Mine return air corner low-oxygen multi-source cause tracing method and early warning system

Abstract

The invention discloses a low-oxygen multi-source tracing method and an early warning system for a mine return air corner. The method comprises the following steps: measuring parameters of a coal seam and physical and mechanical parameters of a coal and rock layer; measuring the isotope value of each coal bed gas body, and establishing an end member gas characteristic database; building a physical similar simulation experiment model, and simulating a mine environment to carry out a coal mining experiment; calculating the flow conductivity and the weight coefficient of the overlying strata fracture; screening out two groups of isotope indexes with the maximum distinction degree from a database, and inputting the isotope indexes, the flow conductivity and the weight coefficient into a multi-end-member mixing equation to obtain a mixed isotope value; based on the concentration, the flow velocity and the flow path of gas, a source and a flow path are analyzed, a numerical model is constructed and compared with field data, and the gas concentration and the flow velocity of the source are predicted and early warned. According to the method, the gas traceability precision and reliability are remarkably improved, prediction and early warning of a gas source and a migration path are realized, and the advancement and comprehensiveness of low-oxygen risk identification are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of coal mine safety, specifically a method and early warning system for tracing the causes of low oxygen levels in the return air corner of a mine. Background Technology

[0002] During coal mining, engineering disturbances alter the stress and displacement fields of the overlying strata, causing overlying rock to collapse and generate numerous fissures. These fissures alter the conductivity between different coal seam working faces in multi-seam mines, enhancing gas transport capacity. In shallow, low-gas mines... , , During coal seam mining, gases migrate from goaf areas of other coal seams through overburden fissures and accumulate in the current coal seam due to factors such as pressure differences between different working faces. As coal seam mining progresses, the migration path, concentration distribution, and source of these gases become time-varying. Affected by these gases, low oxygen levels may occur in the return air corner of the working face.

[0003] Currently, the main approach to addressing this type of hypoxia problem relies on isotope tracing and physical detection methods to identify the gas source. However, existing methods still have significant limitations:

[0004] First, conventional isotope tracing is mostly based on a single gas indicator, which can easily lead to overlapping characteristics when the gas composition of multiple coal seams is similar, resulting in misjudgment of the source.

[0005] Second, most models assume that the isotopic values ​​of the end-member gas are constant and fail to consider the influence of the dynamic evolution of mining-induced fractures on the gas migration path, causing the calculation results to deviate from reality.

[0006] Third, existing monitoring systems are mostly focused on local gas concentration alarms in corners and lack the ability to track gas sources in real time and provide early warnings, making it difficult to prevent and control the risk of hypoxia from the source. Summary of the Invention

[0007] The purpose of this invention is to provide a method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face, comprising the following steps:

[0008] S1. Through geological exploration, the geological structure of the mineable coal seam and its adjacent coal and rock strata is statistically analyzed, relevant geological information is recorded, and the parameters of each coal seam, the physical and mechanical parameters of the coal and rock strata, and the geometry of the coal seam and coal and rock strata are measured.

[0009] S2. Construct a mine shaft and measure the gas content, ventilation, temperature, gas pressure, and coal seam gas composition within the shaft.

[0010] S3. Measure the isotope values ​​of gases in each coal seam and establish an end-member gas characteristic database;

[0011] S4. Based on the coal seam parameters and the physical and mechanical parameters of the coal and rock strata, determine the similarity ratio of the physical similarity simulation experiment, and build a physical similarity simulation experiment model based on the similarity ratio;

[0012] The physical similarity simulation experimental model is equipped with a ventilation module, a sampling loading and gas injection module, a ventilation-loading coordinated control module, a temperature regulation module, a fracture image acquisition module, and a monitoring module.

[0013] S5. Within a physically similar simulation experimental model, a coal seam mining experiment is conducted to simulate the mine environment. During the experiment, a monitoring module is used to record in real time the evolution of overlying rock fissures, as well as the gas concentration and flow rate.

[0014] S6. Based on the evolution process of overburden fractures, analyze the parameter information of overburden fractures, and calculate the conductivity and conductivity weight coefficient of overburden fractures based on the parameter information; then, combine and screen the isotope values ​​of coal seam gas in the end-member gas characteristic database.

[0015] S7. Input the calculated conductivity, conductivity weight coefficient, and selected isotopic indices of each coal seam overburden fracture into the multi-terminal mixing equation to obtain the mixed isotopic values. ;judge Compared with measured isotope values If the deviation is less than or equal to 10%, proceed to step S8; otherwise, return to step S6 and modify the flow guidance weight coefficient.

[0016] S8. Based on the experimental data obtained in step S5) and the multi-isotope combination source calculation method in steps S6) to S7), reverse the deduction of all possible sources of gas in the return wind corner; analyze the migration path, flow rate and concentration of gas at each source.

[0017] S9. Input the source, migration path, velocity and concentration of the return air corner gas obtained from step S8) into the numerical model to simulate the real-time velocity and source concentration of the return air corner gas. Determine whether the deviation between the simulation results and the field data is less than or equal to 5%. If so, the numerical model can be used to predict the gas flow. Otherwise, adjust the key parameters of the numerical model and reconstruct the numerical model.

[0018] Furthermore, in step S3), establishing the end-member gas characteristic database includes the following steps:

[0019] S3.1. Using a core drill bit, obtain coal samples from the upper, middle and lower positions of each coal seam, and place the coal samples into different aluminum foil sampling bags and vacuum them.

[0020] S3.2 After the coal sample releases gas, collect the gas and use gas chromatography-isotope mass spectrometry to determine the stable isotope values ​​of the gas, and establish an end-member gas characteristic database.

[0021] Furthermore, in step S5), the physical similarity simulation experimental model is adjusted using the ventilation module, the mining loading and gas injection module, the ventilation-loading coordinated control module, and the temperature regulation module to simulate the actual mine conditions.

[0022] The ventilation module is used to set ventilation boundary conditions in the physical similarity simulation experimental model to simulate the actual ventilation network of the mine and the gas source boundaries of each coal seam goaf, so as to provide stable ventilation conditions for gas movement in the experiment.

[0023] The mining loading and gas injection module includes a loading head, a jack or hydraulic cylinder, and a gas channel connected to the coal seam and goaf. It is used to apply overlying strata stress and working face advance load to the physical similarity simulation experimental model, and to inject gas into the preset coal seam or goaf through the gas channel, thereby simulating the mining process and the multi-source gas release process simultaneously.

[0024] The ventilation-loading coordinated control module is used to uniformly control the mining loading and ventilation boundary simulation process, so as to realize the coordinated simulation of mining and ventilation conditions.

[0025] The temperature regulation module includes a heating / cooling unit and a temperature sensor, which are used to regulate and monitor the temperature of the environment around the physical similarity simulation experimental model, so that the temperature field inside the experimental model matches the actual working temperature under different burial depths in the well.

[0026] The fracture image acquisition module includes an industrial camera, a light source, and an image acquisition and processing system. It is used to continuously photograph and record the development of model fractures during coal seam mining experiments, and to extract the fracture opening, morphology, orientation, and connectivity through image processing technology, providing basic data for subsequent calculation of the conductivity and connectivity index of overburden fractures.

[0027] The monitoring module includes stress sensors, as well as wind pressure and wind speed sensors, which are used to monitor stress and ventilation parameters at different locations of the physical similarity simulation experimental model, and to obtain time-varying information on the stress state, deformation characteristics and ventilation conditions of the overlying rock during mining.

[0028] Furthermore, in step S6), the parameter information of the overburden fracture includes the fracture aperture. morphology and connectivity;

[0029] Based on the parameter information of the overburden fractures, the conductivity and conductivity weight coefficient of the overburden fractures are calculated; the conductivity of the overburden fractures... The calculation formula is as follows:

[0030] (1)

[0031] In the formula: This refers to the viscosity of the gas.

[0032] The connectivity index is calculated as the total length of the connecting fractures divided by the area of ​​the monitored region.

[0033] The calculation basis of the fracture conductivity weight coefficient is: the fracture conductivity weight coefficient is determined according to the proportion of conductivity of the corresponding fracture region of each coal seam.

[0034] Further, in step S6), the combined screening method is as follows: based on principal component analysis, the isotope values ​​of coal seam gas in the end-member gas feature database are reduced in dimensionality, and then the two sets of isotope indices with the highest discriminative power are screened out.

[0035] Furthermore, in step S7), based on the principle of mass conservation, a multi-terminal component hybrid equation is established; the formula of the multi-terminal component hybrid equation is as follows:

[0036] (2)

[0037] In the formula: This represents the total number of coal seams.

[0038] For the first Coal seam No. Isotope values; =1,2;

[0039] For the first The proportion of conductivity in the fractured region of the coal seam;

[0040] For the first The fracture conductivity weighting coefficient of coal seams.

[0041] Further, in step S7), the measured isotope value The measurement method is as follows: a bundled tube sampling system is installed in the underground return airway. The mixed gas is extracted through the bundled tube sampling system and the isotope values ​​are measured. .

[0042] Furthermore, in step S9), the method for constructing the numerical model includes the following steps:

[0043] S9.1. Based on the coal seam distribution, overlying fracture parameter information, simulation data from the physical similarity simulation experiment model, and field measurement data, a three-dimensional model is established.

[0044] S9.2 Divide the model into multiple calculation units and set the gas source boundary and ventilation boundary;

[0045] S9.3. Based on Darcy's law, construct a mathematical model of gas flow and diffusion;

[0046] S9.4. Obtain multiple sets of data on the source, migration path, flow rate, and concentration of gas in the return air corner as a dataset;

[0047] S9.5. Input the dataset into the mathematical model of gas flow and diffusion to obtain the real-time gas velocity and gas source concentration; determine whether the difference between the real-time gas velocity and gas source concentration obtained by the model simulation and the actual measured real-time gas velocity and gas source concentration meets the threshold. If it does, the numerical model construction is completed; otherwise, return to step S9.1 to modify the overburden fracture parameter information.

[0048] Another objective of this invention is to provide an early warning system for a multi-source attribution method of gas in the return air corner of a fully mechanized mining face, comprising a data acquisition module, a simulation calculation module, a prediction analysis module, and an early warning response module.

[0049] The data acquisition module includes several sensors installed inside the mine, used to collect real-time gas concentration, temperature, wind speed, and pressure in the coal mine.

[0050] The simulation calculation module inputs the data obtained from the physical similarity simulation experiment model into the numerical model to obtain the simulation data.

[0051] The gas flow model of the multi-coal-seam fully mechanized mining face is established based on the coal seam gas content and the data collected by the data acquisition module, and is used to display the migration and enrichment of underground gas in real time.

[0052] The predictive analysis module uses an LSTM prediction model based on simulation data to predict the future gas concentration distribution of each source, the gas concentration distribution of the return air corner of the working face, and to determine whether a low-oxygen area will appear in the return air corner.

[0053] Low-oxygen areas appearing in the return air corner include situations where the levels of nitrogen and carbon dioxide at the source exceed the standards, and where the gas transport path has a strong guiding capacity.

[0054] The early warning response module issues early warnings based on the prediction results of the predictive analysis module.

[0055] The method for building an LSTM prediction model includes the following steps:

[0056] 1) Several sensors are deployed at intervals in the mine to collect real-time gas concentration, temperature, wind speed and pressure in the mine, and the coal seam gas content obtained by pre-drainage through boreholes is used as the initial dataset.

[0057] 2) Preprocess the initial dataset to construct the training set;

[0058] 3) Train the LSTM model using the training set to obtain the LSTM prediction model.

[0059] Furthermore, the concentrations of nitrogen and carbon dioxide at the source correspond to the conductivity, showing a relationship between low-concentration gas and high conductivity, and vice versa.

[0060] Based on the correspondence, the prediction results of the early warning response module include:

[0061] When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas concentration at the source exceeds the standard, but the flow capacity does not exceed the standard.

[0062] When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source exceeds the standard and the flow capacity exceeds the standard.

[0063] When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard, but the flow capacity exceeds the standard.

[0064] When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard and the flow capacity does not exceed the standard.

[0065] The technical effects of this invention are undeniable, and its beneficial effects are as follows:

[0066] 1. By optimizing the combination of multiple isotope indicators through principal component analysis, the ability to distinguish gases from different coal seams is enhanced, effectively overcoming the misjudgment problem caused by the overlap of single isotope characteristics; combined with the dynamic quantification and weight correction of fracture conductivity, the accuracy and reliability of gas source tracing are significantly improved.

[0067] 2. The system integrates downhole real-time monitoring data and dynamic simulation models to construct a "source-path" monitoring and early warning system, realizing the prediction and early warning of gas sources and migration paths, and significantly improving the foresight and comprehensiveness of low oxygen risk identification. Attached Figure Description

[0068] Figure 1This is a schematic diagram of the multi-source attribution method for gas in the return air corner of a fully mechanized mining face according to the present invention.

[0069] Figure 2 This is a schematic diagram illustrating the steps involved in locating the gas source of low oxygen phenomena based on fracture simulation and isotopic gas synergy.

[0070] Figure 3 This is a schematic diagram of a gas monitoring and forecasting early warning system module. Detailed Implementation

[0071] The present invention will be further described below with reference to embodiments, but it should not be construed that the scope of the present invention is limited to the following embodiments. Various substitutions and modifications made based on ordinary technical knowledge and common practices in the art without departing from the above-described technical concept of the present invention should be included within the scope of protection of the present invention.

[0072] Example 1:

[0073] A method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face includes the following steps:

[0074] S1. Through geological exploration, the geological structure of the mineable coal seam and its adjacent coal and rock strata is statistically analyzed, relevant geological information is recorded, and the parameters of each coal seam, the physical and mechanical parameters of the coal and rock strata, and the geometry of the coal seam and coal and rock strata are measured.

[0075] S2. Construct a mine shaft and measure the gas content, ventilation, temperature, gas pressure, and coal seam gas composition within the shaft.

[0076] S3. Measure the isotope values ​​of gases in each coal seam and establish an end-member gas characteristic database;

[0077] S4. Based on the coal seam parameters and the physical and mechanical parameters of the coal and rock strata, determine the similarity ratio of the physical similarity simulation experiment, and build a physical similarity simulation experiment model based on the similarity ratio;

[0078] The physical similarity simulation experimental model is equipped with a ventilation module, a sampling loading and gas injection module, a ventilation-loading coordinated control module, a temperature regulation module, a fracture image acquisition module, and a monitoring module.

[0079] S5. Within a physically similar simulation experimental model, a coal seam mining experiment is conducted to simulate the mine environment. During the experiment, a monitoring module is used to record in real time the evolution of overlying rock fissures, as well as the gas concentration and flow rate.

[0080] S6. Based on the evolution process of overburden fractures, analyze the parameter information of overburden fractures, and calculate the conductivity and conductivity weight coefficient of overburden fractures based on the parameter information; then, combine and screen the isotope values ​​of coal seam gas in the end-member gas characteristic database.

[0081] S7. Input the calculated conductivity, conductivity weight coefficient, and selected isotopic indices of each coal seam overburden fracture into the multi-terminal mixing equation to obtain the mixed isotopic values. ;judge Compared with measured isotope values If the deviation is less than or equal to 10%, proceed to step S8; otherwise, return to step S6 and modify the flow guidance weight coefficient.

[0082] S8. Based on the experimental data obtained in step S5) and the multi-isotope combination source calculation method in steps S6) to S7), reverse the deduction of all possible sources of gas in the return wind corner; analyze the migration path, flow rate and concentration of gas at each source.

[0083] S9. Input the source, migration path, velocity and concentration of the return air corner gas obtained from step S8) into the numerical model to simulate the real-time velocity and source concentration of the return air corner gas. Determine whether the deviation between the simulation results and the field data is less than or equal to 5%. If so, the numerical model can be used to predict the gas flow. Otherwise, adjust the key parameters of the numerical model and reconstruct the numerical model.

[0084] Example 2:

[0085] The main structure of this embodiment is the same as that of Embodiment 1. Further, in step S1), the geological information includes layer height, dip angle, and lithology. Geological structures and information are measured using ground-penetrating radar and well logging instruments. The parameters of the coal seam include stress, deformation, and fracture characteristics. The mechanical parameters of the coal and rock strata include stress, elastic modulus, coefficient of fragmentation, unit weight, compressive strength, etc., measured in the laboratory.

[0086] In step S2), the gas content is measured by a multi-component gas sensor, the wind speed is measured by an anemometer, the temperature is measured by a temperature sensor, the gas pressure is measured by a gas pressure sensor, and the proportion of coal seam gas components is measured by a gas chromatograph.

[0087] Example 3:

[0088] The main structure of this embodiment is the same as any one of embodiments 1 to 2. Further, in step S3), establishing the end-member gas characteristic database includes the following steps:

[0089] S3.1. Using a core drill bit, obtain coal samples from the upper, middle and lower positions of each coal seam, and place the coal samples into different aluminum foil sampling bags and vacuum them.

[0090] S3.2 After the coal sample releases gas, collect the gas and use gas chromatography-isotope mass spectrometry to determine the stable isotope values ​​of the gas, and establish an end-member gas characteristic database.

[0091] Example 4:

[0092] The main structure of this embodiment is the same as any one of embodiments 1 to 3. Further, in step S4), the similarity ratio includes geometric similarity ratio (geometric parameters, length, width and height), stress similarity ratio, elastic modulus similarity ratio, and time similarity ratio (mining time).

[0093] Example 5:

[0094] The main structure of this embodiment is the same as any one of embodiments 1 to 4. Further, in step S5), the physical similarity simulation experimental model is mounted on the experimental platform and equipped with a ventilation module, a sampling loading and gas source injection module, a ventilation-loading coordinated control module, a temperature adjustment module, a crack image acquisition module, and a monitoring module.

[0095] The physical similarity simulation experimental model is adjusted using a ventilation module, a mining loading and gas injection module, a ventilation-loading coordinated control module, and a temperature regulation module to simulate the actual mine conditions.

[0096] Ventilation module: Used to set the air inlet, air outlet and corresponding ventilation boundary conditions such as air volume and negative pressure in the experimental model to simulate the actual ventilation network of the mine and the gas source boundary of each coal seam goaf, so as to provide stable ventilation conditions for gas movement in the experiment.

[0097] Mining loading and gas injection module: including loading head, jack or cylinder and gas channel connected to coal seam and goaf, used to apply overlying strata stress and working face advance load to physical similarity simulation experimental model, and inject gas into preset coal seam or goaf through gas channel, thereby simulating mining process and multi-source gas release process simultaneously.

[0098] Ventilation-Loading Coordinated Control Module: This module is used to uniformly control the simulation process of mining loading and ventilation boundaries. By adjusting key parameters such as loading rate, advance step distance, air intake and return volume, and gas injection flow rate, the experimental conditions correspond to the actual mining face advance behavior in terms of time history and intensity, thereby achieving coordinated simulation of mining and ventilation conditions.

[0099] Temperature control module: including heating / cooling unit and temperature sensor, used to adjust and monitor the temperature of the environment around the physical similarity simulation experimental model, so that the temperature field inside the experimental model matches the actual working temperature under different burial depths in the well, thereby taking into account the influence of temperature on gas diffusion and flow characteristics.

[0100] The fracture image acquisition module includes a high-resolution industrial camera, a light source, and an image acquisition and processing system. It is used to continuously photograph and record the development of fractures in the model during coal seam mining experiments. The module also extracts parameters such as fracture opening, morphology, orientation, and connectivity through image processing technology, providing basic data for subsequent calculation of the conductivity and connectivity index of overburden fractures.

[0101] Monitoring module: Includes stress sensors, wind pressure and wind speed sensors, etc., used to monitor stress and ventilation parameters at different locations of the physical similarity simulation experimental model, and obtain time-varying information on the stress state, deformation characteristics and ventilation conditions of the overlying rock during mining.

[0102] Example 6:

[0103] The main structure of this embodiment is the same as any one of embodiments 1 to 5. Furthermore, in step S5), when the physical similarity simulation experimental model conducts a coal seam mining experiment, the process that needs to be experienced includes the overlying coal seam being mined out and the mining of this coal seam.

[0104] Within a physically similar simulation model, a simulated coal seam mining experiment is conducted in stages according to a pre-set mining schedule. The mining schedule is scaled equivalently to the actual working face advance rate, and typical evolution stages such as the initial mining stage, the stable advancement stage, and the stage near the goaf are set.

[0105] Example 7:

[0106] The main structure of this embodiment is the same as any one of embodiments 1 to 6. Further, in step S6), the parameter information of the overburden fracture includes the fracture aperture. morphology and connectivity;

[0107] Based on the parameter information of the overburden fractures, the conductivity and conductivity weight coefficient of the overburden fractures are calculated; the conductivity of the overburden fractures... The calculation formula is as follows:

[0108] (1)

[0109] In the formula: This refers to the viscosity of the gas.

[0110] The connectivity index is calculated as the total length of the connecting fractures divided by the area of ​​the monitored region.

[0111] The calculation basis of the fracture conductivity weight coefficient is: the fracture conductivity weight coefficient is determined according to the proportion of conductivity of the corresponding fracture region of each coal seam.

[0112] Example 8:

[0113] The main structure of this embodiment is the same as any one of embodiments 1 to 7. Further, in step S6), the combined screening method is as follows: based on principal component analysis, the isotope values ​​of coal seam gas in the end-member gas feature database are reduced in dimensionality, and then the two sets of isotope indices with the highest distinguishability are screened out.

[0114] Example 9:

[0115] The main structure of this embodiment is the same as any one of embodiments 1 to 8. Further, in step S7), based on the principle of mass conservation, a multi-terminal component hybrid equation is established; the formula of the multi-terminal component hybrid equation is as follows:

[0116] (2)

[0117] In the formula: This represents the total number of coal seams.

[0118] For the first Coal seam number Isotope values; =1,2;

[0119] For the first The proportion of conductivity in the fractured region of the coal seam;

[0120] For the first The fracture conductivity weighting coefficient of coal seams.

[0121] Example 10:

[0122] The main structure of this embodiment is the same as any one of embodiments 1 to 9. Further, in step S7), the measured isotope value... The measurement method is as follows: a bundled tube sampling system is installed in the underground return airway. The mixed gas is extracted through the bundled tube sampling system and the isotope values ​​are measured. .

[0123] Example 11:

[0124] The main structure of this embodiment is the same as any one of embodiments 1 to 10. Furthermore, in step S9), the method for constructing the numerical model includes the following steps:

[0125] S9.1. Based on the coal seam distribution, overlying fracture parameter information, simulation data from the physical similarity simulation experiment model, and field measurement data, a three-dimensional model is established.

[0126] S9.2 Divide the model into multiple calculation units and set the gas source boundary and ventilation boundary;

[0127] S9.3. Based on Darcy's law, construct a mathematical model of gas flow and diffusion;

[0128] The functional equations of the mathematical model for gas flow and diffusion (the convection-diffusion governing equations for gas in porous media) are as follows:

[0129] (3)

[0130] The Darcy flow velocity of gas in a porous medium can be expressed as:

[0131] (4)

[0132] In the formula: For gas concentration field; Porosity; For time; The effective diffusion coefficient; For gas source and sink items; This is the Darcy velocity vector for gas seepage. For medium permeability; This refers to the viscosity of the gas. This refers to gas pressure; This is the gradient operator.

[0133] S9.4. Obtain multiple sets of data on the source, migration path, flow rate, and concentration of gas in the return air corner as a dataset;

[0134] S9.5. Input the dataset into the mathematical model of gas flow and diffusion to obtain the real-time gas velocity and gas source concentration; determine whether the difference between the real-time gas velocity and gas source concentration obtained by the model simulation and the actual measured real-time gas velocity and gas source concentration meets the threshold. If it does, the numerical model construction is completed; otherwise, return to step S9.1 to modify the overburden fracture parameter information.

[0135] Example 12:

[0136] The main structure of this embodiment is the same as any one of embodiments 1 to 11. Furthermore, in step S9), the key parameters include fracture permeability, gas source boundary conditions, hydrodynamic parameters, and grid density.

[0137] Permeability: Permeability parameters were obtained through gas seepage experiments using on-site sampling. Combined with fracture morphology images acquired in similar simulation experiments, fracture permeability was estimated using image processing and numerical inversion techniques; the average of these two estimates was then used to obtain the permeability parameters for the model.

[0138] Gas source boundary conditions: This refers to the information on gas sources, composition, and proportions obtained in steps S1-S8, which serves as the initial conditions for the model.

[0139] Fluid dynamic parameters: These mainly include parameters such as gas viscosity, density, and diffusion coefficient, which can be obtained through physical property parameter testing.

[0140] Mesh density: Based on the size of the physical simulation model and the scale of the main fractures, a relatively dense mesh is initially set to ensure that key flow paths can be captured. Subsequently, the mesh can be gradually refined according to the simulation requirements and accuracy.

[0141] Example 13:

[0142] An early warning system for a multi-source attribution method for gas in the return air corner of a fully mechanized mining face, based on any one of embodiments 1 to 12, includes a data acquisition module, a simulation calculation module, a prediction analysis module, and an early warning response module.

[0143] The data acquisition module includes several sensors installed inside the mine, used to collect real-time gas concentration, temperature, wind speed, and pressure in the coal mine.

[0144] The simulation calculation module inputs the data obtained from the physical similarity simulation experiment model into the numerical model to obtain the simulation data.

[0145] The gas flow model of the multi-coal-seam fully mechanized mining face is established based on the coal seam gas content and the data collected by the data acquisition module, and is used to display the migration and enrichment of underground gas in real time.

[0146] The predictive analysis module uses an LSTM prediction model based on simulation data to predict the future gas concentration distribution of each source, the gas concentration distribution of the return air corner of the working face, and to determine whether a low-oxygen area will appear in the return air corner.

[0147] Low-oxygen areas appearing in the return air corner include situations where the levels of nitrogen and carbon dioxide at the source exceed the standards, and where the gas transport path has a strong guiding capacity.

[0148] The early warning response module issues early warnings based on the prediction results of the predictive analysis module.

[0149] The method for building an LSTM prediction model includes the following steps:

[0150] 1) Several sensors are deployed at intervals in the mine to collect real-time gas concentration, temperature, wind speed and pressure in the mine, and the coal seam gas content obtained by pre-drainage through boreholes is used as the initial dataset.

[0151] 2) Preprocess the initial dataset to construct the training set;

[0152] 3) Train the LSTM model using the training set to obtain the LSTM prediction model.

[0153] Example 14:

[0154] The main structure of this embodiment is the same as that of embodiment 13. Furthermore, the concentrations of nitrogen and carbon dioxide at the source correspond to the flow capacity, and there is a correspondence between low concentration gas - high flow capacity and high concentration gas - low flow capacity.

[0155] Based on the correspondence, the prediction results of the early warning response module include:

[0156] When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas concentration at the source exceeds the standard, but the flow capacity does not exceed the standard.

[0157] When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source exceeds the standard and the flow capacity exceeds the standard.

[0158] When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard, but the flow capacity exceeds the standard.

[0159] When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard and the flow capacity does not exceed the standard.

[0160] Defined as a standard value, where, at a specific nitrogen and carbon dioxide concentration, there exists a flow guidance capacity that ensures an oxygen concentration of 20% at the return air corner. If this standard value is exceeded, the oxygen concentration at the return air corner will fall below 20%.

[0161] Example 15:

[0162] The main structure of this embodiment is the same as any one of embodiments 1 to 14. Furthermore, this invention proposes a multi-source tracing method and early warning system for gas in the return air corner of a fully mechanized mining face. The specific steps are as follows:

[0163] S1: Geological information collection.

[0164] This step involves collecting information on the geological structure of the mine and the mechanical parameters of the coal seam.

[0165] (1) Geological structure information acquisition: Through geological exploration, the geological structure of the mineable coal seam and its adjacent coal and rock strata is statistically analyzed, and relevant geological information (such as layer height, dip angle, lithology, etc.) is recorded.

[0166] (2) Measurement of physical and mechanical parameters: Obtain the physical and mechanical parameters of coal and rock strata (such as elastic modulus, coefficient of fragmentation, coal and rock density and compressive strength, etc.), and measure the stress, deformation and fracture characteristics of different coal seams.

[0167] (3) Gas content and ventilation conditions: Collect information on gas content, gas pressure, and coal seam gas composition in the mine (e.g., , , (etc.) and mine ventilation data, providing field data for similar simulation experiments and numerical models.

[0168] S2: Based on crack simulation-isotope gas co-location of the gas source of low oxygen phenomenon.

[0169] (1) Construction of the end-member gas library. This step involves the on-site collection of end-member gases and the desorption of gases in the laboratory.

[0170] a. On-site sampling: Coal samples were taken from the upper, middle and lower parts of each coal seam using a core drill bit. After being slightly crushed, the coal samples were placed in aluminum foil sampling bags and vacuumed once to remove air mixed in with the sampling bags for isotope determination.

[0171] b. Desorbed Gas Experiment: Take 50 μL of desorbed gas samples from each coal seam and analyze them using gas chromatography-isotope mass spectrometry. , , Stable isotope values ​​of gases such as , , and . , , Establish an end-member gas characteristic database.

[0172] (2) Fracture-gas similarity simulation experiment. This step involves the construction of a fracture-gas similarity simulation experiment platform. Based on the coal seam parameter information obtained in step S1, a physical similarity simulation experiment model is built to simulate the working face advancement process.

[0173] a. Model design and similarity ratio determination: Determine the similarity ratio of the physical similarity simulation experiment based on the geometric and mechanical parameters of the coal seam, including geometric similarity ratio, stress similarity ratio, elastic modulus similarity ratio, time similarity ratio, etc.

[0174] b. Installation of Gas Flow Monitoring Device: A gas flow path monitoring device is installed in the experimental model to monitor the gas flow path, velocity, and concentration changes in real time. This device can accurately simulate the flow and accumulation of gases inside a coal mine and is used for isotope gas source location.

[0175] c. Ventilation System Configuration: A ventilation system is configured to simulate the ventilation conditions inside a mine. This system can control the flow direction and speed of gas within the experimental model, and adjust the ventilation volume and gas composition as needed to simulate the actual gas flow conditions in different coal seams.

[0176] d. Temperature Control System: A temperature control device is introduced to adjust the temperature of the experimental model to the actual working temperature of the mine. Temperature changes have a significant impact on gas diffusivity. In the experiment, a precise temperature control system is used to simulate the actual temperature of the mine, ensuring that the experimental environment matches the climatic conditions inside the mine.

[0177] (3) Monitoring of fracture evolution and gas flow. This step involves recording the evolution of overlying fractures in real time during coal seam mining in the model experiment, and monitoring the gas flow path at the same time.

[0178] a. Fracture Evolution Recording and Image Processing: Image processing technology was used to monitor the evolution of overlying fractures during the experiment, analyzing the fracture aperture, morphology, and connectivity. High-precision image recognition technology was used to periodically capture and process image data from the experimental model to obtain the dynamic process of fracture evolution.

[0179] b. Multi-channel gas flow monitoring: A multi-channel gas monitoring device is used to monitor key data such as gas flow path, velocity, and concentration changes in real time, ensuring accurate recording of gas migration paths. Gas concentration and velocity data will be recorded over time at different mining stages to analyze their time-varying patterns.

[0180] c. Temperature Effect Monitoring: The temperature control system adjusts the temperature within the model in real time and monitors the impact of temperature changes on gas flow using temperature sensors. The diffusion rate and concentration distribution of gases change at different temperatures; therefore, temperature control is crucial for simulating gas flow.

[0181] d. Time-varying gas source tracing: As coal seam mining progresses, the formation and expansion of fractures is a gradual process, and the gas migration paths and sources change with different mining stages. Therefore, the gas source paths in the experimental model will be monitored in stages at different mining stages to form time-varying gas source tracing data.

[0182] (4) Quantification and weight allocation of fracture conductivity.

[0183] a. Calculation of fracture backflow coefficient: Based on the fracture aperture in the physical model. Connectivity index ( =Total length of connected fractures / Area of ​​monitored region) and gas viscosity The backflow capacity is calculated according to formula (1). .

[0184] (1)

[0185] in, This refers to the crack aperture; It is the connectivity index. =Total length of connected fractures / Area of ​​the monitored region; This represents the viscosity of the gas.

[0186] b. Coal seam contribution weight allocation: The fracture conductivity weight coefficient is dynamically adjusted according to the proportion of conductivity of the corresponding fracture region of each coal seam.

[0187] (5) Calculation and verification of isotope combination sources.

[0188] a. Combinatorial Screening: Principal component analysis (PCA) was used to reduce the dimensionality of the end-member gas isotope values, and the two groups of isotope indices with the highest discriminative power (e.g., ...) were screened. and );

[0189] b. Constructing source equations: Establishing multi-terminal hybrid equations based on the principle of mass conservation.

[0190] (2)

[0191] in, For the first Coal seam No. Isotope values; The initial contribution ratio is determined by step (4); The fracture conductivity weighting coefficient is determined by step (4).

[0192] c. On-site data verification: Install a bundle tube sampling system in the underground return airway, extract mixed gas and measure isotope values, compare the calculated results with the measured data, and if the deviation is >10%, iteratively correct the flow weight coefficient.

[0193] (6) Gas source tracing and source location.

[0194] This step involves using experimental data, combined with the multi-isotope combination source calculation method and gas flow theory in step (5), to reverse-engineer the gas source.

[0195] a. Gas Flow Path Reverse Deduction: Based on experimental data, the gas flow path is deduced in reverse. Combining changes in gas velocity and concentration, the flow route of the gas from the source to the working face is analyzed. By using gas flow monitoring data and temperature change patterns in the experiment, the gas source causing the low oxygen phenomenon in the return air corner and its migration path are accurately located. The time factor means that the deduction of the gas source not only considers spatial relationships but also the changes in mining progress and different stages.

[0196] b. The impact of time-varying temperature on gas source location: The addition of temperature regulation makes the gas diffusion process in the experimental simulation more closely resemble the actual mine environment. Gas flow and diffusion patterns differ under different temperature conditions; therefore, temperature regulation can further improve the accuracy of source location and provide time-varying data for gas source location at different mining stages.

[0197] S3: Establish a gas monitoring and forecasting early warning system that tracks the "source-path" of a gas.

[0198] By comparing field data with numerical simulation results in space and time, and identifying deviations between the model and the actual situation, key parameters such as fracture permeability, gas source boundary conditions, hydrodynamic parameters, and grid density are dynamically adjusted to achieve real-time model correction. After the model is corrected and verified, the corrected model and field monitoring data are used to conduct advance predictions of gas migration trends in the short term (hours to days) and medium to long term (weeks to months). Early warnings are issued in a timely manner based on the set safety thresholds. Finally, continuous data collection, periodic error analysis, and dynamic updates of model parameters are implemented.

[0199] The system mainly consists of four modules: data acquisition module, simulation calculation module, predictive analysis module, and early warning response module.

[0200] The specific operating steps are as follows:

[0201] (1) Real-time gas concentration in coal mines is collected by deploying sensors. , , Data such as temperature, wind speed, pressure, and coal seam gas content obtained through borehole pre-drainage are collected. Sensor placement is as follows: Figure 2 As shown.

[0202] (2) Establish a gas flow model for multi-coal-seam fully mechanized mining faces, use sensor data to calibrate model parameters, and display the migration and enrichment of underground gas in a timely manner.

[0203] (3) Introduce a predictive analysis module and combine it with the simulation module to predict the gas concentration at a future time in real time.

[0204] (4) Based on the prediction results of machine learning, provide early warning of situations that may lead to low oxygen levels in the return air corner at a certain time in the future, and promptly notify management personnel to take corresponding measures to prevent accidents and ensure the continuity of safe production. Situations that may lead to low oxygen levels in the return air corner: excessive levels of nitrogen, carbon dioxide, and other gases at the source;

[0205] In step (3), the predictive analysis module is mainly completed through a machine learning system (LSTM), and the steps for establishing the machine learning system are as follows:

[0206] a. Receive real-time and historical data from the data acquisition module, and first perform some preprocessing on the data, formatting and standardizing the data to facilitate subsequent calculations.

[0207] b. Establish an LSTM (Long Short-Term Memory) model to predict the transport of low-oxygen gas at a future point in time.

[0208] c. Input the training data into the LSTM network, train the model using the backpropagation algorithm, and adjust the parameters to minimize the prediction error.

[0209] d. After training, the LSTM model can predict values ​​for a future period of time based on the input historical data.

[0210] The predictions in (4) mainly include: the gas concentration distribution of each source at future times, the gas concentration distribution of the return air corner of the working face, and whether a low-oxygen area will appear in the return air corner; if it will appear, it indicates the predicted low-oxygen range of the return air corner and the location and gas causing the low-oxygen occurrence. Based on the range of gas concentration changes, an early warning is issued.

Claims

1. A method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face, characterized in that, Includes the following steps: S1. Through geological exploration, the geological structure of the mineable coal seam and its adjacent coal and rock strata is statistically analyzed, relevant geological information is recorded, and the parameters of each coal seam, the physical and mechanical parameters of the coal and rock strata, and the geometry of the coal seam and coal and rock strata are measured. S2. Construct a mine shaft and measure the gas content, ventilation, temperature, gas pressure, and coal seam gas composition within the shaft. S3. Measure the isotope values ​​of gases in each coal seam and establish an end-member gas characteristic database; S4. Based on the coal seam parameters and the physical and mechanical parameters of the coal and rock strata, determine the similarity ratio of the physical similarity simulation experiment, and build a physical similarity simulation experiment model based on the similarity ratio; The physical similarity simulation experimental model is equipped with a ventilation module, a sampling loading and gas injection module, a ventilation-loading coordinated control module, a temperature regulation module, a fracture image acquisition module, and a monitoring module; S5. Within a physically similar simulation experimental model, a coal seam mining experiment is conducted to simulate the mine environment. During the experiment, a monitoring module is used to record in real time the evolution of overlying rock fissures, as well as the concentration and flow rate of gas during coal seam mining. S6. Based on the evolution process of overburden fractures, analyze the parameter information of overburden fractures, and calculate the conductivity and conductivity weight coefficient of overburden fractures based on the parameter information; then, combine and screen the isotope values ​​of coal seam gas in the end-member gas characteristic database. S7. Input the calculated conductivity, conductivity weight coefficient, and selected isotopic indices of each coal seam overburden fracture into the multi-terminal mixing equation to obtain the mixed isotopic values. ;judge Compared with measured isotope values If the deviation is less than or equal to 10%, proceed to step S8; otherwise, return to step S6 and modify the flow guidance weight coefficient. S8. Based on the experimental data obtained in step S5) and the multi-isotope combination source calculation method in steps S6) to S7), reverse the deduction of all possible sources of gas in the return wind corner; analyze the migration path, flow rate and concentration of gas at each source. S9. Input the source, migration path, velocity and concentration of the return air corner gas obtained from step S8) into the numerical model to simulate the real-time velocity and source concentration of the return air corner gas. Determine whether the deviation between the simulation results and the field data is less than or equal to 5%. If so, the numerical model can be used to predict the gas flow. Otherwise, adjust the key parameters of the numerical model and reconstruct the numerical model.

2. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S3), establishing the end-member gas characteristic database includes the following steps: S3.

1. Using a core drill bit, obtain coal samples from the upper, middle and lower positions of each coal seam, and place the coal samples into different aluminum foil sampling bags and vacuum them. S3.2 After the coal sample releases gas, collect the gas and use gas chromatography-isotope mass spectrometry to determine the stable isotope values ​​of the gas, and establish an end-member gas characteristic database.

3. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S5), the physical similarity simulation experimental model is adjusted using the ventilation module, mining loading and gas injection module, ventilation-loading coordinated control module, and temperature regulation module to simulate the actual mine conditions. The ventilation module is used to set ventilation boundary conditions in the physical similarity simulation experimental model to simulate the actual ventilation network of the mine and the gas source boundary of each coal seam goaf, so as to provide stable ventilation conditions for gas migration in the experiment. The mining loading and gas injection module includes a loading head, a jack or a hydraulic cylinder, and a gas channel connected to the coal seam and goaf. It is used to apply the overlying strata stress and working face advance load to the physical similarity simulation experimental model, and to inject gas into the preset coal seam or goaf through the gas channel, thereby simulating the mining process and the multi-source gas release process simultaneously. The ventilation-loading coordinated control module is used to uniformly control the mining loading and ventilation boundary simulation process, so as to realize the coordinated simulation of mining and ventilation conditions. The temperature regulation module includes a heating / cooling unit and a temperature sensor, which are used to regulate and monitor the temperature of the environment around the physical similarity simulation experimental model, so that the temperature field inside the experimental model matches the actual working temperature under different burial depths in the well. The fracture image acquisition module includes an industrial camera, a light source, and an image acquisition and processing system. It is used to continuously photograph and record the development of model fractures during coal seam mining experiments, and to extract the opening, shape, orientation, and connectivity of fractures through image processing technology, providing basic data for subsequent calculation of the conductivity and connectivity index of overburden fractures. The monitoring module includes stress sensors, as well as wind pressure and wind speed sensors, which are used to monitor stress and ventilation parameters at different locations of the physical similarity simulation experimental model, and to obtain time-varying information on the stress state, deformation characteristics and ventilation conditions of the overlying rock during mining.

4. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S6), the parameter information of the overburden fracture includes the fracture aperture. morphology and connectivity; Based on the parameter information of the overburden fractures, the conductivity and conductivity weight coefficient of the overburden fractures are calculated; the conductivity of the overburden fractures... The calculation formula is as follows: （1） In the formula: This refers to the viscosity of the gas. The connectivity index is calculated as the total length of the connecting fractures divided by the area of ​​the monitored region. The calculation basis of the fracture conductivity weight coefficient is: the fracture conductivity weight coefficient is determined according to the proportion of conductivity of the corresponding fracture region of each coal seam.

5. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S6), the combined screening method is as follows: based on principal component analysis, the isotope values ​​of coal seam gas in the end-member gas characteristic database are reduced in dimensionality, and then the two sets of isotope indices with the highest discriminative power are screened out.

6. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S7), based on the principle of mass conservation, a multi-terminal component hybrid equation is established; the formula of the multi-terminal component hybrid equation is as follows: （2） In the formula: This represents the total number of coal seams. For the first Coal seam number Isotope values; =1,2; For the first The proportion of conductivity in the fractured region of the coal seam; For the first The fracture conductivity weighting coefficient of coal seams.

7. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S7), the measured isotope value The measurement method is as follows: a bundled tube sampling system is installed in the underground return airway. The mixed gas is extracted through the bundled tube sampling system and the isotope values ​​are measured. .

8. The method for tracing the multi-source causes of gas in the return air corner of a fully mechanized mining face according to claim 1, characterized in that: In step S9), the method for constructing the numerical model includes the following steps: S9.

1. Based on the coal seam distribution, overlying fracture parameter information, simulation data from the physical similarity simulation experiment model, and field measurement data, a three-dimensional model is established. S9.2 Divide the model into multiple calculation units and set the gas source boundary and ventilation boundary; S9.

3. Based on Darcy's law, construct a mathematical model of gas flow and diffusion; S9.

4. Obtain multiple sets of data on the source, migration path, flow rate, and concentration of gas in the return air corner as a dataset; S9.

5. Input the dataset into the mathematical model of gas flow and diffusion to obtain the real-time gas velocity and gas source concentration; determine whether the difference between the real-time gas velocity and gas source concentration obtained by the model simulation and the actual measured real-time gas velocity and gas source concentration meets the threshold. If it does, the numerical model construction is completed; otherwise, return to step S9.1 to modify the overburden fracture parameter information.

9. An early warning system based on any one of claims 1 to 8 for a multi-source attribution method of gas in the return air corner of a fully mechanized mining face, characterized in that: It includes a data acquisition module, a simulation calculation module, a predictive analysis module, and the early warning response module; The data acquisition module includes several sensors installed in the mine, used to collect real-time gas concentration, temperature, wind speed and pressure in the coal mine; The simulation calculation module inputs the data obtained from the physical similarity simulation experiment model into the numerical model to obtain the simulation data; The gas flow model of the multi-coal-seam fully mechanized mining face is established based on the coal seam gas content and the data collected by the data acquisition module, and is used to display the migration and enrichment of underground gas in real time. The predictive analysis module uses the LSTM prediction model based on the simulation data to predict the future gas concentration distribution of each source, the gas concentration distribution of the return air corner of the working face, and to determine whether a low-oxygen area will appear in the return air corner. Low-oxygen areas appearing in the corners of the return air include situations where the levels of nitrogen and carbon dioxide at the source exceed the standards, and where the gas transport path has a strong guiding capacity. The early warning response module issues early warnings based on the prediction results of the predictive analysis module; The method for building an LSTM prediction model includes the following steps: 1) Several sensors are deployed at intervals in the mine to collect real-time gas concentration, temperature, wind speed and pressure in the mine, and the coal seam gas content obtained by pre-drainage through boreholes is used as the initial dataset. 2) Preprocess the initial dataset to construct the training set; 3) Train the LSTM model using the training set to obtain the LSTM prediction model.

10. The early warning system for a multi-source attribution method for gas in the return air corner of a fully mechanized mining face according to claim 9, characterized in that, The concentrations of nitrogen and carbon dioxide at the source correspond to the conductivity, with a relationship between low-concentration gas and high conductivity, and vice versa. Based on the correspondence, the prediction results of the early warning response module include: When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas concentration at the source exceeds the standard, but the flow capacity does not exceed the standard. When the nitrogen concentration is greater than or equal to 85%, the carbon dioxide concentration is greater than or equal to 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source exceeds the standard and the flow capacity exceeds the standard. When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity exceeds the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard, but the flow capacity exceeds the standard. When the nitrogen concentration is less than 85%, the carbon dioxide concentration is less than 0.5%, and the flow capacity does not exceed the standard value at the current concentration, it is determined that the gas at the source does not exceed the standard and the flow capacity does not exceed the standard.

Images