reflections of the laser beam on the inner wall of the scene structure, wherein the detection path is composed of multiple path segments. And obtain the laser signal of each path segment; the signal demodulation module is used to obtain the standard absorption spectrum line data of the gas component to be measured, and construct a Bayesian inversion model based on the standard absorption spectrum line data to obtain the concentration estimate of each gas component to be measured; the temperature and pressure disturbance compensation module is used to collect the ambient temperature and pressure parameters of the laser beam in the detection path, and obtain the absorption coefficient of the unit concentration gas through the TDLAS mathematical inversion model, and use the Kalman filter algorithm to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient; the remote leakage positioning module is used to invert the concentration estimate of each gas component to be measured based on the corrected absorption coefficient, through the Beer-Lambert law, and combined with the laser signal of each path segment, to obtain a set of gas concentration estimates of each gas component to be measured in each path segment, and realize the coarse positioning of the long-distance leakage point through the change of the optical path structure to obtain the preliminary positioning area; the precise positioning module is used to call the portable infrared imaging equipment to perform local fine detection on the preliminary positioning area to obtain the precise position coordinates of the leakage point.

[0007] Preferably, the steps for obtaining the center wavelength affected index are: setting a detection time period, obtaining current drive fluctuation data within the detection time period, and obtaining a current drive fluctuation influence coefficient based on the current drive fluctuation data; obtaining temperature data of the laser working shell within the detection time period, and constructing a temperature sequence, performing a first-order decomposition on the temperature sequence to obtain a temperature fluctuation rate sequence; calculating the temperature fluctuation average value of the temperature fluctuation rate sequence, obtaining the maximum temperature fluctuation value of the temperature fluctuation rate sequence, and calculating the temperature change influence coefficient based on the temperature fluctuation average value and the maximum temperature fluctuation value; obtaining output power fluctuation data, and obtaining an output power fluctuation influence coefficient based on the output power fluctuation data; normalizing the current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient to obtain the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient, and calculating the center wavelength affected index based on the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient. The specific obtaining steps are: Where, Expressed as the central wavelength affected index, Expressed as the normalized current drive fluctuation influence coefficient, Expressed as the normalized temperature change influence coefficient, Expressed as the output power fluctuation influence coefficient after normalization, It is expressed as the weight coefficient of the normalized current drive fluctuation influence coefficient, the weight coefficient of the normalized temperature change influence coefficient, and the weight coefficient of the normalized output power fluctuation influence coefficient.

[0008] Preferably, the steps for obtaining the current drive fluctuation influence coefficient are: obtaining the driving current value of the laser in the detection time period under a stable working state through experiments, setting uniform sampling points, and constructing a reference driving current sequence; collecting the real-time driving current value in the current detection time period, and aligning it with the sampling time point of the reference driving current sequence to construct a current driving current sequence; calculating the relative current offset of the reference driving current sequence and the current driving current sequence point by point, and constructing a current offset sequence; calculating the current offset average value according to the current offset sequence, and performing difference calculation on the maximum value and the minimum value in the current offset sequence to obtain the fluctuation amplitude; using fast Fourier transform to perform frequency domain analysis on the current offset sequence, setting a high frequency threshold, obtaining the total energy of the frequency segment higher than the high frequency threshold, and performing ratio calculation with the total energy at all frequencies to obtain the high frequency energy ratio; taking the absolute value of the current offset average value and summing it with the fluctuation amplitude and the high frequency energy ratio to obtain the current drive fluctuation influence coefficient.

[0009] Preferably, the steps for obtaining the output power fluctuation influence coefficient are as follows: within the detection time period, real-time output power values are collected, and an output power sequence is constructed, a sliding window length is set, and the sequence is divided into several sliding windows of equal length according to the sliding window length to obtain several power observation subsequences; for each power observation subsequence, the second-order derivative change value is calculated by fitting the residual sequence, which is recorded as the degree of nonlinear drift; for the nonlinear drift degrees extracted from all sliding windows, the maximum value of the drift degree and the mean value of the drift degree are statistically calculated, and the offset amplitude coefficient is calculated based on the maximum value of the drift degree and the mean value of the drift degree; the standard deviation of the output power sequence is calculated, and the output power fluctuation influence coefficient is calculated based on the offset amplitude coefficient and the standard deviation of the output power sequence.

[0010] Preferably, the step of determining whether center wavelength correction is required based on the center wavelength affected index is as follows: comparing the center wavelength affected index with an affected threshold; if the center wavelength affected index is greater than or equal to the affected threshold, determining that the center wavelength emitted by the current laser has shifted and center wavelength correction is required; if the center wavelength affected index is less than the affected threshold, determining that the center wavelength emitted by the current laser has not shifted and center wavelength correction is not required.

[0011] Preferably, the step of correcting the center wavelength set by the laser according to the center wavelength affected index to obtain the center wavelength that should be set for each laser after correction is as follows: obtaining the standard center wavelength of each gas component to be measured, and recording the center wavelength currently set by the laser as the system set wavelength value; collecting the center wavelength actually emitted by the current laser through a wavelength monitor, recording it as the actual emitted center wavelength, and comparing the actual emitted center wavelength with the standard center wavelength; if the actual emitted center wavelength is greater than the standard center wavelength, it is determined that the currently emitted center wavelength is too long, and the affected threshold value is ratio-calculated with the center wavelength affected index to obtain a correction factor; multiplying the system set wavelength value and the correction factor to obtain the center wavelength that should be set after correction; if the actual emitted center wavelength is less than the standard center wavelength, it is determined that the currently emitted center wavelength is too short, and the system set wavelength value is divided by the correction factor to obtain the center wavelength that should be set after correction.

[0012] Preferably, the steps for constructing the detection path with an equivalent absorption light path length exceeding one kilometer are as follows: obtaining the structural boundary information of the scene to be measured, the boundary information including the geometric dimensions of the scene, the diffuse reflectivity of the inner wall material and the main plane position coordinates, and setting the initial emission position and incident angle of the laser beam according to the structural boundary information; using the geometric optical path tracing method, with the set incident angle and initial emission position as the starting point, combined with the structural boundary information, calculating the multiple diffuse reflection paths of the laser beam in the scene, determining the spatial position, reflection angle and propagation direction of each reflection point, and obtaining the detection path with an equivalent absorption light path length exceeding one kilometer.

[0013] Preferably, the steps of achieving coarse positioning of a long-distance leak point by changing the optical path structure and obtaining a preliminary positioning area are as follows: constructing a concentration space mapping matrix based on the path segments based on the set of gas concentration estimates of each gas component to be measured in each path segment and the spatial position of the path segment; performing a three-dimensional gradient analysis on the concentration space mapping matrix to obtain the concentration change between adjacent spatial grids, and obtaining a gas concentration gradient distribution by a central difference method based on the concentration change between adjacent spatial grids; setting a concentration threshold, and screening continuous grid areas with concentration values higher than the concentration threshold and with the same concentration gradient direction based on the gas concentration gradient distribution, and recording them as candidate leakage areas; for each candidate leakage area, counting the number of path segments passing through or covering the area, and recording them as the number of supporting paths; at the same time, calling the three-dimensional gradient distribution results in the concentration space mapping matrix, extracting the concentration gradient vector modulus of each grid in the area, and calculating their maximum and average values, which are recorded as the concentration gradient intensity of the area; weightedly fusing the number of supporting paths and the concentration gradient intensity according to a preset weight to construct a coarse positioning confidence score value, setting a coarse positioning confidence threshold, and marking candidate areas with scores greater than or equal to the threshold as preliminary positioning areas.

[0014] Preferably, a non-contact detection method for multiple gas components using ultra-long-distance lasers comprises the following steps: Step 1: obtaining the affected parameters of the center wavelength of each laser, the affected parameters of the center wavelength including current drive fluctuation data, temperature change data and output power fluctuation data, evaluating the affected parameters of the center wavelength to obtain the affected index of the center wavelength, and determining whether center wavelength correction is required based on the affected index of the center wavelength; Step 2: if it is determined that center wavelength correction is required, obtaining the center wavelength of the current laser setting, correcting the center wavelength of the laser setting based on the affected index of the center wavelength, obtaining the center wavelength that should be set for each laser after correction, setting the center wavelength value of each laser according to the center wavelength that should be set after correction, and emitting center wavelength laser beams in sequence; Step 3: constructing a detection path with an equivalent absorption light path length exceeding one kilometer by multiple diffuse reflections of the laser beam on the inner wall of the scene structure, and the detection path is composed of multiple path segments. , and obtain the laser signal of each path segment; Step 4: Obtain the standard absorption spectrum line data of the gas component to be measured, and construct a Bayesian inversion model based on the standard absorption spectrum line data to obtain the concentration estimate of each gas component to be measured; Step 5: Collect the ambient temperature and pressure parameters of the laser beam in the detection path, and obtain the absorption coefficient of the unit concentration gas through the TDLAS mathematical inversion model; Use the Kalman filter algorithm to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient; Step 6: According to the corrected absorption coefficient, through the Beer-Lambert law and combined with the laser signal of each path segment, the concentration estimate of each gas component to be measured is inverted to obtain a set of gas concentration estimates of each gas component to be measured in each path segment, and the coarse positioning of the long-distance leakage point is achieved through the change of the optical path structure to obtain the preliminary positioning area; Step 7: Call the portable infrared imaging equipment to perform local fine detection on the preliminary positioning area to obtain the precise position coordinates of the leakage point.

[0015] The technical effects and advantages of the present invention are as follows: The affected parameters of the central wavelength of each laser are obtained, the affected index of the central wavelength is evaluated, and it is determined whether central wavelength correction is required. If it is determined that central wavelength correction is required, the central wavelength set by the laser is corrected according to the affected index of the central wavelength, and a detection path is constructed to obtain a concentration estimate. The absorption coefficient of the unit concentration gas is obtained through the TDLAS mathematical inversion model, and the absorption coefficient is dynamically corrected to obtain the corrected absorption coefficient, which effectively improves the accuracy of gas concentration detection and reduces the inaccuracy rate of gas signal recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a structural diagram of an ultra-long-distance laser multi-gas component non-contact detection system provided in an embodiment of the present application.

[0017] Figure 2 A flow chart of an ultra-long-distance laser non-contact detection method for multiple gas components provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The technical solutions of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. In addition, the forms of the various structures described in the following embodiments are merely examples. The ultra-long-distance laser multi-gas component non-contact detection system and method involved in the present invention are not limited to the various structures described in the following embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] The present invention provides an ultra-long-distance laser multi-gas component non-contact detection system, such as Figure 1 As shown, the system includes: The center wavelength offset judgment module is used to obtain the parameters affected by the center wavelength of each laser. The parameters affected by the center wavelength include current drive fluctuation data, temperature change data, and output power fluctuation data. The center wavelength affected index is obtained based on the center wavelength affected parameters. The center wavelength affected index is used to determine whether center wavelength correction is required. It should be noted that one laser is responsible for one center wavelength. In this embodiment, it should be specifically explained that the steps for obtaining the center wavelength affected index are: Set a detection time period, obtain current drive fluctuation data within the detection time period, and evaluate the current drive fluctuation influence coefficient based on the current drive fluctuation data. It should be noted that the detection time period can be adjusted according to actual conditions. For example, the detection time period can be 5 minutes or 1 minute; Obtain the temperature data of the laser working housing during the detection period, construct a temperature sequence, and perform first-order decomposition on the temperature sequence to obtain a temperature fluctuation rate sequence. The larger the value in the temperature fluctuation rate sequence, the stronger the temperature fluctuation. Calculate the temperature fluctuation average value of the temperature fluctuation rate sequence, obtain the maximum temperature fluctuation value of the temperature fluctuation rate sequence, and calculate the temperature change influence coefficient based on the temperature fluctuation average value and the maximum temperature fluctuation value. The specific acquisition steps are as follows: ; Where, Expressed as the temperature change influence coefficient, Expressed as the maximum temperature fluctuation value, It is expressed as the average value of temperature fluctuation. The formula processes the maximum temperature change rate through a logarithmic function to enhance the recognition sensitivity of small temperature disturbances and suppress the influence of extreme values. At the same time, the average value of temperature fluctuation is introduced to measure the overall temperature control stability. The product of the two constitutes a nonlinear amplification mechanism, so that when instantaneous disturbances and global fluctuations exist at the same time, the influence coefficient is significantly improved, thereby more accurately reflecting the combined interference effect of temperature fluctuations on the stability of the laser center wavelength. The addition of 1 in the formula is to avoid the special case where the average value of temperature fluctuation and the maximum temperature fluctuation value are 0, so as to avoid the situation where the formula is meaningless. Obtain output power fluctuation data, and evaluate the output power fluctuation influence coefficient based on the output power fluctuation data; The current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient are normalized to obtain the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient. The center wavelength influence index is calculated based on the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient. The specific acquisition steps are as follows: ; Where, Expressed as the central wavelength affected index, It is expressed as the normalized current drive fluctuation influence coefficient, which is used to characterize the fluctuation amplitude of the actual laser drive current relative to the target set current. Since the laser output wavelength is highly sensitive to the drive current, when the drive current has nonlinear perturbations or short-term abnormal jumps, it will cause a non-negligible small shift in the laser center wavelength. Therefore, the larger the current drive fluctuation influence coefficient, the more significant the drive current fluctuation, the stronger the perturbation effect on the center wavelength, and the higher the center wavelength influence index. It is expressed as the normalized temperature variation coefficient, which is used to measure the severity of temperature changes and the intensity of rate fluctuations during laser operation. Since the center emission wavelength of the laser has a thermal tuning characteristic to temperature changes (generally about 0.1 nm / °C), when the temperature fluctuates greatly, it will directly cause the emission wavelength to drift, resulting in a decrease in the matching degree with the target gas absorption line. Therefore, the larger the temperature variation coefficient, the more severe the temperature disturbance, the more obvious the impact on the stability of the center wavelength, and the larger the center wavelength influence index. It is expressed as the normalized output power fluctuation influence coefficient, which is used to measure the dynamic stability of the laser output optical power. When the laser power fluctuates, it is usually accompanied by changes in the carrier density distribution in the laser cavity and local thermal field disturbances in the chip, which indirectly affect the effective refractive index and resonance conditions of the laser, causing the center wavelength to shift. Especially in medium and long-distance detection scenarios, small wavelength changes will be amplified into differences in absorption intensity. Therefore, the larger the output power fluctuation influence coefficient, the more significant the power instability, and the more likely the center wavelength is to be shifted by its disturbance. It is expressed as the weight coefficient of the normalized current drive fluctuation influence coefficient, the weight coefficient of the normalized temperature change influence coefficient, and the weight coefficient of the normalized output power fluctuation influence coefficient, and , It can be obtained by analytic hierarchy process, such as The values can be 0.4, 0.2, or 0.4. The Analytic Hierarchy Process (AHP) is a decision analysis method used to determine the relative importance of multiple factors. It constructs a judgment matrix to compare each factor pairwise and calculates the weight coefficient for each factor using eigenvalue analysis. In this invention, the AHP is used to assess the relative impact of current drive fluctuations, temperature variations, and output power fluctuations on the stability of the laser's central wavelength. Expert or empirical data can be used to construct a judgment matrix. After a consistency check, normalized weights are obtained, which serve as the source of the weight coefficients for each factor influencing the central wavelength impact index, thereby enhancing the objectivity and rationality of the impact assessment.

[0020] In this embodiment, it should be specifically explained that the steps for obtaining the current driving fluctuation influence coefficient are: Through experiments, the driving current value of the laser in the detection period under stable working state is obtained, uniform sampling points are set, and a benchmark driving current sequence is constructed; Collect the real-time driving current value within the current detection time period and align it with the sampling time point of the reference driving current sequence to construct the current driving current sequence; Calculating the relative current offset of the reference drive current sequence and the current drive current sequence point by point, and constructing a current offset sequence; The current offset average value is calculated based on the current offset sequence, and the difference between the maximum and minimum values in the current offset sequence is calculated to obtain the fluctuation amplitude, which reflects the fluctuation range of the current offset; Use fast Fourier transform to perform frequency domain analysis on the current offset sequence, set a high-frequency threshold, obtain the total energy of the frequency segment above the high-frequency threshold, and calculate the ratio with the total energy at all frequencies to obtain the high-frequency energy ratio; It should be specifically noted that the use of fast Fourier transform to perform frequency domain analysis on the current offset sequence is a prior art, and this embodiment does not provide a detailed description of the specific steps.

[0021] The absolute value of the current offset average is taken and then summed with the fluctuation amplitude and high-frequency energy ratio to obtain the current drive fluctuation influence coefficient.

[0022] In this embodiment, it should be specifically explained that the steps for obtaining the output power fluctuation influence coefficient are: During the detection period, real-time output power values are collected and an output power sequence is constructed. The sliding window length is set, for example, 5 seconds. The sequence is divided into several sliding windows of equal length based on the sliding window length to obtain several power observation subsequences. Each sliding window represents an independent time segment, which is used to compare local power change behavior to ensure that the analysis covers the entire working cycle. For each power observation subsequence, the second-order derivative change value is calculated by fitting the residual sequence, which is recorded as the degree of nonlinear drift. The greater the drift degree, the stronger the interference to the laser output power. For the nonlinear drift degrees extracted from all sliding windows, the maximum drift degree and the mean drift degree are statistically calculated, and the offset amplitude coefficient is calculated based on the maximum drift degree and the mean drift degree. The specific acquisition steps are as follows: ; Where, Expressed as the offset amplitude coefficient, Expressed as the maximum drift degree, Expressed as the mean value of the drift degree; Calculate the standard deviation of the output power sequence. The output power fluctuation influence coefficient is calculated based on the offset amplitude coefficient and the standard deviation of the output power sequence. The specific steps for obtaining the coefficient are as follows: ; Where, Expressed as the output power fluctuation influence coefficient, Expressed as the offset amplitude coefficient, Expressed as the standard deviation of the output power series, it combines the overall dispersion of power fluctuations (measured by the standard deviation) with the relative offset, and uses a logarithmic function to enhance the resolution of small fluctuations and suppress the influence of extreme values. This makes the final output power fluctuation influence coefficient more robust and interpretable, and is used to quantify the actual impact of output power fluctuations on the stability of the central wavelength.

[0023] In this embodiment, it should be specifically explained that the steps of determining whether center wavelength correction is required according to the center wavelength impact index are: The center wavelength affected index is compared with the affected threshold. If the center wavelength affected index is greater than or equal to the affected threshold, it is determined that the center wavelength emitted by the current laser has shifted and center wavelength correction is required; if the center wavelength affected index is less than the affected threshold, it is determined that the center wavelength emitted by the current laser has not shifted and center wavelength correction is not required. The affected threshold is obtained by the adaptive threshold method. The adaptive threshold method refers to a parameter update method that automatically adjusts the judgment threshold based on the dynamic change characteristics of historical data during system operation. In the present invention, the adaptive threshold method is used to calculate the judgment threshold that best distinguishes "normal fluctuations" from "actual offsets" in real time based on the statistical distribution characteristics of the center wavelength affected index under different environmental conditions, laser states, and working hours. This method is usually based on the mean and standard deviation of the affected index within a sliding time window, combined with the dynamic offset amplitude change trend, to set a judgment boundary that can self-adjust with time and state changes, thereby improving the sensitivity and robustness of wavelength offset identification.

[0024] If the TDLAS laser emission module determines that center wavelength correction is required, the center wavelength of the current laser setting is obtained, and the center wavelength of the laser setting is corrected according to the center wavelength affected index to obtain the center wavelength that each laser should be set to after correction. The center wavelengths are matched with the characteristic absorption bands of the gas components to be measured (CH4, CO2 and CO). The center wavelength value of each laser is set according to the center wavelength that should be set after correction, and the center wavelength laser beam is emitted in sequence, and its echo signal is collected to ensure that the echo information of the gas component corresponding to each laser is complete and distinguishable; It should be specifically noted that although laser beams of different central wavelengths share the same or similar propagation path in structural design, since gas components such as CH4, CO2, and CO have their own specific absorption spectra in the infrared band, they need to be detected using laser wavelengths that match their absorption characteristics, thereby realizing the identification and concentration inversion of multiple gas components.

[0025] By dynamically correcting the laser's set center wavelength, the system ensures that it can accurately emit a laser beam that matches the characteristic absorption band of each gas component under test, even under the influence of environmental disturbances (such as current fluctuations, temperature changes, and power drift). This effectively prevents gas absorption signal distortion or decreased sensitivity caused by deviations from the set value due to the actual output wavelength, improves the system's ability to recognize weak absorption features, and enhances stability and detection accuracy under complex operating conditions.

[0026] In this embodiment, it should be specifically explained that the steps of correcting the central wavelength set for the laser according to the central wavelength affected index to obtain the central wavelength to be set for each laser after correction are as follows: Obtain the standard center wavelength of each gas component to be measured, which represents the theoretical characteristic absorption wavelength of the gas. Record the center wavelength currently set by the laser as the system setting wavelength value. It should be noted that in this embodiment, the initial system setting wavelength value is the standard center wavelength. The wavelength monitor is used to collect the actual center wavelength emitted by the current laser, which is recorded as the actual center wavelength. The actual center wavelength is compared with the standard center wavelength. If the actual center wavelength is greater than the standard center wavelength, it is determined that the current center wavelength is too long. The affected threshold value is then compared with the affected index of the center wavelength to obtain the correction factor. The specific acquisition steps are as follows: ; Where, Expressed as a correction factor, Expressed as the affected threshold, Expressed as the central wavelength affected index; Multiply the system set wavelength value and the correction factor to obtain the center wavelength that should be set after correction; If the actual emitted center wavelength is shorter than the standard center wavelength, it is determined that the currently emitted center wavelength is too short. The system set wavelength value is divided by the correction factor to obtain the center wavelength that should be set after correction.

[0027] The ultra-long-distance optical path construction module is used to construct a detection path with an equivalent absorption optical path length exceeding one kilometer through multiple diffuse reflections of a laser beam on the inner walls of scene structures (such as tunnels and pipe corridors). The detection path refers to the equivalent absorption path formed by the laser beam in the scene structure. The detection path is composed of multiple path segments and the laser signal of each path segment is obtained. The laser signal includes the incident path, reflection angle, and signal intensity variation. This module can complete the construction of a high equivalent optical path in passive scenes without the need for a mirror reflection device, significantly enhancing the strength of the gas absorption signal, which is the key to achieving ultra-long-distance detection. It's important to note that the core purpose of the ultra-long-distance optical path building module is to create a stable, repeatable long-distance propagation path within the detection scenario. This path relies on the multiple diffuse reflections of laser light within a spatial structure, rather than the specific central wavelength of the laser beam. The various central wavelength laser beams used in the system exhibit similar optical behavior in spatial propagation and reflection paths, and can all be used to form equivalent absorption optical paths. Therefore, to avoid repetition and unnecessary clarification, the term "through a laser beam" can be used rather than "laser beams of different central wavelengths."

[0028] In this embodiment, it should be specifically explained that the steps for constructing a detection path with an equivalent absorption light path length exceeding one kilometer are as follows: Obtain structural boundary information for the scene to be measured (e.g., a tunnel or pipe gallery). This boundary information includes the scene's geometric dimensions (length, height, width), the diffuse reflectivity of the inner wall material, and the coordinates of the main planes. This boundary information can be pre-modeled using construction drawings to provide a basic spatial framework for subsequent laser propagation path simulation and reflection behavior modeling. Based on the acquired structural boundary information, the initial launch position and incident angle of the laser beam are set. The incident angle must meet the geometric condition of "enabling multiple consecutive wall reflections" to prevent the laser beam from leaving the enclosed space during propagation or causing excessive reflection energy loss. Using geometric optical path tracing, starting from a set incident angle and initial emission position, combined with structural boundary information, the laser beam's multiple diffuse reflection paths within the scene are calculated, determining the spatial position, reflection angle, and propagation direction of each reflection point. This path simulation results in an "equivalent absorption light path" consisting of several continuous path segments. The total propagation length is the optical path length, and the incident path, reflection angle, and length information of each path segment are recorded. Geometric optics path tracing is a method for calculating light propagation paths based on the principles of geometric optics. It simulates the propagation of light within a specific boundary structure. This method treats the laser beam as an ideal ray with a propagation direction. It calculates the reflection direction based on the incident point and the surface normal vector, and iteratively updates the reflection path each time it encounters a boundary until the specified conditions are met.

[0029] After the reflection path is constructed, the energy attenuation of each path segment is evaluated in combination with the diffuse reflectivity and scattering loss model of the wall material, and the actual geometric propagation length corresponding to the path is obtained by cumulative summing based on the recording of its spatial geometric parameters. If the cumulative length of the actual propagation path does not meet the set requirements (such as the equivalent absorption light path length exceeds one kilometer), the incident angle of the laser beam is adjusted to reconstruct the reflection path. Ultimately, a complete sequence of path segments that meets the optical path requirements is returned. The path segment sequence constitutes a detection path with an equivalent absorption light path length exceeding one kilometer, providing a spatial reference for subsequent multi-path signal acquisition and gas concentration distribution estimation. It should be specifically noted that the energy attenuation evaluation of each path segment in combination with the diffuse reflectivity and scattering loss model of the wall material is a prior art, and this embodiment does not provide a detailed description of its specific process.

[0030] It should be further pointed out that in high-dust environments such as railway tunnels, suspended particles can interfere with laser signals by scattering, blocking, or attenuating them. Traditional single-path detection methods are prone to increased concentration inversion errors due to localized blockage. The multi-path diffuse reflection equivalent long path constructed by this invention enables redundant acquisition of laser signals within a spatial range, preventing the impact of single-path failure on overall detection accuracy. Furthermore, multiple reflections along the path segments help average out local intensity fluctuations caused by dust distribution, thereby improving the system's optical path stability and concentration estimation accuracy in high-dust conditions and demonstrating excellent resistance to dust interference.

[0031] The signal demodulation module is used to obtain the standard absorption spectrum data of the gas components to be measured and build a Bayesian inversion model based on the standard absorption spectrum data. The Bayesian inversion model is used to decouple the multiple gas absorption components present in the received signal to obtain the concentration estimate of each gas component to be measured, thereby achieving effective separation of multi-component mixed absorption signals; The Bayesian inversion model is a parameter estimation method based on Bayesian statistical theory. It combines observation data with prior information to construct a posterior probability distribution, thereby performing inverse inference on unknown physical parameters.

[0032] The temperature and pressure disturbance compensation module is used to collect the ambient temperature and pressure parameters of the laser beam in the detection path, and input the ambient temperature and pressure parameters into the TDLAS mathematical inversion model to calculate the absorption coefficient of the unit concentration gas; The Kalman filter algorithm is used to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient, thereby eliminating the fitting deviation caused by temperature and pressure disturbances, thereby improving the stability and accuracy of the gas concentration estimation results; The TDLAS mathematical inversion model is an absorption spectroscopy analysis method based on the Beer-Lambert law. It uses the absorption characteristics of laser light at a specific central wavelength as it passes through a gas-containing sample to infer the concentration of the gas being measured. The model uses the incident laser intensity, transmission intensity, path length, and gas absorption coefficient as input variables to calculate the theoretical absorption value per unit gas concentration and, based on this, quantitatively analyze the echo signal.

[0033] The Kalman filter algorithm is a recursive state estimation method based on the minimum mean square error principle, which can dynamically optimize measurement parameters in the presence of noise and interference.

[0034] The remote leak location module is used to invert the concentration estimate of each gas component to be measured based on the corrected absorption coefficient, the Beer-Lambert law, and the laser signal of each path segment. This module obtains a set of gas concentration estimates for each gas component to be measured in each path segment, and achieves coarse positioning of the remote leak point by changing the optical path structure to obtain a preliminary positioning area. The Beer-Lambert law is a fundamental physical law that describes the absorption of light as it propagates through a homogeneous medium. Its core principle is that as light passes through a sample containing a gas, its intensity decays exponentially due to absorption by the gas molecules. The law states that the intensity of absorbed light is proportional to the gas concentration, the absorption coefficient, and the optical path length.

[0035] It's important to note that while both the temperature and pressure disturbance compensation module and the remote leak location module perform concentration inversion calculations based on the Beer-Lambert law, they perform different functions within the system. The temperature and pressure disturbance compensation module focuses on correcting for absorption coefficient perturbations caused by temperature and pressure variations before concentration inversion, improving the stability and accuracy of single-path concentration calculations. The remote leak location module, on the other hand, estimates the location of gas leaks based on concentration inversion results from multiple paths, combining the laser propagation path and reflection structure, achieving coarse spatial localization. Therefore, the two modules function as a front-to-back processing mechanism, complementing each other's functions and ensuring a non-duplicative design.

[0036] In this embodiment, it should be specifically explained that the steps for obtaining a set of gas concentration estimation values of each gas component to be measured in each path segment are: The ultra-long-distance optical path construction module is called to obtain the coordinates of the starting and ending points of each path segment and the corresponding path segment length. The signal demodulation module is called to obtain the echo signal of each central wavelength laser beam in each path segment. The echo signal includes the laser incident intensity at the starting point of the path segment and the transmission intensity at the end point. All echo signals are paired with the path segment position according to the timestamp to form a multi-wavelength, multi-path segment signal data set. For each path segment, the temperature and pressure disturbance compensation module is called to obtain the corrected absorption coefficient under the environmental conditions of the current path segment. The corrected absorption coefficient represents the absorption capacity of the gas to the laser at a unit concentration under the current path conditions. For each gas component to be measured, the echo signal of the corresponding central wavelength is called, and the Beer-Lambert law is applied in each path segment to calculate the estimated gas concentration of the component in the current path segment. The specific acquisition steps are as follows: ; Where, It is expressed as the estimated gas concentration of the component in the current path segment. Expressed as the corrected absorption coefficient, Expressed as path segment length, Expressed as the laser incident intensity, Expressed as the transmittance intensity at the end point; The above inversion calculation results are sorted out separately to obtain a set of concentration estimation values of all gas components to be measured in each path segment.

[0037] In this embodiment, it should be specifically explained that the steps of achieving coarse positioning of a long-distance leakage point by changing the optical path structure and obtaining a preliminary positioning area are as follows: Based on the set of gas concentration estimates for each gas component to be measured in each path segment, combined with the coordinates of the path segment's start and end points and spatial position, a spatial discrete grid model is constructed. The concentration estimates for each path segment are projected onto the corresponding spatial grid cells, resulting in a path segment-based concentration spatial mapping matrix. The concentration values are cumulatively recorded as voxels within each grid cell, reflecting the gas distribution trend within the spatial region. Specifically, the spatial position is obtained by invoking the ultra-long-distance optical path construction module using geometric optical path tracing. Perform three-dimensional gradient analysis on the concentration space mapping matrix to obtain the concentration variation between adjacent spatial grids. Based on the concentration variation between adjacent spatial grids, the gas concentration gradient distribution is obtained by the central difference method. The gradient direction points to the location where the gas concentration increases fastest. Set a concentration threshold and, based on the gas concentration gradient distribution, screen out continuous grid areas with concentration values higher than the concentration threshold and with the same concentration gradient direction. These areas are recorded as candidate leakage areas and their spatial coordinate ranges are marked as the candidate area set for coarse leakage location. For each candidate leakage area, the number of path segments that pass through or cover the area is counted, recorded as the number of supported paths, which is used to characterize the path support. At the same time, the three-dimensional gradient distribution results in the concentration space mapping matrix are called to extract the concentration gradient vector modulus of each grid in the area, and its maximum and average values are calculated. These are recorded as the concentration gradient intensity of the area, which is used to characterize the local concentration mutation characteristics. The number of supporting paths and the concentration gradient intensity are weightedly fused according to preset weights to construct a coarse position confidence score value, and a coarse position confidence threshold is set. The candidate areas with scores greater than or equal to the threshold are marked as preliminary leakage positioning areas; the spatial coordinate range of the preliminary leakage positioning area is output as the coarse positioning result to form a "preliminary positioning area". It should be noted that the weight coefficients used for weighted fusion are obtained through the hierarchical analysis method.

[0038] The precise positioning module is used to call the portable infrared imaging equipment to perform local fine detection on the preliminary positioning area, obtain the local concentration distribution gradient and the concentration change slope, screen out the point with the maximum concentration change slope, and obtain the precise location coordinates of the leakage point based on the local concentration distribution gradient and the point with the maximum concentration change slope.

[0039] It should be specifically noted that obtaining the precise position coordinates of the leakage point based on the local concentration distribution gradient and the point with the maximum concentration change slope is an existing technology, and this embodiment does not provide a detailed description of the specific process.

[0040] In this embodiment, it is necessary to specifically explain that a non-contact detection method of multiple gas components by ultra-long-distance laser is provided. Figure 2 As shown, the following steps are included: Step 1: Obtain the parameters affected by the center wavelength of each laser. The parameters affected by the center wavelength include current drive fluctuation data, temperature change data, and output power fluctuation data. The center wavelength affected index is obtained based on the center wavelength affected parameters. The center wavelength affected index is used to determine whether center wavelength correction is required. Step 2: If it is determined that center wavelength correction is required, the center wavelength of the current laser setting is obtained, and the center wavelength of the laser setting is corrected according to the center wavelength affected index to obtain the center wavelength that should be set for each laser after correction. The center wavelength value of each laser is set according to the center wavelength that should be set after correction, and the center wavelength laser beams are emitted in sequence; Step 3: By multiple diffuse reflections of the laser beam on the inner wall of the scene structure, a detection path with an equivalent absorption optical path length of more than one kilometer is constructed. The detection path consists of multiple path segments, and the laser signal of each path segment is obtained; Step 4: Obtain the standard absorption spectrum data of the gas components to be measured, and build a Bayesian inversion model based on the standard absorption spectrum data to obtain the concentration estimate of each gas component to be measured; Step 5: Collect the ambient temperature and pressure parameters of the laser beam in the detection path, and obtain the absorption coefficient of the unit concentration gas through the TDLAS mathematical inversion model; use the Kalman filter algorithm to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient; Step 6: Based on the corrected absorption coefficient, the Beer-Lambert law is used, and combined with the laser signal of each path segment, the concentration estimate of each gas component to be measured is inverted to obtain a set of gas concentration estimates for each gas component to be measured in each path segment. The long-distance leak point is roughly located by changing the optical path structure, and the preliminary positioning area is obtained; Step 7: Use portable infrared imaging equipment to perform local fine detection on the preliminary positioning area to obtain the precise location coordinates of the leak point.

[0041] Finally: 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 in the scope of protection of the present invention.

[0042] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

Claims

1. An ultra-long-distance laser multi-gas component non-contact detection system, characterized in that: The system comprises: The center wavelength offset judgment module is used to obtain the parameters affected by the center wavelength of each laser. The parameters affected by the center wavelength include current drive fluctuation data, temperature change data, and output power fluctuation data. The center wavelength affected index is obtained based on the center wavelength affected parameters. The center wavelength affected index is used to determine whether center wavelength correction is required. If the TDLAS laser emission module determines that center wavelength correction is required, it obtains the center wavelength of the current laser setting, corrects the center wavelength of the laser setting according to the center wavelength affected index, obtains the center wavelength that each laser should be set to after correction, sets the center wavelength value of each laser according to the center wavelength that should be set after correction, and emits center wavelength laser beams in sequence; The ultra-long-distance optical path construction module is used to construct a detection path with an equivalent absorption optical path length of more than one kilometer through multiple diffuse reflections of the laser beam on the inner wall of the scene structure. The detection path is composed of multiple path segments and the laser signal of each path segment is obtained; The signal demodulation module is used to obtain the standard absorption spectrum data of the gas components to be measured, and to construct a Bayesian inversion model based on the standard absorption spectrum data to obtain the concentration estimate of each gas component to be measured; The temperature and pressure disturbance compensation module is used to collect the ambient temperature and pressure parameters of the laser beam in the detection path, and obtain the absorption coefficient of the unit concentration gas through the TDLAS mathematical inversion model. The Kalman filter algorithm is used to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient; The remote leak location module is used to invert the concentration estimate of each gas component to be measured based on the corrected absorption coefficient, the Beer-Lambert law, and the laser signal of each path segment. This module obtains a set of gas concentration estimates for each gas component to be measured in each path segment, and achieves coarse positioning of the remote leak point by changing the optical path structure to obtain a preliminary positioning area. The precise positioning module is used to call the portable infrared imaging equipment to perform local fine detection on the preliminary positioning area to obtain the precise location coordinates of the leakage point.

2. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 1, characterized in that: The steps for obtaining the center wavelength affected index are: Setting a detection time period, obtaining current drive fluctuation data within the detection time period, and evaluating the current drive fluctuation influence coefficient based on the current drive fluctuation data; Obtain the temperature data of the laser working shell during the detection period, construct a temperature sequence, perform first-order decomposition on the temperature sequence, and obtain a temperature fluctuation rate sequence; Calculate the temperature fluctuation average value of the temperature fluctuation rate sequence, obtain the maximum temperature fluctuation value of the temperature fluctuation rate sequence, and calculate the temperature change influence coefficient based on the temperature fluctuation average value and the maximum temperature fluctuation value; Obtain output power fluctuation data, and evaluate the output power fluctuation influence coefficient based on the output power fluctuation data; The current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient are normalized to obtain the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient. The center wavelength influence index is calculated based on the normalized current drive fluctuation influence coefficient, the temperature change influence coefficient, and the output power fluctuation influence coefficient. The specific acquisition steps are as follows: ； Where, Expressed as the central wavelength affected index, Expressed as the normalized current drive fluctuation influence coefficient, Expressed as the normalized temperature change influence coefficient, Expressed as the output power fluctuation influence coefficient after normalization, It is expressed as the weight coefficient of the normalized current drive fluctuation influence coefficient, the weight coefficient of the normalized temperature change influence coefficient, and the weight coefficient of the normalized output power fluctuation influence coefficient.

3. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 2, characterized in that: The steps for obtaining the current drive fluctuation influence coefficient are: Through experiments, the driving current value of the laser in the detection period under stable working state is obtained, uniform sampling points are set, and a benchmark driving current sequence is constructed; Collect the real-time driving current value within the current detection time period and align it with the sampling time point of the reference driving current sequence to construct the current driving current sequence; Calculating the relative current offset of the reference drive current sequence and the current drive current sequence point by point, and constructing a current offset sequence; Calculate the current offset average value according to the current offset sequence, and calculate the difference between the maximum and minimum values in the current offset sequence to obtain the fluctuation amplitude; Use fast Fourier transform to perform frequency domain analysis on the current offset sequence, set a high-frequency threshold, obtain the total energy of the frequency segment above the high-frequency threshold, and calculate the ratio with the total energy at all frequencies to obtain the high-frequency energy ratio; The absolute value of the current offset average is taken and then summed with the fluctuation amplitude and high-frequency energy ratio to obtain the current drive fluctuation influence coefficient.

4. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 2, characterized in that: The steps for obtaining the output power fluctuation influence coefficient are as follows: During the detection period, real-time output power values are collected and an output power sequence is constructed. The sliding window length is set and the sequence is divided into several sliding windows of equal length according to the sliding window length to obtain several power observation subsequences. For each power observation subsequence, the second-order derivative change value is calculated by fitting the residual sequence, which is recorded as the degree of nonlinear drift; For the nonlinear drift degrees extracted from all sliding windows, the maximum drift degree and the mean drift degree are statistically calculated, and the offset amplitude coefficient is calculated based on the maximum drift degree and the mean drift degree; The standard deviation of the output power sequence is calculated, and the output power fluctuation influence coefficient is calculated based on the offset amplitude coefficient and the standard deviation of the output power sequence.

5. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 1, characterized in that: The step of determining whether center wavelength correction is required based on the center wavelength impact index is as follows: The affected index of the central wavelength is compared with the affected threshold. If the affected index of the central wavelength is greater than or equal to the affected threshold, it is determined that the central wavelength emitted by the current laser has shifted and central wavelength correction is required. If the affected index of the central wavelength is less than the affected threshold, it is determined that the central wavelength emitted by the current laser has not shifted and central wavelength correction is not required.

6. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 5, characterized in that: The step of correcting the central wavelength of the laser according to the central wavelength affected index to obtain the central wavelength that should be set for each laser after correction is as follows: Obtain the standard center wavelength of each gas component to be measured, and record the center wavelength of the current laser setting as the system setting wavelength value; The wavelength monitor is used to collect the center wavelength actually emitted by the current laser, which is recorded as the actual center wavelength. The actual center wavelength is compared with the standard center wavelength. If the actual center wavelength is greater than the standard center wavelength, it is determined that the current center wavelength is too long. The affected threshold value is then compared with the affected index of the center wavelength to obtain a correction factor. Multiply the system set wavelength value and the correction factor to obtain the center wavelength that should be set after correction; If the actual emitted center wavelength is shorter than the standard center wavelength, it is determined that the currently emitted center wavelength is too short. The system set wavelength value is divided by the correction factor to obtain the center wavelength that should be set after correction.

7. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 5, characterized in that: The steps for constructing the detection path with an equivalent absorption light path length exceeding one kilometer are as follows: Obtaining structural boundary information of the scene to be measured, the boundary information includes the geometric dimensions of the scene, the diffuse reflectivity of the inner wall material, and the coordinates of the main plane positions, and setting the initial emission position and incident angle of the laser beam based on the structural boundary information; Using the geometric optical path tracing method, with the set incident angle and initial emission position as the starting point, combined with the structural boundary information, the multiple diffuse reflection paths of the laser beam in the scene are calculated, the spatial position, reflection angle and propagation direction of each reflection point are determined, and a detection path with an equivalent absorption light path length of more than one kilometer is obtained.

8. The ultra-long-distance laser multi-gas component non-contact detection system according to claim 1, characterized in that: The steps of achieving coarse positioning of a long-distance leakage point by changing the optical path structure and obtaining a preliminary positioning area are as follows: According to the set of gas concentration estimates of each gas component to be measured in each path segment and the spatial position of the path segment, a concentration space mapping matrix based on the path segment is constructed; Perform three-dimensional gradient analysis on the concentration space mapping matrix to obtain the concentration variation between adjacent spatial grids. Based on the concentration variation between adjacent spatial grids, the gas concentration gradient distribution is obtained by the central difference method. Set a concentration threshold and, based on the gas concentration gradient distribution, screen out continuous grid areas with concentration values higher than the concentration threshold and with the same concentration gradient direction, and record them as candidate leakage areas; For each candidate leakage area, the number of path segments that pass through or cover the area is counted and recorded as the number of supported paths. At the same time, the three-dimensional gradient distribution results in the concentration space mapping matrix are called to extract the concentration gradient vector modulus of each grid in the area, and the maximum and average values are calculated, which are recorded as the concentration gradient intensity of the area. The number of supporting paths and the concentration gradient strength are weighted and fused according to the preset weights to construct a rough position confidence score value. A rough position confidence threshold is set, and candidate areas with a score greater than or equal to the threshold are marked as preliminary positioning areas.

9. A method for ultra-long-distance laser non-contact detection of multiple gas components, for implementing an ultra-long-distance laser non-contact detection system for multiple gas components according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step 1: Obtain the parameters affected by the center wavelength of each laser. The parameters affected by the center wavelength include current drive fluctuation data, temperature change data, and output power fluctuation data. The center wavelength affected index is obtained based on the center wavelength affected parameters. The center wavelength affected index is used to determine whether center wavelength correction is required. Step 2: If it is determined that center wavelength correction is required, the center wavelength of the current laser setting is obtained, and the center wavelength of the laser setting is corrected according to the center wavelength affected index to obtain the center wavelength that should be set for each laser after correction. The center wavelength value of each laser is set according to the center wavelength that should be set after correction, and the center wavelength laser beams are emitted in sequence; Step 3: By multiple diffuse reflections of the laser beam on the inner wall of the scene structure, a detection path with an equivalent absorption optical path length of more than one kilometer is constructed. The detection path consists of multiple path segments, and the laser signal of each path segment is obtained; Step 4: Obtain the standard absorption spectrum data of the gas components to be measured, and build a Bayesian inversion model based on the standard absorption spectrum data to obtain the concentration estimate of each gas component to be measured; Step 5: Collect the ambient temperature and pressure parameters of the laser beam in the detection path, and obtain the absorption coefficient of the unit concentration gas through the TDLAS mathematical inversion model; use the Kalman filter algorithm to dynamically correct the absorption coefficient to obtain the corrected absorption coefficient; Step 6: Based on the corrected absorption coefficient, the Beer-Lambert law is used, and combined with the laser signal of each path segment, the concentration estimate of each gas component to be measured is inverted to obtain a set of gas concentration estimates for each gas component to be measured in each path segment. The long-distance leak point is roughly located by changing the optical path structure, and the preliminary positioning area is obtained; Step 7: Use portable infrared imaging equipment to perform local fine detection on the preliminary positioning area to obtain the precise location coordinates of the leak point.

Images