Method for constructing and updating oxygen concentration dynamic distribution model of stope face of shallow coal seam in real time

Abstract

The invention relates to a shallow coal seam stope face oxygen concentration dynamic distribution model construction and real-time updating method, belongs to the technical field of coal mine safety monitoring, and aims to solve the technical problems that a traditional monitoring means cannot obtain an oxygen concentration space full view and a static model is difficult to track a dynamic mining environment. According to the technical scheme, at least four sensors are arranged on a coal face to form a gradient monitoring network, and oxygen concentration data of multiple positions are collected in real time; constructing an exponential decay function to characterize the distribution rule of the concentration along the tendency; solving an undetermined coefficient based on a nonlinear least square method with space-time constraint; and dynamically calculating a model updating period t through an environment complexity index ECI, and triggering model iteration. Working face oxygen concentration global dynamic sensing, low-oxygen risk area accurate early warning and ventilation decision intelligence are achieved, and the mine safety prevention and control capacity is remarkably improved.

Description

Technical Field

[0001] This invention belongs to the field of coal mine safety monitoring technology, and relates to a method for constructing and real-time updating a dynamic distribution model of oxygen concentration at the mining face of a shallow coal seam. Background Technology

[0002] Shallow coal seam mining faces unique geological challenges. When the coal seam is buried at a depth of less than 150 meters and the interlayer spacing is small, surface fissures caused by mining activities create air leakage channels. Once the surface connects with the overlying goaf, adjacent goaf, and the goaf of the recovery face, the air leakage increases dramatically, directly leading to abnormal fluctuations in oxygen content at the working face. Residual coal in the goaf undergoes oxidation reactions in environments with varying oxygen concentrations, releasing heat and harmful gases such as carbon monoxide and carbon dioxide. These gases diffuse and flow within the complex goaf area, seriously threatening the lives of underground personnel and the normal operation of equipment.

[0003] Traditional oxygen concentration monitoring methods have significant limitations. Existing technologies mostly rely on intermittent, single-point measurements, capturing only instantaneous data from isolated locations at specific times, failing to reflect the overall spatial distribution of oxygen concentration at the working face. Some numerical simulation methods, lacking real-time data, struggle to track dynamic changes in oxygen concentration. When ventilation systems are adjusted, surface air pressure changes, or mining progress fluctuates, existing technologies cannot provide timely warnings of low-oxygen or oxygen-rich risk areas, resulting in insufficient response capabilities for coal mining enterprises to sudden disasters.

[0004] More importantly, static models are fundamentally incompatible with dynamic mining environments. Factors such as the operating status of the ventilation system, the speed of the mining face advance, and geological tectonic movements continuously alter the gas distribution pattern at the working face, and traditional fixed-parameter models cannot adapt to these real-time changes. Oxygen concentration imbalances not only affect personnel breathing safety but also accelerate the wear and tear on electromechanical equipment and can even induce asphyxiation accidents. The lag and one-sidedness of existing technologies have become a core bottleneck restricting safe production in coal mines.

[0005] Therefore, there is an urgent need for an oxygen concentration distribution analysis method that integrates multi-source real-time monitoring data and is capable of dynamic modeling and autonomous updating. This method needs to accurately characterize the attenuation law of oxygen along the working face, track the concentration change trend under mining disturbances in real time, and provide quantitative decision-making basis for ventilation optimization and disaster prevention. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a method for constructing and updating a dynamic distribution model of oxygen concentration at the mining face of a shallow coal seam in real time.

[0007] To achieve the above objectives, the present invention provides the following technical solution:

[0008] A method for constructing and real-time updating a dynamic distribution model of oxygen concentration at a shallow coal seam mining face includes the following steps:

[0009] S1: Real-time monitoring and acquisition of oxygen concentration at multiple locations: At least four oxygen concentration monitoring sensors are deployed at the coal face, with the sensor location set R = (r1, r2, ..., r...). m ), satisfying r1 < r2 < ... < r m The increasing order relationship and the arrangement position conform to the D-optimal design criterion to maximize the determinant of the Fisher information matrix of parameter estimation; among which, no less than 2 sensors are set within 100m of the return air corner;

[0010] Each sensor collects oxygen concentration data at preset time intervals and transmits it to the ground monitoring center in real time. The time interval is dynamically adjusted according to the mining stage: once every 20 seconds in the initial mining stage and once every 30 seconds in the stable mining stage.

[0011] S2: Set the initial update period of the oxygen concentration distribution model to t1, process the oxygen concentration in the time period t1, and form the basic data set V for oxygen concentration fitting.

[0012] S3: Oxygen Concentration Distribution Function Construction: Defines the function that governs the distribution of oxygen concentration along the working face:

[0013] y = a - be c(l-x)

[0014] Where: a is the oxygen concentration in the intake airway, in %, and is the maximum value of the distribution curve; l is the working face dip length, in meters, determined based on geological surveys; c<0, characterizing the attenuation coefficient of oxygen concentration decreasing along the working face dip; b is the fitting parameter, calculated by the formula b=ad when data is insufficient; d is the oxygen concentration in the return air, in %.

[0015] S4: Solving for undetermined coefficients: Based on the real-time data set U = [(r i ,z ij )] Calculate the average oxygen concentration at each location. The basic dataset V = [(r1,c1),...,(r m ,c m The undetermined coefficients a, b, c, l in the function are solved using the nonlinear least squares method with time and space constraints.

[0016] S5: Preliminary static model formation: Substitute the solved coefficients into the distribution function y = a - be c(l-x) Generate a static distribution model of oxygen concentration at the working face;

[0017] S6: Dynamic update cycle calculation: The update cycle t is dynamically calculated based on the environment complexity index.

[0018] ECI = w1c z +w2v f +w3q

[0019] Among them, c z =s1 / v1, representing the coefficient of variation of oxygen concentration, where s1 is the standard deviation of oxygen concentration and v1 is the average oxygen concentration; v f =s2 / v2, representing the air volume fluctuation coefficient, where s2 is the standard deviation of air volume and v2 is the average air volume; This represents the normalized coal production, where q0 is the real-time coal production, and q min q max This is a historical extreme value;

[0020] The weighting coefficients are w1 = 0.7, w2 = 0.2, and w3 = 0.1.

[0021] The update cycle is calculated using the formula t = t0e. -k·ECI The baseline period t0 = 10 min, the adjustment coefficient k = 2, and t ∈ [t min ,t max ];

[0022] S7: Dynamic Model Update: When the duration of the new monitoring data change reaches period t, repeat steps S2 to S4 to update the coefficients and reconstruct the model.

[0023] Furthermore, the installation position, angle, and height of the oxygen concentration monitoring sensor in S1 are determined based on the ventilation system diagram and mining engineering plan of the coal mining face to ensure that the sensor is located in the core airflow zone of the area to be measured.

[0024] The installation location satisfies:

[0025] The sensors in the air intake area are positioned along the main airflow path axis.

[0026] The upper corner sensor is ≤0.5m away from the roof and tilted 30°~45° toward the direction of air leakage in the goaf.

[0027] The sensor in the central area is suspended vertically from the support column, at a height level with the breathing zone, approximately 1.5m ± 0.2m.

[0028] Furthermore, in S4, the nonlinear least squares method incorporates a compensation mechanism for sensor accuracy errors, electromagnetic interference, and instantaneous fluctuations caused by mining disturbances during the solution process.

[0029] Furthermore, in S6, the minimum value of the update period t is tmin. min =2min, maximum value t max =10min.

[0030] A dynamic monitoring system for oxygen concentration at a shallow coal seam mining face includes:

[0031] Sensor array module: Multiple oxygen concentration monitoring sensors deployed at the coal mining face, with their location set R = (r1, r2, ..., r m ), satisfying r1 < r2 < ... < r m Furthermore, the density within 100m of the return air corner should not be less than 2 units;

[0032] Data transmission module: Connects to the sensor array module via an industrial wireless network to send the collected oxygen concentration data to the processing center at dynamically adjusted time intervals;

[0033] Model processing module: Connected to the data transmission module via a fiber optic ring network, used for execution of:

[0034] (a) Construct the distribution function y = a - be c(l-x) ;

[0035] (b) Solve for coefficients a, b, c, l based on dataset V;

[0036] (c) Calculate the Environment Complexity Index (ECI) and its update period t;

[0037] (d) Trigger dynamic model updates;

[0038] Early warning output module: Connects to the model processing module via API interface to generate oxygen concentration distribution map of the working face and early warning signals for low oxygen risk areas, and outputs them to the ground monitoring terminal.

[0039] Furthermore, the data transmission module adopts a redundant communication protocol, which automatically switches to the mine emergency communication link when the main wireless network is interrupted.

[0040] Furthermore, the model processing module has a built-in error compensation unit to eliminate sensor accuracy errors and electromagnetic interference noise.

[0041] Furthermore, the concentration distribution map generated by the early warning output module is presented in the form of a heat map, and red warning areas with concentrations below 18% are marked.

[0042] The beneficial effects of this invention are as follows:

[0043] (1) Breaking through the limitations of traditional single-point intermittent monitoring, a full-space distribution map of oxygen concentration at the working face is constructed through multi-sensor gradient deployment and real-time data transmission. The decay law of oxygen along the working face is accurately characterized by the exponential decay function. Combined with the spatiotemporal constraint parameter solving algorithm, a continuous quantitative description of the concentration from the intake airway to the upper corner is realized for the first time.

[0044] (2) The system autonomously calculates the model update cycle based on the environmental complexity index and tracks dynamic factors such as ventilation fluctuations and mining disturbances in real time. When the oxygen concentration changes abnormally, the system automatically triggers model reconstruction and marks low-oxygen disaster areas, providing early warning of asphyxiation risks and gas explosion hazards, thus buying critical time for personnel evacuation and ventilation control.

[0045] (3) The generated heat map visually displays the concentration gradient distribution and risk hotspots, guiding the precise allocation of ventilation system airflow. By analyzing the correlation between oxygen concentration and mining activities, it supports the optimization of coal mining technology and the formulation of emergency plans, promoting the shift of coal mine safety management from passive response to proactive prevention and control.

[0046] (4) The sensor deployment adopts the D-optimal design criterion and gradient monitoring strategy, which significantly improves the data reliability under complex interference environments. The modular architecture is compatible with different geological conditions and mining scales, and the adaptive update mechanism of model parameters ensures that the system remains effective in high-risk scenarios such as shallow coal seams and close-range coal seams.

[0047] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0048] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0049] Figure 1 This is a flowchart of the present invention;

[0050] Figure 2 Layout diagram of oxygen concentration sensors at the working face;

[0051] Figure 3 The curve is a fitting curve for oxygen concentration;

[0052] Figure 4 The distribution pattern of oxygen concentration at the working face;

[0053] Figure 5 Steps for calculating the period t to update the oxygen concentration distribution model. Detailed Implementation

[0054] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0055] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0056] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0057] like Figure 1 As shown, the present invention was implemented in the field at the 85218 working face of a shallow coal seam in a coal mine. Based on the ventilation system diagram and the mining engineering plan, four oxygen concentration monitoring sensors were installed at the intake corner, the middle of the working face, and the upper corner. The set of sensor locations is R = (r1, r2, r3, r4), where r1 represents the distance between sensor T1 and the intake airway, r2 represents the distance between sensor T2 and the intake airway, r3 represents the distance between sensor T3 and the intake airway, and r4 represents the distance between sensor T4 and the intake airway.

[0058] The 85218 working face has a dip length of 210m. The predicted oxygen concentration in the intake airway is 20.1%, and the expected oxygen concentration in the return air is 18%. Four oxygen concentration sensors are required. Since the sensor placement on the working face meets the D-optimal design criterion, the determinant of the Fisher information matrix for parameter estimation is maximized. Simultaneously, at least two oxygen concentration sensors are installed within a 100m radius of the return air corner to form a gradient monitoring network for capturing areas of abrupt changes in oxygen concentration. The final set of oxygen concentration sensor locations on the 85218 working face is R = (0, 82, 157, 210). Figure 2 As shown. After installation, a comprehensive calibration is performed using a standard gas source, and parameters such as sensitivity are finely adjusted to ensure accurate initial readings.

[0059] The initial update cycle of the oxygen concentration distribution model is set to t = 5 min. The oxygen concentration over the 5 min period is processed to form the basic data set V for oxygen concentration fitting. Using the data acquisition system, oxygen concentration data is collected every 20 seconds according to the initial settings. Therefore, the total number of oxygen concentration samples over the 5 min period is n = 15.

[0060] The average gas concentration at each location for time period t in the measured gas concentration data U is calculated to obtain the gas concentration sequence U = [20.5, 20.3, 19.5, 15.9], where 20.5 is the average value of time period t for sensor T1, 20.3 is the average value of time period t for sensor T2, 19.5 is the average value of time period t for sensor T3, and 15.9 is the average value of time period t for sensor T4. Finally, the basic data set for fitting oxygen concentration is obtained as V = [(0, 20.5), (82, 20.3), (157, 19.5), (210, 15.9)].

[0061] The oxygen concentration distribution along the working face is assumed to follow the function y = a - be. c(l-x) The parameter c is less than 0, and its physical meaning is to characterize the characteristic that the oxygen concentration gradually decreases along the working face as it moves away from the intake airway. This decrease trend is closely related to factors such as the diffusion capacity of the ventilation airflow, the influx of harmful gases from the goaf to the working face, and the spatial structure of the stope. Parameter a is explicitly the oxygen concentration in the intake airway, presented intuitively as a percentage. As the maximum value of the oxygen concentration distribution curve described by this function, it represents the initial oxygen content when fresh air enters the working face and is an important benchmark value for the entire model. Parameter l, based on precise geological surveys and mining design, corresponds to the dip length of the working face, in meters, providing a crucial spatial scale for subsequent calculations of oxygen concentration based on the location variable x.

[0062] Since the average oxygen concentration at position T1, the starting end of the working face, is 20.5%, the parameter a = 20.5%; the dip length of the 85218 working face is 210m, therefore l = 210m; using the nonlinear least squares method, the oxygen concentration distribution is fitted to the basic data set V, and the distribution of oxygen concentration along the dip of the working face follows the function y = a - be. c(l-x) The fitting yielded undetermined coefficients b = 4.87 and c = -0.0231, and the fitted curve is shown below. Figure 3 As shown.

[0063] The coefficients a = 20.5, b = 4.87, c = -0.0231, and l = 210 are embedded into the distribution function y = ab*e. c(l-x) A preliminary model of gas concentration distribution was constructed, y = 20.5 - 4.87e -0.0231(l-x) Based on this basic model of gas concentration distribution, the curve of oxygen concentration along the dip direction at the 85218 working face can be obtained as follows: Figure 4 As shown in the figure, the oxygen concentration gradually decreases along the dip of the working face, from a maximum of 20.5% in the intake airway to a minimum of 15.9% at the return air corner. The oxygen concentration is below 18% from 182m from the intake airway to the return air corner, constituting a low-oxygen hazard zone. Therefore, the obtained oxygen concentration distribution pattern at the working face provides an important basis for the prevention and control of low-oxygen hazards. Although the preliminary gas concentration distribution model is based on data from a specific time period and currently lacks dynamic tracking capabilities, it lays a solid foundation for the evolution of subsequent dynamic models. It can preliminarily estimate the corresponding predicted oxygen concentration value based on any variable x at any location on the working face, achieving a static quantitative description of the oxygen concentration distribution at the working face.

[0064] As coal mining operations continue, the working face resembles a constantly changing "ecosystem." The amount of coal seam gas and dust generated by mining activities fluctuates continuously; the decrease in surface atmospheric pressure leads to a drop in pressure at the upper corner of the working face, causing toxic and harmful gases from the goaf to flow towards the working face, resulting in hypoxia disasters; mine pressure can also occasionally cause local spatial deformation in the mining area, affecting ventilation. The coupling of all these factors means that the oxygen concentration at the intake, upper corner, and central locations of the working face is constantly changing. Therefore, the deployed oxygen concentration monitoring sensors are continuously used to collect real-time oxygen concentration values ​​at these locations in an uninterrupted real-time monitoring mode.

[0065] Based on the real-time oxygen concentration data monitored by S5, when the duration of the newly monitored data change reaches the update cycle t, S3 and S4 are re-executed. An advanced data fitting algorithm is used again, combined with the latest oxygen concentration data, to resolve the undetermined coefficients b and c. The newly solved coefficients are then substituted into the distribution function to update the entire model architecture. The update cycle t is dynamically calculated using the environmental complexity index to achieve adaptive updates of the model parameters. The dynamic update cycle t has a negative exponential relationship with the environmental complexity index (ECI), meaning the more complex the environment (higher ECI), the shorter the update cycle (higher real-time requirements); the more stable the environment (lower ECI), the longer the update cycle.

[0066] The steps for calculating the update period t are as follows:

[0067] S41: Computational Environment Complexity Index (ECI)

[0068] ECI = w1c z +w2v f +w3q

[0069] In the formula: c z c is the coefficient of variation of oxygen concentration. z =s1 / v1, where s1 is the standard deviation of oxygen concentration and v1 is the average oxygen concentration;

[0070] v f The air volume fluctuation coefficient v f =s2 / v2, where s2 is the standard deviation of the working face air volume and v2 is the average air volume;

[0071] q is the normalized value of coal production per unit time. q0 represents the coal production per unit time, q min Minimum coal production, q max This represents the maximum coal production.

[0072] w1, w2, and w3 are weighting coefficients, with values ​​of 0.7, 0.2, and 0.1 respectively.

[0073] S42: The update period t is calculated based on the Environmental Complexity Index (ECI). The formula is: t = t0e -k*ECI

[0074] Where t0 is the baseline update period, t0 = 10 min; k is the adjustment coefficient, k = 2;

[0075] S43: Set the minimum period t for updating oxygen concentration distribution. min and the maximum period t max ;

[0076] S44: Determine the final update period t. If the oxygen concentration distribution update period t is calculated... <t minLet t = t min If t>t max Let t = t max If t min ≤t≤t max If the update period is t, then the update period is taken as t.

[0077] To more clearly demonstrate the calculation method of the oxygen concentration update cycle t, we will use a monitoring cycle of the 85218 working face as an example to illustrate the calculation method of t (e.g., Figure 5 As shown):

[0078] (1) Parameter acquisition

[0079] The system collects data on oxygen concentration, air volume, and coal production every 30 seconds, and calculates the current environmental complexity index (ECI). The formula for calculating the Environmental Complexity Index (ECI) is as follows:

[0080] ECI = w1c z +w2v f +w3q

[0081] In the formula: c z c is the coefficient of variation of oxygen concentration. z =s1 / v1, where s1 is the standard deviation of oxygen concentration and v1 is the average oxygen concentration;

[0082] v f The air volume fluctuation coefficient v f = s² / v², where s² is the standard deviation of the working face airflow, v² is the average airflow, and q is the normalized value of coal production per unit time. q0 represents the coal production per unit time, q min Minimum coal production, q max The maximum coal production is represented by w1, w2, and w3, which are weighting coefficients of 0.7, 0.2, and 0.1, respectively.

[0083] (2) Calculation of key parameters

[0084] Oxygen concentration coefficient of variation: Monitoring showed an average oxygen concentration v1 = 19.5% and a standard deviation s1 = 0.3%. Based on the formula for calculating the oxygen concentration coefficient of variation, c... z =s1 / v1, so c z =0.3% / 19.5% = 0.0154; this coefficient reflects the dispersion of oxygen concentration data.

[0085] Air volume fluctuation coefficient: The average air volume obtained from monitoring is v2 = 1200m 3 / min, standard deviation s² = 60m 3 / min, calculated using the airflow fluctuation coefficient formula v f =s² / v², to get v f=60 / 100=0.05; This coefficient reflects the fluctuation of air volume data.

[0086] Normalized coal production: q0 = 300 t / h, normalized according to the following formula. Because q max =500t / h, q min =100t / h, and q = 0.5 is calculated, which reflects the change in output.

[0087] (3) Calculation of Environmental Complexity Index (ECI)

[0088] Based on the Environmental Complexity Index calculation formula ECI=w1c z +w2v f Substituting the calculated values ​​into ECI, we get ECI = 0.7 × 0.0154 + 0.3 × 0.05 + 0.1 × 0.5 = 0.065. This index comprehensively reflects the complexity of the working environment.

[0089] (4) Preliminary calculation of update cycle

[0090] According to the formula for calculating the update cycle, t = t0e -k*ECI Where the baseline update period t0 = 10 min and the adjustment coefficient k = 2, we get t = 10 × e -2×0.065 =10 × 0.877 = 8.77 min

[0091] (5) Update cycle adjustment

[0092] To prevent the system from overloading due to excessively frequent updates, the minimum period t for updating the oxygen concentration distribution is set based on the system's computing capacity. min =2min; Maximum period t max =10 minutes, to ensure the timeliness and responsiveness of safety monitoring. According to the set rules, when the calculated oxygen concentration distribution update cycle t... <t min Let t = t min If t>t max Let t = t max If t min ≤t≤t max Therefore, the update cycle is taken as t. Since 2≤8.77≤10, the final update cycle for the oxygen concentration distribution at working face 85218 is determined to be t=8.77min.

[0093] The calculated oxygen concentration distribution model is updated every 8.7 minutes. The oxygen concentration data for this 8.7-minute period is processed. Since the oxygen concentration data collection time is 20 seconds, the total number of oxygen concentration samples for the 8.7-minute period remains 26. The average value of the oxygen concentration at each location of the measured oxygen concentration data U over the 8.7-minute period is calculated, resulting in the gas concentration sequence U = [20.4, 20.1, 19.6, 15.3]. Finally, the basic data set for fitting the oxygen concentration for the next period is obtained as V = [(0, 20.4), (82, 20.1), (157, 19.6), (210, 15.3)].

[0094] Using an advanced data fitting algorithm again, combined with the latest oxygen concentration data, the undetermined coefficients a, b, and c were resolved, yielding b = 5.09, c = -0.0339, a = 20.4, and l = 210. These newly solved coefficients were then substituted into the distribution function to update the entire model architecture. The resulting oxygen concentration distribution model is y = 20.4 - 5.09e^(-1 / 2). -0.0339(210-x) Based on this model, the oxygen concentration distribution curve during this period can be obtained as follows: Figure 5 As shown in the figure, the oxygen concentration gradually decreases along the dip of the working face, from a maximum of 20.4% in the intake airway to a minimum of 15.3% at the return air corner. The oxygen concentration is below 18% from 189m from the intake airway to the return air corner, which is considered a low-oxygen hazard zone. Therefore, the obtained oxygen concentration distribution pattern at the working face provides an important basis for the prevention and control of low-oxygen hazards. It allows for the dynamic calculation of the corresponding predicted oxygen concentration value based on any variable x at any location on the working face, achieving a dynamic quantitative description of the oxygen concentration distribution at the working face.

[0095] This dynamic, cyclical update mechanism ensures that the constructed oxygen concentration distribution model can track every minute change in oxygen concentration at the working face in real time and accurately, maintaining a high degree of synchronization with the actual underground working conditions, thus providing a rock-solid guarantee for safe coal mine production. As mining progresses, the oxygen concentration monitoring data is updated in real time, obtaining the gas concentration data for time period t again. The above steps are repeated, the data is refitted, and the solution is obtained to obtain a dynamic oxygen concentration distribution model with an update cycle of t.

[0096] Coal mine personnel at all levels can access a visual interactive interface via a monitoring screen and a mobile app to view the oxygen "map" at the working face in real time. They can quickly determine the concentration level by using color indicators and utilize query and retrospective functions to provide key decision support for ventilation control and disaster early warning. In the following months of mining, safety accidents caused by abnormal oxygen concentrations "disappeared", and production efficiency steadily improved.

[0097] Example 1: Dynamic Modeling and Initial Deployment of Oxygen Concentration at the Working Face

[0098] 1.1 Sensor Deployment: In the 85218 working face of the shallow coal seam, based on the ventilation system diagram and mining engineering plan, and following the D-optimal design criteria, four electrochemical oxygen sensors were installed at the intake corner (0m), the middle (82m), near the upper corner (157m), and the return corner (210m). Sensor installation requirements:

[0099] The air inlet corner sensor is suspended vertically on the main airflow axis;

[0100] The upper corner sensor is 0.4m from the roof and tilted at 40° toward the goaf.

[0101] The central sensor is fixed to the support column at a height of 1.5m;

[0102] The oxygen concentration sensor proposed in this invention is positioned at the axial position of the main airflow path, solving the monitoring error problem that may be caused by edge airflow interference in traditional installation methods. In shallow coal seams, the main airflow path is significantly affected by air leakage from surface fissures, and the selection of the axial position can more accurately capture the oxygen concentration changes in the core airflow zone.

[0103] The design of the upper corner sensor, with a distance of ≤0.5m from the roof, more closely approximates the actual path of air leakage in the goaf compared to the conventional requirement of ≤0.3m. Combined with an angle of 30°–45° towards the direction of air leakage, it effectively overcomes the challenges of gas accumulation in the upper corner and the variable direction of air leakage. Because the air leakage intensity is high in shallow, closely spaced coal seams, traditional vertically installed sensors are easily disturbed by air leakage, while the tilted design allows for more direct monitoring of oxygen concentration along the leakage path.

[0104] The design of the sensor height in the central area being level with the breathing zone (1.5m ± 0.2m) meets both ergonomic requirements (the breathing zone height of the worker's operating area) and reflects the actual oxygen concentration that needs to be monitored at the working face. For example, in a fully mechanized mining face, the workers' activity height is concentrated at around 1.5m, and the oxygen concentration at this height is directly related to personnel safety. However, existing standards focus more on equipment protection (such as water splash protection) and do not take into account the breathing needs of personnel.

[0105] This invention, through the coordinated design of axial position, tilt angle, and height, ensures that the oxygen sensor is always located in the critical region of oxygen concentration change, providing reliable data support for dynamic models. For example, in shallow coal seams, surface fissures and air leakage can create complex three-dimensional airflow fields, and arbitrary sensor placement is easily affected by local eddies. The installation parameters of this invention can effectively avoid such interference.

[0106] 1.2 Data Acquisition and Processing:

[0107] Data was collected at 20-second intervals during the initial mining phase, and 15 sets of samples were obtained within 5 minutes.

[0108] Calculate the average concentration at each location: inlet corner 20.5%, middle 20.3%, near the upper corner 19.5%, return corner 15.9%;

[0109] Generate the basic dataset V = [(0,20.5),(82,20.3),(157,19.5),(210,15.9)]; if sensor failure leads to insufficient data, such as fewer than 4 valid data points, calculate b using the formula b = ad, where d is the oxygen concentration in the return air, which is 15.9% here, taken from the return air corner data; otherwise, use the nonlinear least squares method to solve the coefficients.

[0110] 1.3 Static Model Construction:

[0111] Define the distribution function as y = a - be c(l-x) Substitute the parameters a = 20.5 (inlet air concentration) and l = 210 (tendency length);

[0112] The solution obtained by nonlinear least squares method is b = 4.87, c = -0.0231;

[0113] The initial model is formed as y = 20.5 - 4.87e. -0.0231(210-x) .

[0114] Example 2: Dynamic Model Update under Environmental Disturbance

[0115] 2.1 Monitoring anomaly triggers:

[0116] During the mining process, a sudden drop in surface atmospheric pressure caused gas to gush out from the mined-out area;

[0117] The sensor at the top corner detected a sharp drop in concentration from 19.5% to 18.1% over a period of 3 minutes;

[0118] 2.2 Update cycle calculation:

[0119] Collect current parameters:

[0120] oxygen concentration variation coefficient c z =0.3% / 19.5% = 0.0154;

[0121] Air volume fluctuation coefficient v f =60 / 1200 = 0.05;

[0122] Normalized coal production q = (300-100) / (500-100) = 0.5;

[0123] Calculate the environmental complexity index: ECI = 0.7 × 0.0154 + 0.2 × 0.05 + 0.1 × 0.5 = 0.065

[0124] Dynamic update cycle: t = 10 × e -2×0.065 = 8.77 minutes;

[0125] 2.3 Iterative Model Reconstruction:

[0126] 26 new sets of data were collected in 8.77 minutes.

[0127] Update the dataset V = [(0,20.4),(82,20.1),(157,19.6),(210,15.3)]

[0128] The refit yielded b = 5.09 and c = -0.0339.

[0129] A new model y = 20.4 - 5.09e is generated. -0.0339(210-x)

[0130] The early warning module marks the area between 189-210m as a low-oxygen risk zone.

[0131] Example 3: Application of Ventilation Decision Support System

[0132] 3.1 Data Fusion Processing:

[0133] The sensor array module transmits concentration data at four locations in real time.

[0134] The model processing module performs the following operations every 8.77 minutes: verify data validity and remove electromagnetic interference anomalies; update the calculation cycle by calling the current ECI index; and distribute the new model to the early warning module through the fiber optic ring network.

[0135] 3.2 Risk visualization output:

[0136] The early warning output module converts the model into a heatmap;

[0137] Areas marked in red with a concentration below 18%;

[0138] Alarms are automatically pushed to the ground monitoring screen and the mine manager's mobile terminal.

[0139] 3.3 Ventilation Control Response:

[0140] Engineers located the air leaks using heat maps;

[0141] Increase the opening of the return airway damper by 10%, and retest after 20 minutes;

[0142] The concentration in the upper corner rose to 18.5%, and the system lifted the red alert.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for constructing and real-time updating a dynamic distribution model of oxygen concentration at a shallow coal seam mining face, characterized in that: Includes the following steps: S1: Real-time monitoring and acquisition of oxygen concentration at multiple locations: At least four oxygen concentration monitoring sensors are deployed at the coal face, with the sensor location set R = (r1, r2, ..., r...). m ), satisfying r1 < r2 < ... < r m The increasing order relationship and the arrangement position conform to the D-optimal design criterion to maximize the determinant of the Fisher information matrix of parameter estimation; among which, no less than 2 sensors are set within 100m of the return air corner; S2: Set the initial update period of the oxygen concentration distribution model to t1, process the oxygen concentration in the time period t1, and form the basic data set V for oxygen concentration fitting. S3: Oxygen Concentration Distribution Function Construction: Defines the function that governs the distribution of oxygen concentration along the working face: y=a-be c(l-x) Where: a is the oxygen concentration in the intake airway, in %, and is the maximum value of the distribution curve; l is the working face dip length, in meters, determined based on geological surveys; c<0, characterizing the attenuation coefficient of oxygen concentration decreasing along the working face dip; b is the fitting parameter, calculated by the formula b=ad when data is insufficient; d is the oxygen concentration in the return air, in %. S4: Solving for undetermined coefficients: Based on the real-time data set U = [(r i ,z ij )] Calculate the average oxygen concentration at each location. The basic dataset V = [(r1,c1),...,(r m ,c m The undetermined coefficients a, b, c, l in the function are solved using the nonlinear least squares method with time and space constraints. S5: Preliminary static model formation: Substitute the solved coefficients into the distribution function y = a - be c(l-x) Generate a static distribution model of oxygen concentration at the working face; S6: Dynamic update cycle calculation: The update cycle t is dynamically calculated based on the Environment Complexity Index (ECI). ECI=w1c z +w2v f +w3q Among them, c z =s1 / v1, representing the coefficient of variation of oxygen concentration, where s1 is the standard deviation of oxygen concentration and v1 is the average oxygen concentration; v f =s2 / v2, representing the air volume fluctuation coefficient, where s2 is the standard deviation of air volume and v2 is the average air volume; This represents the normalized coal production, where q0 is the real-time coal production, and q min q max This is a historical extreme value; The weighting coefficients are w1 = 0.7, w2 = 0.2, and w3 = 0.

1. The update cycle is calculated using the formula t = t0e. -k·ECI The baseline period t0 = 10 min, the adjustment coefficient k = 2, and t ∈ [t min ,t max ]; S7: Dynamic Model Update: When the duration of the new monitoring data change reaches period t, repeat steps S2 to S4 to update the coefficients and reconstruct the model.

2. The method for constructing and real-time updating a dynamic distribution model of oxygen concentration at a shallow coal seam mining face according to claim 1, characterized in that: In S1, the installation position, angle, and height of the oxygen concentration monitoring sensor are determined based on the ventilation system diagram and mining engineering plan of the coal mining face to ensure that the sensor is located in the core airflow zone of the area to be measured. The installation location satisfies: The sensors in the air intake area are positioned along the main airflow path axis. The upper corner sensor is ≤0.5m away from the roof and tilted 30°~45° toward the direction of air leakage in the goaf. The sensor in the central area is suspended vertically from the support column at a height that is level with the breathing zone, at 1.5m ± 0.2m.

3. The method for constructing and real-time updating a dynamic distribution model of oxygen concentration at a shallow coal seam mining face according to claim 1, characterized in that: In S4, the nonlinear least squares method incorporates a compensation mechanism for instantaneous fluctuations caused by sensor accuracy errors, electromagnetic interference, and mining disturbances during the solution process.

4. The method for constructing and real-time updating a dynamic distribution model of oxygen concentration at a shallow coal seam mining face according to claim 1, characterized in that: In S6, the minimum value of the update period t is tmin min =2min, maximum value t max =10min.

5. A dynamic monitoring system for oxygen concentration at a shallow coal seam mining face, characterized in that: include: Sensor array module: Multiple oxygen concentration monitoring sensors deployed at the coal mining face, with their location set R = (r1, r2, ..., r m ), satisfying r1 < r2 < ... < r m Furthermore, the density within 100m of the return air corner should not be less than 2 units; Data transmission module: Connects to the sensor array module via an industrial wireless network to send the collected oxygen concentration data to the processing center at dynamically adjusted time intervals; Model processing module: Connected to the data transmission module via a fiber optic ring network, used for execution of: (a) Construct the distribution function y = a - be c(l-x) ; (b) Solve for coefficients a, b, c, l based on dataset V; (c) Calculate the Environment Complexity Index (ECI) and its update period t; (d) Trigger dynamic model updates; Early warning output module: Connects to the model processing module via API interface to generate oxygen concentration distribution map of the working face and early warning signals for low oxygen risk areas, and outputs them to the ground monitoring terminal.

6. The dynamic monitoring system for oxygen concentration at shallow coal seam mining faces according to claim 5, characterized in that: The data transmission module adopts a redundant communication protocol and automatically switches to the mine emergency communication link when the main wireless network is interrupted.

7. The dynamic monitoring system for oxygen concentration at shallow coal seam mining faces according to claim 5, characterized in that: The model processing module has a built-in error compensation unit to eliminate sensor accuracy errors and electromagnetic interference noise.

8. The dynamic monitoring system for oxygen concentration at shallow coal seam mining faces according to claim 5, characterized in that: The concentration distribution map generated by the early warning output module is presented in the form of a heat map, and red warning areas with concentrations below 18% are marked.

Images