Mine full laser spontaneous combustion sign gas rapid determination device and prediction method

Abstract

The application is suitable for the technical field of coal mine safety monitoring, and provides a mine full-laser spontaneous combustion sign gas rapid determination device and a prediction method, wherein the mine full-laser spontaneous combustion sign gas rapid determination device comprises an explosion-proof shell, a gas detection unit, a gas path control unit, an environmental parameter detection unit, a data processing and transmission unit and a power module are integrated in the explosion-proof shell; the flow regulation module is used for calculating a target detection flow according to the accuracy evaluation index, the dangerous gas concentration and the dangerous gas concentration change rate; each device transmits detection data and the accuracy evaluation index to the ground monitoring system in real time through a wireless communication module. In the application, the flow regulation module dynamically calculates the target detection flow according to the accuracy evaluation index, the dangerous gas concentration and the change rate thereof, and automatically increases the flow when the detection quality decreases or the dangerous gas concentration rapidly rises, so that the sampling representativeness and the response speed are improved.

Description

Technical Field

[0001] This invention belongs to the field of coal mine safety monitoring technology, and in particular relates to a rapid detection device and prediction method for mine-use all-laser spontaneous combustion indicator gas. Background Technology

[0002] Spontaneous combustion of coal seams is a frequent and significant safety hazard during coal mining. Its early identification relies on the accurate monitoring of various marker gases, including key components such as methane, carbon monoxide, and ethylene. Existing monitoring methods face multiple technical bottlenecks in practical applications.

[0003] Manual sampling and gas chromatography analysis require downhole personnel to periodically collect gas samples and transport them to a surface laboratory. This process is too time-consuming, causing the test results to lag significantly behind the dynamic development of the spontaneous combustion process, failing to meet the needs of real-time early warning. Furthermore, human interference during the operation can easily introduce data deviations. Bundle tube monitoring systems extract gas to the surface for analysis via long-distance pipelines, but the wide pipeline coverage and long gas transmission path result in a significant increase in response time. At the same time, the pipelines are susceptible to blockages or leaks due to the downhole environment, making maintenance difficult. More importantly, single-point detection data lacks spatial correlation, making it difficult to accurately determine the location of the fire source. Single gas sensors, such as electrochemical or infrared types, have poor stability in the high humidity, high dust, and flammable and explosive environments underground. The sensors are easily contaminated, leading to a decrease in sensitivity. Furthermore, they can only monitor a single gas indicator and cannot comprehensively assess the risk of spontaneous combustion of coal by integrating the concentration change trends of multiple marker gases.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] The purpose of this invention is to provide a rapid detection device for spontaneous combustion marking gases using all-laser technology in mining, thereby solving the aforementioned problems.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a rapid gas detection device for mine-use all-laser spontaneous combustion marking, comprising an explosion-proof housing, wherein the explosion-proof housing integrates a gas detection unit, a gas path control unit, an environmental parameter detection unit, a data processing and transmission unit, and a power module; The gas detection unit includes multiple laser sensor modules based on the principle of tunable semiconductor laser absorption spectroscopy, used to detect various spontaneous combustion marker gases; The gas circuit control unit includes a flow sensor and a proportional control valve for real-time monitoring and adjustment of the detection gas flow rate; The data processing and transmission unit includes an accuracy evaluation module and a flow control module; The accuracy evaluation module is used to calculate the accuracy evaluation index based on the measured values ​​of the gas detection unit, the detected values ​​of the environmental parameter detection unit, and the preset interference correction model. The flow control module is used to calculate the target detection flow rate based on the accuracy evaluation index, the concentration of hazardous gas, and the rate of change of hazardous gas concentration. The device is configured to: perform an initial detection with a default flow rate; calculate the accuracy evaluation index of the initial detection using the accuracy evaluation module; if the index exceeds the accuracy evaluation index threshold, trigger the flow control module to adjust the flow rate and perform a second detection with the adjusted target detection flow rate; when the accuracy evaluation index of the second detection still exceeds the accuracy evaluation index threshold, add a low detection confidence indicator to the output result. Multiple devices can be networked together. Each device transmits the detection data and accuracy evaluation index to the ground monitoring system in real time through a wireless communication module. The ground monitoring system uses the reciprocal of the accuracy evaluation index as the data quality weight of each detection point and uses a weighted triangulation positioning algorithm to estimate the location of the fire source.

[0007] A further technical solution is that the accuracy evaluation module calculates an accuracy evaluation index. The formula is: ; in The target gas concentration after correction by the interference correction model; The expected regression value of the interference correction model; To prevent division by zero of small constants; To correct the current regression residuals of the interference correction model; This represents the maximum value of the regression residuals; This represents the actual detection limit at the current background gas concentration. This represents the ideal detection limit under interference-free conditions. It is a concentration of combined interference from multiple gases. gas For gas The interference coefficient; For the first The concentration of the gas; This represents the maximum combined interference concentration. As the deviation weight, For residual weights, To detect the lower limit weight, For joint interference weights.

[0008] Further technical solutions, , , as well as All are linear functions of environmental parameters, and their general expression is: ; in Represents any weight ( , , as well as one of the), , This is the baseline value for the weight under standard conditions. Standard ambient temperature, Standard relative humidity, This refers to the actual ambient temperature. This represents the actual relative humidity. This represents the actual dust concentration. This represents the maximum dust concentration. This is the temperature influence coefficient. Humidity influence coefficient This represents the influence coefficient of dust concentration.

[0009] A further technical solution is that the flow control module calculates the target detection flow. The formula is: ; in Traffic is detected by default; The accuracy assessment index threshold; This represents the current concentration of the hazardous gas. The threshold for hazardous gas warnings; This represents the rate of change in the concentration of hazardous gases. This refers to a single detection sampling cycle; For accuracy weighting, For hazardous gas concentration weights, Weighted by the rate of change of hazardous gas concentration. , .

[0010] A further technical solution is that the accuracy evaluation index threshold... The adaptive threshold is dynamically adjusted based on historical detection results. Specifically, the mean of the historical accuracy evaluation index is calculated, and the product of the confidence coefficient and the standard deviation of the historical accuracy evaluation index is added to obtain the adaptive threshold at the current moment. The historical data is updated using a sliding window method.

[0011] A further technical solution is that the interference correction model adopts a binary quadratic polynomial regression model, which linearly combines the measured values ​​of the methane channel, the original measured values ​​of the ethylene channel, and their square terms and cross terms to obtain the corrected ethylene concentration; wherein each regression coefficient is obtained by least squares fitting through calibration experiments.

[0012] A further technical solution is that the device adopts a two-stage detection lower limit judgment strategy: (1) Detection limit calibration: For each background gas concentration gradient, the detection limit of the target gas at that concentration is determined experimentally, and the detection limit is linearly interpolated in the direction of background gas concentration to obtain the detection limit at any background gas concentration. (2) Detection result judgment: When the corrected target gas concentration is lower than the detection limit of the current background gas concentration, output the lower detection limit indicator; otherwise, output the corrected concentration value and perform amplitude limiting processing to ensure that the concentration value is non-negative.

[0013] In a further technical solution, the gas detection unit includes seven laser sensor modules, which are used to detect methane, carbon monoxide, carbon dioxide, ethylene, acetylene, ethane and oxygen, respectively. Seven laser sensor modules are connected in series via gas pipelines. Each sensor module has an independent air inlet and outlet, and the outlet is connected to the air inlet of the next sensor module.

[0014] A rapid method for determining and predicting spontaneous combustion indicator gases using a full-laser system in mining applications, applied to the aforementioned device, includes the following steps: Step S1, Initial Detection: The control gas circuit control unit draws the gas to be tested at the default flow rate, obtains the concentration measurement values ​​of multiple gases through the gas detection unit, and corrects them through the interference correction model to obtain the corrected concentration; Step S2, calculate the accuracy evaluation index: The accuracy evaluation module calculates the accuracy evaluation index based on the corrected concentration, the regression residual of the interference correction model, the detection limit under the current background gas concentration, the combined interference concentration of multiple gases, and environmental parameters. Step S3, Accuracy Judgment: Compare the accuracy assessment index with the accuracy assessment index threshold. If the accuracy assessment index does not exceed the accuracy assessment index threshold and the concentrations of each hazardous gas are lower than the hazardous gas warning threshold, output the first detection result and proceed to step S6. If the accuracy assessment index exceeds the accuracy assessment index threshold, or the concentration of any hazardous gas reaches the hazardous gas warning threshold and the accuracy assessment index exceeds half of the accuracy assessment index threshold, proceed to step S4. Step S4, Flow Control and Secondary Detection: The flow control module calculates the target detection flow rate based on the current accuracy assessment index, hazardous gas concentration, and hazardous gas concentration change rate, controls the proportional control valve to adjust to the target flow rate, performs secondary detection, reacquires gas concentration data, and recalculates the accuracy assessment index. Step S5, Result Output: If the accuracy evaluation index of the second detection does not exceed the accuracy evaluation index threshold, the second detection result is output; if the accuracy evaluation index of the second detection still exceeds the accuracy evaluation index threshold, the detection result is output and a low detection reliability label is attached; after outputting the result, proceed to step S6. Step S6, fire source location: When multiple devices are networked, each device uploads the detection data and accuracy evaluation index to the ground monitoring system. The ground monitoring system uses the reciprocal of the accuracy evaluation index as the data quality weight of each detection point and uses a weighted triangulation positioning algorithm to estimate the fire source location. Step S7, Fire Trend Prediction: Based on the gas concentration data from multiple consecutive detections, the trend of key indicator gas concentrations is predicted using the exponential smoothing method or the grey prediction model; the prediction results are weighted by the accuracy evaluation index; when the predicted concentration will reach the hazardous gas warning threshold within a preset time window, a fire risk increase warning is output; wherein the key indicator gases include carbon monoxide and ethylene.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: In this invention, an accuracy assessment module is set up to comprehensively correct multiple indicators, including the deviation between the concentration and the expected regression value, the regression residual, the actual detection limit, and the combined interference concentration of multiple gases. The weighting coefficients are dynamically adjusted in conjunction with environmental parameters to achieve a quantitative assessment of the quality of the detection results. This mechanism can effectively identify data with severe interference or low reliability, providing a scientific basis for subsequent flow control and result output, and significantly improving the accuracy and reliability of gas measurement results.

[0016] In this invention, the flow control module dynamically calculates the target detection flow rate based on the accuracy assessment index, the concentration of hazardous gas, and its rate of change. When detection quality declines or the concentration of hazardous gas rises rapidly, the flow rate is automatically increased, improving sampling representativeness and response speed. This intelligent control mechanism enables the device to adapt to complex and changing downhole environments, ensuring detection accuracy while enhancing its responsiveness to sudden risks.

[0017] In this invention, the device integrates seven independent laser sensor modules based on the principle of tunable semiconductor laser absorption spectroscopy, enabling the simultaneous and rapid determination of multiple spontaneous combustion marker gases such as methane, carbon monoxide, and ethylene. By constructing interference correction models such as a binary quadratic polynomial regression model, the device effectively corrects the cross-interference of high-concentration background gases on low-concentration target gases, ensuring detection accuracy under complex multi-component atmospheres.

[0018] In this invention, multiple devices can be networked together to transmit detection data and accuracy assessment indices to a ground monitoring system in real time via wireless communication. The system uses the reciprocal of the accuracy assessment index as the data quality weight and employs a weighted triangulation algorithm to accurately estimate the location of the fire source. By combining continuous detection data with exponential smoothing or a grey prediction model, the system can predict the trends of key indicator gases such as carbon monoxide and ethylene, providing timely warnings before the fire risk increases, and significantly improving the early identification and prevention capabilities of spontaneous combustion in coal mines. Attached Figure Description

[0019] Figure 1 A schematic diagram of a rapid detection device for spontaneous combustion indicator gases using a full laser in mining applications; Figure 2 A schematic diagram illustrating the steps of a rapid determination and prediction method for spontaneous combustion indicator gases using all-laser in mining applications; Figure 3 This is a schematic diagram of the gas circuit control unit; Figure 4 This is a schematic diagram of the data processing and transmission unit. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0021] The specific implementation of the present invention will be described in detail below with reference to specific embodiments.

[0022] like Figure 1 , Figure 3 as well as Figure 4 As shown, this is a mine-use all-laser spontaneous combustion indicator gas rapid detection device provided by an embodiment of the present invention. Its overall structure is integrated within an explosion-proof enclosure. Inside the explosion-proof enclosure, multiple functional modules are integrated, including a gas detection unit, a gas path control unit, an environmental parameter detection unit, a data processing and transmission unit, and a power supply module. These modules work together to achieve rapid and accurate detection and intelligent early warning of spontaneous combustion indicator gases.

[0023] The gas detection unit is the core detection component of the device, containing multiple laser sensor modules based on the principle of tunable semiconductor laser absorption spectroscopy. Tunable semiconductor laser absorption spectroscopy is a method that uses a laser of a specific wavelength to pass through the gas to be measured, determining the gas concentration by measuring the degree to which the laser is absorbed by the gas. This principle has advantages such as high resolution, strong anti-interference capability, and fast response speed, making it suitable for multi-component gas detection in complex environments. These sensor modules are configured to detect multiple spontaneous combustion indicator gases; for example, they can be configured to detect multiple gases such as methane, carbon monoxide, carbon dioxide, and ethylene. As one implementation, each laser sensor module can be designed to detect only one specific gas. By integrating multiple such single-gas sensor modules in parallel or series, simultaneous detection of multiple target gases can be achieved.

[0024] The gas path control unit is responsible for sampling and delivering the gas to be tested. It includes a flow sensor and a proportional control valve. The flow sensor monitors the gas flow rate entering the gas detection unit in real time; for example, a thermal mass flow sensor or a differential pressure flow sensor can be used to obtain accurate flow data. The proportional control valve adjusts the gas flow rate according to a control signal; for example, an electric proportional valve or a pneumatic proportional valve can be used. By receiving instructions from the data processing and transmission unit, it precisely controls the valve opening, thereby achieving real-time monitoring and regulation of the gas flow rate. In one implementation, the flow sensor may only provide the instantaneous flow rate value, while the proportional control valve performs simple open-loop control based on a preset fixed flow rate value.

[0025] The data processing and transmission unit is the intelligent control center of the device, integrating an accuracy assessment module and a flow control module. This unit is responsible for receiving measurement data from the gas detection unit and environmental data from the environmental parameter detection unit, and performing complex calculations and logical judgments. As one implementation method, the data processing and transmission unit can employ a high-performance microcontroller or embedded processor, along with appropriate memory and communication interfaces, to achieve rapid data processing and transmission.

[0026] The accuracy assessment module is a key component of the data processing and transmission unit. It calculates the accuracy assessment index based on measurements from the gas detection unit, environmental parameter detection units, and a pre-defined interference correction model. For example, this module can receive raw gas concentration data from the laser sensor module, as well as temperature and humidity data from the environmental parameter detection units. The interference correction model is used to initially correct the raw measurements based on the environmental parameters. The accuracy assessment index is calculated based on the deviation of the corrected concentration value from a fixed reference value, or by weighted summation based on simple indicators such as the residuals of the interference correction model.

[0027] The flow control module is also integrated into the data processing and transmission unit. It calculates the target detection flow rate based on the accuracy assessment index, the concentration of hazardous gases, and the rate of change of hazardous gas concentration. The target detection flow rate refers to the gas flow rate calculated and set by the flow control module under specific detection conditions to optimize detection accuracy or response speed. By dynamically adjusting the flow rate, the representativeness of sampling and the timeliness of detection can be improved when detection quality declines or hazardous situations occur. For example, this module can receive the index value output by the accuracy assessment module, as well as the concentration of currently detected hazardous gases such as methane and carbon monoxide and their short-term trends. In one implementation, the flow control module can preset a simple lookup table that directly corresponds to a fixed target detection flow rate value based on the range of the accuracy assessment index, the level of the hazardous gas concentration, and the sign of the concentration change rate.

[0028] The device's detection process is configured as follows: First, an initial detection is performed using the default flow rate. After the initial detection, the accuracy evaluation module calculates the accuracy evaluation index of the initial detection. If this index exceeds the accuracy evaluation index threshold, the flow control module is triggered to adjust the flow rate, and a second detection is performed using the adjusted target detection flow rate. If the accuracy evaluation index of the second detection still exceeds the accuracy evaluation index threshold, a "low detection reliability" flag is appended to the output result. For example, a fixed accuracy evaluation index threshold can be set. When the index of the initial detection exceeds this fixed threshold, the system determines that there may be a problem with the detection quality, thereby initiating flow control and a second detection. If the index is still higher than the fixed threshold after the second detection, the user is prompted that the reliability of the current detection result is low.

[0029] Furthermore, this device is designed for multi-unit network operation. Each device transmits detection data and accuracy evaluation index to the ground monitoring system in real time via its built-in wireless communication module. For example, the wireless communication module can use wireless communication technologies such as Wi-Fi, ZigBee, or LoRa to encrypt the data before sending it to the ground monitoring center. After receiving data from different downhole devices, the ground monitoring system uses the reciprocal of the accuracy evaluation index as the data quality weight for each detection point and employs a weighted triangulation algorithm to estimate the location of the fire source. (Weighted triangulation is an algorithm used to estimate the location of a fire source. When multiple devices are networked, this algorithm uses data from different detection points and combines the quality weights of each detection point's data (e.g., the reciprocal of the accuracy evaluation index) to determine the relative position of the fire source through geometric calculations, thereby improving the accuracy and reliability of the positioning.) For example, a ground monitoring system can simply use the reciprocal of the accuracy evaluation index as a weighting factor and then directly apply these weighting factors to the traditional triangulation algorithm to weight the data from different detection points. This allows for the priority acceptance of data from detection points with higher accuracy evaluation indices when there are discrepancies among the data from multiple detection points, thus improving the accuracy of fire source positioning.

[0030] In a preferred embodiment of the present invention, the accuracy evaluation module calculates the accuracy evaluation index. The formula is: ; in The target gas concentration after correction by the interference correction model; The expected regression value of the interference correction model; To prevent the use of a small constant that divides by zero, 0.1% of the sensor's range can be used. (This value ensures that when the expected regression value approaches zero, the contribution of the deviation term will not be abnormally amplified due to the small denominator, while also ensuring that the relative deviation between the corrected concentration and the expected regression value still effectively reflects the detection quality.) As an alternative, It can also be determined through calibration experiments to balance numerical stability and evaluation sensitivity. To correct the current regression residuals of the interference correction model; This represents the maximum value of the regression residuals; This represents the actual detection limit at the current background gas concentration. This represents the ideal detection limit under interference-free conditions. It is a concentration of combined interference from multiple gases. gas For gas The interference coefficient; For the first The concentration of the gas; This represents the maximum combined interference concentration. As the deviation weight, For residual weights, To detect the lower limit weight, For joint interference weights.

[0031] In this embodiment, the accuracy evaluation index Used to quantify the reliability or quality of gas detection results. The higher the value, the greater the interference or uncertainty in the detection results, and the lower the reliability.

[0032] Target gas concentration corrected by interference correction model This is the concentration value of the target gas obtained by the gas detection unit after processing by a preset interference correction model. This value aims to eliminate or reduce the cross-interference effects of background gases, environmental factors, etc., on the target gas measurement, so as to more accurately reflect the actual concentration.

[0033] Regression expected value of the interference correction model This represents the target gas concentration predicted by the perturbation correction model under the current input conditions, based on its internally established regression relationship. It serves as the target gas concentration corrected by the perturbation correction model. A benchmark or reference used to assess the degree of deviation of the correction effect. A small constant to prevent division by zero. When calculating the relative deviation, to avoid the regression expected value A division-by-zero error occurs when the value is close to or equal to zero, introducing a very small positive number. This ensures the numerical stability of the calculation process.

[0034] The current regression residuals of the interference correction model This represents the difference between the predicted and actual observations of the interference correction model during the current detection period. Larger residuals usually indicate a poor fit of the model to the current environment or gas composition, potentially suggesting interference factors that were not adequately considered by the model.

[0035] Maximum value of regression residuals It is the maximum regression residual recorded during the historical operation or calibration of the interference correction model. It is used to evaluate the current regression residual. Normalization is performed to make their weights in the evaluation index comparable.

[0036] Actual detection limit at current background gas concentration This refers to the lowest gas concentration that the sensor can reliably detect. In actual downhole environments, the presence of background gases (such as high concentrations of methane) can affect the detection sensitivity of the target gas, leading to an increase in the actual detection limit. This value reflects the impact of the current environment on detection sensitivity.

[0037] Ideal detection limit under interference-free conditions This represents the lowest detection limit achievable by a target gas sensor in ideal, interference-free environments such as a laboratory. It serves as a benchmark for measuring the current practical detection limit and is used to assess the degree to which environmental interference degrades detection sensitivity.

[0038] Multi-gas combined interference concentration This study quantifies the combined interference effect of multiple coexisting gases, besides the target gas itself, on the detection of the target gas. It does this by analyzing each interfering gas... concentration Rather than the target gas interference coefficient The product is obtained by multiplying and summing, which fully reflects the complexity of cross-interference between multi-component gases.

[0039] Indicates a specific gas For target gas The degree to which measurement results are interfered with. These coefficients are usually obtained through prior experimental calibration and reflect the strength of interactions such as overlap of absorption spectra or chemical reactions between different gases. Concentration of the gas This refers to the real-time concentration of each detected gas other than the target gas in a multi-gas detection scenario. These concentration values ​​are used to calculate the combined interference concentration of multiple gases. Basic data. Maximum combined interference concentration. Multi-gas combined interference concentration The maximum value recorded during historical operation or calibration. It is used to determine the current combined interference concentration. Normalization is performed to make their weights in the evaluation index comparable.

[0040] Deviation weight Residual weights Detection lower limit weight Joint interference weights These weighting coefficients are used to adjust the accuracy assessment index. The relative importance of each component (deviation term, residual term, detection limit term, and joint interference term) is determined. By adjusting these weights, the sensitivity of the evaluation index can be flexibly configured according to the actual application scenario, target gas characteristics, or the degree of attention paid to different interference sources. These weights can be preset to fixed values ​​or dynamically adjusted according to environmental conditions or operating modes.

[0041] The solution proposed in this application introduces a comprehensive accuracy evaluation index. The calculation formula addresses the lack of comprehensive consideration in the quantitative evaluation of detection quality in rapid detection devices for spontaneous combustion marker gases used in mines. This formula organically integrates multiple key factors, including the measured values ​​from the gas detection unit, the detected values ​​from the environmental parameter detection unit, and a pre-set interference correction model, forming a unified evaluation index. Specifically, the formula first calculates the target gas concentration after correction by the interference correction model. Regression expected value of the interference correction model The relative deviation between the two values ​​is used to measure the degree of agreement between the corrected result and the model prediction. A larger deviation indicates that there may be uncertainty in the effect of the correction. Simultaneously, the current regression residuals that interfere with the correction model are considered. This directly reflects the model's goodness of fit to the current detection data; the larger the residual, the lower the model's reliability. Furthermore, this formula also incorporates the actual detection limit at the current background gas concentration. The ideal detection limit under interference-free conditions The ratio is used to quantify the impact of environmental factors (such as background gas concentration) on detection sensitivity, ensuring a true reflection of detection capability under harsh conditions. Furthermore, this is achieved by calculating the combined interference concentration of multiple gases. This formula can comprehensively evaluate the cross-interference effect of multiple coexisting gases on the detection of the target gas, where the interference coefficient is... and the concentration of each gas The summation of the products precisely characterizes this complex interaction.

[0042] After normalization to their respective maximum or benchmark values, the above indicators are then weighted by deviation. Residual weights Detection lower limit weight and joint interference weights By performing a weighted summation, the accuracy assessment index is finally obtained. This weighting mechanism allows the system to flexibly adjust the importance of different interference factors in the overall evaluation according to actual needs. Through the above comprehensive evaluation, the accuracy evaluation module can generate an accuracy evaluation index that comprehensively reflects the current detection data quality and reliability. This index not only considers the accuracy of single-gas measurements but also delves into complex factors such as model fitting effectiveness, environmental influences on sensitivity, and cross-interference among multiple gases. This provides a scientific and quantitative basis for the device's subsequent flow control module, enabling it to intelligently adjust the target detection flow rate based on the actual detection quality, thereby optimizing the detection process. For example, when the accuracy assessment index... A higher value indicates that the detection quality may be compromised, and the flow control module can adjust the flow rate accordingly to attempt to improve detection conditions. Simultaneously, this index provides an objective standard for adding a "low detection reliability" label to the output results, thereby improving the reliability and decision support capabilities of the entire device for rapid determination of spontaneous combustion marker gases in complex downhole environments.

[0043] As a preferred embodiment of the present invention , , as well as All are linear functions of environmental parameters, and their general expression is: ; in Represents any weight ( , , as well as ), , This is the baseline value for the weight under standard conditions. Standard ambient temperature, Standard relative humidity, This refers to the actual ambient temperature. This represents the actual relative humidity. This represents the actual dust concentration. This represents the maximum dust concentration. This is the temperature influence coefficient. Humidity influence coefficient This represents the influence coefficient of dust concentration.

[0044] It should be noted that, to ensure the physical rationality of each weighting coefficient, the calculated weighting coefficient values ​​need to be truncated. Specifically, if the calculated value is less than zero, it is set to zero; if the calculated value is greater than one, it is set to one. This process ensures that each weighting coefficient always remains within the range of zero to one, conforming to its basic definition as a weighting coefficient in the accuracy assessment index. This avoids negative values ​​or exceeding reasonable ranges in the weighting coefficients due to extreme changes in environmental parameters, thereby ensuring the stability and reliability of the accuracy assessment index calculation results.

[0045] In this embodiment, the aforementioned weighting coefficient ( , , as well as In the calculation of the accuracy assessment index, the weights are used to measure the importance of different influencing factors (such as the deviation of the corrected concentration from the expected regression value, regression residual, detection limit, and combined interference concentration of multiple gases). Designing it as a linear function of environmental parameters means that these weights are no longer fixed values, but are dynamically adjusted according to actual temperature, humidity, and dust concentration. This design allows the accuracy assessment index to more realistically reflect the detection quality under the current environment, avoiding assessment distortion caused by fixed weights in complex and variable downhole environments. This general expression provides a unified mathematical model for calculating all four weight coefficients (…). , , as well as It will include environmental parameters (). , , The changes in these environmental factors are compared with benchmark values ​​under standard conditions, and the impact of these environmental changes on the weights is quantified using their respective influence coefficients. This universality simplifies model construction and maintenance and ensures the consistency of the evaluation logic.

[0046] in, Represents any weight ( , , as well as This indicates that the linear function model is applicable to all weight terms in the accuracy evaluation index formula. This is the baseline value of the weighting coefficient under standard conditions, which is the value when the environmental parameters are under standard conditions ( , , When = 0, it corresponds to the initial or ideal value of the weight coefficient. It provides a stable reference point for the dynamic adjustment of the weight coefficient, ensuring that the evaluation model can provide a reasonable weight allocation under standard or near-standard conditions.

[0047] Standard ambient temperature, Standard relative humidity is a preset value representing ideal or typical working conditions for temperature and relative humidity, serving as a benchmark for calculating relative changes in environmental parameters. This refers to the actual ambient temperature. This represents the actual relative humidity. These are the actual dust concentrations, which are real-time environmental condition data acquired by the environmental parameter detection unit (including temperature sensor, humidity sensor and dust concentration sensor), and are the input variables for dynamic adjustment of the weighting coefficients. This represents the maximum dust concentration, used to normalize the dust concentration term and make it comparable in the expression.

[0048] This is the temperature influence coefficient. Humidity influence coefficient These are the dust concentration influence coefficients, which are key parameters for quantifying the degree of influence of changes in temperature, humidity, and dust concentration on specific weighted coefficients. They are usually obtained through experimental calibration or historical data analysis.

[0049] The scheme in this application uses the weighting coefficients of each item in the accuracy evaluation index ( , , as well as Designed as a linear function of environmental parameters, this design solves the problem of inaccurate evaluation in the complex and variable underground environment caused by traditional fixed weights. Specifically, when the mine-use all-laser spontaneous combustion indicator gas rapid detection device is running underground, the environmental parameter detection unit will acquire the current actual ambient temperature in real time. Actual relative humidity and actual dust concentration These real-time environmental parameters are then input into a general expression for the weighting coefficients. This expression uses a preset standard ambient temperature. Standard relative humidity and maximum dust concentration Based on the baseline, and combined with the temperature influence coefficient obtained through experimental calibration. Humidity Influence Coefficient and dust concentration influence coefficient Dynamically calculate the weight coefficients under the current environmental conditions. , , as well as For example, when the ambient temperature rises, if the temperature effect coefficient... If it is a positive value, then The weights will increase accordingly, indicating that temperature is becoming more important in this assessment. These dynamically adjusted weighting coefficients are then used by the accuracy assessment module to calculate the accuracy assessment index. In this way, the accuracy assessment index This allows for a more accurate reflection of the true reliability of gas detection under current environmental conditions. For example, in humid or dusty environments, the weighting coefficients related to humidity or dust may be increased, making the evaluation index more sensitive to these interfering factors and thus identifying the risk of decreased detection quality earlier. This dynamic adjustment mechanism transforms the accuracy evaluation index from a static value into an indicator that adapts to environmental changes. This provides a more reliable basis for the subsequent flow control module to calculate the target detection flow rate and further improves the accuracy of data quality weighting in the ground monitoring system when locating fire sources. Consequently, it enhances the reliability and accuracy of the entire device in rapidly determining and predicting spontaneous combustion marker gases in complex downhole environments.

[0050] In a preferred embodiment of the present invention, the flow control module calculates the target detection flow. The formula is: ; in Traffic is detected by default; The accuracy assessment index threshold; This represents the current concentration of the hazardous gas. The threshold for hazardous gas warnings; This represents the rate of change in the concentration of hazardous gases. This refers to a single detection sampling cycle; For accuracy weighting, For hazardous gas concentration weights, Weighted by the rate of change of hazardous gas concentration. , .

[0051] In this embodiment, the flow control module is the core component of the device responsible for dynamically managing the gas sampling flow rate. Its main function is to intelligently calculate the optimal target detection flow rate based on various parameters monitored in real time, and to precisely regulate the gas flow rate by controlling the proportional control valve in the gas path control unit. This module can be an embedded processor running a preset flow control algorithm, or a programmable logic controller (PLC) that implements complex logical judgments and flow calculations through programming.

[0052] Target detection traffic This is the gas sampling flow rate dynamically determined by the device under specific detection conditions to ensure detection accuracy and response speed. It is not a fixed value but is adjusted in real time according to the environment and detection requirements. This flow rate can drive a proportional control valve by outputting a voltage or current signal through a digital-to-analog converter, or control the opening of the proportional control valve through a pulse width modulation signal.

[0053] Default traffic detection This is the preset baseline gas sampling flow rate for the device under normal or initial detection conditions. It serves as the starting point for flow rate control calculations and represents the detection flow rate typically used by the device when no special circumstances occur. This default flow rate can be experimentally calibrated and embedded in the firmware at the factory, based on typical application scenarios and sensor characteristics, or it can be manually set by the user in the device's configuration interface.

[0054] Accuracy assessment index It is a comprehensive indicator for measuring the reliability of current gas detection results. It reflects the degree of deviation between the detected value and the expected value, the residual of the interference correction model, the detection limit, and the impact of factors such as multi-gas joint interference on the detection quality. This index can be calculated in real time by the accuracy assessment module according to a preset formula, or it can be predicted in real time by statistically analyzing historical data and establishing a machine learning model.

[0055] Accuracy assessment index threshold This is a critical value used to determine whether the current detection accuracy is acceptable. When the accuracy assessment index exceeds this threshold, it indicates that the reliability of the detection results may be affected, and further measures are needed. This threshold can be determined during the device design phase through extensive experiments and data analysis, or it can be dynamically adjusted based on historical detection data and environmental changes using an adaptive algorithm.

[0056] Current concentration of hazardous gas This refers to the actual concentration value of gaseous components identified as potentially hazardous, detected in real time by the device. This concentration value is a key parameter for assessing the current environmental safety status and determining whether the detection strategy needs to be adjusted. It is usually measured directly by the laser sensor module in the gas detection unit and corrected by an interference correction model, or obtained by fusing measurements from multiple sensors.

[0057] Hazardous gas warning threshold This is a safety concentration limit set for a specific hazardous gas. When the current concentration of a hazardous gas reaches or exceeds this threshold, it usually indicates a potential hazard. This threshold can be preset and stored according to national or industry safety standards, or it can be configured by the user based on the specific conditions of the mine.

[0058] Rate of change of hazardous gas concentration This refers to the rate at which the concentration of a hazardous gas changes over time. This parameter reflects the trend and urgency of the gas concentration change. It can be calculated by differentially analyzing multiple consecutively detected hazardous gas concentration values, or by smoothing the concentration data using algorithms such as Kalman filtering.

[0059] Single detection sampling cycle Accuracy weighting refers to the time required for the device to complete a full gas sampling, detection, data processing, and result output cycle. This cycle determines the device's response speed to environmental changes. It is typically determined by the device's hardware performance and software algorithms, but can also be dynamically adjusted according to detection requirements. Hazardous gas concentration weighting Weighting of the rate of change of hazardous gas concentration These are parameters used to balance the impact of different factors on target detection traffic. They can be calibrated and optimized through extensive field experiments and expert experience, and then embedded in firmware, or dynamically adjusted using adaptive control or machine learning methods.

[0060] When the accuracy assessment index is lower than the accuracy assessment index threshold, the first adjustment term is negative, and the target detection flow rate may be lower than the default flow rate. This situation applies when the detection quality is excellent and there are no risk signals; appropriately reducing the flow rate can reduce power consumption and sampling gas consumption. To ensure detection sensitivity, the target detection flow rate is limited by a preset minimum flow rate and will not fall below the minimum flow rate required for stable sensor operation.

[0061] Minimum and maximum flow rates are upper and lower limits set to ensure normal operation and detection effectiveness of the device. The minimum flow rate ensures that the gas can effectively enter the sensor for detection, while the maximum flow rate prevents damage to the sensor due to excessive flow. These limits can be determined based on the physical characteristics of the gas path system and the sensor's operating range, or adjusted through software configuration.

[0062] This application's solution provides a specific calculation formula that enables the flow control module to dynamically adjust the detection flow rate precisely based on multiple key factors, thereby optimizing the gas detection process. Specifically, the formula is based on the default flow rate. Adjustments are made, among which The item is evaluated based on the accuracy index. Accuracy assessment index threshold The flow rate is adjusted according to the deviation ratio; when the accuracy is low, the flow rate is increased to improve the detection accuracy. Based on the concentration of hazardous gases Hazardous gas warning threshold Adjust the flow rate proportionally, and increase the flow rate at high concentrations to enhance detection sensitivity; The item is based on the rate of change of hazardous gas concentration. With standardized rate of change The flow rate is adjusted proportionally to quickly respond to dynamic changes in gas concentration and prevent missed detections. Weighting coefficient. , , The value ranges from 0.2 to 0.8, ensuring a reasonable adjustment range and avoiding detection instability caused by over-adjustment. Furthermore, the target detection flow rate is limited by preset minimum and maximum flow rates, ensuring the flow rate remains within a safe and effective range and preventing extreme values ​​from affecting device performance. This comprehensive adjustment mechanism effectively solves the problem of inaccurate flow rate regulation, improving the adaptability and reliability of the detection. The synergistic effect of this flow rate control mechanism with the gas detection unit, gas path control unit, environmental parameter detection unit, and data processing and transmission unit within the device enables the device to achieve real-time evaluation of detection quality and intelligent optimization of sampling strategies within a unified framework, thereby maintaining a high level of detection performance in complex and changing environments.

[0063] As a preferred embodiment of the present invention, the accuracy evaluation index threshold The threshold is adaptive and dynamically adjusted based on historical detection results, specifically as follows: ; in Let t be the accuracy evaluation index threshold. This is the average of the historical accuracy assessment index. The standard deviation of the historical accuracy assessment index. , is the confidence coefficient Historical data for the historical accuracy assessment index is updated using a sliding window method.

[0064] In this embodiment, the aforementioned adaptive threshold is a threshold that can automatically adjust its value based on the system's operating status, environmental conditions, or historical data. Its purpose is to address the problem of poor adaptability of fixed thresholds in dynamic environments, thereby improving the accuracy and robustness of the judgment. This adaptive threshold can be based on statistical methods, such as calculating it using the mean and standard deviation of historical data as in this application; alternatively, it can be based on machine learning algorithms to predict the optimal threshold through model training; or it can be based on fuzzy logic control to dynamically adjust the threshold according to multiple input parameters (such as environmental change rate and gas concentration fluctuations).

[0065] The dynamic adjustment based on historical detection results aims to ensure that the threshold can reflect changes in actual detection conditions in real time, avoiding misjudgments caused by environmental fluctuations. Besides using a sliding window mechanism to retain only recent data for statistical calculations to quickly respond to environmental changes, methods such as exponentially weighted moving averages can also be used to assign different weights to historical data, making recent data have a greater impact.

[0066] The mean of the historical accuracy assessment index This refers to the arithmetic mean of all accuracy assessment indices over a period of time. Its purpose is to provide a benchmark accuracy level, reflecting the average performance of the system under normal operating conditions. In implementation, a queue of historical accuracy assessment indices can be maintained in the data storage module of the data processing and transmission unit, and the average of all values ​​in the queue can be taken for each calculation; alternatively, an accumulation and counting method can be used to update the mean in real time.

[0067] The standard deviation of the historical accuracy assessment index It is an indicator that measures the dispersion of historical accuracy assessment index data. Its function is to quantify the volatility of the data, so that the accuracy assessment index threshold can accommodate normal variation and avoid oversensitivity. Similar to mean calculation, the data can be maintained in the data storage module and its standard deviation can be calculated; alternatively, an online algorithm can be used to update the mean and standard deviation in real time to reduce storage requirements.

[0068] The confidence coefficient This is a configurable parameter used to adjust the stringency of the accuracy assessment index threshold. Its function is to balance the sensitivity and reliability of the threshold and can be adjusted according to actual application requirements (such as tolerance for false alarms and false negatives). This coefficient can be manually set in the device's configuration interface or ground monitoring system, usually selected within the range of [2,3]; alternatively, it can be automatically recommended or adjusted by an expert system or based on historical data analysis.

[0069] The historical accuracy assessment index uses a sliding window method to update historical data. This is a data processing mechanism that retains only the most recent N data points for calculation, removing older data points. Its purpose is to ensure that historical data promptly reflects the current environment and system state, avoiding interference from older data in current judgments, thereby improving system response speed. In implementation, a first-in, first-out (FIFO) queue can be used to store historical data, removing the oldest data as new data arrives; alternatively, a fixed-size buffer can be allocated in the memory of the data processing and transmission unit, managing data access and updates through pointers or indexes.

[0070] The solution proposed in this application addresses the problem of insufficient adaptability of fixed thresholds in dynamic mining environments by designing the accuracy assessment index threshold as an adaptive threshold. Specifically, the data processing and transmission unit in the device continuously collects the accuracy assessment index and dynamically calculates the adaptive accuracy assessment index threshold using this historical data. The dynamically adjusted threshold is then compared to the current accuracy assessment index by the flow control module. When the current accuracy assessment index exceeds this adaptive threshold, the flow control module triggers flow regulation. This dynamic adjustment mechanism ensures that the threshold for triggering flow regulation is always correlated with current operating conditions and historical performance, thereby avoiding unnecessary adjustments in stable but slightly changing environments and ensuring timely adjustments in the event of actual anomalies. In this way, this scheme significantly improves the decision-making quality of the flow control module, thereby enhancing the measurement accuracy and reliability of the entire device in complex and variable environments.

[0071] In a preferred embodiment of the present invention, the interference correction model adopts a bivariate quadratic polynomial regression model: ; in This is the corrected ethylene concentration. This is the measurement value for the methane channel. These are the original measurements from the ethylene channel. These are the regression coefficients obtained by fitting the data using the least squares method through calibration experiments.

[0072] In this embodiment, the solution of this application solves the problem of accurate correction of cross-interference of specific gases by using a binary quadratic polynomial regression model as the interference correction model. This model can effectively capture the nonlinear relationship and complex interaction between methane and ethylene concentrations, including linear, quadratic, and cross-terms, thereby more accurately describing the interference effect and avoiding correction bias caused by simple linear models. Specifically, the formula incorporates the measured value of the methane channel. and raw measurements of the ethylene channel As a core variable, the model directly targets the interference source of methane on ethylene, ensuring that it specifically addresses the impact of overlapping spectral lines of high-concentration methane on the detection of low-concentration ethylene. Simultaneously, including quadratic and cross-terms of these variables, the model comprehensively reflects the nonlinear contribution of concentration changes to the interference, enhancing its adaptability in complex environments. Regression coefficients were obtained through calibration experiments using the least squares method. Parameter optimization based on actual data enhances the model's accuracy and generalization ability, avoiding overfitting or underfitting. Overall, this technical solution accurately compensates for cross-interference through mathematical modeling, making the corrected ethylene concentration closer to the true value, providing a reliable foundation for subsequent accuracy assessment.

[0073] The bivariate quadratic polynomial regression model is a statistical model used to describe the nonlinear relationship between two independent variables and one dependent variable. By including linear, quadratic, and interaction terms of the independent variables, it captures more complex interactions, thus enabling more accurate fitting of nonlinear data and handling complex interactive interference between variables. This model can be implemented using a programming language within a data processing and transmission unit, or constructed using specialized mathematical modeling software. Regression coefficients These are the parameters in the model, representing the weight or degree of influence of each term on the dependent variable, determining the specific shape and fitting effect of the model. These coefficients can be obtained through preset empirical or theoretical values, but a more precise method is to calculate and optimize them through data fitting methods. Calibration experiments are controlled experiments that measure the device's response under known conditions to obtain a dataset for model training and parameter determination, thus providing real and reliable data for model training and validation. Calibration experiments can use a high-precision gas mixing system to prepare standard gases with different concentration gradients and mixing ratios, and measure them under controlled environments (such as temperature, humidity, and pressure). Least squares fitting is a commonly used mathematical optimization technique that finds the best function fit for the data by minimizing the sum of squared residuals. It is used to calculate the optimal regression coefficients from calibration experiment data, maximizing the model's fit to the data and minimizing error. This method can be used to solve normal equations through matrix operations or to approximate the optimal solution using iterative optimization algorithms such as gradient descent.

[0074] The proposed solution in this application, within a rapid gas detection device for spontaneous combustion indicators in mines using a full-laser process, involves processing the measured values ​​from the gas detection unit via a data processing and transmission unit. When the gas detection unit detects methane and ethylene, the data processing and transmission unit first acquires the measured value from the methane channel. and raw measurements of the ethylene channel Subsequently, the data processing and transmission unit substitutes these measurements into a bivariate quadratic polynomial regression model, pre-fitted through calibration experiments and the least squares method. This model, utilizing its linear, quadratic, and cross terms, accurately calculates the nonlinear interference effect of methane on ethylene detection, and accordingly corrects the original measurements in the ethylene channel, thus obtaining the corrected ethylene concentration. This correction mechanism effectively compensates for the spectral overlap interference of high-concentration methane on the detection of low-concentration ethylene, making the corrected ethylene concentration closer to the true value. The corrected ethylene concentration, as a more accurate input, is used by the accuracy evaluation module to calculate the accuracy evaluation index, thereby improving the reliability of the entire device's gas detection accuracy assessment. In this way, the scheme of this application is closely integrated with the overall architecture of the mine-use all-laser spontaneous combustion marker gas rapid determination device, enabling the device to provide more reliable gas concentration data in complex underground environments, thus supporting more accurate fire source location and fire trend prediction.

[0075] In a preferred embodiment of the present invention, the device employs a two-stage detection lower limit determination strategy: (1) Detection limit calibration: For each background gas concentration gradient, the detection limit of the target gas at that concentration is determined experimentally, and the detection limit is linearly interpolated in the direction of background gas concentration to obtain the detection limit at any background gas concentration. (2) Detection result judgment: When the corrected target gas concentration is lower than the detection limit of the current background gas concentration, output the "lower than detection limit" flag; otherwise, output the corrected concentration value and perform amplitude limiting processing to ensure that the concentration value is non-negative.

[0076] In this embodiment, the two-stage detection limit determination strategy aims to address the inaccuracy of traditional fixed detection limits in complex and variable environments (such as mines). By separating the determination of the detection limit from the judgment of the actual detection result and introducing a dynamic adjustment mechanism, the reliability of low-concentration gas detection is improved. Its core lies in dynamically adjusting the detection limit based on environmental conditions (especially background gas concentration) and rigorously validating the final output result.

[0077] The detection limit calibration stage is used to establish a detection limit model for the target gas under different background gas concentrations. Specifically, under laboratory conditions, a series of background gases with known concentrations (e.g., methane, carbon dioxide, etc.) are configured, and the concentration of the target gas (e.g., ethylene, carbon monoxide, etc.) is gradually reduced at each background gas concentration gradient until the gas detection unit can no longer reliably detect the target gas, thereby determining the detection limit at that background gas concentration. This process can be repeated multiple times to ensure the accuracy and statistical significance of the data. After obtaining discrete background gas concentrations and corresponding detection limit data points, linear interpolation methods, such as Lagrange interpolation, Newton interpolation, or simple piecewise linear interpolation, can be used to construct a continuous functional relationship, enabling the real-time calculation of the corresponding target gas detection limit for any given background gas concentration. This dynamic calibration method ensures that the detection limit can adapt to changes in background gas concentrations under actual operating conditions.

[0078] The detection result judgment stage is used to determine and process the validity of the actual measurement results based on the dynamically determined detection lower limit. When the target gas concentration measured by the gas detection unit and corrected by the interference correction model is lower than the current detection lower limit obtained from the detection lower limit calibration stage corresponding to the current background gas concentration, the data processing and transmission unit will output a clear "below detection lower limit" flag. This indicates that under these environmental conditions, the actual concentration of the target gas may be too low to be reliably detected by the device, thus avoiding inaccurate output values. Conversely, if the corrected target gas concentration is not lower than the current detection lower limit, the data processing and transmission unit will output the corrected concentration value. To ensure the physical rationality of the data output, i.e., the concentration value cannot be negative, the data processing and transmission unit will also perform amplitude limiting processing on the concentration value. For example, if the corrected concentration value is slightly less than zero due to measurement noise or other factors, it will be forcibly set to zero to ensure the non-negativity of the output result.

[0079] The device described in this application effectively solves the problems of insufficient accuracy and easy misjudgment of traditional fixed detection limits in complex and variable environments, especially when detecting low-concentration gases, through the aforementioned two-stage detection limit determination strategy. Before the device is put into use or during periodic maintenance, the actual detection limit of the target gas is accurately determined under multiple preset background gas concentration gradients through controlled experiments. These discrete experimental data points are then used to construct a continuous functional relationship, for example, through a linear interpolation algorithm, enabling the device to dynamically calculate the target gas detection limit under the current environmental conditions based on the real-time monitored background gas concentration. This process makes the detection limit no longer fixed but can respond to environmental changes in real time, thereby improving the adaptability of the detection. In actual operation, after the gas detection unit detects the target gas concentration and corrects it through the interference correction model, the data processing and transmission unit obtains the background gas concentration provided by the current environmental parameter detection unit and calculates the dynamic detection limit under the current background gas concentration using the model established in the calibration stage. Subsequently, the corrected target gas concentration is compared with this dynamic detection limit. If the corrected concentration is below the dynamic detection lower limit, it indicates that the reliability of the detection result is low. The data processing and transmission unit will output a "below detection lower limit" flag, rather than a potentially inaccurate value, thus effectively avoiding false alarms or underestimation of risk. Conversely, if the corrected concentration is not below the dynamic detection lower limit, the corrected concentration value will be output. To ensure the physical meaning of the data output, the data processing and transmission unit will also perform amplitude limiting processing on the concentration value, such as forcing any negative value to zero, to ensure the rationality and non-negativity of the output data. Through this two-stage strategy, the device of this application can overcome the limitations of traditional fixed detection lower limits, enabling it to provide more accurate and reliable detection results in complex and variable mine environments, especially when detecting low-concentration target gases. This, combined with the gas detection unit's ability to detect multiple spontaneous combustion marker gases and the interference correction model's handling of cross-interference, jointly improves the accuracy and reliability of the entire device for early warning of coal spontaneous combustion.

[0080] In a preferred embodiment of the present invention, the gas detection unit includes seven laser sensor modules, each used to detect methane (…). ), carbon monoxide ( ),carbon dioxide( ), ethylene ( ), acetylene ( ), ethane ( ) and oxygen ( ); Seven laser sensor modules are connected in series via gas pipelines. Each sensor module has an independent air inlet and outlet, and the outlet is connected to the air inlet of the next sensor module.

[0081] In this embodiment, these independent laser sensor modules are designed based on the principle of Tunable Semiconductor Laser Absorption Spectroscopy (TDLAS), with each module specifically designed to detect a particular spontaneous combustion marker gas. Independence is achieved through each module having its own independent laser source, detector, and gas absorption cell, or through optical multiplexing technology to create physically independent detection channels. This design ensures high selectivity and sensitivity for specific gases, avoiding direct optical or electronic interference between different gas detection channels. The seven selected gases (methane, carbon monoxide, carbon dioxide, ethylene, acetylene, ethane, and oxygen) are key marker gases produced during spontaneous combustion in coal mines, comprehensively covering the early, middle, and late stages of combustion, providing a comprehensive data foundation for accurately determining combustion trends. As one implementation, each independent laser sensor module can include a dedicated distributed feedback laser whose emission wavelength is precisely locked to the characteristic absorption peak of the target gas, and equipped with a corresponding photodetector and signal processing circuitry. Alternatively, each independent laser sensor module can employ a vertical-cavity surface-emitting laser (VCSEL), which has a smaller size and lower power consumption, while still achieving accurate detection of specific gases.

[0082] Series connection refers to the gas to be measured flowing sequentially through the absorption cells of each laser sensor module. This connection method ensures that the gas has been detected by the previous module before entering the next module, allowing the order of gas flow through each module to be optimized based on preset detection strategies or gas characteristics. For example, the detection module for high-concentration or easily interfering gases (such as methane) can be placed earlier to reduce its potential impact on the detection of subsequent low-concentration or sensitive gases (such as ethylene). As one implementation method, the gas pipeline can be made of corrosion-resistant, low-adsorption polytetrafluoroethylene (PTFE) or stainless steel tubing, connecting the inlets and outlets of each sensor module through standard connectors (such as compression fittings or threaded fittings). Alternatively, an integrated gas flow path board can be designed, with each sensor module directly mounted on it. Precision channels within the flow path board enable series flow of gas between modules, reducing external piping and connectors and improving the system's compactness and reliability.

[0083] Independent inlets and outlets ensure that each sensor module can independently receive and discharge gas, maintaining a stable gas environment within its internal absorption cell. The design of the outlet connecting to the inlet of the next sensor module clearly defines the unidirectional flow path of gas in the series system, preventing gas backflow or unnecessary mixing between modules. This ensures that the gas composition detected by each module is the result of processing by its upstream module, effectively avoiding cross-contamination and the accumulation of detection errors. As one implementation, the inlets and outlets can be controlled using miniature solenoid valves or miniature pneumatic valves to achieve precise control of the gas flow to and from each module, and to perform purging or calibration when necessary. Alternatively, the inlets and outlets can be designed as interfaces with specific dimensions and shapes, tightly connected to the gas pipeline via quick-connect or threaded joints, ensuring airtightness and facilitating maintenance.

[0084] The proposed solution configures the gas detection unit to include seven laser sensor modules connected in series via gas pipelines, allowing the target gas to flow through each module sequentially in a preset order. Each module is dedicated to detecting a specific spontaneous combustion marker gas, such as methane, carbon monoxide, carbon dioxide, ethylene, acetylene, ethane, and oxygen. This independent module design, combined with the high selectivity and high sensitivity based on the principle of tunable semiconductor laser absorption spectroscopy, ensures specificity for the detection of each target gas. By precisely connecting the independent inlet and outlet of each sensor module—that is, connecting the outlet of one module to the inlet of the next—a clear, unidirectional gas flow path is constructed. This series structure ensures that the composition of the gas has already been detected by the preceding module as it flows through subsequent modules, allowing for strategic arrangement of the detection order. For example, the detection module for high-concentration gases (such as methane) that may cause strong cross-interference can be placed at the front of the flow path to reduce its impact on the detection of subsequent low-concentration or easily interfered-with gases (such as ethylene). This meticulous gas flow path design effectively reduces cross-interference issues when measuring multiple components of gas simultaneously, improving the accuracy of concentration measurements for each marker gas. Furthermore, this configuration works collaboratively with other units in the device (such as the gas path control unit and data processing and transmission unit). For instance, the gas path control unit precisely controls the gas flow rate, ensuring stable gas flow within the series modules, while the data processing and transmission unit receives and processes detection data from each module, further correcting it using an interference correction model. Together, they achieve rapid, accurate, and low-interference measurement of various spontaneous ignition marker gases.

[0085] like Figure 2 As shown, the rapid determination and prediction method for spontaneous combustion marker gases using a full laser in mining, applied to the aforementioned device, includes the following steps: Step S1, Initial Detection: The control gas circuit control unit draws the gas to be tested at the default flow rate, obtains the concentration measurement values ​​of multiple gases through the gas detection unit, and corrects them through the interference correction model to obtain the corrected concentration; the interference correction model can effectively correct the cross-interference of multiple components of gas, such as the interference correction of the ethylene detection channel in the presence of high concentration of methane, to ensure the reliability of the initial data.

[0086] Step S2, calculate the accuracy assessment index: The accuracy assessment module calculates the accuracy assessment index based on the corrected concentration, the regression residual of the interference correction model, the detection limit under the current background gas concentration, the combined interference concentration of multiple gases, and environmental parameters. This index comprehensively quantifies the reliability of the detection quality and provides a scientific basis for subsequent decision-making.

[0087] Step S3, Accuracy Judgment: Compare the accuracy assessment index with the accuracy assessment index threshold. If the accuracy assessment index does not exceed the accuracy assessment index threshold and the concentrations of each hazardous gas are lower than the hazardous gas warning threshold, output the first detection result and proceed to step S6. If the accuracy assessment index exceeds the accuracy assessment index threshold, or the concentration of any hazardous gas reaches the hazardous gas warning threshold and the accuracy assessment index exceeds half of the accuracy assessment index threshold, proceed to step S4. Step S4, Flow Control and Secondary Detection: The flow control module calculates the target detection flow rate based on the current accuracy assessment index, hazardous gas concentration, and hazardous gas concentration change rate, controls the proportional control valve to adjust to the target flow rate, performs secondary detection, reacquires gas concentration data, and recalculates the accuracy assessment index. By dynamically optimizing the flow rate, the system can adapt to detection needs when environmental changes or hazardous signals occur. For example, when the carbon monoxide concentration rises rapidly, the target detection flow rate is increased to accelerate the sampling speed, thereby improving the representativeness and timeliness of the data.

[0088] Step S5, Result Output: If the accuracy evaluation index of the second detection does not exceed the accuracy evaluation index threshold, the second detection result is output; if the accuracy evaluation index of the second detection still exceeds the accuracy evaluation index threshold, the detection result is output and a "low detection reliability" label is added; after outputting the result, proceed to step S6. Step S6, Fire Source Location: When multiple devices are networked, each device uploads its detection data and accuracy evaluation index to the ground monitoring system. The ground monitoring system uses the reciprocal of the accuracy evaluation index as the data quality weight of each detection point and uses a weighted triangulation positioning algorithm to estimate the fire source location. This weighted mechanism prioritizes high-quality data, which significantly improves the accuracy of fire source location.

[0089] It also includes step S7, fire trend prediction: based on the gas concentration data from multiple consecutive detections, the trend prediction of the concentration of key indicator gases is performed using the exponential smoothing method or the grey prediction model; the prediction results are weighted by the confidence level in conjunction with the accuracy evaluation index; when the predicted concentration will reach the dangerous gas warning threshold within a preset time window, a "fire risk increase" warning is output; wherein the key indicator gases include carbon monoxide and ethylene.

[0090] Through the above technical solutions, this application achieves rapid synchronous detection of multiple gases, cross-interference correction, quantitative evaluation of detection quality and adaptive sampling optimization, as well as intelligent early warning and remote collaboration. Compared with existing technologies, this method significantly improves the real-time performance, accuracy, and early warning capabilities of detection, providing reliable technical support for coal mine safety management.

[0091] 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, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A rapid detection device for spontaneous combustion marking gases using all-laser technology in mining, characterized in that, The system includes an explosion-proof enclosure, which integrates a gas detection unit, a gas path control unit, an environmental parameter detection unit, a data processing and transmission unit, and a power module. The gas detection unit includes multiple laser sensor modules based on the principle of tunable semiconductor laser absorption spectroscopy, used to detect various spontaneous combustion marker gases; The gas circuit control unit includes a flow sensor and a proportional control valve for real-time monitoring and adjustment of the detection gas flow rate; The data processing and transmission unit includes an accuracy evaluation module and a flow control module; The accuracy evaluation module is used to calculate the accuracy evaluation index based on the measured values ​​of the gas detection unit, the detected values ​​of the environmental parameter detection unit, and the preset interference correction model. The flow control module is used to calculate the target detection flow rate based on the accuracy evaluation index, the concentration of hazardous gas, and the rate of change of hazardous gas concentration. Multiple units of the device can be networked together, and each device transmits the detection data and accuracy evaluation index to the ground monitoring system in real time through a wireless communication module.

2. The apparatus according to claim 1, characterized in that, The accuracy evaluation module calculates the accuracy evaluation index. The formula is: ； in The target gas concentration after correction by the interference correction model; The expected regression value of the interference correction model; To prevent division by zero of small constants; To correct the current regression residuals of the interference correction model; This represents the maximum value of the regression residuals; This represents the actual detection limit at the current background gas concentration. This represents the ideal detection limit under interference-free conditions. For the combined interference concentration of multiple gases, , gas For gas The interference coefficient; For the first The concentration of the gas; This represents the maximum combined interference concentration. For deviation weight, For residual weights, To detect the lower limit weight, For joint interference weights.

3. The apparatus according to claim 2, characterized in that, , , as well as All are linear functions of environmental parameters, and their general expression is: ； in represent , , , one of the, , This is the baseline value for the weight under standard conditions. Standard ambient temperature, Standard relative humidity, This refers to the actual ambient temperature. This represents the actual relative humidity. This represents the actual dust concentration. This represents the maximum dust concentration. This is the temperature influence coefficient. Humidity influence coefficient This represents the influence coefficient of dust concentration.

4. The apparatus according to claim 2, characterized in that, The flow control module calculates the target detection flow. The formula is: ； in Traffic is detected by default; The accuracy assessment index threshold; This represents the current concentration of the hazardous gas. The threshold for hazardous gas warnings; The rate of change of hazardous gas concentration; This refers to the sampling cycle for a single test. For accuracy weighting, For hazardous gas concentration weights, Weighted by the rate of change of hazardous gas concentration. , .

5. The apparatus according to claim 4, characterized in that, The accuracy evaluation index threshold is an adaptive threshold that is dynamically adjusted based on historical detection results. Specifically, the mean of the historical accuracy evaluation index is calculated, and the product of the confidence coefficient and the standard deviation of the historical accuracy evaluation index is added to obtain the adaptive threshold at the current moment.

6. The apparatus according to claim 1, characterized in that, The interference correction model adopts a binary quadratic polynomial regression model, which linearly combines the measured values ​​of the methane channel, the original measured values ​​of the ethylene channel, and their square and cross terms to obtain the corrected ethylene concentration; the regression coefficients are obtained by least squares fitting through calibration experiments.

7. The apparatus according to claim 1, characterized in that, The device employs a two-stage detection lower limit judgment strategy: Detection limit calibration: For each background gas concentration gradient, the detection limit of the target gas at that concentration is determined experimentally, and the detection limit is linearly interpolated in the direction of background gas concentration to obtain the detection limit at any background gas concentration. Detection result determination: When the corrected target gas concentration is lower than the detection limit of the current background gas concentration, output a lower detection limit indicator; otherwise, output the corrected concentration value and perform amplitude limiting to ensure that the concentration value is non-negative.

8. The apparatus according to claim 1, characterized in that, The gas detection unit includes seven laser sensor modules, which are used to detect methane, carbon monoxide, carbon dioxide, ethylene, acetylene, ethane and oxygen, respectively. Seven laser sensor modules are connected in series via gas pipelines. Each sensor module has an independent air inlet and outlet, and the outlet is connected to the air inlet of the next sensor module.

9. A rapid method for determining and predicting spontaneous combustion marker gases using a full-laser method in mining, applied to the apparatus described in any one of claims 1-8, characterized in that, include: S1 controls the gas path control unit to draw the gas to be tested at the default flow rate, obtains the concentration measurement values ​​of multiple gases through the gas detection unit, and obtains the corrected concentration through the interference correction model; S2, the accuracy assessment module calculates the accuracy assessment index based on the corrected concentration, the regression residual of the interference correction model, the detection limit under the current background gas concentration, the combined interference concentration of multiple gases, and environmental parameters. S3. Compare the accuracy assessment index with the accuracy assessment index threshold. If the accuracy assessment index does not exceed the accuracy assessment index threshold and the concentrations of each hazardous gas are lower than the hazardous gas warning threshold, output the first detection result and proceed to step S6. If the accuracy assessment index exceeds the accuracy assessment index threshold, or the concentration of any hazardous gas reaches the hazardous gas warning threshold and the accuracy assessment index exceeds half of the accuracy assessment index threshold, proceed to step S4. S4, the flow control module calculates the target detection flow rate based on the current accuracy assessment index, hazardous gas concentration, and hazardous gas concentration change rate, controls the proportional control valve to adjust to the target flow rate, performs secondary detection, reacquires gas concentration data, and recalculates the accuracy assessment index. S5. If the accuracy evaluation index of the second test does not exceed the accuracy evaluation index threshold, the second test result is output; if the accuracy evaluation index of the second test still exceeds the accuracy evaluation index threshold, the test result is output and a "low test reliability" label is added; after outputting the result, proceed to step S6. S6 When multiple devices are networked, each device uploads its detection data and accuracy evaluation index to the ground monitoring system. The ground monitoring system uses the reciprocal of the accuracy evaluation index as the data quality weight of each detection point and uses a weighted triangulation positioning algorithm to estimate the location of the fire source.

Images