Tunnel Multi-Component High-Risk Hazardous Gas Detection System and Method Based on Spectral Technology

Abstract

The present invention provides a tunnel multi-component high-risk harmful gas detection system and method based on spectral technology, which relates to the technical field of tunnel high-risk harmful gas detection and includes: a data acquisition module for acquiring an infrared spectral image of a gas sample in a tunnel with a wavelength range of [], and real-time collecting environmental impact parameters in the tunnel to be detected, where the environmental impact parameters include temperature data and air pressure data in the tunnel. The present invention extracts the detected values of the absorption peak wavelength, the detected value of the light absorption frequency, and the detected value of the light absorption intensity of the gas sample in the spectral image, analyzes and processes the collected temperature data, air pressure data, and gas attribute parameters, corrects the gas attribute parameters of the sample, and improves the detection accuracy of the gas concentration; accurately determines the specific components of the gas sample by comparing the spectral images and fitting the absorption peaks, and determines the harm risk degree of the high-risk harmful gas in the tunnel.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel high-risk and harmful gas detection, and specifically to a tunnel multi-component high-risk and harmful gas detection system and method based on spectral technology. Background Art

[0002] During the road and bridge construction process, it is often necessary to excavate tunnels in mountains. The newly excavated tunnels are relatively enclosed environments. When the content of a certain gas in the tunnel exceeds the standard, if it cannot be detected in time, it will cause production safety accidents.

[0003] In the prior art, a toxic and harmful gas monitoring device for tunnels with the publication number of CN212904792U is provided. The device includes: a plurality of gas detection sensors, a data transmission module, a grading and labeling unit, an LED display, and a duty display; each gas detection sensor is arranged at intervals on the side walls of each detection point in the tunnel; each gas detection sensor is electrically connected to the data transmission module respectively; the data transmission module is electrically connected to the grading and labeling unit; the grading and labeling unit is electrically connected to the LED display and the duty display respectively; by arranging a plurality of gas detection sensors at intervals on the inner side walls of the tunnel and electrically connecting the measurement data with the grading and labeling unit through the data transmission module electrically connected to the sensors, the hierarchical display of the measurement results is realized, and the real-time and accurate display of various gas components in each area of the tunnel can be realized, improving the measurement accuracy and timeliness, and realizing a reminder function through the LED display arranged on the side wall of the tunnel entrance, reducing the labor cost and improving the detection effect.

[0004] However, there are still the following deficiencies. From the above statements, when the gas detection sensor detects the gas components and concentrations, since the numerical changes of temperature and air pressure will both affect the measurement accuracy of gas concentration, if the environmental impact parameters are not incorporated into the calculation of gas concentration, the detection accuracy of gas concentration will be reduced;

[0005] And when detecting harmful gases existing in tunnels, such as methane, hydrogen sulfide, and carbon monoxide, due to the similarity of molecular structures and bonding modes, they show similar absorption characteristics in the spectrum, making the spectral images of them have overlapping areas, and it is difficult to judge the gas components and concentration values. When the concentration of high-risk and harmful gases is too high, it poses a threat to the safety of construction workers.

[0006] The above information disclosed in the background art section is only used to strengthen the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to provide a tunnel multi-component high-risk harmful gas detection system and method based on spectral technology to solve the problems raised in the above-mentioned background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] A tunnel multi-component high-risk harmful gas detection system based on spectral technology, comprising:

[0010] A data acquisition module for acquiring an infrared spectral image of a gas sample in a tunnel within a wavelength range of , and real-time collecting environmental impact parameters in the tunnel to be detected, where the environmental impact parameters include tunnel temperature data and air pressure data;

[0011] An image division module for receiving the acquired infrared spectral image, dividing the infrared spectral image into several identification intervals with different wavelengths, and respectively collecting gas attribute parameters for each interval, where the gas attribute parameters include detected values of the absorption peak wavelength, detected values of the light absorption frequency, and detected values of the light absorption intensity;

[0012] A data processing module for analyzing and processing the collected temperature, air pressure, and gas attribute parameters, correcting the gas attribute parameters of the sample, generating corrected values of the absorption peak wavelength, corrected values of the light absorption frequency, and corrected values of the light absorption intensity, and generating a new gas spectral image based on the corrected values of the absorption peak wavelength, corrected values of the light absorption frequency, and corrected values of the light absorption intensity;

[0013] A first data analysis module for comparing the new gas spectral image with the standard spectral image of the sample gas, evaluating the similarity degree of the new gas spectral image, and when the similarity between the new gas spectral image and the standard spectral image exceeds a set similarity threshold, adjusting the position, height, and width of the absorption peak according to the peak fitting technology, fitting the actual spectral data, and obtaining the actual value of the absorption peak wavelength, the actual value of the light absorption frequency, and the actual value of the light absorption intensity according to the actual spectral data;

[0014] A component judgment module for judging the specific components of the gas sample according to the interval ranges where the actual value of the absorption peak wavelength and the actual value of the light absorption frequency are located;

[0015] A second data analysis module for converting the actual value of the light absorption intensity of each gas in the sample into a gas concentration according to the Beer-Lambert law, performing dimensionless processing on the gas concentration and the proportion coefficients of each gas in the tunnel, and performing correlation analysis to generate a comprehensive coefficient for evaluating the risk degree of hazards in the tunnel, and comparing the comprehensive coefficient with a preset comprehensive coefficient risk threshold;

[0016] A risk judgment module, configured to judge the risk degree of harmful gases in the tunnel according to the comparison result between the comprehensive coefficient and a preset comprehensive coefficient risk threshold.

[0017] Further, by the different positions of the significant absorption peaks generated by the infrared spectral image, the wavelength range can be divided into several different intervals, where 0 , , within the infrared spectral range, within the wavelength range , methane has a significant absorption peak; within the infrared spectral range, within the wavelength range , hydrogen sulfide has a significant absorption peak; within the infrared spectral range, within the wavelength range , carbon monoxide has a significant absorption peak; within the infrared spectral range, within the wavelength range , carbon dioxide has a significant absorption peak.

[0018] Further, the data processing module combines the temperature and air pressure with the detected values of the absorption peak wavelength, light absorption frequency, and light absorption intensity respectively, corrects the values of the detected absorption peak wavelength, light absorption frequency, and light absorption intensity, and generates a corrected value of the absorption peak wavelength, a corrected value of the light absorption frequency, and a corrected value of the light absorption intensity. The basis formula is as follows:

[0019] ;

[0020] Wherein, is the proportional factor coefficient of temperature, is the proportional factor coefficient of air pressure, is the detected value of the absorption peak wavelength, is the detected value of the light absorption frequency, is the detected value of the light absorption intensity, is the corrected value of the absorption peak wavelength, is the corrected value of the light absorption frequency, is the corrected value of the light absorption intensity, , and , is the constant correction coefficient.

[0021] Further, the process by which the first data analysis module compares the new gas spectral image with the standard spectral image of the gas to determine the gas components contained in the sample is as follows:

[0022] S11. Collect standard spectral data: First, it is necessary to collect standard spectral data containing gas components, and these standard data come from a professional database;

[0023] S12. Calibration and Alignment: Ensure that the new gas spectral image and the standard spectral image are aligned in wavelength. Wavelength calibration and spectral alignment operations are required to ensure that their peaks and valleys correspond.

[0024] S13. Comparison and Matching: Compare the new gas spectral image with the standard spectral image, which can be done by calculating the similarity index between them. Common methods include cosine similarity.

[0025] Furthermore, in spectral image comparison, the new gas spectrum and the standard spectrum can be represented as vectors, and then the cosine similarity is used to measure the similarity between them. The following are the steps to describe the process in detail:

[0026] S131. Represent Spectral Data as Vectors: Represent the new sample spectral image and the standard spectral image as one-dimensional vectors, which can be achieved by arranging the spectral intensity values of each pixel into a vector. For example, if the spectral image has N pixel points, then each vector will have N elements.

[0027] S132. Calculate Cosine Similarity: Use the cosine similarity formula to calculate the similarity between the new sample spectrum and the standard spectrum. The cosine similarity formula is as follows:

[0028] ;

[0029] where, is the vector of the new sample spectral image, is the vector of the standard spectral image, is the norm of the vector of the new sample spectral image, is the norm of the vector of the standard spectral image, is the degree of similarity between the new sample spectrum and the standard spectrum;

[0030] The obtained cosine similarity value is in the range of [-1, 1], where 1 indicates and are completely similar, -1 indicates and are completely dissimilar, and 0 indicates that the included angle is 90 degrees, and there is no similarity between the vectors;

[0031] S133. Analysis result: According to the calculation result of cosine similarity, the similarity degree between the new gas spectrum and the standard spectrum can be judged. When the similarity between the new gas spectrum image and the standard spectrum image exceeds the set similarity threshold, assuming the similarity threshold is set to 0.8, and the similarity between the new gas spectrum image and the standard spectrum image exceeds 0.8, then the similarity between the new gas spectrum image and the standard spectrum image is very high, and the gas components of the sample can be roughly judged.

[0032] Further, after judging the similarity degree between the new gas spectrum image and the standard spectrum image, according to the peak fitting technology, adjust the position, height and width of the absorption peak, fit the actual spectrum data, and the process of obtaining the actual value of the wavelength of the absorption peak, the actual value of the light absorption frequency and the actual value of the light absorption intensity according to the actual spectrum data is as follows:

[0033] The specific process of peak fitting usually involves the non - linear least - squares method. One of the commonly used algorithms is the Levenberg - Marquardt algorithm. The following is the general process of the Levenberg - Marquardt algorithm, taking the fitting of a single peak as an example:

[0034] S21. Initialize parameters: Initially, give the initial values of parameters such as the position, height, and width of the peak;

[0035] S22. Construct the model function: Select the Gaussian function to construct the model function describing the peak shape. For the Gaussian function, the model function can be expressed as:

[0036] + B f(X)= （-2 ） + B ;

[0037] where, is the amplitude of the Gaussian peak, representing the maximum height of the peak, is the center position of the Gaussian peak, that is, the actual value of the wavelength of the peak, is the standard deviation of the Gaussian peak, which determines the width of the peak, is the abbreviation of the exponential function, usually expressed as e^x, B is the background signal;

[0038] S23. Define the error function: Measure the difference between the model function and the actual spectrum data. The typical error function is the sum of squared residuals, and the formula is as follows:

[0039] ;

[0040] where, is the error value, which measures the difference between the model function and the actual spectral data. is the number of data points. and are the actual data points.

[0041] S24. Iteration process: Use the Levenberg-Marquardt algorithm for iteration. The goal of the iteration is to minimize the error function by adjusting the model parameters. The algorithm will update the parameters in each iteration to make the model better fit the actual data.

[0042] This process is repeated until the convergence condition is reached, such as the parameter change is small enough, the error is small enough, or the predetermined number of iterations is reached.

[0043] S25. Extract peak parameters: Once the iteration is completed, extract the relevant information of the peak from the final parameters, including the actual wavelength value, the actual frequency value of light absorption, and the actual light absorption intensity value.

[0044] Furthermore, the process of the component judgment module judging the specific components of the gas sample according to the range of the actual wavelength value of the absorption peak and the actual frequency value of light absorption is as follows:

[0045] When the actual wavelength value of the absorption peak is between 3.3μm ≤ BC ≤ 3.4μm and the actual frequency value of light absorption is between 8800Hz ≤ PL ≤ 9100Hz, and only when the above conditions are all met, it is judged that there is methane gas in the sample.

[0046] When the actual wavelength value of the absorption peak is between 2.5μm ≤ BC ≤ 2.6μm and the actual frequency value of light absorption is between 3800Hz ≤ PL ≤ 4000Hz, and only when the above conditions are all met, it is judged that there is hydrogen sulfide gas in the sample.

[0047] When the actual wavelength value of the absorption peak is between 4.6μm ≤ BC ≤ 4.7μm and the actual frequency value of light absorption is between 2000Hz ≤ PL ≤ 2220Hz, and only when the above conditions are all met, it is judged that there is carbon monoxide gas in the sample.

[0048] When the actual wavelength value of the absorption peak is between 2.7μm ≤ BC ≤ 4.3μm and the actual frequency value of light absorption is between 400Hz ≤ PL ≤ 2500Hz, and only when the above conditions are all met, it is judged that there is carbon dioxide gas in the sample.

[0049] Among them, BC is the actual wavelength value of the absorption peak, and PL is the actual frequency value of light absorption.

[0050] Further, the second data analysis module converts the actual value of the light absorption intensity of each gas into the gas concentration according to the Beer-Lambert law, and the formula is as follows:

[0051] ;

[0052] Among them, is the actual value of the light absorption intensity of methane gas, is the actual value of the light absorption intensity of hydrogen sulfide gas, is the actual value of the light absorption intensity of carbon monoxide gas, is the actual value of the light absorption intensity of carbon dioxide gas, is the concentration value of methane gas, is the concentration value of hydrogen sulfide gas, is the concentration value of carbon monoxide gas, is the concentration value of carbon dioxide gas, is the molar absorption coefficient of methane gas, is the molar absorption coefficient of hydrogen sulfide gas, is the molar absorption coefficient of carbon monoxide gas, is the molar absorption coefficient of carbon dioxide gas, is the optical path;

[0053] The concentration values of methane gas, hydrogen sulfide gas, carbon monoxide gas, carbon dioxide gas and the proportionality coefficients of each gas in the tunnel are dimensionless processed and correlation analyzed to generate a comprehensive coefficient for evaluating the hazard risk degree in the tunnel , and the comprehensive coefficient is compared with the preset comprehensive coefficient risk threshold . Among them, the formula for generating the comprehensive coefficient is as follows:

[0054] ;

[0055] Among them, is the proportionality coefficient of , is the proportionality coefficient of , is the proportionality coefficient of , is the proportionality coefficient of , , and .

[0056] Further, the risk judgment module is based on the comprehensive coefficient and the preset comprehensive coefficient risk threshold The process of judging the harm risk degree of harmful gases in the tunnel based on the comparison result is as follows:

[0057] When it is the case, the harm risk of harmful gases in the tunnel is relatively high;

[0058] When it is the case, the harm risk of harmful gases in the tunnel is relatively low.

[0059] A method for detecting multi-component high-risk harmful gases in tunnels based on wide-spectrum technology, the specific steps include:

[0060] S1. Obtain the infrared spectral image of the gas sample in the tunnel with a wavelength range of , and collect the environmental impact parameters in the tunnel to be detected in real time. The environmental impact parameters include the temperature data and air pressure data in the tunnel;

[0061] S2. Receive the collected infrared spectral image, divide the infrared spectral image into several identification intervals with different wavelengths, and collect gas attribute parameters for each interval respectively. The gas attribute parameters include the detected value of the absorption peak wavelength, the detected value of the light absorption frequency, and the detected value of the light absorption intensity;

[0062] S3. Analyze and process the collected temperature, air pressure and gas attribute parameters, correct the gas attribute parameters of the sample, generate the corrected value of the absorption peak wavelength, the corrected value of the light absorption frequency, and the corrected value of the light absorption intensity, and generate a new gas spectral image based on the corrected value of the absorption peak wavelength, the corrected value of the light absorption frequency, and the corrected value of the light absorption intensity;

[0063] S4. Compare the new gas spectral image with the standard spectral image of the sample gas, evaluate the similarity degree of the new gas spectral image. When the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, according to the peak fitting technology, adjust the position, height and width of the absorption peak, fit the actual spectral data, and obtain the actual value of the absorption peak wavelength, the actual value of the light absorption frequency, and the actual value of the light absorption intensity according to the actual spectral data;

[0064] S5. Judge the specific components of the gas sample according to the interval range where the actual value of the absorption peak wavelength and the actual value of the light absorption frequency are located;

[0065] S6. Convert the actual value of the light absorption intensity of each gas in the sample into gas concentration according to the Beer-Lambert law, perform dimensionless processing on the gas concentration and the proportionality coefficient of each gas in the tunnel, and perform correlation analysis to generate a comprehensive coefficient for evaluating the harm risk degree in the tunnel, and compare the comprehensive coefficient with the preset comprehensive coefficient risk threshold;

[0066] S7. Determine the hazard risk level of harmful gases in the tunnel according to the comparison result between the comprehensive coefficient and the preset comprehensive coefficient risk threshold.

[0067] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0068] The present invention obtains an infrared spectral image of a gas sample in a tunnel with a wavelength range of , and collects the temperature data and air pressure data in the tunnel to be detected, and extracts the detection values of the absorption peak wavelength, the light absorption frequency, and the light absorption intensity of the gas sample in the spectral image. The collected temperature data, air pressure data, and gas attribute parameters are analyzed and processed to correct the gas attribute parameters of the sample, improving the detection accuracy of the gas concentration;

[0069] Compare the new gas spectral image with the standard spectral image of the sample gas to evaluate the similarity degree of the new gas spectral image. When the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, according to the peak fitting technology, adjust the position, height, and width of the absorption peak, and fit and obtain the actual values of the wavelength, the actual value of the light absorption frequency, and the actual value of the light absorption intensity of the absorption peak based on the actual spectral data. And according to the interval range where the actual value of the absorption peak wavelength and the actual value of the light absorption frequency are located, accurately judge the specific components of the gas sample;

[0070] Convert the actual value of the light absorption intensity of each gas in the sample into the gas concentration according to the Beer-Lambert law, process and analyze the gas concentration and the proportional coefficient of each gas in the tunnel to generate a comprehensive coefficient for evaluating the hazard risk level in the tunnel, and compare the comprehensive coefficient with the preset comprehensive coefficient risk threshold. According to the comparison result, judge the hazard risk level of high-risk harmful gases in the tunnel. Description of the Drawings

[0071] Figure 1 It is a block diagram of the module composition of the present invention;

[0072] Figure 2 It is a schematic diagram of the overall method flow of the present invention. Detailed Embodiments

[0073] To make the objectives, technical solutions, and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below in conjunction with specific embodiments.

[0074] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second" and similar terms used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left" and "right" are only used to represent relative positional relationships, and when the absolute position of the object to be described changes, the relative positional relationship may also change accordingly.

[0075] Embodiment:

[0076] Please refer to Figures 1 to 2 , the present invention provides a technical solution:

[0077] A multi-component high-risk harmful gas detection system for tunnels based on spectral technology, as Figure 1 shown, includes:

[0078] A data acquisition module, configured to obtain an infrared spectral image of a gas sample in a tunnel within a wavelength range of , and to collect in real time the environmental impact parameters in the tunnel to be detected, where the environmental impact parameters include temperature data and air pressure data in the tunnel;

[0079] An image division module, configured to receive the collected infrared spectral image, divide the infrared spectral image into several identification intervals with different wavelengths, and collect gas attribute parameters for each interval respectively, where the gas attribute parameters include the detected value of the absorption peak wavelength, the detected value of the light absorption frequency, and the detected value of the light absorption intensity;

[0080] A data processing module, configured to analyze and process the collected temperature, air pressure and gas attribute parameters, correct the gas attribute parameters of the sample, generate corrected values of the absorption peak wavelength, the light absorption frequency, and the light absorption intensity, and generate a new gas spectral image based on the corrected values of the absorption peak wavelength, the light absorption frequency, and the light absorption intensity;

[0081] The first data analysis module is used to compare the new gas spectral image with the standard spectral image of the sample gas, evaluate the similarity degree of the new gas spectral image, and when the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, according to the peak fitting technology, adjust the position, height and width of the absorption peak, fit the actual spectral data, and obtain the actual value of the wavelength of the absorption peak, the actual value of the light absorption frequency and the actual value of the light absorption intensity according to the actual spectral data;

[0082] The component judgment module is used to judge the specific components of the gas sample according to the interval range where the actual value of the wavelength of the absorption peak and the actual value of the light absorption frequency are located;

[0083] The second data analysis module is used to convert the actual value of the light absorption intensity of each gas in the sample into the gas concentration according to the Beer-Lambert law, perform dimensionless processing on the gas concentration and the proportional coefficient of each gas in the tunnel, and conduct correlation analysis to generate a comprehensive coefficient for evaluating the hazard risk degree in the tunnel, and compare the comprehensive coefficient with the preset comprehensive coefficient risk threshold;

[0084] The risk judgment module is used to judge the hazard risk degree of the harmful gas in the tunnel according to the comparison result between the comprehensive coefficient and the preset comprehensive coefficient risk threshold.

[0085] In this embodiment, for gases, their molecules usually show specific absorption peaks in the infrared spectrum, corresponding to specific vibration or rotation modes within the molecule. Different gases have different infrared absorption characteristics, which makes the infrared spectrum a tool for identifying and measuring gas components. First, select an infrared spectrometer and a sample collection system applicable to the wavelength range [λ1, λ2], connect the infrared spectrometer to the sample collection system to ensure that the infrared spectral image can be obtained within the wavelength range [λ1, λ2], and pass the infrared lamp emitted by the infrared spectrometer through the gas. The radiation of a specific wavelength will be absorbed to obtain the infrared spectral image of the gas sample.

[0086] In this embodiment, through the different positions of the significant absorption peaks generated by the spectral image, the interval with the wavelength range of can be divided into several different interval ranges, 0 , , within the infrared spectrum range, within the wavelength range , methane has a significant absorption peak; within the infrared spectrum range, within the wavelength range , hydrogen sulfide has a significant absorption peak; within the infrared spectrum range, within the wavelength range , carbon monoxide has a significant absorption peak; within the infrared spectrum range, within the wavelength range , carbon dioxide has a significant absorption peak.

[0087] In this embodiment, by comparing the new gas spectral image with the standard spectral image of the gas, the gas components of the sample can be roughly judged. In order to determine the specific components of the sample, the peak fitting technology is adopted to adjust the position, height and width of the absorption peak, fit the actual spectral data, and according to the actual spectral data, obtain the actual values of the wavelength of the absorption peak, the actual value of the light absorption frequency and the actual value of the light absorption intensity, and transmit them to the component judgment module. Using the component judgment module, according to the interval ranges of the actual value of the wavelength of the absorption peak, the actual value of the light absorption frequency and the actual value of the light absorption intensity, the specific components of the sample are judged.

[0088] When the temperature coefficients in the tunnel are different, the spectral characteristic data of the gas sample are different. The following are the specific reasons:

[0089] Change in molecular vibration state: The change in temperature will affect the vibration state of gas molecules. At different temperatures, the vibrational energy level distribution of molecules will change, resulting in changes in the position (wavelength) and intensity of the absorption peak. This involves the vibrations of bonds such as C-H and O-H in the infrared spectrum.

[0090] Change in molecular rotation state: High temperature may increase the rotational state of gas molecules. This may lead to changes in the rotational spectral lines observed in the infrared spectrum.

[0091] Change in energy level spacing: An increase in temperature will cause the energy level spacing of gas molecules to become larger. This may cause a change in the wavelength of the absorption peak because the wavelength in the spectrum is related to the energy level spacing.

[0092] Change in absorption intensity: The change in temperature may affect the electric dipole moment and vibration intensity of gas molecules, thereby affecting the light absorption intensity.

[0093] Therefore, it is particularly important to detect the temperature coefficient in the tunnel for detecting the specific components of multi-component high-risk harmful gases in the tunnel. For example, the following effects can be achieved:

[0094] Accurate gas identification: The change in temperature may cause changes in the infrared spectral characteristics of different gases. By detecting the temperature coefficient, different gas components can be more accurately identified and distinguished according to the spectral characteristics of the gas at different temperatures.

[0095] Adaptability and stability: By real-time monitoring the temperature coefficient, the system can achieve adaptability, that is, automatically adjust the gas identification and measurement algorithms under different temperature conditions, improving the stability and reliability of the system.

[0096] Adaptability to environmental changes: The temperature inside the tunnel may be affected by seasonal variations, weather conditions, and other factors. By considering the temperature coefficient, the system can better adapt to these environmental changes and improve the accuracy of detection.

[0097] When the air pressure coefficients inside the tunnel are different, the spectral characteristic data of the gas sample are different. The following are the specific reasons:

[0098] Change in the collision frequency of molecules: Under high air pressure conditions, the collision frequency between gas molecules increases. This may cause changes in the shape and width of the absorption peak, affecting the line width of the spectrum.

[0099] Change in the intensity of the absorption peak: The collisions of gas molecules under high air pressure conditions may affect their vibrational and rotational states. This may cause changes in the intensity of the absorption peak because the absorption intensity is related to the change in the molecular state.

[0100] Effect of gas density: Under high air pressure conditions, the gas density increases. Due to the increase in the number of gas molecules in the optical path, the intensity of the absorption characteristics may increase.

[0101] Effect of Brownian motion: The Brownian motion of gas molecules may weaken under high air pressure. This may have an impact on the line width and shape of the spectral characteristics.

[0102] Combined effect of temperature and air pressure: Temperature and air pressure are usually interrelated because high temperature is usually accompanied by high air pressure and vice versa. The change in air pressure may be combined with the change in temperature to jointly affect the spectral characteristics.

[0103] Therefore, it is particularly important to detect the air pressure coefficient inside the tunnel through a pressure sensor for detecting the specific components of multi-component high-risk harmful gases inside the tunnel. For example, the following effects can be achieved:

[0104] Accurate gas identification: The change in air pressure may affect the vibrational, rotational states, and collision frequency of gas molecules, thereby changing their infrared spectral characteristics. By monitoring the air pressure coefficient, different gas components can be more accurately identified and distinguished.

[0105] Accurate concentration measurement: The change in air pressure may affect the intensity of the absorption peak. Therefore, by monitoring the air pressure coefficient, gas concentration measurement can be more accurate. This is crucial for detecting the concentration change of harmful gases.

[0106] Adaptability and stability: By real-time monitoring of the air pressure coefficient, the system can achieve self-adaptability, that is, automatically adjust the gas identification and measurement algorithms under different air pressure conditions, improving the stability and reliability of the system.

[0107] Adaptability to environmental changes: The air pressure in the tunnel may be affected by factors such as weather changes. By considering the air pressure coefficient, the system can better adapt to these environmental changes and improve the accuracy of detection.

[0108] In summary, the data acquisition module plays a crucial role in detecting the specific components of multi-component high-risk harmful gases in the tunnel by collecting the detection values of the absorption peak wavelength, light absorption frequency, light absorption intensity, temperature, and air pressure. The following are the specific implementation methods for extracting the detection values of the absorption peak wavelength, light absorption frequency, light absorption intensity, temperature coefficient, and air pressure coefficient in this embodiment.

[0109] Based on the above embodiments, the data processing module combines the temperature and air pressure coefficients with the detection values of the absorption peak wavelength, light absorption frequency, and light absorption intensity respectively, corrects the values of the detection values of the absorption peak wavelength, light absorption frequency, and light absorption intensity, and generates the corrected values of the absorption peak wavelength, light absorption frequency, and light absorption intensity. The formulas are as follows:

[0110] ;

[0111] Among them, is the proportional factor coefficient of temperature, is the proportional factor coefficient of air pressure, is the detection value of the absorption peak wavelength, is the detection value of the light absorption frequency, is the detection value of the light absorption intensity, is the corrected value of the absorption peak wavelength, is the corrected value of the light absorption frequency, is the corrected value of the light absorption intensity. Through the correlation analysis of the temperature coefficient , air pressure coefficient and the detection value of the absorption peak wavelength , detection value of the light absorption frequency , detection value of the light absorption intensity , the correlation coefficient can be calculated. By alternately changing one parameter and keeping the other parameter unchanged, it can be known that the correlation intensity between temperature , air pressure and the detection value of the absorption peak wavelength , detection value of the light absorption frequency , detection value of the light absorption intensity decreases in turn. Therefore , and , is a constant correction coefficient.

[0112] As can be seen from the above formula, within a certain range, when the temperature is higher, the correction value of the absorption peak wavelength is higher. When the air pressure is higher, the correction value of the absorption peak wavelength is lower. Then, the temperature and the correction value of the absorption peak wavelength show a positive correlation. The air pressure and the correction value of the absorption peak wavelength show a negative correlation. Within a certain range, when the temperature is higher, the correction value of the light absorption frequency is lower. When the air pressure is higher, the correction value of the light absorption frequency is higher. Then, the temperature and the correction value of the light absorption frequency show a negative correlation. The air pressure and the correction value of the light absorption frequency show a positive correlation. Within a certain range, when the coefficient and the air pressure coefficient are both higher, the correction value of the light absorption intensity is higher. Then, the temperature , the air pressure and the correction value of the light absorption intensity all show a positive correlation. The factor coefficient in the formula is used to balance the proportion of each item of data in the formula, so as to improve the accuracy of the calculation result.

[0113] Based on the above embodiments, the process by which the data analysis module compares the new gas spectral image with the standard spectral image of the sample gas to determine the gas components contained in the sample is as follows:

[0114] S11. Collect standard spectral data: First, it is necessary to collect standard spectral data containing gas components, and these standard data come from a professional database;

[0115] S12. Calibration and alignment: Ensure that the new gas spectral image and the standard spectral image are aligned in wavelength. It is necessary to perform wavelength calibration and spectral alignment operations to ensure that their peaks and valleys correspond;

[0116] S13. Comparison and matching: Compare the new gas spectral image with the standard spectral image, which can be completed by calculating the similarity index or distance between them. Common methods include cosine similarity.

[0117] In the comparison of spectral images, the new gas spectrum and the standard spectrum can be represented as vectors, and then the cosine similarity is used to measure the similarity between them. The following are the steps to describe the process in detail:

[0118] S131. Represent spectral data as vectors: Represent the new gas spectral image and the standard spectral image as one-dimensional vectors, which can be achieved by arranging the spectral intensity values of each pixel into a vector. For example, if the spectral image has N pixel points, then each vector will have N elements;

[0119] S132. Calculate cosine similarity: Use the cosine similarity formula to calculate the similarity between the new gas spectrum and the standard spectrum. The formula for cosine similarity is as follows:

[0120] ;

[0121] where, is the vector of the new gas spectral image, is the vector of the standard spectral image, is the norm of the vector of the new gas spectral image, is the norm of the vector of the standard spectral image, is the degree of similarity between the new gas spectrum and the standard spectrum;

[0122] The obtained cosine similarity value is in the range of [-1, 1], where 1 indicates that and are completely similar, -1 indicates that and are completely dissimilar, 0 indicates that the included angle is 90 degrees, and there is no similarity between the vectors;

[0123] S133. Analyze the results: According to the calculation results of the cosine similarity, the degree of similarity between the new gas spectrum and the standard spectrum can be judged. When the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, assuming the similarity threshold is 0.8, and the similarity between the new gas spectral image and the standard spectral image exceeds 0.8, then the similarity between the new gas spectral image and the standard spectral image is very high, and the gas components of the sample can be roughly judged.

[0124] Based on the above embodiments, after judging the degree of similarity between the new gas spectral image and the standard spectral image, according to the peak fitting technology, adjust the position, height and width of the absorption peak, fit the actual spectral data, and the process of obtaining the actual wavelength value of the absorption peak, the actual frequency value of light absorption and the actual value of light absorption intensity according to the actual spectral data is as follows:

[0125] The specific process of peak fitting usually involves the non - linear least - squares method. One of the commonly used algorithms is the Levenberg - Marquardt algorithm. The following is the general process of the Levenberg - Marquardt algorithm, taking the fitting of a single peak as an example:

[0126] S21. Initialize parameters: Initially, give the initial values of parameters such as the position, height, and width of the peak;

[0127] S22. Construct the model function: Select the Gaussian function to construct the model function describing the peak shape. For the Gaussian function, the model function can be expressed as:

[0128] + B f(X)= （-2 ） + B ;

[0129] Where, is the amplitude of the Gaussian peak, representing the maximum height of the peak, is the center position of the Gaussian peak, that is, the actual value of the wavelength of the peak, is the standard deviation of the Gaussian peak, which determines the width of the peak, B is the background signal;

[0130] S23. Define the error function: Measure the difference between the model function and the actual spectral data. A typical error function is the sum of squared residuals, and the formula is as follows:

[0131] ;

[0132] Where, is the error value, measuring the size of the difference between the model function and the actual spectral data, is the number of data points, and are the actual data points;

[0133] S24. Iterative process: Use the Levenberg - Marquardt algorithm for iteration. The goal of iteration is to minimize the error function by adjusting the model parameters. The algorithm will update the parameters in each iteration to make the model better fit the actual data.

[0134] This process is repeated until the convergence condition is reached (for example, the parameter change is small enough and the error is small enough) or the predetermined number of iterations is reached;

[0135] S25. Extract peak parameters: Once the iteration is completed, extract the relevant information of the peak from the final parameters, including the actual value of the wavelength, the actual value of the frequency of light absorption, and the actual value of the light absorption intensity.

[0136] Based on the above embodiments, the process by which the component determination module determines the specific components of the gas sample according to the ranges of the actual value of the wavelength of the absorption peak and the actual value of the frequency of light absorption is as follows:

[0137] When the actual value of the wavelength of the absorption peak is between 3.3 μm ≤ BC ≤ 3.4 μm and the actual value of the frequency of light absorption is between 8800 Hz ≤ PL ≤ 9100 Hz, and only when all the above conditions are satisfied, it is determined that methane gas exists in the sample;

[0138] When the actual value of the wavelength of the absorption peak is between 2.5 μm ≤ BC ≤ 2.6 μm and the actual value of the frequency of light absorption is between 3800 Hz ≤ PL ≤ 4000 Hz, and only when all the above conditions are satisfied, it is determined that hydrogen sulfide gas exists in the sample;

[0139] When the actual value of the wavelength of the absorption peak is between 4.6 μm ≤ BC ≤ 4.7 μm and the actual value of the frequency of light absorption is between 2000 Hz ≤ PL ≤ 2220 Hz, and only when all the above conditions are satisfied, it is determined that carbon monoxide gas exists in the sample;

[0140] When the actual value of the wavelength of the absorption peak is between 2.7 μm ≤ BC ≤ 4.3 μm and the actual value of the frequency of light absorption is between 400 Hz ≤ PL ≤ 2500 Hz, and only when all the above conditions are satisfied, it is determined that carbon dioxide gas exists in the sample;

[0141] Wherein, BC is the actual value of the wavelength of the absorption peak, and PL is the actual value of the frequency of light absorption.

[0142] Based on the above embodiments, the second data analysis module converts the actual value of the light absorption intensity of each gas into a gas concentration according to the Beer-Lambert law, and the formula is as follows:

[0143] ;

[0144] Wherein, is the actual value of the light absorption intensity of methane gas, is the actual value of the light absorption intensity of hydrogen sulfide gas, is the actual value of the light absorption intensity of carbon monoxide gas, is the actual value of the light absorption intensity of carbon dioxide gas, is the concentration value of methane gas, is the concentration value of hydrogen sulfide gas, is the concentration value of carbon monoxide gas, is the concentration value of carbon dioxide gas, is the molar extinction coefficient of methane gas, is the molar extinction coefficient of hydrogen sulfide gas, is the molar extinction coefficient of carbon monoxide gas, is the molar extinction coefficient of carbon dioxide gas, is the optical path;

[0145] Dimensionalize the concentration values of methane gas, hydrogen sulfide gas, carbon monoxide gas, carbon dioxide gas and the proportion coefficients of each gas in the tunnel, and perform correlation analysis to generate a comprehensive coefficient for evaluating the degree of hazard risk in the tunnel and compare the comprehensive coefficient with the preset comprehensive coefficient risk threshold Among them, the formula for generating the comprehensive coefficient is as follows:

[0146] ;

[0147] Among them, is the proportion coefficient of , is the proportion coefficient of , is the proportion coefficient of , is the proportion coefficient of . At the same concentration, the hazard levels of hydrogen sulfide gas, carbon monoxide gas, methane gas and carbon dioxide gas decrease in turn, so and .

[0148] On the basis of the above embodiments, the risk judgment module determines the degree of hazard risk of harmful gases in the tunnel according to the comparison result of the comprehensive coefficient and the preset comprehensive coefficient risk threshold The process is as follows:

[0149] When , the hazard risk of harmful gases in the tunnel is relatively high;

[0150] When , the hazard risk of harmful gases in the tunnel is relatively low.

[0151] In the formula, , , , and The specific value of is generally determined by those skilled in the art according to actual conditions. The essence of this formula is a comprehensive analysis by weighted summation. The technicians in this field collect multiple groups of sample data, set corresponding preset proportional coefficients for each group of sample data, substitute the preset proportional coefficients and the collected sample data into the formula, observe the accuracy of the model output and the rationality of the results through repeated experiments and parameter adjustments, gradually adjust these factor coefficients, and compare the performance and effect of the model under different parameter settings to find the optimal coefficient combination, screen the calculated factor coefficients and take the average, and obtain , , , and The value of .

[0152] In addition, the size of the preset factor coefficient is to quantify each parameter to obtain a specific value. In order to facilitate subsequent comparison, the size of the coefficient depends on the amount of sample data and the preset proportional coefficient initially set by technical personnel in this field for each set of sample data. It is not unique as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0153] A tunnel multi-component high-risk harmful gas detection method based on wide spectrum technology, such as Figure 2 As shown, the specific steps include:

[0154] S1. Get the wavelength range as [ ] and collects the infrared spectrum image of the gas sample in the tunnel to be tested in real time, and the environmental impact parameters in the tunnel to be tested include the temperature data and air pressure data in the tunnel;

[0155] S2. Receive the collected infrared spectrum image, divide the infrared spectrum image into several identification intervals with different wavelengths, and collect gas attribute parameters for each interval, wherein the gas attribute parameters include the detection value of the absorption peak wavelength, the detection value of the light absorption frequency, and the detection value of the light absorption intensity;

[0156] S3. Analyze and process the collected temperature, air pressure and gas property parameters, calibrate the gas property parameters of the sample, generate correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity, and generate a new gas spectrum image based on the correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity;

[0157] S4. Compare the new gas spectral image with the standard spectral image of the sample gas to evaluate the similarity degree of the new gas spectral image. When the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, according to the peak fitting technique, adjust the position, height, and width of the absorption peak, fit the actual spectral data, and based on the actual spectral data, obtain the actual value of the wavelength of the absorption peak, the actual value of the frequency of light absorption, and the actual value of the light absorption intensity;

[0158] S5. Determine the specific components of the gas sample according to the range of the actual value of the wavelength of the absorption peak and the actual value of the frequency of light absorption;

[0159] S6. Convert the actual value of the light absorption intensity of each gas in the sample into gas concentration according to the Beer-Lambert law, perform dimensionless processing on the gas concentration and the proportionality coefficient of each gas in the tunnel, and conduct a correlation analysis to generate a comprehensive coefficient for evaluating the hazard risk degree in the tunnel, and compare the comprehensive coefficient with the preset comprehensive coefficient risk threshold;

[0160] S7. Judge the hazard risk degree of the harmful gas in the tunnel according to the comparison result between the comprehensive coefficient and the preset comprehensive coefficient risk threshold.

[0161] The above formulas are all dimensionless and take their numerical values for calculation. The formula is obtained by collecting a large amount of data for software simulation to get a formula closest to the actual situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0162] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by computer software, electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0163] The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0164] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. A tunnel multi-component high-risk and harmful gas detection system based on spectral technology, characterized in that: include: A data acquisition module is used to obtain infrared spectrum images of gas samples in the tunnel with a wavelength range of [λ1, λ2], and to collect environmental impact parameters in the tunnel to be detected in real time, wherein the environmental impact parameters include temperature data and air pressure data in the tunnel; An image division module is used to receive the collected infrared spectrum image, divide the infrared spectrum image into a plurality of identification intervals with different wavelengths, and respectively collect gas attribute parameters for each interval, wherein the gas attribute parameters include the detection value of the absorption peak wavelength, the detection value of the light absorption frequency, and the detection value of the light absorption intensity; A data processing module, used for analyzing and processing the collected temperature, air pressure and gas property parameters, correcting the gas property parameters of the sample, generating correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity, and generating a new gas spectrum image according to the correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity; A first data analysis module is used to compare the new gas spectrum image with the standard spectrum image of the sample gas, evaluate the similarity of the new gas spectrum image, and when the similarity between the new gas spectrum image and the standard spectrum image exceeds a set similarity threshold, adjust the position, height and width of the absorption peak according to the peak fitting technology, fit the actual spectrum data, and obtain the actual value of the wavelength of the absorption peak, the actual value of the frequency of the light absorption, and the actual value of the light absorption intensity according to the actual spectrum data; A component judgment module, used to judge the specific components of the gas sample according to the interval range of the actual value of the wavelength of the absorption peak and the actual value of the frequency of light absorption; The second data analysis module is used to convert the actual value of the light absorption intensity of each gas in the sample into a gas concentration according to the Beer-Lambert law, perform dimensionless processing on the gas concentration and the proportional coefficient of each gas in the tunnel, and perform correlation analysis to generate a comprehensive coefficient for evaluating the risk degree of hazards in the tunnel, and compare the comprehensive coefficient with a preset comprehensive coefficient risk threshold; A risk judgment module, used to judge the risk level of harmful gases in the tunnel according to the comparison result of the comprehensive coefficient and a preset comprehensive coefficient risk threshold; The first data analysis module compares the new gas spectrum image with the standard gas spectrum image to determine the gas components contained in the sample in the following process: S11. Collect standard spectrum data: First, it is necessary to collect standard spectrum data containing gas components, which comes from a professional database; S12. Calibration and alignment: Ensure that the new gas spectrum image and the standard spectrum image are aligned in wavelength. Wavelength calibration and spectrum alignment operations are required to ensure that their peaks and troughs correspond; S13. Comparison and matching: comparing the new gas spectrum image with the standard spectrum image by calculating the similarity index between them, including cosine similarity; The second data analysis module converts the actual value of the light absorption intensity of each gas into gas concentration according to the Beer-Lambert law, and the formula is as follows: Among them, A JW is the actual value of the light absorption intensity of methane gas, A LQ is the actual value of the light absorption intensity of hydrogen sulfide gas, A YT is the actual value of the light absorption intensity of carbon monoxide gas, A ET is the actual value of the light absorption intensity of carbon dioxide gas, C JW is the concentration of methane gas, C LQ is the concentration of hydrogen sulfide gas, C YT is the concentration of carbon monoxide gas, C ET is the concentration of carbon dioxide gas, ε JW is the molar absorption coefficient of methane gas, ε LQ is the molar absorption coefficient of hydrogen sulfide gas, ε YT is the molar absorption coefficient of carbon monoxide gas, ε EY is the molar absorption coefficient of carbon dioxide gas, L is the optical path; The process by which the risk judgment module determines the risk level of harmful gases in the tunnel based on the comparison result between the comprehensive coefficient XS and the preset comprehensive coefficient risk threshold YZ is as follows: When XS ≥ YZ, the risk of harmful gases in the tunnel is high; When XS < YZ, the risk of harmful gases in the tunnel is low.

2. According to claim 1, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: Based on the different positions of the prominent absorption peaks generated by the infrared spectral image, the interval with a wavelength range of [λ1, λ2] is divided into several different intervals, where 0 ≤ λ1 ≤ 2.5 and λ2 ≥ 4.

7. In the infrared spectral range, methane has a prominent absorption peak within the wavelength range [3.3, 3.4]; in the infrared spectral range, hydrogen sulfide has a prominent absorption peak within the wavelength range [2.5, 2.6]; in the infrared spectral range, carbon monoxide has a prominent absorption peak within the wavelength range [4.6, 4.7]; in the infrared spectral range, carbon dioxide has a prominent absorption peak within the wavelength range [2.7, 4.3].

3. According to claim 1, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: The data processing module combines the temperature and air pressure with the detected values of the absorption peak wavelength, light absorption frequency, and light absorption intensity respectively, and corrects the values of the detected values of the absorption peak wavelength, light absorption frequency, and light absorption intensity to generate corrected values of the absorption peak wavelength, light absorption frequency, and light absorption intensity. The formulas are as follows: Where α is the proportional factor coefficient of temperature, β is the proportional factor coefficient of air pressure, BJC is the detected value of the absorption peak wavelength, PJC is the detected value of the light absorption frequency, QJC is the detected value of the light absorption intensity, BJZ is the corrected value of the absorption peak wavelength, PJZ is the corrected value of the light absorption frequency, QJZ is the corrected value of the light absorption intensity, α > β > 0, and α + β = 1, and C1 is a constant correction coefficient.

4. According to claim 1, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: In the spectral image comparison, the new gas spectrum and the standard spectrum are represented as vectors, and then the cosine similarity is used to measure the similarity between them. The following are the detailed steps of the process: S131. Represent the spectral data as vectors: Represent the new sample spectral image and the standard spectral image as one-dimensional vectors by arranging the spectral intensity values of each pixel into a vector. If the spectral image has N pixel points, then each vector will have N elements; S132. Calculate the cosine similarity: Use the cosine similarity formula to calculate the similarity between the new sample spectrum and the standard spectrum. The formula for cosine similarity is as follows: in, is the new sample spectral image vector, is the standard spectral image vector, is the norm of the new sample spectral image vector, is the norm of the standard spectrum image vector, CS is the similarity between the new sample spectrum and the standard spectrum; The resulting cosine similarity values ​​are in the range of [-1,1], where 1 means and Completely similar, -1 means and Completely dissimilar, 0 means the angle is 90 degrees, and There is no similarity between vectors; S133. Analyze the results: Based on the calculation result of the cosine similarity, judge the similarity degree between the new gas spectrum and the standard spectrum. When the similarity between the new gas spectral image and the standard spectral image exceeds the set similarity threshold, then the similarity between the new gas spectral image and the standard spectral image is very high, and the gas components of the sample can be roughly judged.

5. According to claim 4, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: After determining the degree of similarity between the new gas spectrum image and the standard spectrum image, the position, height and width of the absorption peak are adjusted according to the peak fitting technology to fit the actual spectrum data. According to the actual spectrum data, the actual value of the wavelength of the absorption peak, the actual value of the frequency of light absorption and the actual value of the light absorption intensity are obtained as follows: The specific process of peak fitting involves nonlinear least squares method, one of which is the Levenberg-Marquardt algorithm. The following is the general process of the Levenberg-Marquardt algorithm, taking fitting a single peak as an example: S21. Initialization parameters: Initially, the initial values ​​of the peak position, height, and width parameters are given; S22. Constructing a model function: Select a Gaussian function to construct a model function that describes the peak shape. For a Gaussian function, the model function is expressed as: Among them, A is the amplitude of the Gaussian peak, which indicates the maximum height of the peak, μ is the center position of the Gaussian peak, that is, the actual value of the wavelength of the peak, σ is the standard deviation of the Gaussian peak, which determines the width of the peak, exp is the abbreviation of the exponential function, expressed as e^x, and B is the background signal; S23. Define the error function: It measures the difference between the model function and the actual spectral data. The typical error function is the residual sum of squares, and the formula is as follows: Where WC is the error value, which measures the difference between the model function and the actual spectral data, N is the number of data points, and xi and yi are the actual data points; S24. Iteration process: Use the Levenberg-Marquardt algorithm to iterate. The goal of the iteration is to minimize the error function by adjusting the model parameters. The algorithm will update the parameters in each iteration so that the model can better fit the actual data. This process is repeated until the convergence condition is reached, the parameter change is small enough, the error is small enough, or the predetermined number of iterations is reached; S25. Extract peak parameters: Once the iteration is completed, extract relevant information of the peak from the final parameters, including the actual value of the wavelength, the actual value of the frequency of light absorption, and the actual value of the light absorption intensity.

6. According to claim 5, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: The process of the component judgment module judging the specific components of the gas sample according to the interval range of the actual value of the wavelength of the absorption peak and the actual value of the frequency of the light absorption is as follows: When the actual value of the wavelength of the absorption peak is between 3.3μm≤BC≤3.4μm, and the actual value of the frequency of light absorption is between 8800Hz≤PL≤9100Hz, if and only if the above conditions are met, it is judged that methane gas exists in the sample; When the actual value of the wavelength of the absorption peak is between 2.5μm≤BC≤2.6μm, and the actual value of the frequency of light absorption is between 3800Hz≤PL≤4000Hz, if and only if the above conditions are met, it is judged that hydrogen sulfide gas exists in the sample; When the actual value of the wavelength of the absorption peak is between 4.6μm≤BC≤4.7μm, and the actual value of the frequency of light absorption is between 2000Hz≤PL≤2220Hz, if and only if the above conditions are met, it is judged that carbon monoxide gas exists in the sample; When the actual value of the wavelength of the absorption peak is between 2.7μm≤BC≤4.3μm, and the actual value of the frequency of light absorption is between 400Hz≤PL≤2500Hz, if and only if the above conditions are met, it is judged that carbon dioxide gas exists in the sample; Among them, BC is the actual value of the wavelength of the absorption peak, and PL is the actual value of the frequency of light absorption.

7. According to claim 5, a tunnel multi-component high-risk and harmful gas detection system based on spectral technology is characterized in that: The concentration values ​​of the methane gas, the concentration values ​​of the hydrogen sulfide gas, the concentration values ​​of the carbon monoxide gas, the concentration values ​​of the carbon dioxide gas and the proportion coefficients of the gases in the tunnel are dimensionlessly processed, and correlation analysis is performed to generate a comprehensive coefficient XS for evaluating the risk level of hazards in the tunnel, and the comprehensive coefficient XS is compared with a preset comprehensive coefficient risk threshold value YZ, wherein the comprehensive coefficient XS is generated according to the following formula: XS=γC JW +δC LQ +μC YT +ρC ET Where γ is C JW The proportionality coefficient, δ is C LQ The proportionality coefficient, μ is C YT The proportionality coefficient, ρ is C ET The proportional coefficient of , δ>μ>γ>ρ>0, and δ+μ+γ+ρ=1.

8. A method for detecting multi-component high-risk and harmful gases in tunnels based on spectral technology, the method is generated based on a multi-component high-risk and harmful gas detection system for tunnels based on spectral technology as described in any one of claims 1-7, characterized in that: The specific steps include: S1. Obtain an infrared spectrum image of a gas sample in a tunnel in a wavelength range of [λ1, λ2], and collect environmental impact parameters in the tunnel to be detected in real time, wherein the environmental impact parameters include temperature data and air pressure data in the tunnel; S2. receiving the collected infrared spectrum image, dividing the infrared spectrum image into a plurality of identification intervals with different wavelengths, and collecting gas attribute parameters for each interval, wherein the gas attribute parameters include the detection value of the absorption peak wavelength, the detection value of the light absorption frequency, and the detection value of the light absorption intensity; S3. Analyze and process the collected temperature, air pressure and gas property parameters, calibrate the gas property parameters of the sample, generate correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity, and generate a new gas spectrum image based on the correction values ​​of absorption peak wavelength, light absorption frequency and light absorption intensity; S4. Compare the new gas spectrum image with the standard spectrum image of the sample gas to evaluate the similarity of the new gas spectrum image. When the similarity between the new gas spectrum image and the standard spectrum image exceeds a set similarity threshold, adjust the position, height and width of the absorption peak according to the peak fitting technology, fit the actual spectrum data, and obtain the actual value of the wavelength of the absorption peak, the actual value of the frequency of light absorption and the actual value of the light absorption intensity according to the actual spectrum data; S5. Determine the specific composition of the gas sample according to the interval range of the actual value of the wavelength of the absorption peak and the actual value of the frequency of light absorption; S6. Convert the actual value of the light absorption intensity of each gas in the sample into a gas concentration according to the Beer-Lambert law, perform dimensionless processing on the gas concentration and the proportional coefficient of each gas in the tunnel, perform correlation analysis, generate a comprehensive coefficient for evaluating the risk level of hazards in the tunnel, and compare the comprehensive coefficient with a preset comprehensive coefficient risk threshold; S7. Determine the risk level of harmful gases in the tunnel based on the comparison result between the comprehensive coefficient and the preset comprehensive coefficient risk threshold.

Images