A coal mine underground hydrogen sulfide intelligent grading prevention and control and comprehensive treatment method

Abstract

The present application relates to the technical field of coal mine safety mining and mine disaster prevention and control, and particularly relates to a coal mine underground hydrogen sulfide intelligent grading prevention and control and comprehensive management method. The method first collects target mine geology, coal quality, hydrogen sulfide emission and field data, and builds a basic data set after preprocessing and standardization; based on the cause type and concentration distribution, a sensitive area is divided and a grading early warning index is established; a four-in-one comprehensive management model is constructed; through real-time feedback of online monitoring data, dynamic optimization of absorption liquid formula, injection parameters, ventilation system and plugging process is realized; the management effect is evaluated in multiple dimensions, and if it is unqualified, parameter adaptive adjustment and global iterative optimization are started; and finally an intelligent, visual and economic and environmentally friendly hydrogen sulfide whole-process prevention and control system is formed. The present application solves the problems of poor injection effect and low intelligent level in the prior art, and is suitable for coal mine hydrogen sulfide disaster prevention and control, and provides support for safe mining of deep high-sulfur coal seams.

Description

Technical Field

[0001] This invention relates to the fields of safe coal mining, mine gas control, and coalfield geological disaster prevention and control, and in particular to a method for intelligent hierarchical prevention and comprehensive management of hydrogen sulfide in underground coal mines. Background Technology

[0002] In recent years, the depth of coal mining in my country has continued to increase. The complex geological conditions of deep coal seams have led to increasingly prominent problems with abnormal hydrogen sulfide (H2S) emissions from high-sulfur coal, goaf areas, and abandoned roadways. These emissions have caused numerous miner accidents and injuries, and safety risks are rapidly increasing. More than 30 coal mines in provinces and autonomous regions such as Henan, Shandong, Inner Mongolia, Shanxi, Gansu, and Sichuan have experienced abnormal H2S emissions.

[0003] Hydrogen sulfide is a colorless, highly toxic gas with a rotten egg odor. It tends to accumulate in low-lying areas of tunnel floors, corroding metal equipment and endangering the health of miners. A concentration of 0.40 ppm (H2S) is detectable by smell; concentrations exceeding 0.76 ppm can cause bronchitis, pneumonia, and pulmonary edema; and concentrations exceeding 1.00 ppm can induce electrocution. Furthermore, the distribution of H2S concentrations in coal mines is extremely uneven, and miners often lack sufficient awareness and attention to its hazards. A sudden surge of high-concentration H2S can easily trigger a major safety accident.

[0004] Currently, H2S prevention and control in coal mines has developed a series of technologies, including monitoring and control, ventilation and dilution, spraying alkaline solution, coal seam injection, borehole extraction, fracture sealing, and personal protective equipment. These technologies have been applied in multiple mining areas such as Binchangting South, Wudong, Longtan, Xiqu, Jinniu, and Shuangma. However, existing technologies still have significant shortcomings, making it difficult to meet the needs of precise, efficient, economical, environmentally friendly, and sustainable H2S control in deep coal mines: First, commonly used alkaline absorbents (lime water, Na2CO3, NaOH, etc.) are highly corrosive, easily damaging underground equipment and significantly increasing control costs; second, absorbents containing K and Na, when injected into the coal seam, alter coal quality, affecting industrial production such as metallurgical coke, and even causing blast furnace perforation and shortening equipment lifespan; third, the fluid diffusion range during coal seam injection is limited, the workload is large, and the time consumption is long, the permeability enhancement and penetration mechanisms are unclear, and the injection effect is poor; fourth, absorbents cannot be fully utilized, resulting in serious resource waste; fifth, the level of intelligence in H2S monitoring and control is low, lacking an automatic detection, graded response, dynamic control, and full-process optimization system, relying excessively on manual operation, and the response is lagging; sixth, passive control models are mostly adopted, failing to achieve integrated prevention and control of "cause identification—zonal early warning—source control—process blocking—end absorption—intelligent optimization," resulting in weak control targeting and unstable effects. Therefore, there is an urgent need for an intelligent, hierarchical, and comprehensive method for controlling hydrogen sulfide, to address the pain points of existing technologies, and to provide reliable technical support for the safe mining of deep high-sulfur coal seams. Summary of the Invention

[0005] To overcome the problems of strong corrosiveness of absorbent liquid, impact on coal quality, poor injection effect, low level of intelligence, resource waste, and lack of dynamic optimization in existing coal mine H2S prevention and control technologies, this invention provides an intelligent hierarchical prevention and control and comprehensive management method for hydrogen sulfide in coal mines. This method achieves H2S source inhibition, process blocking, end-point absorption, intelligent regulation and full-process optimization, providing technical support for the safe and efficient mining of deep high-sulfur coal seams.

[0006] The present invention adopts the following technical solution: A method for intelligent hierarchical prevention and comprehensive management of hydrogen sulfide in coal mines, comprising: S1: Collect geological, coal quality, hydrological, ventilation, H2S concentration and existing treatment data of the target coal mine, construct a basic dataset and perform preprocessing and standardization to obtain a standardized dataset; S2, based on the identification of organic / inorganic causes of H2S, emission intensity and spatial distribution, divide the prevention and control areas and establish a concentration-level early warning system; S3, based on the cause type and risk level, constructs a four-in-one comprehensive prevention and control model of "source inhibition - process interruption - end absorption - individual protection"; S4 automatically performs high-pressure injection, spray absorption, fissure sealing, air volume control and directional extraction according to the risk level to achieve coordinated treatment; S5 uses downhole multi-point H2S sensors to monitor in real time and transmit concentration data back to form a closed-loop feedback. S6. Establish multi-dimensional governance effectiveness evaluation indicators to determine whether the standards are met; S7, when the standard is not met, the absorbent formula, injection parameters, ventilation system and sealing position are adjusted in a graded adaptive manner based on the deviation coefficient; S8, based on the cumulative governance error and the effect decay coefficient, perform global iterative optimization of the model; S9 outputs the final governance solution and builds an intelligent monitoring, visualization and data management system.

[0007] Further, step S1 includes: S11, the collected data includes: burial depth, formation temperature, sulfur content in coal, sulfate-reducing bacteria, groundwater pH and mineralization, coal seam fracture development characteristics, real-time H2S concentration, air volume and wind speed, and injection / spraying / plugging / extraction process parameters. S12 employs wavelet denoising and the 3σ criterion to remove outliers, and Z-score standardization to eliminate dimensional differences.

[0008] Furthermore, step S2, the graded early warning, includes: S21, Identifying the origin of H2S: biodegradation, microbial sulfate reduction (BSR), thermochemical sulfate reduction (TSR), and magmatic volcanic origin; S22 is classified into four levels according to concentration: Level I: <0.40 ppm (routine control); Level II: 0.40~0.76 ppm (enhanced monitoring); Level III: 0.76~1.00 ppm (active treatment); Level IV: >1.00 ppm (immediate evacuation + intensive treatment); S23, combining the outflow location, spatial aggregation characteristics, and operational intensity, divides the area into key prevention and control areas, secondary key prevention and control areas, and routine prevention and control areas, and establishes a zoned early warning model.

[0009] Furthermore, step S3, the four-in-one model, includes: S31, Source Suppression: High-pressure pre-injection of alkaline absorbent solution in both deep and shallow wells; S32, process interruption: grout sealing of fractures, goaf areas, and seepage channels in abandoned roadways; S33, terminal absorption: online capture of H2S using spray, water curtain, and foam absorbent liquid; S34, Personal Protective Equipment: Protective gear, metal corrosion protection, emergency evacuation and ventilation systems; S35 forms a systematic governance system of "removal, drainage, blocking, dredging, and pumping".

[0010] Furthermore, step S4 intelligent execution includes: S41, automatically prepares corrosion-inhibiting alkaline absorbent solution with pH 9.0~10.0; S42 automatically adjusts the injection pressure, flow rate, orifice diameter, and spacing based on the coal permeability. S43 automatically controls the air curtain, water curtain, and local ventilation fan to achieve directional dilution and drainage; In S44, high-risk areas, drilling extraction and sealing, as well as fluid injection, are carried out in a coordinated manner.

[0011] Furthermore, step S5, real-time monitoring feedback, includes: S51, H2S sensors are deployed in the working face, upper corner, goaf, return airway and low-lying areas of the floor; S52, uploads concentration data in real time, and automatically triggers spraying, alarm, power cut-off, and evacuation when concentration exceeds limits; S53 generates time-space-concentration dynamic evolution curves, providing data support for model optimization.

[0012] Furthermore, the evaluation of the effect of step S6 includes: S61, Evaluation indicators: H2S concentration reduction rate, gas absorption rate, treatment coverage, degree of impact on coal quality, material consumption, equipment corrosion rate, labor cost, and safety risk level. S62. Construct a comprehensive evaluation index. If the index reaches a preset threshold, the evaluation is considered qualified; otherwise, it is considered unqualified.

[0013] Furthermore, step S7, hierarchical adaptive adjustment, includes: S71, Calculate the governance deviation coefficient; S72, deviation ≤0.2: fine-tune spray flow rate, injection time and ventilation volume; S73, 0.2 < Deviation ≤ 0.5: Adjust the concentration of the absorbent and add corrosion inhibitor and penetration enhancer; S74, Deviation > 0.5: Reassess the cause and risk level, and adjust the injection, extraction, and blocking strategies accordingly; S75 uses incremental learning to update the local prevention and control model.

[0014] Furthermore, step S8, global iterative optimization, includes: S81, set the statistical window, calculate the cumulative error and effect attenuation coefficient; S82, slow decay: Expanded field data, optimized parameters and formulation; S83, severe attenuation: reconstruct the prevention and control model, and change the absorption system and process route; S84 takes into account safety, coal quality, corrosion prevention, cost, and environmental protection, achieving sustainable optimization.

[0015] Furthermore, the S9 intelligent visualization system includes: S91, Establish an H2S prevention and control database, linking geological, causal, concentration, measures, and effect data; The S92 uses cloud maps, graphs, early warning interfaces, and prevention and control engineering distribution maps to enable remote monitoring, intelligent decision-making, historical tracing, and report output.

[0016] In summary, this invention integrates H2S cause identification, graded early warning, comprehensive treatment, and intelligent control, solving the problems of passive, fragmented, and poor treatment effects of traditional technologies. It employs corrosion-inhibiting absorbent liquid with intelligent parameter matching to reduce equipment corrosion and minimize the impact on coal quality. The coordinated operation of injection, extraction, plugging, drainage, and spraying improves absorption efficiency and coverage. It establishes an automatic monitoring-feedback-adjustment-optimization closed loop with a high degree of intelligence. Considering safety, economy, environmental protection, and sustainability, it is suitable for widespread application in multiple mining areas. Attached Figure Description

[0017] Figure 1 A flowchart illustrating the steps of an intelligent hierarchical prevention and comprehensive management method for hydrogen sulfide in underground coal mines; Figure 2 This is the logic diagram for hydrogen sulfide formation identification and regional early warning in this invention; Figure 3 This is a schematic diagram of the four-in-one integrated governance structure of the present invention; Figure 4 This is a flowchart illustrating the governance effect evaluation and adaptive adjustment process of this invention. Figure 5 This is a block diagram of the global iterative optimization and intelligent visualization system of this invention. Detailed Implementation

[0018] To make the objectives and advantages of the present invention clearer, the present invention will be further described below in conjunction with embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0019] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0020] It should be noted that in the description of this invention, the terms "upper," "lower," "left," "right," "inner," and "outer," etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is merely for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0021] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0022] Specifically, the basic data collected in step S1 of this invention covers all dimensions of data required for H2S control in the target coal mine, ensuring the integrity and representativeness of the dataset and providing reliable data support for subsequent model construction, early warning and optimization. Specifically, it includes: formation temperature, burial depth, sulfur content of coal and rock (organic sulfur and inorganic sulfur content), distribution density of sulfate-reducing bacteria, groundwater salinity, groundwater pH value, coal seam fracture development characteristics (fracture density, aperture, connectivity), real-time H2S concentration (instantaneous concentration and average concentration), air volume and wind speed, existing injection / spraying / plugging / extraction process parameters (injection pressure, flow rate, borehole spacing, spray flow rate and pressure, plugging material ratio, extraction negative pressure and flow rate), and historical treatment records (treatment measures, treatment effects, equipment wear and tear, cost data), etc.

[0023] Data preprocessing employs a full-process approach of "denoising—outlier removal—completion—standardization" to ensure data quality: A wavelet denoising algorithm (using the db4 wavelet basis with a 4-level decomposition scale) is used to smooth spikes and jumps in time-series data such as H2S concentration, air volume, and wind speed, effectively preserving the data's trend characteristics while removing random noise (mainly caused by sensor malfunctions, downhole electromagnetic interference, and operational disturbances); linear interpolation is used to complete missing data (applicable when the missing rate is below 5%, and when the missing rate is above 5%, completion is performed by combining the trends of similar parameters in adjacent areas), avoiding the impact of missing values ​​on subsequent analysis; outlier removal uses the 3σ criterion, calculating the mean μ and standard deviation σ of each parameter, identifying values ​​exceeding μ ± 3σ as outliers (such as instantaneous changes in H2S concentration, abnormal fluctuations in air volume and wind speed, etc.), and correcting them using the mean of 10 adjacent data points; standardization uses the Z-score standardization method, calculated using the formula... x *=( x - m ) / s (in x The original data values, m This is the mean of the parameter. s The standard deviation of this parameter normalizes all types of data to the order of magnitude of mean 0 and variance 1, effectively eliminating the differences in data dimensions between different mines (differences in geological conditions and mining depth), different monitoring equipment (differences in instrument model and accuracy), and different construction techniques (differences in injection and extraction parameters), improving the universality of the data and ensuring that cross-mine data can be directly compared and reused.

[0024] Please see Figure 2 The diagram shown illustrates the logic diagram for hydrogen sulfide formation identification and regional early warning in this invention. The formation identification and graded early warning process in step S2 of this invention is a core prerequisite for achieving precise H2S control. Through a three-level logic of "formation identification—concentration grading—regional division," a scientifically sound early warning system is constructed, specifically implemented as follows: First, the etiology of H2S was identified to clarify its source and provide a basis for matching subsequent remediation measures. The identification indicators covered four core etiologies: biodegradation (mainly generated by the decomposition of organic sulfur by microorganisms in coal seams), microbial sulfate reduction (BSR) (sulfate-reducing bacteria reduce sulfate in coal seams to H2S under suitable temperature and pH conditions, which is the most common type of H2S formation underground), thermochemical sulfate reduction (TSR) (sulfate reacts thermochemically with organic matter to generate H2S under high temperature and pressure conditions in deep formations), and magmatic / volcanic origin (H2S released during magmatic activity flows underground along formation fissures; this type of formation is relatively rare in Chinese coal mines). The identification process combined parameters such as formation temperature, sulfate-reducing bacteria distribution density, sulfur isotope composition, and groundwater pH, employing a multi-indicator comprehensive identification method to ensure an accuracy rate of ≥90% in etiology identification.

[0025] Secondly, the risk level is divided into four levels based on H2S concentration, according to the toxicity and hazard characteristics of H2S and coal mine safety operation standards. The specific levels are as follows: Level I (Routine Prevention Level): H2S concentration < 0.40 ppm, below the odor threshold, no obvious harm to the human body, only routine monitoring is required, no special treatment is needed; Level II (Warning Level): H2S concentration 0.40~0.76 ppm, can be clearly smelled by the human body, long-term exposure may cause mild respiratory discomfort, real-time monitoring should be strengthened, and concentration changes should be closely monitored; Level III (Hazard Level): H2S concentration 0.76~1.00 ppm, has obvious toxicity, exposure can cause bronchitis, pneumonia and other diseases, active treatment measures should be initiated immediately, and the working time of workers should be restricted; Level IV (Lethal Level): H2S concentration > 1.00 ppm, extremely toxic, short-term exposure can cause electric shock death, workers should be evacuated immediately, power to the work area should be cut off, and the highest level of emergency treatment should be initiated.

[0026] Finally, based on the location of H2S outbursts, spatial aggregation characteristics, and work intensity, sensitive control areas were divided into three categories: key control areas (H2S concentration ≥ 0.76 ppm, and densely populated areas, such as the working face head, upper corner, and goaf exit), secondary key control areas (H2S concentration 0.40~0.76 ppm, or areas not densely populated but key channels for H2S transport, such as return airways and connecting roadways), and routine control areas (H2S concentration < 0.40 ppm, and no significant risk of H2S outbursts, such as the bottom of the shaft and electromechanical chambers). For different control areas, a zoned early warning model was established, setting differentiated monitoring frequencies and early warning thresholds. Concentration data was collected every 10 seconds in key control areas, and a Level II or higher early warning immediately triggered a coordinated response to ensure timely and targeted early warnings.

[0027] Single-cause identification methods are easily affected by a single parameter, leading to identification bias (e.g., judging TSR formation solely based on formation temperature can easily be confused with BSR formation). Single concentration grading, without considering the operational scenario, can easily result in wasted resources or inadequate control (e.g., the same concentration may have different risk levels in densely populated areas versus non-operational areas). This invention, through a collaborative model of "cause identification + concentration grading + regional division," clarifies the source of H2S generation, providing a scientific basis for matching control measures, and also divides risk levels and control areas based on actual operational scenarios, achieving "precise early warning and tiered policy implementation." Compared to traditional single-cause early warning models, early warning response efficiency is improved by more than 40%, and the targeted nature of control is significantly enhanced.

[0028] Please see Figure 3 The diagram shown illustrates the four-in-one integrated governance structure of this invention. The four-in-one integrated governance model of "source suppression—process interruption—end-of-pipe absorption—individual protection" constructed in step S3 of this invention breaks through the limitations of traditional single governance models, achieving coordinated prevention and control of H2S across the entire process and multiple pathways. The modules cooperate with each other and progress layer by layer. The specific structure is as follows: 1. Source Suppression Module: The core objective is to suppress the generation and release of H2S, reducing H2S emissions at the source. This is achieved by injecting alkaline absorbent solution into the coal seam under high pressure in shallow / deep pores. The absorbent solution reacts with sulfur and sulfate-reducing bacteria in the coal seam, inhibiting H2S generation and simultaneously sealing micro-fractures within the coal seam, reducing H2S release into the roadway. For BSR (Biosulfate Reduction) formation, sulfate-reducing bacteria inhibitors can be added to the absorbent solution to further enhance the source suppression effect.

[0029] 2. Process Interception Module: Focusing on the migration path of H2S, this module uses methods such as slurry sealing and airtight isolation to block the migration of H2S from coal seams, goafs, and abandoned roadways to the working area. Specifically, it includes: using cement slurry and polyurethane slurry to seal coal and rock fissures, abandoned roadway interfaces, and goaf seepage channels to form an airtight isolation layer; sealing goafs to reduce H2S leakage; and setting up isolation walls at key H2S migration channels to guide H2S to accumulate in designated areas for subsequent extraction and absorption.

[0030] 3. End-of-line absorption module: For H2S that has entered the working area, a synergistic approach of "spray absorption + foam capture + extraction treatment" is adopted to achieve rapid removal of H2S. Specifically, this includes: deploying medium- and high-pressure spray systems in areas such as the working face, upper corner, and return airway to spray alkaline absorbent liquid, achieving H2S absorption through gas-liquid contact; spraying foam absorbent liquid in H2S accumulation areas (such as low-lying areas on the roadway floor) to expand the gas-liquid contact area and improve absorption efficiency; and implementing directional drilling extraction in areas with high concentrations of H2S to extract H2S to the surface treatment device for compliant discharge.

[0031] 4. Personal Protective Equipment (PPE): As the last line of defense, this module ensures the personal safety of personnel working underground. Specifically, it includes: equipping personnel with acid-resistant masks, protective goggles, and chemical protective suits, and regularly testing the effectiveness of the protective equipment; performing anti-corrosion treatment on underground metal equipment (such as applying emulsified oil or wrapping with an anti-corrosion layer) to reduce H2S corrosion; and establishing an emergency evacuation and ventilation system to guide personnel to quickly evacuate to a safe area when H2S concentration exceeds the standard, while simultaneously activating the emergency ventilation system to dilute the H2S concentration in the work area.

[0032] The aforementioned four-in-one model does not operate with each module operating independently, but rather forms a systematic prevention and control strategy of "removal, elimination, blocking, dredging, and extraction." Based on the type and risk level of H2S formation, the operating parameters and priorities of each module are dynamically matched (e.g., for BSR-causing diseases, the focus is on source suppression and process interruption; for TSR-causing diseases, the focus is on extraction and end-of-pipe absorption), ensuring the stability and targeted nature of the treatment effect. Compared to traditional single-mode treatment, this model improves the H2S removal rate by more than 30% and the treatment stability by 50%.

[0033] The core of the intelligent governance execution process in step S4 of this invention is to achieve "automatic matching of risk level and governance parameters," reducing manual intervention and improving governance efficiency and accuracy. Specific execution details are as follows: Based on the risk level and cause type determined in step S2, the treatment plan and process parameters are automatically configured: First, the absorbent formulation is automatically matched, prioritizing the use of corrosion-inhibiting alkaline absorbent to avoid equipment corrosion and coal quality impact. The pH value is controlled between 9.0 and 10.0 (this range ensures H2S absorption efficiency while reducing corrosivity). The specific formulation is adjusted according to the cause type: for BSR, a composite formulation of sodium carbonate + corrosion inhibitor + sulfate-reducing bacteria inhibitor is used; for TSR, a composite formulation of sodium bicarbonate + penetration aid is used; and for conventional control, lime slurry dilution is used (economical and environmentally friendly). Second, the injection parameters are automatically adjusted. Based on the coal permeability (obtained through previous geological data), the injection pressure, flow rate, borehole diameter, and spacing are automatically adjusted. When the permeability is low (<1.0 mD), the injection pressure is increased (3.5 mD). The system employs several measures: first, reducing the injection pressure (2.5-3.0 MPa) and increasing the borehole spacing (2.0-2.5 m) to improve the diffusion range of the absorbent; second, reducing the injection pressure (2.5-3.0 MPa) and increasing the borehole spacing (3.0-3.5 m) when the permeability is high (>1.5 mD) to lower treatment costs; third, automatically adjusting ventilation parameters based on the location and concentration of H2S, including adjusting the angle of the air curtain, the number of water curtains opened, and the speed of local ventilators to achieve directional dilution and drainage of H2S, with wind speeds controlled at 2.5-3.5 m / s in key control areas to ensure that H2S does not accumulate; and fourth, implementing collaborative operation control, automatically initiating collaborative operations of "injection + sealing + extraction + spraying" in high-risk areas (Level III and above), and initiating single or combined treatment measures in medium- and low-risk areas according to the actual situation to achieve a reasonable allocation of treatment resources.

[0034] Please see Figure 4 The diagram shown illustrates the flowchart for evaluating and adaptively adjusting the governance effect of this invention. The real-time monitoring and feedback process in step S5 of this invention is crucial for achieving closed-loop control of the governance effect. Through a "multi-point monitoring—real-time transmission—over-limit linkage" model, it ensures that the H2S concentration remains within a controllable range. The specific implementation is as follows: The monitoring points are arranged according to the principle of "full coverage with emphasis on key areas." Fixed H2S sensors are deployed in areas prone to H2S accumulation, such as the working face head, upper corner, inside the goaf and exit, return airway, and low-lying areas of the roadway floor, with one sensor every 50m. In key control areas, one sensor is deployed every 30m. Additionally, 2-3 mobile H2S sensors are provided for temporary monitoring and hazard identification in the work area. The sensors use high-precision detection modules with a detection range of 0-5.0ppm and a detection accuracy of ±0.01ppm, ensuring the accuracy of concentration data.

[0035] Real-time monitoring data is transmitted back to the ground monitoring center via a wireless transmission module (adapted to downhole explosion-proof requirements), forming a dynamic evolution curve of H2S concentration over time, space, and time, visually displaying the changing trend and spatial distribution characteristics of H2S concentration. When the monitored concentration exceeds the warning threshold for the corresponding risk level, the system automatically activates relevant treatment equipment and emergency devices: when the concentration reaches Level II warning, the alarm device is activated to alert workers; when it reaches Level III warning, the spray flow rate is automatically increased, the directional extraction device is activated, and the working time of workers is limited; when it reaches Level IV warning, the power supply to the work area is immediately cut off, an emergency evacuation prompt is activated, and the emergency ventilation system is activated to ensure that workers quickly evacuate to a safe area. Simultaneously, real-time concentration data is stored in association with treatment parameters, providing data support for subsequent treatment effect evaluation and model optimization.

[0036] The multi-dimensional evaluation process of the treatment effect in step S6 breaks through the limitations of traditional single-concentration evaluation. It combines the needs of safety, economy, environmental protection, and coal quality protection to construct a comprehensive evaluation system, ensuring the comprehensiveness and rationality of the treatment effect. The specific evaluation indicators and judgment criteria are as follows: The evaluation indicators cover eight core categories, divided into three main categories: safety indicators, economic indicators, and environmental and coal quality indicators. Safety indicators include H2S concentration reduction rate, gas absorption rate, treatment coverage area, and safety risk level. Economic indicators include material consumption, equipment corrosion rate, and labor costs. Environmental and coal quality indicators include the degree of impact on coal quality (mainly detecting changes in K and Na content in coal). Specifically, the H2S concentration reduction rate = (average concentration before treatment - average concentration after treatment) / average concentration before treatment × 100%; the gas absorption rate = (amount of absorbed H2S / amount of emitted H2S) × 100%; and the equipment corrosion rate = (corrosion area / total equipment surface area) × 100%.

[0037] A comprehensive evaluation index is constructed and calculated using a weighted summation method. The weighting is adjusted according to the risk level (high-risk areas have higher weights for safety indicators, while medium- and low-risk areas have higher weights for economic indicators). A passing threshold of 0.75 is set for the comprehensive evaluation index. (This threshold is determined based on statistical analysis of 2000 sets of treatment data from eight coal mines with different geological conditions. When the comprehensive evaluation index ≥ 0.75, the H2S concentration is stably controlled below the warning threshold, the equipment corrosion rate is ≤ 0.3%, the coal quality shows no significant change, and the treatment cost is reasonable. Below 0.75, there are problems such as concentration rebound, severe equipment corrosion, and affected coal quality.) If the comprehensive evaluation index ≥ 0.75, the treatment effect is deemed satisfactory, and the process proceeds to the subsequent global optimization stage. If the comprehensive evaluation index < 0.75, the treatment effect is deemed unsatisfactory, the treatment deviation coefficient is calculated, and a tiered adaptive adjustment is initiated.

[0038] The core advantage of this treatment effect evaluation is that traditional single concentration evaluation only focuses on the H2S removal effect, ignoring key factors such as equipment corrosion, coal quality impact, and treatment cost, which can easily lead to problems such as "treatment meets standards but costs are too high" and "concentration decreases but equipment is damaged". The multi-dimensional evaluation system of this invention takes into account safety, economy, environmental protection and coal quality protection, ensuring the sustainability of treatment effect, while providing a clear direction for subsequent adjustments, making the optimized treatment plan more practical.

[0039] The hierarchical adaptive adjustment process in step S7 adopts differentiated adjustment strategies based on the magnitude of the governance deviation coefficient to achieve precise optimization of governance parameters. Simultaneously, an incremental learning strategy is used to update the local model, improving the stability of the governance effect. The specific adjustment logic and implementation details are as follows: First, calculate the treatment deviation coefficient. The formula is: Treatment deviation coefficient = (target concentration - measured concentration) / target concentration, where the target concentration is determined according to the risk level (Level I ≤ 0.40 ppm, Level II ≤ 0.76 ppm, Level III ≤ 0.76 ppm, Level IV ≤ 0.40 ppm). The deviation coefficient ranges from 0 to +∞. The larger the deviation coefficient, the greater the gap between the treatment effect and the target, and the greater the adjustment is required.

[0040] Adjustment levels are determined based on the deviation coefficient, which is determined by the correlation between the deviation coefficient and the treatment effect (through statistical analysis of multiple sets of experiments, when the deviation coefficient is ≤0.2, the treatment effect is close to the target and only minor adjustments are needed; when 0.2 < deviation coefficient ≤0.5, there is a significant gap in the treatment effect, and the core parameters need to be adjusted; when the deviation coefficient >0.5, the treatment effect is seriously substandard, and the treatment plan needs to be rematched). The specific adjustment strategies are as follows: 1. When the treatment deviation coefficient is ≤0.2, fine-tune the basic parameters—no need to adjust the absorbent formula and treatment mode, only fine-tune the spray flow rate (adjustment range is ±10%~15%), injection time (adjustment range is ±2~3 hours / day) and ventilation volume (adjustment range is ±5%~10%) to ensure that the concentration is stable within the target range. This adjustment method is fast and efficient and does not require interruption of treatment operations.

[0041] 2. When 0.2 < treatment deviation coefficient ≤ 0.5, adjust the core parameters—focus on adjusting the concentration and formula of the absorbent, and add corrosion inhibitors (0.3%~0.8%) and penetration enhancers (0.2%~0.5%) according to the cause type to improve absorption efficiency and diffusion range; at the same time, fine-tune parameters such as injection pressure and extraction negative pressure to optimize the treatment effect. There is no need to reconstruct the treatment model, only update some parameters.

[0042] 3. When the governance deviation coefficient is >0.5, reconstruct the governance plan—re-determine the cause type and risk level of H2S (there may be deviations in cause identification or unreasonable risk area division), adjust the injection location, sealing range and extraction plan, replace the absorbent system, and ensure that the governance measures match the actual situation to achieve a fundamental improvement in governance effect.

[0043] The incremental learning strategy is implemented as follows: Data from areas with substandard treatment effects (measured concentrations, treatment parameters, geological conditions) are combined with supplementary monitoring data from the corresponding areas (supplementary sensor deployment to obtain more detailed concentration distribution data) to form an incremental learning dataset (the incremental dataset is 1.5 times the size of the substandard area data, supplementing with data from similar compliant areas to avoid incremental data bias). An incremental learning algorithm based on SGD (Stochastic Gradient Descent) is used to update the local control model. The specific steps are as follows: ① Fix the basic structure of the local prevention and control model, and only unlock the parameter update permissions (such as absorbent formula coefficient, injection parameter threshold, ventilation control logic, etc.). ② Divide the incremental learning dataset into an incremental training set and an incremental validation set in an 8:2 ratio. Use the SGD algorithm with a learning rate of 0.0005 (lower than the initial model training learning rate to avoid parameter mutations) and 100 iterations to locally update the unlocked parameters. ③ After incremental training is completed, the model accuracy is verified using an incremental validation set. If the comprehensive evaluation index of the incremental validation set is ≥0.80, the local optimization is complete; if the comprehensive evaluation index is <0.80, repeat the incremental training steps and adjust the learning rate (±0.0001) until the accuracy is achieved.

[0044] This incremental learning strategy does not require full retraining of the entire prevention and control model, but only updates local parameters. Compared with full retraining, the optimization efficiency is improved by more than 60%, and it can realize real-time correction of the treatment process, adapt to subtle changes in local geological conditions and H2S concentration, and ensure the stability of the treatment effect.

[0045] Please see Figure 5 The diagram shown is a block diagram of the global iterative optimization and intelligent visualization system implemented in this invention. The core objective of the global iterative optimization process in step S8 of this invention is to improve the cross-regional universality and long-term operational stability of the prevention and control model, avoiding the model being adapted only to the geological conditions of a single mine or region. The specific implementation process is as follows: The statistical window is set to 7 days (which can be adjusted to 5-10 days depending on the actual mining pace). The cumulative treatment error coefficient and effect attenuation coefficient are calculated for each window. The formula for calculating the cumulative treatment error coefficient is as follows: (in C i For the first iThe measured average concentration over the days C o For the target concentration, n (Number of days within the statistical window), cumulative governance error coefficient E The value range is 0~5.0. E The larger the value, the greater the governance error within that window, and the worse the model's adaptability; the formula for calculating the effect decay coefficient is: or = E k / E k-1 (in E k This represents the cumulative governance error coefficient for the current statistical window. E k-1 This is the cumulative governance error coefficient from the previous statistical window. k ≥2, the first window has no effect decay coefficient and does not trigger hierarchical optimization). or The range of values ​​for is 0 to +∞. or Reflecting the changing trend of governance effectiveness between the current window and the previous window: or When the error is ≤1, the current window's governance error is ≤ the previous window's error, and the effect shows a decreasing or stable trend. or When the value is greater than 1, the governance error of the current window is greater than that of the previous window, the effect shows a downward trend, and the global adaptability of the model continues to decline.

[0046] The cumulative governance error coefficient threshold was set at 0.15 (this threshold was determined based on statistical analysis of experiments conducted in three coal mines with different geological conditions). E When the value is ≤0.15, the comprehensive evaluation index within the window is ≥0.78, indicating good model adaptability; E When the coefficient is greater than 0.15, the comprehensive evaluation index within the window is less than 0.75, indicating insufficient global adaptability of the model. If the cumulative governance error coefficient is greater than 0.15, global iterative optimization is triggered. E If the value is ≤0.15, the model has good global adaptability and no global optimization is required. Continue with the subsequent governance steps.

[0047] Based on the magnitude of the effect decay coefficient, a hierarchical global optimization strategy is executed, with the specific logic as follows: ① When orWhen the value is ≤1, a data expansion and optimization strategy is triggered. At this time, the model error shows a decay or stabilization trend, indicating that the model itself has a reasonable structure and only has the problem of insufficient training data (especially the lack of cross-regional data). There is no need to reconstruct the model. Just add the newly collected cross-regional mine data (covering at least one new mine, with a data volume of no less than 20% of the existing training set) and the corresponding governance data to the training set, re-divide the training set and the test set (still divided at 7:3), use the hyperparameter settings of the initial training, and perform light training on the model (the number of iterations is 1 / 3 of the initial training) to supplement the cross-regional data features, improve the model's generalization ability, and adapt to the differences in geological conditions of different mines.

[0048] ② When or When the error is greater than 1, a full retraining optimization strategy is triggered. At this point, the model error shows an upward trend, indicating that the existing model structure or governance parameters can no longer adapt to the global data (this may be due to large differences in geological conditions across regions, changes in the H2S formation type, or insufficient universality of governance process parameters). It is necessary to return to steps S2-S3, re-identify the H2S formation type, divide the risk areas, adjust the four-in-one model structure, and perform full retraining based on the expanded cross-regional standardized dataset (original data + newly collected cross-regional data). Iteratively adjust hyperparameters (such as absorbent formulation ratio, injection parameter range, warning threshold, etc.) until the model's comprehensive evaluation index on the test set is ≥0.80, thereby realizing the model's self-learning and evolution, and significantly improving the model's cross-regional generality and long-term operational stability.

[0049] The intelligent visualization system construction in step S9 of this invention realizes intelligent management of the entire process of H2S prevention and control, taking into account data storage, real-time monitoring, intelligent decision-making and historical traceability. The specific implementation is as follows: First, a smart database for the prevention and control of hydrogen sulfide in mines is constructed. A relational database (such as MySQL) is used to store data in all dimensions. H2S concentration data, geological condition data, cause identification results, risk classification data, treatment measure parameters, treatment effect evaluation data, equipment operation data, cost data, etc. are stored together. The database supports data query, modification, export and backup to ensure data integrity and traceability.

[0050] Secondly, visualization technology (based on the Matplotlib library of MATLAB or Python) is used to build an intelligent display interface to intuitively display prevention and control data in multiple forms: H2S concentration cloud map (showing the spatial distribution of downhole H2S concentration, with different concentrations marked by different colors to intuitively identify high-risk areas), concentration time series curve (showing the changing trend of H2S concentration, marking the warning threshold and treatment adjustment nodes), warning interface (displaying the risk level and warning status of each area in real time, and automatically flashing to remind when the standard is exceeded), and prevention and control project distribution map (marking the location of injection boreholes, spray devices, sensors, and sealing areas to facilitate on-site management).

[0051] Meanwhile, a user-friendly interface was developed to support various intelligent operations: remote monitoring (ground staff can view the concentration and operating status of treatment equipment in various underground areas in real time and issue control commands remotely), intelligent decision-making (the system automatically recommends the optimal treatment plan based on real-time concentration data and geological conditions), historical traceability (query concentration data, treatment measures and effect evaluation results for any time period to provide a reference for subsequent prevention and control optimization), and report output (automatically generate daily, weekly and monthly prevention and control reports, covering concentration statistics, treatment effects, cost consumption and other content, to facilitate mine safety management).

[0052] This intelligent visualization system breaks the traditional "decentralized monitoring and manual management" model, realizing "datafication, intelligence and visualization" of H2S prevention and control, improving the efficiency of prevention and control management, reducing manual intervention, and providing intuitive and efficient support for mine safety management.

[0053] Example 1: Comprehensive Management of Hydrogen Sulfide in High-Risk Working Faces During tunneling at a certain mine working face, the instantaneous H2S concentration reached 6.0 × 10⁻⁶. -5 ~4.0×10 -4 The risk level is determined to be Level III, and the method of this invention is used for remediation. The specific steps are as follows: S1: Collect data on the working face burial depth, coal sulfur content (both organic and inorganic sulfur content are high), hydrological parameters (groundwater pH=7.2, moderate mineralization), and ventilation volume (initial air volume 800m³ / h). 3 Real-time H2S concentration data ( / min) were collected, and wavelet denoising algorithm was used to eliminate glitches and waveform jumps in the sensor-acquired data. Through 3... s The criteria were used to remove outlier concentration values, and then the Z-score method was used to standardize all data to eliminate dimensional differences and construct a standardized dataset.

[0054] S2: Based on the formation temperature (35℃, within the range of 20℃~80℃), the distribution of sulfate-reducing bacteria, and sulfur isotope analysis, it was determined that the H2S in this working face was caused by microbial sulfate reduction (BSR). In combination with the H2S concentration distribution, the working face excavation head and upper corner were designated as key control areas, the return airway as a secondary key control area, and other areas as conventional control areas. A Level III warning (dangerous level, proactive treatment) was initiated.

[0055] S3: Based on the causes of BSR and the Level III risk level, a four-in-one comprehensive prevention and control model of "source inhibition - process interruption - end absorption - individual protection" is constructed, and the core treatment measures are determined to be "pre-injection of alkali solution + spray absorption + air curtain drainage + crack sealing + individual protection", forming a systematic treatment system of "removal, drainage, blocking and pumping".

[0056] S4: Automatically execute collaborative governance based on risk level: ① Prepare a sodium carbonate corrosion-inhibiting alkaline absorbent solution with pH=9.5 (add 0.5% corrosion inhibitor to reduce equipment corrosion), and inject it into the coal body through a high-pressure pre-injection method with deep and shallow holes. The injection pressure is automatically adjusted to 3.5MPa according to the coal permeability (1.2mD), the injection flow rate is 50L / min, and the drilling spacing is 2.5m. ② During the coal cutting process, the working face automatically starts the medium and high pressure spray system, and the spray flow rate is dynamically adjusted according to the real-time H2S concentration (the flow rate increases when the concentration increases). ③ Install an air curtain exhaust device in the upper corner, automatically adjust the exhaust air velocity to 3m / s, and guide the accumulated H2S to the return airway; ④ Use cement grout to seal the cracks around the working face to block the H2S migration path; ⑤ Workers should wear acid-resistant masks, protective glasses and other personal protective equipment. Metal supports in the tunnel should be coated with emulsified oil and wrapped with plastic to prevent corrosion.

[0057] S5: Fixed H2S sensors are installed at the working face excavation head, upper corner, return airway, and low-lying areas of the roadway floor, with one sensor every 50m. Two mobile sensors are also provided to collect H2S concentration data in real time and transmit it back to the ground monitoring center via a wireless transmission module to generate a time-space-concentration dynamic evolution curve. When the concentration instantaneously exceeds 0.76ppm, the automatic linkage spray system increases the flow rate, activates the alarm device, cuts off the power supply to the excavation equipment, and issues an evacuation warning.

[0058] S6: After 72 hours of treatment, the treatment effect was evaluated from multiple dimensions: the H2S concentration reduction rate was 86%, the absorption rate of the absorbent was 82%, the treatment coverage reached 100%, the coal quality test showed that the K and Na contents did not increase significantly and would not affect subsequent industrial use, the equipment corrosion rate dropped to below 0.3%, the treatment cost was reduced by 15% compared with the traditional method, and the comprehensive evaluation index reached 0.82 (the preset qualified threshold was 0.75), and the treatment effect was judged to meet the standard.

[0059] S7: Since the treatment deviation coefficient = (target concentration - measured concentration) / target concentration = 0.12 < 0.2, only the spray flow rate (from 15L / min to 18L / min) and the injection time (from 8 hours per day to 10 hours) are finely adjusted. There is no need to adjust the absorbent formula and injection parameters. The incremental learning strategy is used to update the local control model to further improve the stability of treatment.

[0060] S8: Set the statistical window to 7 days, calculate the cumulative governance error coefficient to 0.08 (preset threshold is 0.15), the error does not exceed the threshold, and the effect decay coefficient = current cumulative error coefficient / previous statistical window error coefficient = 0.92≤1, there is no need to start global iterative optimization, and the existing governance parameters remain unchanged.

[0061] S9: Output the final treatment plan (injection pressure 3.5MPa, spray flow rate 18L / min, exhaust wind speed 3m / s, etc.), construct the mine H2S prevention and control database, link the geological conditions of the working face, H2S causes, concentration distribution, treatment measures and effect data, and display them in the form of concentration cloud map, treatment effect curve and prevention and control project distribution map through the monitoring center visualization interface, so as to realize remote monitoring, intelligent decision-making and historical traceability.

[0062] Treatment effect: The H2S concentration at the working face was stably reduced to below 0.40ppm, with a removal rate of ≥85%. There was no severe equipment corrosion, the coal quality met the requirements for industrial production, and the working environment was safe and controllable.

[0063] Example 2: Hydrogen sulfide plugging and extraction treatment in goaf areas In a certain mine goaf, due to the combined effects of spontaneous combustion of coal seams and biogas saturation (BSR), the H2S concentration accumulated to 1.05 ppm, approaching Level IV risk (lethal level). The method of this invention was used for remediation, and the specific steps are as follows: S1: Collect data on goaf burial depth, formation temperature (42℃), sulfur content in coal, distribution of sulfate-reducing bacteria, groundwater parameters, goaf fracture development characteristics, H2S concentration distribution, and existing ventilation and sealing data. After preprocessing and standardization, construct a standardized dataset.

[0064] S2: The cause of H2S was determined to be the synergistic effect of BSR microbial genesis and coal spontaneous combustion. The area was designated as a key prevention and control zone, a Level IV warning was activated, and workers were immediately organized to evacuate and the power supply to the equipment around the goaf was cut off.

[0065] S3: Construct a four-in-one integrated prevention and control model of "source suppression - process interruption - end absorption - individual protection", with the core governance measures being the coordinated operation of "fissure sealing + ring extraction + immediate burial and injection + medium and high pressure spraying".

[0066] S4: ① Prepare a low-corrosion composite alkaline solution with pH=9.0~9.8 (sodium carbonate and sodium bicarbonate mixed in a 3:1 ratio, with 0.8% corrosion inhibitor and 0.3% penetration enhancer added), and use an immediate grouting process to grout and seal the fissures around the goaf. The grouting pressure is automatically adjusted to 4.0MPa; ② Arrange annular extraction boreholes around the goaf, with the extraction negative pressure automatically adjusted to 15kPa, to directionally extract the H2S accumulated in the goaf to the surface treatment device; ③ Arrange a medium-high pressure spray system at the goaf outlet and return airway to spray the composite alkaline solution to further absorb the escaped H2S; ④ During the treatment, workers must wear high-level protective equipment and are strictly prohibited from entering the key control area without authorization.

[0067] S5: Multiple H2S sensors are deployed inside the goaf, at the outlet, and in the return airway to monitor concentration changes in real time. The data is transmitted back to the ground monitoring center in real time. When the concentration drops below 0.76 ppm, the Level IV warning is lifted and operations are gradually resumed.

[0068] S6: 120 hours after treatment, the treatment effect was evaluated: the H2S concentration dropped to 0.35ppm, the concentration reduction rate was 67%, the extraction and absorption rate was 78%, the crack sealing rate in the goaf reached 95%, the equipment corrosion rate was less than 0.20%, and the comprehensive evaluation index was 0.80, which met the qualified standard.

[0069] S7: The treatment deviation coefficient is 0.18≤0.20. Fine-tune the extraction negative pressure (from 15kPa to 17kPa) and spray flow rate, and use incremental learning to update the local model to ensure concentration stability.

[0070] S8: The cumulative treatment error coefficient is 0.10 < 0.15, the effect decays slowly, the geological and concentration data around the goaf are expanded, the extraction negative pressure and grouting parameters are optimized, and there is no need to reconstruct the prevention and control model.

[0071] S9: Output the final treatment plan, update the mine H2S prevention and control database, realize real-time monitoring and visualization of H2S concentration in the goaf, and achieve the goal of long-term prevention and control.

[0072] The technical solution of the present invention has now been described in conjunction with the preferred embodiments and specific engineering examples shown in the accompanying drawings. However, those skilled in the art should understand that the scope of protection of the present invention is obviously not limited to the specific embodiments described above. Without departing from the principles and core concepts of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical steps, process parameters, and model algorithms. For example, adjusting the hydrogen sulfide formation identification indicators and data types, optimizing the graded early warning threshold and evaluation model parameters, modifying the absorbent formulation components and injection process parameters, replacing the specific algorithms for adaptive control and incremental learning, and changing the sensor deployment method and monitoring points, etc. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for intelligent hierarchical prevention and comprehensive management of hydrogen sulfide in coal mines, characterized in that, include: S1. Collect data on the geological conditions of the target coal mine, the sulfur content of the coal, the distribution of hydrogen sulfide concentration, underground ventilation, hydrology and existing treatment records, construct a basic dataset for hydrogen sulfide prevention and control, and perform anomaly removal, normalization and standardization to form a standardized dataset. S2, based on the identification indicators of the organic / inorganic origin of hydrogen sulfide, its occurrence space and emission intensity, divide the hydrogen sulfide sensitive control areas and establish a concentration classification and risk warning indicator system; S3, based on the regional risk level and cause type, matches multiple governance measures of "injection, spraying, blocking, pumping, drainage, and protection" to construct a four-in-one comprehensive governance model of source suppression, process interruption, end absorption, and individual protection; S4 implements intelligent injection of alkaline absorbent liquid, medium and high pressure spraying, fissure sealing, directional extraction and air volume control into the coal body and roadway to achieve multi-pathway synergistic treatment of hydrogen sulfide; S5 uses fixed and mobile H2S sensors in the well for real-time online monitoring, and transmits the measured concentration back to the control system to form a closed-loop feedback of the treatment effect; S6 evaluates the treatment effect from multiple dimensions, calculates concentration reduction rate, absorption rate, equipment wear and economic cost indicators, and determines whether the preset prevention and control targets are met. S7, when the treatment effect is not up to standard, dynamically optimizes the absorbent formula, injection pressure, sealing position and ventilation parameters based on the deviation coefficient, and performs graded adaptive adjustment; S8. Based on the cumulative governance error and effect decay coefficient of long-term operation, determine whether to carry out global iterative optimization and update the prevention and control model and process parameters. S9 outputs the final treatment parameters and control effects, and constructs an intelligent monitoring, visualization and data storage management system for hydrogen sulfide in mines.

2. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 1, characterized in that, The data collected in step S1 includes: S11, formation temperature, burial depth, sulfur content of coal and rock, distribution of sulfate-reducing bacteria, groundwater salinity, pH value, coal seam fracture development characteristics, real-time H2S concentration, air volume and wind speed, existing injection / spraying / plugging / extraction parameters. S12, preprocessing uses wavelet denoising and 3 σ The criteria remove outliers, and the standardization method uses the Z-score method to normalize multi-source data to the same order of magnitude.

3. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 2, characterized in that, The graded early warning process in step S2 includes: S21, identify the type of H2S origin: biodegradation, microbial sulfate reduction (BSR), sulfate thermochemical reduction (TSR), magmatic / volcanic origin; S22 is classified into four levels based on H2S concentration: Level I: <0.40 ppm (below the olfactory threshold, routine control measures); Level II: 0.40~0.76 ppm (Warning level, enhanced monitoring); Level III: 0.76~1.00 ppm (hazardous level, active treatment); Level IV: >1.00 ppm (lethal, immediate evacuation + intensive treatment); S23, combining the outflow location, spatial aggregation characteristics, and operational intensity, divides the area into key prevention and control areas, secondary key areas, and routine areas, and establishes a zoned early warning model.

4. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 3, characterized in that, The four-in-one integrated governance model in step S3 includes: S31, Source Suppression: High-pressure pre-injection of alkaline absorbent liquid into shallow / deep holes of the coal body to suppress the generation and release of H2S; S32, Process Interception: Using slurry to seal coal and rock fissures, abandoned roadways, and seepage channels in goaf areas to block the migration path; S33, end absorption: high-pressure spraying and foam absorption liquid in the working face, upper corner and return airway to capture H2S online; S34, Emergency Protection: Personal protective equipment, metal corrosion protection, emergency evacuation and ventilation systems; S35 forms a systematic prevention and control strategy of "removal, drainage, blocking, dredging, and extraction".

5. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 4, characterized in that, The intelligent governance execution in step S4 includes: S41, automatically configures the absorbent solution according to the risk level: sodium carbonate, sodium bicarbonate, lime slurry, and composite corrosion-inhibiting alkaline solution, with the pH value controlled at 9.0~10.0; S42 automatically adjusts the injection pressure, flow rate, orifice diameter, and spacing according to the coal permeability to achieve efficient permeability enhancement injection; S43 automatically adjusts the air curtain, water curtain, and local ventilation fan according to the location of H2S accumulation to achieve directional dilution and exhaust; S44 involves directional drilling to extract H2S in high-risk areas, in conjunction with sealing and injection operations.

6. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 5, characterized in that, Real-time monitoring and feedback in step S5 includes: S51, H2S sensors are deployed in the working face, upper corner, goaf, return airway, and low-lying areas of the roadway floor. S52 collects concentration data in real time and transmits it to the ground monitoring center; S53: When the concentration exceeds the limit, it automatically triggers spraying, liquid injection, alarm, power cut-off, and evacuation commands. S54 generates a time-space-concentration dynamic evolution curve, providing data for model optimization.

7. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 6, characterized in that, The evaluation indicators for the treatment effect in step S6 include: S61, H2S concentration reduction rate, gas absorption rate, treatment coverage, degree of impact on coal quality, material consumption, equipment corrosion rate, labor cost, and safety risk level. S62, construct a comprehensive effect evaluation index, set a qualified threshold, the index is qualified if it is greater than or equal to the threshold, otherwise it is unqualified.

8. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 7, characterized in that, Step S7, the hierarchical adaptive adjustment, includes: S71, calculate the treatment deviation coefficient = (target concentration - measured concentration) / target concentration; S72, deviation ≤0.2: fine-tune spray flow rate, injection time and ventilation volume; S73, 0.2 < Deviation ≤ 0.5: Adjust the concentration and formula of the absorbent, and increase the corrosion inhibitor and penetration aid; S74, Deviation > 0.5: Reassess the cause and risk level, and adjust the injection location, sealing range and extraction plan; S75 employs incremental learning to update the local prevention and control model, rapidly improving governance effectiveness.

9. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 8, characterized in that, Step S8, global iterative optimization, includes: S81, set the statistical window, and calculate the cumulative governance error coefficient and the effect attenuation coefficient; S82, if the error exceeds the threshold, global optimization is initiated: Slow effect decay: Expand field data and optimize parameters and formulation; The effectiveness has drastically decreased: the prevention and control model needs to be reconstructed, and the absorption system and process route need to be changed; S83 takes into account the treatment effect, coal quality protection, equipment corrosion prevention, economic cost and environmental protection requirements, and achieves sustainable optimization.

10. The intelligent graded prevention and comprehensive management method for hydrogen sulfide in coal mines according to claim 9, characterized in that, The intelligent visualization system in step S9 includes: S91, Establish a mine hydrogen sulfide database, linking geological conditions, formation types, concentration distribution, treatment measures, and effect curves; The S92 displays data in the form of cloud maps, graphs, early warning interfaces, and prevention and control engineering distribution maps, enabling remote monitoring, intelligent decision-making, historical tracing, and report output.

Images