Y-type ventilation gas extraction method for close coal seams group

Abstract

The present application provides a kind of close coal seam group Y type ventilation gas extraction method, belong to coal mining technical field, the present application is first laid out the main borehole of vertical coal seam and the branch borehole extending upstream and downstream at 45 ° angle, form Y type three-dimensional extraction network;Then install gas concentration monitoring sensor network, set monitoring point and collect real-time data;Build gas concentration regulation buffer system and intelligent mixed gas distribution device, maintain the extraction gas concentration between 30% and 45%;Establish prediction control model based on long short-term memory neural network, predict gas emission law in advance and dynamically adjust extraction parameters;Implement gas concentration grading utilization system, different concentration gas is introduced into corresponding utilization system respectively;Deploy gas extraction intelligent monitoring platform to realize remote monitoring and fault early warning;Solve the technical problems that extraction concentration is unstable and difficult to predict control in the process of coal seam group gas extraction.

Description

Technical Field

[0001] This invention belongs to the field of coal mining technology, and specifically relates to a Y-type ventilation gas extraction method for closely spaced coal seam groups. Background Technology

[0002] Coal mine gas drainage is a key technology for ensuring safe production in coal mines. Traditional gas drainage methods mainly use single linear or horizontal boreholes for extraction, using a negative pressure system to extract gas from the coal seam and transport it to the surface. These traditional methods have achieved certain results in single coal seam extraction, but their application is limited to simple geological conditions and stable gas emission environments.

[0003] However, under complex coal seam conditions, existing gas drainage technologies face problems such as large fluctuations in drainage concentration and poor stability. Traditional borehole layouts cannot effectively cover the goaf and gas-rich areas of adjacent coal seams, resulting in a limited drainage range. At the same time, drainage parameters such as negative pressure and flow rate are mostly fixed values ​​or manually adjusted, and cannot be adjusted in real time according to the gas emission patterns, resulting in low drainage efficiency and even the phenomenon of "air leakage" during drainage.

[0004] Especially during rapid advance of the working face, the gas emission rate and concentration fluctuate drastically. Existing technologies lack the ability to accurately predict gas emission patterns and have dynamic control mechanisms, resulting in unstable extracted gas concentrations that are difficult to maintain within a suitable utilization range. This severely restricts gas extraction efficiency and gas resource utilization rate. In other words, existing technologies suffer from the technical problem of unstable and unpredictable extracted concentrations during coal seam gas extraction. Summary of the Invention

[0005] In view of this, the present invention provides a Y-type ventilation gas extraction method for close-range coal seam groups, which can solve the technical problem of unstable extraction concentration and difficulty in predicting and controlling the gas extraction concentration in the existing technology.

[0006] This invention is implemented as follows: It provides a method for Y-type ventilation gas extraction in close-range coal seam groups, comprising: deploying a group of Y-type ventilation extraction boreholes; installing a gas concentration monitoring sensor network; constructing a gas concentration regulation and buffer system; installing an intelligent gas mixing and distribution device; establishing a dynamic control system for gas extraction parameters; applying a predictive control model to analyze the gas emission patterns of the working face based on historical data, receiving data on the working face advance speed, borehole outlet gas concentration, key node gas concentration data of the main pipeline, and negative pressure data, outputting predicted values ​​for gas emission and concentration change trends for the next 12 hours, and adjusting the extraction negative pressure and flow rate 12 hours in advance; implementing a gas concentration tiered utilization system; deploying an intelligent gas extraction monitoring platform; and performing a gas extraction effect evaluation.

[0007] The Y-type ventilation and extraction borehole group is a special borehole structure pre-arranged at the front of coal seam mining. It consists of a main borehole and two branch boreholes. The main borehole is drilled perpendicular to the coal seam direction to the depth of the coal seam to be mined. The two branch boreholes extend from the bottom of the main borehole in the upstream and downstream directions respectively, with an inclination angle of 45°, covering the goaf and adjacent coal seams to form a three-dimensional extraction network.

[0008] Specifically, the gas concentration monitoring sensor network is configured with a monitoring point every 50 meters at the borehole outlet and key nodes of the main pipeline to collect real-time gas concentration data at the borehole outlet, gas concentration data at key nodes of the main pipeline, and negative pressure data.

[0009] The gas concentration regulating and buffering system is specifically a device composed of a main buffer tank, an auxiliary buffer tank, a flow regulating valve, a pressure sensor, and a concentration sensor. The main buffer tank has a cylindrical structure and a volume of 10. 3 m 3 The auxiliary buffer tank has a volume of 5×10 2 m 3 Both ends are equipped with electric regulating valves.

[0010] The intelligent gas mixing and distribution device specifically comprises a mixing chamber, multiple air inlet pipes, an electric proportional valve, a gas analyzer, and a controller. The mixing chamber is a sealed container with a volume of 5m³. 3 It is equipped with a turbine stirring device inside, and multiple air intake pipes are connected to high-concentration gas sources, medium-concentration gas sources and low-concentration gas sources respectively.

[0011] Specifically, the dynamic control system for gas extraction parameters automatically adjusts the extraction negative pressure and flow rate based on data such as the working face advance speed, borehole outlet gas concentration, key node gas concentration in the main pipeline, and negative pressure value.

[0012] The predictive control model is specifically a gas outburst prediction model based on a long short-term memory neural network. It consists of three parts: an input layer, a hidden layer, and an output layer. The input layer receives data on the working face advance speed, borehole outlet gas concentration, key node gas concentration in the main pipeline, and negative pressure. The hidden layer contains four layers of long short-term memory units, each with 128 neurons.

[0013] The predictive control model is a multivariate temporal prediction network structure, consisting of a data preprocessing unit, a feature extraction unit, a temporal learning unit, and a prediction unit. The feature extraction unit employs a bidirectional long short-term memory network structure, and the temporal learning unit uses an attention mechanism to enhance the model's ability to capture long-term temporal dependencies. Specifically, the temporal learning unit of the predictive control model has an attention layer composed of multi-head self-attention modules with 8 heads and an attention dimension of 64. Residual connections and layer normalization are used to improve model stability. The prediction unit consists of two fully connected layers, with the first layer containing 64 neurons.

[0014] The steps for establishing the training dataset of the predictive control model specifically include four parts: data acquisition, data cleaning, feature extraction, and dataset partitioning. The data acquisition stage obtains 90 consecutive days of historical monitoring data from the gas concentration monitoring sensor network.

[0015] This invention utilizes a unique Y-shaped borehole structure to form a three-dimensional extraction network, effectively expanding the extraction coverage area and solving the problem of insufficient coverage in traditional borehole layouts. Simultaneously, by employing a gas concentration regulation and buffer system and an intelligent mixing and distribution device, stable control of the extracted gas concentration is achieved, effectively eliminating concentration fluctuations in traditional extraction methods.

[0016] The predictive control model based on a long short-term memory neural network can accurately analyze the correlation between face advancement and gas emission, predicting gas emission patterns 12 hours in advance, providing a scientific basis for dynamic adjustment of extraction parameters. The system automatically adjusts the extraction negative pressure and flow rate based on the prediction results, enabling the extraction process to proactively adapt to changes in gas emission, thus realizing a shift from passive extraction to proactive pre-control.

[0017] The technical solution of this invention effectively solves the core technical problem of unstable extraction concentration and difficulty in predicting and controlling it during the gas extraction process of coal seam groups, significantly improving the gas extraction efficiency and gas resource utilization rate, and providing a new technical approach for safe coal mine production and efficient utilization of gas resources. Attached Figure Description

[0018] Figure 1 This is a flowchart of the method of the present invention.

[0019] Figure 2 This is a schematic diagram of the overall structure of the Y-type ventilation gas extraction method in Example 2.

[0020] Figure 3 This is a detailed diagram of the Y-type drilling structure in Example 2.

[0021] Figure 4 This is a structural diagram of the gas concentration regulating buffer system in Example 2.

[0022] Figure 5 This is a structural diagram of the intelligent gas mixing and distribution device in Example 2.

[0023] Figure 6 This is a structural diagram of the predictive control model in Example 2.

[0024] Figure 7 This is a system architecture diagram of gas concentration grading and utilization in Example 2.

[0025] Figure 8 This is a structural diagram of the intelligent monitoring platform for gas extraction in Example 2. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0027] like Figure 1 The diagram shown is a flowchart of a Y-type ventilation gas extraction method for close-range coal seam groups provided by the present invention. This method includes the following steps:

[0028] S01. Deploy a Y-shaped ventilation and extraction borehole group, in which the main borehole is arranged vertically to the coal seam, and the branch boreholes extend upward and downward at a 45° angle to cover the goaf and coal seam group.

[0029] S02. Install a gas concentration monitoring sensor network, setting up a monitoring point every 50 meters at the borehole outlet and key nodes of the main pipeline to collect real-time data on gas concentration at the borehole outlet, gas concentration data at key nodes of the main pipeline, and negative pressure value data.

[0030] S03. Construct a gas concentration regulation and buffer system, including a main buffer tank, an auxiliary buffer tank, a flow regulating valve, and a mixing device, to stabilize the extracted gas concentration between 30% and 45%.

[0031] S04. Install an intelligent gas mixing and distribution device to mix low-concentration methane with high-concentration methane at a ratio of 4:6 through a negative pressure extraction system to maintain a stable methane concentration in the pipeline network.

[0032] S05. Establish a dynamic control system for gas extraction parameters, and automatically adjust the extraction negative pressure and flow rate based on the working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data and negative pressure value data.

[0033] S06. Apply a predictive control model, analyze the gas emission pattern of the working face based on historical data, receive working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data and negative pressure value data, output the predicted value of gas emission volume and concentration change trend for the next 12 hours, and adjust the extraction negative pressure and flow rate 12 hours in advance.

[0034] S07. Implement a gas concentration-based utilization system, dividing gas into high-concentration, medium-concentration, and low-concentration zones, and introducing them into different utilization systems respectively.

[0035] S08. Deploy an intelligent monitoring platform for gas extraction, integrating extraction negative pressure data, flow data, borehole outlet gas concentration data, key node gas concentration data of main pipeline, and system status data to achieve remote monitoring and fault early warning.

[0036] S09. Conduct gas extraction effect evaluation. Quantitatively evaluate the extraction effect and guide the optimization of extraction negative pressure and flow rate by measuring residual gas content, calculating extraction rate, and statistically analyzing gas utilization rate.

[0037] Among them, the Y-type ventilation and extraction borehole group is a special borehole structure that is pre-arranged at the front of coal seam mining. It consists of a main borehole and two branch boreholes, resembling the English letter Y. The main borehole is drilled perpendicular to the coal seam direction, reaching the depth of the coal seam to be mined. The two branch boreholes extend from the bottom of the main borehole in the upstream and downstream directions respectively, with an inclination angle of 45°, covering the goaf and adjacent coal seams, forming a three-dimensional extraction network.

[0038] The gas concentration regulation and buffer system is specifically composed of a main buffer tank, an auxiliary buffer tank, flow regulating valves, a pressure sensor, and a concentration sensor. The main buffer tank has a cylindrical structure and a volume of 10. 3 m 3 It has multiple layers of baffles inside, and the auxiliary buffer tank is connected to the main buffer tank through pipelines, with a volume of 5×10 2 m 3 Both ends are equipped with electric regulating valves. When the extracted gas concentration fluctuates, the system controls the mixing ratio of different gas concentrations by adjusting the valve opening to keep the output gas concentration stable.

[0039] The intelligent gas mixing and distribution device specifically consists of a mixing chamber, multiple air inlet pipes, an electric proportional valve, a gas analyzer, and a controller. The mixing chamber is a closed container with a volume of 5m³. 3 It is equipped with a turbine stirring device inside, and multiple air intake pipes are connected to high-concentration gas sources, medium-concentration gas sources and low-concentration gas sources respectively. Each air intake pipe is equipped with an electric proportional valve. The controller precisely controls the opening of each valve according to the real-time concentration data measured by the gas analyzer, so as to achieve precise mixing of gas of different concentrations.

[0040] The predictive control model is specifically a gas emission prediction model built on a long short-term memory neural network. It consists of three parts: an input layer, a hidden layer, and an output layer. Data is processed by connecting weights and activation functions. The input layer receives data on the working face advance speed, borehole outlet gas concentration, key node gas concentration in the main pipeline, and negative pressure. The hidden layer contains four layers of long short-term memory units, each with 128 neurons. The output layer outputs the predicted gas emission volume and concentration change trend for the next 12 hours.

[0041] The specific structure of the predictive control model is a multivariate time-series prediction network structure, consisting of a data preprocessing unit, a feature extraction unit, a time-series learning unit, and a prediction unit. The data preprocessing unit is responsible for normalizing and denoising the raw data collected by the sensors. The feature extraction unit adopts a bidirectional long short-term memory network structure, containing two forward propagation layers and two backward propagation layers, each with 128 neurons, used to extract deep features from the time-series data. The time-series learning unit uses an attention mechanism to enhance the model's ability to capture long-term time-series dependencies. The attention layer consists of multi-head self-attention modules with 8 heads and an attention dimension of 64, combined with residual connections and layer normalization to improve model stability. The prediction unit consists of two fully connected layers. The first layer contains 64 neurons, and the output dimension of the second layer is consistent with the prediction target, namely, the predicted hourly gas emission and concentration for the next 12 hours. The model input layer is directly connected to the gas concentration monitoring sensor network, receiving four key parameters: working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data. The working face advance speed data comes from the working face advance speed monitoring system, the borehole outlet gas concentration data and the main pipeline key node gas concentration data come from the gas concentration monitoring sensor network in step S02, and the negative pressure value data comes from the gas concentration monitoring sensor network in step S02. The model output layer is connected to the gas extraction parameter dynamic control system, providing predicted values ​​for gas emission and concentration change trends for the next 12 hours, which are used for adjusting extraction negative pressure and flow rate in step S05. The model as a whole adopts a serial structure, with information transmission between units through fully connected layers. At the same time, a dropout mechanism is set between layers to prevent overfitting.

[0042] The steps for establishing the training dataset for the predictive control model specifically include four parts: data acquisition, data cleaning, feature extraction, and dataset partitioning. In the data acquisition stage, historical monitoring data for 90 consecutive days is obtained from the gas concentration monitoring sensor network in step S02, including borehole outlet gas concentration data, gas concentration data at key nodes of the main pipeline, and negative pressure data. Simultaneously, historical data on working face advance speed and gas emission are collected from the working face advance speed monitoring system for the corresponding time period. The sampling frequency is set to once every 10 minutes to form the initial dataset. In the data cleaning stage, outlier detection and processing are performed on the initial dataset. Outliers are identified using the three-standard-deviation method, and missing data is filled using the forward imputation method. Then, the moving average method is used to partition the data. To reduce the impact of noise, smoothing was performed. In the feature extraction stage, time-series features were extracted based on the time window method, with a window length of 24 hours and a step size of 1 hour. For each time window, the average working face advance speed, average gas concentration, standard deviation of gas concentration, average negative pressure value, and total gas emission were extracted as features. At the same time, the correlation coefficient between each feature was calculated as an additional feature. In the dataset partitioning stage, the processed dataset was divided into training set, validation set, and test set in a ratio of 7:2:1. The training set was used for model parameter learning, the validation set was used for model hyperparameter tuning and early stopping strategy implementation, and the test set was used to evaluate the final model performance. To enhance the model's generalization ability, data augmentation techniques were applied to the training set, including adding Gaussian noise and time shifting to generate additional samples.

[0043] The training steps of the predictive control model specifically include four parts: parameter initialization, model training, parameter optimization, and model evaluation. In the parameter initialization stage, the Xavier initialization method is used to initialize the network weights, ensuring consistent variance between the input and output of each layer to avoid gradient vanishing or exploding problems. The bias term is initialized to zero, and the forget gate bias of the Long Short-Term Memory network is initialized to 1 to promote information transfer. In the model training stage, the mini-batch stochastic gradient descent algorithm is used for parameter updates, with a batch size of 64, 200 training epochs, an initial learning rate of 0.001, and mean squared error as the loss function. An L2 regularization term is also introduced to prevent overfitting, with a regularization coefficient set to 0.00. 01. During training, a learning rate decay strategy is used, halving the learning rate every 50 rounds. In the parameter optimization phase, Bayesian optimization is used to tune the model's hyperparameters. The optimization targets include the number of hidden layer neurons, the number of long short-term memory layers, the number of attention heads, and the dropout ratio. The optimization objective is the prediction error on the validation set. A Gaussian process regression model is constructed to estimate the relationship between the hyperparameters and the objective function, guiding the hyperparameter search direction. In the model evaluation phase, three indicators are used to evaluate the model's performance on the test set: root mean square error, mean absolute error, and coefficient of determination. The stability and adaptability of the model are analyzed by comparing the prediction results at different time scales. Finally, the model with the best overall performance is selected as the core component of the predictive control system.

[0044] The gas concentration classification and utilization system specifically divides the extracted gas into three levels according to its concentration: in the high-concentration zone, the gas concentration is greater than 30%, and it is directly used for power generation or liquefaction; in the medium-concentration zone, the gas concentration is between 10% and 30%, and it is converted into thermal energy through a catalytic oxidation device; in the low-concentration zone, the gas concentration is between 5% and 10%, and it is treated by thermal oxidation technology before being discharged.

[0045] The intelligent monitoring platform for gas extraction is a comprehensive monitoring system that integrates a data acquisition system, a transmission network, a server, and application software. The data acquisition system consists of sensors distributed at various monitoring points. It transmits extraction negative pressure data, flow data, borehole outlet gas concentration data, gas concentration data at key nodes of the main pipeline, and system status data to the central server via industrial Ethernet. The application software includes a data processing module, a visualization display module, an early warning module, and a decision support module. The system displays extraction negative pressure data, flow data, borehole outlet gas concentration data, gas concentration data at key nodes of the main pipeline, and system status data in real time. When the monitored data exceeds the set threshold, an early warning is automatically triggered.

[0046] The working face advance speed monitoring system is a monitoring device composed of a displacement sensor, a data acquisition unit, and a data processing unit. The displacement sensor is installed on the coal mining machine to monitor the position changes of the coal mining machine in real time. The data acquisition unit collects the displacement sensor signal and converts it into a digital signal. The data processing unit calculates the working face advance speed data based on the position change and time.

[0047] The evaluation of gas extraction effectiveness is specifically carried out through three aspects: measuring residual gas content, calculating extraction rate, and statistically analyzing gas utilization rate. The residual gas content is measured by sampling at different locations on the working face using the borehole sampling method. The extraction rate is calculated by dividing the difference between the original gas content and the residual gas content by the original gas content. The gas utilization rate is the ratio of the actual amount of gas utilized to the total amount of gas extracted.

[0048] The specific implementation methods of the above steps are described in detail below.

[0049] The specific implementation of step S01 involves determining the coal seam distribution using ground-penetrating radar and 3D modeling techniques. First, a main borehole is drilled above the coal seam, perpendicular to the coal seam direction, reaching the depth of the coal seam to be mined. The borehole diameter is 120 mm, and the borehole wall is reinforced with high-strength steel pipes with a wall thickness of 10 mm. Then, two branch boreholes are drilled from the bottom of the main borehole in both upstream and downstream directions. The branch boreholes maintain an inclination angle of 45° to the horizontal plane, have a diameter of 90 mm, and their length is determined based on the coal seam thickness, generally 3 to 5 times the thickness. Steel pipes are also laid inside the branch boreholes for support. After drilling, a pressure test is conducted to ensure the borehole's sealing performance. When deploying the Y-shaped ventilation and drainage borehole group, the spacing between adjacent main boreholes is 30 meters, and a row of boreholes is laid every 50 meters along the coal seam strike, forming a three-dimensional drainage network covering the entire mining area. This step is based on fluid mechanics principles, increasing the gas drainage coverage area and improving gas drainage efficiency through a three-dimensional, intersecting borehole network. The purpose of this step is to form a three-dimensional gas drainage channel network, creating conditions for subsequent gas drainage.

[0050] The specific implementation of step S02 involves constructing a monitoring network using catalytic combustion gas sensors and thermal conductivity gas sensors. Catalytic combustion gas sensors are installed at the Y-shaped borehole outlet, with a measurement range of 0 to 5% and an accuracy of ±0.1%. Thermal conductivity gas sensors are installed every 50 meters at key nodes along the main pipeline, with a measurement range of 0 to 100% and an accuracy of ±0.5%. Simultaneously, negative pressure sensors are installed at each monitoring point, with a measurement range of 0 to 25 kPa and an accuracy of ±0.2 kPa. All sensors are intrinsically safe, with an explosion-proof rating of Exia I, and the sensor acquisition frequency is set to once every 10 seconds. The sensors are connected to the nearest data acquisition unit via an RS485 bus, and the data acquisition unit transmits the collected data to the monitoring center via an industrial Ethernet network. Furthermore, a data anomaly self-checking mechanism is implemented in the system; when sensor readings show sudden changes or remain unchanged for an extended period, the system automatically flags the anomaly and issues an alarm. This step, based on sensor network technology and the principles of the Industrial Internet of Things (IIoT), constructs a real-time monitoring network for gas concentration and negative pressure values, providing data support for the gas extraction process. Key parameters of the sensor include measurement range, accuracy, and acquisition frequency. These parameters determine the reliability and real-time nature of the monitoring data, directly affecting the decision-making quality of subsequent steps. The purpose of this step is to achieve real-time monitoring of gas concentration and negative pressure values, providing a data foundation for subsequent dynamic adjustments.

[0051] The specific implementation of step S03 is as follows: First, a main buffer tank is constructed using Q345R steel, with a wall thickness of 20 mm, a cylindrical shape, a diameter of 10 meters, a height of 12.8 meters, and a volume of 10... 3 m 3 The internal structure features five layers of baffles arranged in an evenly spaced spiral pattern at a 30° angle. The baffle surfaces are made of a hydrophilic material to absorb moisture from the gas. The auxiliary buffer tank is also constructed of Q345R steel, with a diameter of 8 meters, a height of 10 meters, and a volume of 5 × 10⁶ m³. 2 m 3The system is internally packed with foamed plastic filler, boasting a specific surface area of ​​1000 square meters per cubic meter and a filler layer height of 5 meters. An electrically operated regulating valve is installed between the main and auxiliary buffer tanks. This valve is an angle-stroke electrically operated regulating valve with a maximum diameter of 800 mm, a rated pressure of 1.6 MPa, a regulation ratio of 100:1, and an on / off time of less than 30 seconds. An online gas concentration analyzer is installed throughout the system, employing infrared absorption principles, with a measurement range of 0 to 100%, an accuracy of ±0.2%, and a response time of less than 10 seconds. When the extracted gas concentration is detected to be below 30%, the control system automatically reduces the opening of the low-concentration inlet valve and increases the opening of the high-concentration inlet valve, and vice versa. This flow rate ratio adjustment ensures that the mixed gas concentration is maintained between 30% and 45%. This step, based on fluid dynamics and automatic control principles, stabilizes the gas concentration through buffer devices and an automatic adjustment system, thus resolving the problem of gas concentration fluctuations. The key parameter in this step is maintaining a stable gas concentration within the range of 30% to 45%. This range meets the concentration requirements for comprehensive gas utilization while avoiding the risk of gas entering the explosive concentration range. The purpose of this step is to stabilize the extracted gas concentration, creating conditions for subsequent comprehensive utilization.

[0052] The specific implementation of step S04 involves constructing a mixing cavity made of 304 stainless steel with a wall thickness of 15 mm and a volume of 5 m³. 3 The system is equipped with an internal turbine agitator with a diameter of 0.8 meters and an adjustable speed from 0 to 500 rpm. Four air inlets are installed in the mixing chamber for connecting to high-concentration, medium-concentration, and low-concentration methane sources, with one inlet for emergency air dilution. Each inlet is fitted with an electric proportional valve, a linear-stroke electric regulating valve with a maximum diameter of 200 mm, a rated pressure of 1.0 MPa, and a control ratio of 50:1. A gas analyzer based on thermal conductivity is installed at the mixing chamber outlet, measuring from 0 to 100% with an accuracy of ±0.3% and a response time of less than 5 seconds. The control system employs a fuzzy PID control algorithm, adjusting the valve openings based on real-time concentration data to maintain a constant 4:6 mixing ratio of low-concentration to high-concentration methane, ensuring the mixed methane concentration remains stable within the required range. The control system has a sampling period of 1 second, with adaptive PID parameter adjustment, resulting in a response time of less than 3 seconds to concentration fluctuations. This step is implemented based on mixed gas theory and fuzzy control technology. By precisely controlling the mixing ratio of different concentrations of methane, stable control of the methane concentration is achieved. The key parameter in this step is a mixing ratio of 4:6. This ratio is determined based on the optimal operating conditions of the methane utilization system, ensuring stable system operation and maximizing energy conversion efficiency. The purpose of this step is to improve methane utilization efficiency by achieving precise control of the methane concentration through intelligent mixing and distribution.

[0053] The specific implementation of step S05 involves constructing a dynamic control system for gas extraction parameters using a multi-input multi-output control system. The system receives four key parameters: working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data. Data preprocessing, including filtering, noise reduction, and normalization, is performed by a data processing unit. The processed data is then input into a fuzzy neural network controller, which automatically calculates the optimal extraction negative pressure and flow rate parameters based on a preset fuzzy rule base. The fuzzy rule base contains 50 inference rules, covering combinations of working face advance speed, gas concentration, and negative pressure under different operating conditions. The controller outputs extraction negative pressure adjustment commands and flow rate adjustment commands, which adjust the speed of the extraction pump and the valve opening on the pipeline through the actuator, achieving dynamic adjustment of extraction negative pressure and flow rate. The system's control cycle is 5 minutes, which can be shortened to 1 minute in extreme cases. The system also has a safety protection mechanism; when the gas concentration exceeds 45% or falls below 25%, an early warning is automatically triggered, and the system switches to a safe operating mode. This step is implemented based on fuzzy neural network control theory, transforming expert experience into fuzzy rules and achieving adaptive control of the system through the self-learning capability of the neural network. Key parameters in this step include a negative pressure range of 15 to 25 kPa and a flow rate adjustment range of 50 to 200 cubic meters per minute. These parameter ranges are determined based on the actual needs of coal mine gas drainage, ensuring drainage effectiveness while avoiding excessive air intake caused by excessive negative pressure. The purpose of this step is to dynamically adjust the drainage parameters based on real-time data, improving drainage efficiency and safety.

[0054] The specific implementation of step S06 is to construct a gas outburst prediction model based on a long short-term memory neural network. The model adopts a hierarchical structure, including four functional modules: a data preprocessing unit, a feature extraction unit, a time-series learning unit, and a prediction unit. The data preprocessing unit standardizes the input working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data, using the Z-score method to ensure a mean of 0 and a standard deviation of 1, while simultaneously using wavelet transform for signal denoising. The feature extraction unit adopts a bidirectional long short-term memory network structure, containing two forward propagation layers and two backward propagation layers, each with 128 neurons. It uses the tanh activation function to capture the contextual features of the time-series data in both forward and backward directions. The time-series learning unit employs a multi-head self-attention mechanism with 8 heads and an attention dimension of 64. It combines residual connections and layer normalization to improve model stability and enhances the model's ability to capture long-term temporal dependencies by calculating the correlation weights between different time steps. The prediction unit consists of two fully connected layers. The first layer contains 64 neurons and uses the ReLU activation function. The second layer outputs the same dimension as the prediction target, providing the predicted hourly gas emission rate and concentration trend for the next 12 hours. Based on the prediction results, the system adjusts the extraction negative pressure and flow rate 12 hours in advance. When the predicted gas emission rate increases, the extraction negative pressure and flow rate are increased in advance; conversely, they are appropriately decreased. This step is implemented based on deep learning and time series prediction theory. By analyzing historical data to uncover gas emission patterns, it achieves advance prediction of gas emission and pre-adjustment of extraction parameters. The key parameter in this step is the 12-hour prediction duration, which is determined based on the actual needs of mine production. This parameter satisfies the accuracy requirements of the prediction while allowing sufficient time for adjusting extraction parameters. The purpose of this step is to achieve advance prediction of gas emission and provide decision support for the dynamic adjustment of extraction parameters.

[0055] The specific implementation of step S07 involves establishing a three-tiered utilization system based on gas concentration. First, an online gas concentration analysis system is installed, using an infrared absorption gas analyzer with a measurement range of 0 to 100%, an accuracy of ±0.2%, and a sampling frequency of once every 5 seconds. The system classifies gas into three levels based on the real-time measured gas concentration: a high concentration zone with a gas concentration greater than 30%, a medium concentration zone with a gas concentration between 10% and 30%, and a low concentration zone with a gas concentration between 5% and 10%. An electrically operated three-way diverter valve is installed in the pipeline system. This valve is an angle-stroke electrically adjustable valve with a maximum diameter of 400 mm, a rated pressure of 1.0 MPa, a control ratio of 50:1, and a response time of less than 5 seconds. The control system automatically adjusts the direction of the three-way valve according to the gas concentration, guiding gas of different concentrations into the corresponding utilization system. High-concentration methane is directly introduced into a gas power plant or liquefaction unit. Medium-concentration methane is introduced into a catalytic oxidation unit, where the oxidation process takes place at 350℃ using a palladium-platinum catalyst at a loading of 1.5 kg / m³, achieving a catalytic conversion rate greater than 95%. The heat released during oxidation is used for heating or hot water supply in the mining area. Low-concentration methane is introduced into a thermal oxidation unit for complete oxidation at 850℃, achieving an oxidation efficiency greater than 99%, with methane content in the emissions below 0.1%, meeting environmental protection requirements. This step is based on the theory of graded utilization of methane resources, employing different utilization methods according to the characteristics of different methane concentrations, thus improving the comprehensive utilization rate of methane. Key parameters in this step are the methane concentration grading thresholds of 30% and 10%, determined based on the applicable conditions of different methane utilization technologies to ensure the most effective utilization of methane at each level. The purpose of this step is to achieve tiered utilization of methane resources, improve the comprehensive utilization efficiency of methane, and reduce greenhouse gas emissions.

[0056] The specific implementation of step S08 involves constructing an intelligent monitoring platform for gas extraction based on an industrial Internet of Things (IoT) architecture. The platform comprises four layers: a sensing layer, a network layer, a platform layer, and an application layer. The sensing layer consists of sensors distributed at various monitoring points, including gas concentration sensors, negative pressure sensors, flow sensors, and status sensors, with a sampling frequency of once per second. The network layer adopts a redundant star network topology. The backbone network uses industrial Ethernet with a transmission rate of 1000 Mbps, while the backup network uses an industrial wireless network with a transmission rate of 100 Mbps and a network latency of less than 10 milliseconds. The platform layer employs a distributed architecture, including a data acquisition server, a data storage server, and an application server. The servers utilize a dual-machine hot backup mode, and data storage uses a time-series database, supporting a data write speed of 10,000 points per second and a millisecond-level query response time. The application layer includes a data processing module, a visualization module, an early warning module, and a decision support module. The data processing module adopts a stream processing architecture to support real-time data analysis. The visualization module is implemented using web technology, supporting multi-terminal access. The early warning module is based on a rule engine and contains 100 early warning rules covering abnormal gas concentration, negative pressure, and flow rates. The decision support module uses case-based reasoning technology, storing 500 historical cases, and can match similar cases based on the current situation to provide decision suggestions. This step, based on Industrial Internet of Things (IIoT) and big data analytics, constructs an intelligent monitoring system for gas extraction, achieving real-time data monitoring, analysis, and early warning. Key parameters in this step are a system response time of less than 1 second and a data storage period of 1 year. These parameters ensure the system's real-time performance and data integrity, providing a guarantee for the safe operation and subsequent analysis of gas extraction. The purpose of this step is to achieve intelligent monitoring of the gas extraction process, improving the safety and reliability of system operation.

[0057] The specific implementation of step S09 involves evaluating the gas extraction effect through three aspects of measurement and calculation. First, the residual gas content is determined using a borehole sampling method. A sampling point is placed every 20 meters along the strike of the working face and every 10 meters along the dip, forming a sampling point network. A sampling borehole is drilled at each sampling point to a depth of 8 meters and a diameter of 42 millimeters. A sampler, a sealed container with a volume of 500 cubic centimeters, is placed in the borehole. Gas is extracted from the coal sample using a vacuum pump, and the gas content is determined. Then, the extraction rate is calculated using the formula η = (q0 - q1) / q0 × 100%, where η is the extraction rate, q0 is the original gas content (in cubic meters per ton), and q1 is the residual gas content (in cubic meters per ton). Finally, the gas utilization rate is calculated using the formula ξ = V1 / V0 × 100%, where ξ is the gas utilization rate, V1 is the actual amount of gas utilized (in cubic meters), and V0 is the total amount of gas extracted (in cubic meters). Based on the evaluation results, when the extraction rate is below 70% or the gas utilization rate is below 80%, the system automatically optimizes the extraction negative pressure and flow parameters, improving the extraction effect by increasing the extraction negative pressure or adjusting the extraction time. This step is based on scientific evaluation and closed-loop optimization principles, providing a basis for optimizing extraction parameters through quantitative evaluation of the extraction effect. The key parameters in this step are the target extraction rate of 70% and the target gas utilization rate of 80%. These target values ​​are determined comprehensively based on the actual needs and technical feasibility of coal mine gas management, ensuring both safe coal mine production and improved gas resource utilization. The purpose of this step is to guide the optimization of extraction parameters through quantitative evaluation, forming a closed-loop control and continuously improving the gas extraction effect and utilization efficiency.

[0058] Specifically, the principle of this invention is as follows: The technical principle of this invention to solve the problem of unstable gas concentration in coal seam drainage is mainly reflected in three aspects: spatial layout optimization, concentration control mechanism, and intelligent predictive control. From the perspective of spatial layout, the Y-shaped borehole structure innovatively combines the main borehole with branch boreholes to form a three-dimensional drainage network. The main borehole provides a stable drainage channel perpendicular to the coal seam, while the two branch boreholes at a 45° angle extend upward and downward respectively, effectively covering the goaf and the gas-rich area of ​​adjacent coal seams, solving the problem of insufficient coverage by traditional single boreholes.

[0059] Regarding the concentration control mechanism, this invention designs a buffer system consisting of a main buffer tank, an auxiliary buffer tank, and a flow regulating valve. It utilizes the principle of physical mixing to buffer and regulate gas concentrations of different levels. When high-concentration gas enters the system, an appropriate amount of low-concentration gas is introduced through the regulating valve for dilution; when low-concentration gas enters, high-concentration gas is introduced for enrichment. Through this dynamic balance mechanism, the output gas concentration is maintained within a stable range of 30% to 45%, meeting the concentration requirements for resource utilization.

[0060] Intelligent predictive control is the core innovation of this invention. The predictive model, built upon a long short-term memory neural network, can capture the complex nonlinear relationship between factors such as working face advance speed, extraction negative pressure, and gas emission. The model employs a multi-head self-attention mechanism to enhance its learning ability for long-term time-series dependencies. By analyzing time patterns and correlations in historical data, it accurately predicts the gas emission trend for the next 12 hours. The prediction results directly drive the dynamic adjustment system for extraction parameters, enabling advance adjustment of extraction negative pressure and flow rate. This allows the extraction process to proactively adapt to impending changes in gas emission, rather than reacting with lag as in traditional technologies, fundamentally improving the extraction system's adaptability and control accuracy to changes in gas emission.

[0061] The organic combination of these three technical principles forms a closed-loop gas extraction control system. From spatial coverage and concentration regulation to predictive control, it comprehensively solves the technical problem of unstable gas concentration in coal seam extraction, and realizes efficient, stable and intelligent gas extraction.

[0062] The following provides a specific embodiment 1 of the present invention, and the specific implementation of each step in this embodiment 1 is described in detail below.

[0063] The specific implementation of step S01 involves determining the coal seam distribution using ground-penetrating radar (GPR) and 3D modeling technologies. First, a main borehole is drilled above the coal seam, perpendicular to the coal seam direction, reaching the depth of the coal seam to be mined. The borehole diameter is 120 mm, and the borehole wall is reinforced with high-strength steel pipes with a wall thickness of 10 mm. Then, two branch boreholes are drilled from the bottom of the main borehole in both upstream and downstream directions. The branch boreholes maintain an inclination angle of 45° to the horizontal plane, have a diameter of 90 mm, and their length is determined based on the coal seam thickness, generally 3 to 5 times the thickness. Steel pipes are also laid inside the branch boreholes for support. After drilling is completed, a pressure test is conducted to ensure the borehole's sealing performance. When deploying a Y-shaped ventilation and extraction borehole group, the spacing between adjacent main boreholes is 30 meters, and a row of boreholes is deployed every 50 meters along the coal seam strike, forming a three-dimensional extraction network covering the entire mining area. The formula for calculating the length of a branch borehole is: L = k × H, where L is the length of the branch borehole in meters; k is a length coefficient, ranging from 3 to 5; and H is the coal seam thickness in meters. The formula for calculating the coverage area of ​​a Y-type borehole is: S = π × (L × sin(α)) 2 Where S is the coverage area of ​​a single Y-shaped borehole, in square meters; L is the length of the branch borehole, in meters; and α is the angle between the branch borehole and the horizontal plane, taken as 45°. This step is implemented based on fluid mechanics principles, increasing the gas drainage coverage area and improving gas drainage efficiency through a three-dimensional, intersecting borehole network. The purpose of this step is to form a three-dimensional gas drainage channel network, creating conditions for subsequent gas drainage.

[0064] The specific implementation of step S02 involves constructing a monitoring network using catalytic combustion gas sensors and thermal conductivity gas sensors. Catalytic combustion gas sensors are installed at the Y-shaped borehole exit, with a measurement range of 0 to 5% and an accuracy of ±0.1%. Thermal conductivity gas sensors are installed every 50 meters at key nodes along the main pipeline, with a measurement range of 0 to 100% and an accuracy of ±0.5%. Simultaneously, negative pressure sensors are installed at each monitoring point, with a measurement range of 0 to 25 kPa and an accuracy of ±0.2 kPa. All sensors are intrinsically safe, with an explosion-proof rating of Exia I, and the sensor acquisition frequency is set to once every 10 seconds. The sensors are connected to the nearest data acquisition unit via an RS485 bus, and the data acquisition unit transmits the collected data to the monitoring center via an industrial Ethernet network. During sensor data acquisition, the raw data is filtered using a digital low-pass filter algorithm. The filter function is: y(n) = α × x(n) + (1-α) × y(n-1), where y(n) is the current output value; x(n) is the current input value; y(n-1) is the previous output value; and α is the filter coefficient, ranging from 0 to 1, and is 0.3 in this embodiment. During data transmission, a CRC check algorithm is used to ensure data integrity. The check function is: CRC(M) = R(M × x r The expression mod G(x) is used, where M is the data to be transmitted; G(x) is the generator polynomial; r is the order of the generator polynomial; and R represents the remainder operation. This step, based on sensor network technology and the principles of the Industrial Internet of Things, constructs a real-time monitoring network for gas concentration and negative pressure, providing data support for the gas extraction process. The purpose of this step is to achieve real-time monitoring of gas concentration and negative pressure, providing a data foundation for subsequent dynamic adjustments.

[0065] The specific implementation of step S03 is as follows: First, a main buffer tank is constructed using Q345R steel, with a wall thickness of 20 mm, a cylindrical shape, a diameter of 10 meters, a height of 12.8 meters, and a volume of 10... 3 m 3 The internal structure features five layers of baffles arranged in an evenly spaced spiral pattern at a 30° angle. The baffle surfaces are made of a hydrophilic material to absorb moisture from the gas. The auxiliary buffer tank is also constructed of Q345R steel, with a diameter of 8 meters, a height of 10 meters, and a volume of 5 × 10⁶ m³. 2 m 3The system is internally filled with foamed plastic packing, with a specific surface area of ​​1000 square meters per cubic meter and a packing layer height of 5 meters. An electrically operated regulating valve is installed between the main buffer tank and the auxiliary buffer tank. This valve is an angle-stroke electrically operated regulating valve with a maximum diameter of 800 mm, a rated pressure of 1.6 MPa, a regulation ratio of 100:1, and an opening / closing time of less than 30 seconds. An online gas concentration analyzer is installed throughout the system, employing infrared absorption principles, with a measurement range of 0 to 100%, an accuracy of ±0.2%, and a response time of less than 10 seconds. When the extracted gas concentration is detected to be below 30%, the control system automatically reduces the opening of the low-concentration inlet valve and increases the opening of the high-concentration inlet valve, and vice versa. This flow rate proportional adjustment ensures that the mixed gas concentration is maintained between 30% and 45%. The valve opening adjustment uses a proportional-integral-derivative control algorithm, and the control function is: Where u(t) is the control output; e(t) is the error signal, i.e., the difference between the target concentration and the actual concentration; K p K is the proportionality constant, with a value of 1.5. i K is the integral coefficient, with a value of 0.3; d τ is the differential coefficient, with a value of 0.05; t is the time variable; τ is the integral variable. This step, based on fluid dynamics and automatic control principles, stabilizes the gas concentration through a buffer device and an automatic adjustment system, thus solving the problem of gas concentration fluctuations. The purpose of this step is to stabilize the extracted gas concentration, creating conditions for subsequent comprehensive utilization.

[0066] The specific implementation of step S04 involves constructing a mixing cavity made of 304 stainless steel with a wall thickness of 15 mm and a volume of 5 m³. 3 The mixing chamber is equipped with an internal turbine agitator with a diameter of 0.8 meters and an adjustable speed from 0 to 500 rpm. Four air inlets are installed on the mixing chamber, for connecting to high-concentration, medium-concentration, and low-concentration gas sources, respectively, with one inlet for emergency air dilution. Each inlet is equipped with an electric proportional valve, a linear-stroke electric regulating valve with a maximum diameter of 200 mm, a rated pressure of 1.0 MPa, and a control ratio of 50:1. A gas analyzer is installed at the outlet of the mixing chamber, using a thermal conductivity principle, with a measurement range of 0 to 100%, an accuracy of ±0.3%, and a response time of less than 5 seconds. The control system employs a fuzzy PID control algorithm, adjusting the valve openings based on real-time concentration data to maintain a constant mixing ratio of low-concentration to high-concentration gas at 4:6, ensuring the mixed gas concentration remains stable within the required range. The formula for calculating the mixed gas concentration is: Where C mC1 represents the concentration of the mixed gas as a percentage; C2 represents the concentration of the low-concentration gas as a percentage; V1 represents the volumetric flow rate of the low-concentration gas as cubic meters per hour; C2 represents the concentration of the high-concentration gas as a percentage; and V2 represents the volumetric flow rate of the high-concentration gas as cubic meters per hour. The fuzzy control rules adopt an IF-THEN structure, for example: IF (current concentration is "low") AND (concentration change rate is "negative") THEN (high-concentration valve opening is "increased") AND (low-concentration valve opening is "decreased"). The membership function of the fuzzy set uses a combination of triangles and trapezoids. Fuzzy inference uses the Mamdani algorithm, and defuzzification uses the centroid method. The calculation formula is as follows: Where u is the precise value after defuzzification; μ(x) i ) represents the membership degree; x i is the corresponding universe of discourse value; n is the number of discrete points. This step is implemented based on mixed gas theory and fuzzy control technology. By precisely controlling the mixing ratio of different concentrations of methane, stable control of the methane concentration is achieved. The purpose of this step is to improve methane utilization efficiency by achieving precise control of methane concentration through intelligent mixing and distribution.

[0067] The specific implementation of step S05 involves constructing a dynamic control system for gas extraction parameters using a multi-input multi-output control system. The system receives four key parameters: working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data. Data preprocessing, including filtering, noise reduction, and normalization, is performed by a data processing unit. The processed data is then input into a fuzzy neural network controller, which automatically calculates the optimal extraction negative pressure and flow rate parameters based on a preset fuzzy rule base. The normalization process uses a maximum-minimum value normalization method, calculated using the following formula: Where x norm x is the normalized value; x is the original value; x min The minimum value of this parameter; x max This is the maximum value of the parameter. The structure of a fuzzy neural network consists of five parts: an input layer, a fuzzy layer, a rule layer, a defuzzification layer, and an output layer. The input layer has four neurons, corresponding to four input parameters. The fuzzy layer performs fuzzification processing on the input parameters, converting precise values ​​into fuzzy sets. Each input parameter is divided into three fuzzy sets: "low," "medium," and "high." The rule layer contains 3... 4 =81 neurons, corresponding to 81 IF-THEN rules; the defuzzification layer performs a weighted average of the outputs of the rule layer to obtain the precise control quantity; the output layer has 2 neurons, outputting the sampling negative pressure adjustment value and the flow rate adjustment value respectively. The weight update of the fuzzy neural network adopts the backpropagation algorithm, with a learning rate of 0.05, a momentum factor of 0.8, and the error function being the mean square error, calculated using the following formula: Where E is the error; y i The expected output; This represents the actual output; m is the number of samples. The weight update formula is: Where w(t+1) is the updated weight; w(t) is the current weight; and η is the learning rate. Let be the partial derivative of the error with respect to the weights; α be the momentum factor; and Δw(t) be the current change in weights. This step is implemented based on fuzzy neural network control theory, transforming expert experience into fuzzy rules and achieving adaptive control of the system through the self-learning capability of the neural network. The purpose of this step is to dynamically adjust the sampling parameters based on real-time data, thereby improving sampling efficiency and safety.

[0068] The specific implementation of step S06 is based on constructing a gas outburst prediction model using a long short-term memory neural network. The model adopts a hierarchical structure, including four functional modules: a data preprocessing unit, a feature extraction unit, a time-series learning unit, and a prediction unit. The data preprocessing unit standardizes the input working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data. The Z-score method is used to ensure the data mean is 0 and the standard deviation is 1. The calculation formula is as follows: Where z is the standardized value; x is the original value; μ is the mean of the parameter; and σ is the standard deviation of the parameter. Wavelet transform is also used for signal denoising. The wavelet transform formula is: Among them W f (a, b) are wavelet transform coefficients; f(t) is the original signal; ψ * denoted by , where is the conjugate of the wavelet function; 'a' is the scale parameter; 'b' is the translation parameter; and 't' is the time variable. The feature extraction unit employs a bidirectional long short-term memory network structure, containing two forward propagation layers and two backward propagation layers, each with 128 neurons. It uses the tanh activation function to capture the contextual features of time-series data in both forward and backward directions. The core of the long short-term memory unit consists of three gate structures: the forget gate, the input gate, and the output gate, calculated using the following formula:

[0069] f t =σ(W f ·[h t-1 x t ]+b f )

[0070] i t =σ(W i ·[h t-1 x t ]+b i )

[0071]

[0072] o t =σ(W o ·[h t-1 x t ]+b o )

[0073] h t =o t ×tanh(C t )

[0074] Where f t Output for the forget gate; i t For input gate output; Candidate memory units; C t For the current memory unit; o t For output gate output; h t Hidden state; x t For the current input; h t-1 The previous hidden state; C t-1 For the memory unit of the previous moment; W f w i W C W o b is the weight matrix; f b i b C b o σ is the bias term; σ is the sigmoid activation function; tanh is the hyperbolic tangent activation function. The temporal learning unit employs a multi-head self-attention mechanism with 8 heads and a 64-dimensional attention dimension. Residual connections and layer normalization are combined to improve model stability. The model's ability to capture long-term temporal dependencies is enhanced by calculating the correlation weights between different time steps. The calculation formula for the self-attention mechanism is:

[0075] Where Q is the query matrix; K is the key matrix; V is the value matrix; d k is the dimension of the key vector; softmax is the softmax normalization function. The prediction unit consists of two fully connected layers. The first layer contains 64 neurons and uses the ReLU activation function. The output dimension of the second layer is consistent with the prediction target, outputting the predicted hourly gas outburst volume and concentration trend for the next 12 hours. The model is trained using the backpropagation algorithm, with the mean squared error loss function and the Adam optimizer as the optimization algorithm. The initial learning rate is 0.001, which decays exponentially with the number of training epochs, with a decay rate of 0.95. This step is based on deep learning and time series prediction theory, using historical data analysis to mine gas outburst patterns and achieve early prediction of gas outbursts and pre-adjustment of extraction parameters. The purpose of this step is to achieve early prediction of gas outbursts and provide decision support for dynamic adjustment of extraction parameters.

[0076] The specific implementation of step S07 involves establishing a three-tiered utilization system based on gas concentration. First, an online gas concentration analysis system is installed, using an infrared absorption gas analyzer with a measurement range of 0 to 100%, an accuracy of ±0.2%, and a sampling frequency of once every 5 seconds. The system classifies gas into three levels based on the real-time measured gas concentration: a high-concentration zone with a gas concentration greater than 30%, a medium-concentration zone with a gas concentration between 10% and 30%, and a low-concentration zone with a gas concentration between 5% and 10%. An electrically operated three-way diverter valve is installed in the pipeline system. This valve is an angle-stroke electrically adjustable valve with a maximum diameter of 400 mm, a rated pressure of 1.0 MPa, a control ratio of 50:1, and a response time of less than 5 seconds. The control system automatically adjusts the direction of the three-way valve according to the gas concentration, guiding gas of different concentrations into the corresponding utilization system. The formula for calculating the energy conversion efficiency of the tiered utilization system is: Where η i E represents the energy conversion efficiency of the i-th stage utilization system. out，i Energy output is measured in megajoules (E). in，i Input energy is measured in megajoules (MJ). The formula for calculating input energy is: E in，i =V i ×C i ×Q, where V i C represents the volume of the i-th level gas, in cubic meters; i The concentration of methane in the i-th level is expressed as a percentage; Q is the calorific value of pure methane, taken as 35.9 MJ / m³. High-concentration methane is directly introduced into a methane power plant or liquefaction unit, with a power generation efficiency of no less than 40%. Medium-concentration methane is introduced into a catalytic oxidation unit, where the oxidation process is carried out at 350℃ using a palladium-platinum catalyst with a catalyst loading of 1.5 kg / m³, achieving a catalytic conversion rate greater than 95%. The heat released during oxidation is used for heating or hot water supply in the mining area, with a heat recovery efficiency of no less than 70%. Low-concentration methane is introduced into a thermal oxidation unit, where complete oxidation is performed at 850℃, achieving an oxidation efficiency greater than 99%. The methane content in the emitted gas is less than 0.1%, meeting environmental protection requirements, and the heat recovery efficiency is no less than 60%. This step is based on the theory of graded utilization of methane resources, employing different utilization methods according to the characteristics of different methane concentrations, thereby improving the comprehensive utilization rate of methane. The purpose of this step is to achieve tiered utilization of methane resources, improve the comprehensive utilization efficiency of methane, and reduce greenhouse gas emissions.

[0077] The specific implementation of step S08 involves constructing an intelligent monitoring platform for gas extraction based on an industrial Internet of Things (IoT) architecture. The platform comprises four layers: a sensing layer, a network layer, a platform layer, and an application layer. The sensing layer consists of sensors distributed at various monitoring points, including gas concentration sensors, negative pressure sensors, flow sensors, and status sensors, with a sampling frequency of once per second. The network layer adopts a redundant star network topology. The backbone network uses industrial Ethernet with a transmission rate of 1000 Mbps, while the backup network uses an industrial wireless network with a transmission rate of 100 Mbps and a network latency of less than 10 milliseconds. The platform layer employs a distributed architecture, including a data acquisition server, a data storage server, and an application server. The servers utilize a dual-machine hot backup mode, and data storage uses a time-series database, supporting a data write speed of 10,000 points per second and a millisecond-level query response time. The application layer includes a data processing module, a visualization module, an early warning module, and a decision support module. The data processing module adopts a stream processing architecture to support real-time data analysis. The visualization module is implemented using web technology and supports multi-terminal access. The early warning module is based on a rule engine and contains 100 early warning rules covering situations such as abnormal gas concentration, abnormal negative pressure, and abnormal flow. The decision support module is based on case-based reasoning technology, storing 500 historical cases, and can match similar cases based on the current situation to provide decision suggestions. The system reliability calculation formula is: R(t) = e -λt Where R(t) represents the system reliability within time t; λ represents the system failure rate, in times per hour; and t represents the operating time, in hours. The system availability calculation formula is: Where A represents system availability; MTBF is the Mean Time Between Failures (MTBF) in hours; and MTTR is the Mean Time To Repair (MTTR) in hours. This system has an MTBF of no less than 10,000 hours, an MTTR of no more than 2 hours, and an availability of no less than 99.98%. This step, based on Industrial Internet of Things (IIoT) and big data analytics, constructs an intelligent monitoring system for gas extraction, enabling real-time data monitoring, analysis, and early warning. The purpose of this step is to achieve intelligent monitoring of the gas extraction process, improving the safety and reliability of system operation.

[0078] The specific implementation of step S09 involves evaluating the gas extraction effect through three aspects of measurement and calculation. First, the residual gas content is determined using a borehole sampling method. A sampling point is placed every 20 meters along the strike of the working face and every 10 meters along the dip, forming a sampling point network. A sampling borehole with a depth of 8 meters and a diameter of 42 millimeters is drilled at each sampling point. A sampler, a sealed container with a volume of 500 cubic centimeters, is placed in the borehole. Gas is extracted from the coal sample using a vacuum pump, and the gas content is determined. Then, the extraction rate is calculated using the formula... Where η is the extraction rate; q0 is the original gas content, in cubic meters per ton; and q1 is the residual gas content, in cubic meters per ton. Finally, the gas utilization rate is calculated using the formula... Where ξ is the gas utilization rate; V1 is the actual amount of gas utilized, in cubic meters; and V0 is the total amount of gas extracted, in cubic meters. The comprehensive evaluation of the extraction effect adopts a weighted scoring method, calculated using the formula: S = w1 × η + w2 × ξ + w3 × γ, where S is the comprehensive score; η is the extraction rate; ξ is the gas utilization rate; and γ is the safety index, calculated using the formula... N represents the number of gas exceedances after gas extraction, and N0 represents the number of gas exceedances before extraction; w1, w2, and w3 are weighting coefficients, with values ​​of 0.4, 0.3, and 0.3, respectively. Based on the evaluation results, when the extraction rate is below 70% or the gas utilization rate is below 80%, the system automatically optimizes the extraction negative pressure and flow parameters, improving the extraction effect by increasing the extraction negative pressure or adjusting the extraction time. This step is based on scientific evaluation and closed-loop optimization principles, providing a basis for optimizing extraction parameters through quantitative evaluation of the extraction effect. The purpose of this step is to guide the optimization of extraction parameters through quantitative evaluation, forming closed-loop control and continuously improving the gas extraction effect and utilization efficiency.

[0079] Optionally, in existing technologies, the Y-type ventilation and drainage borehole group is a special borehole structure pre-arranged at the coal seam mining front. It consists of a main borehole and two branch boreholes, resembling the letter Y. The main borehole is drilled perpendicular to the coal seam, reaching the depth of the coal seam to be mined. The two branch boreholes extend from the bottom of the main borehole in the upstream and downstream directions respectively, with an inclination angle of 45°, covering the goaf and adjacent coal seams, forming a three-dimensional drainage network. The advantage of the Y-type borehole is that its three-dimensional structure can increase the contact area with the coal seam, improve gas drainage efficiency, and at the same time reduce the number of boreholes, thus lowering construction costs. The formula for calculating the penetration radius of the Y-type borehole is: Where R is the permeability radius in meters; k is the coal seam permeability in millidarcy; ΔP is the pressure difference in Pascals; and t is the extraction time in seconds.

[0080] The following is an example 2 of a specific application scenario of the present invention: The Y-type ventilation gas extraction method was applied in a large coal mine working face. This coal mine is located in a deep mining area with an average coal seam thickness of 3.5 meters and an initial gas content as high as 15.8 cubic meters per ton, classifying it as a high-gas mine. The working face is 240 meters long, with an average advance speed of 5.2 meters per day. Due to the large and unstable gas outburst from the working face, posing a serious threat to the safety of coal mining operations, researchers decided to implement the Y-type ventilation gas extraction method, such as... Figure 2 As shown, the problem of gas control is solved through systematic design and implementation.

[0081] During implementation, firstly according to Figure 3 The design shown depicts a Y-shaped ventilation and extraction borehole group. A total of eight main boreholes are laid out, one every 30 meters along the working face. The main boreholes are drilled perpendicular to the coal seam, reaching a depth of 150 meters and a diameter of 120 mm. The borehole walls are reinforced with 10 mm thick high-strength steel pipes. At the bottom of each main borehole, two branch boreholes are drilled in the upward and downward directions, with an inclination angle of 45°. The branch boreholes have a diameter of 90 mm and a length of 15 meters, covering the goaf and adjacent coal seams. The borehole layout parameters are shown in Table 1.

[0082] Table 1. Y-type borehole layout parameters

[0083]

[0084]

[0085] Gas concentration monitoring sensors and negative pressure sensors were installed every 50 meters at the borehole exit and key nodes of the main pipeline, totaling 16 gas concentration sensors and 8 negative pressure sensors. The gas concentration sensors combine catalytic combustion and thermal conductivity methods, with a measurement range of 0 to 100%, an accuracy of ±0.2%, and a sampling frequency of once every 10 seconds. The negative pressure sensors have a measurement range of 0 to 25 kPa and an accuracy of ±0.2 kPa. Actual monitoring data are shown in Table 2.

[0086] Table 2. Partial Data on Gas Concentration and Negative Pressure at Borehole Outlet

[0087] time Drill hole number Gas concentration (%) Negative pressure value (kPa) <![CDATA[Flow rate (m 3 / min)]]> 2025-01-10 08:00 YZK-01 42.5 18.6 85.3 2025-01-10 08:00 YZK-02 38.7 17.8 78.4 2025-01-10 08:00 YZK-03 45.2 19.2 92.7 2025-01-10 08:00 YZK-04 40.1 18.5 83.2 2025-01-10 14:00 YZK-01 36.8 17.5 76.3 2025-01-10 14:00 YZK-02 33.2 16.9 72.8 2025-01-10 14:00 YZK-03 39.6 18.3 85.1 2025-01-10 14:00 YZK-04 35.4 17.2 77.5 2025-01-10 20:00 YZK-01 44.3 19.4 89.7 2025-01-10 20:00 YZK-02 41.5 18.7 84.6 2025-01-10 20:00 YZK-03 47.8 20.2 96.4 2025-01-10 20:00 YZK-04 43.2 19.1 87.5

[0088] according to Figure 4 The gas concentration regulating and buffering system shown comprises one main buffer tank and one auxiliary buffer tank. The main buffer tank is constructed of Q345R steel, with a wall thickness of 20 mm, a diameter of 10 meters, a height of 12.8 meters, and a volume of 10... 3 m 3 The auxiliary buffer tank is also made of Q345R steel, with a diameter of 8 meters, a height of 10 meters, and a volume of 5 × 10⁶ m. 2 m 3 The key parameters of the buffer system are shown in Table 3:

[0089] Table 3. Parameters of Gas Concentration Adjustment and Buffer System

[0090] Parameter name Main buffer tank Auxiliary buffer tank Material Q345R steel Q345R steel Wall thickness (mm) 20 18 Diameter (m) 10 8 Height (m) 12.8 10 <![CDATA[Volume (m 3 )]]> <![CDATA[10 3 ]]> <![CDATA[5×10 2 ]]> Internal structure 5 layers of deflector Foamed plastic filler <![CDATA[Specific surface area of filler (m 2 / m 3 )]]> - 1000

[0091] like Figure 5 As shown, an intelligent gas mixing and distribution device is installed. This device uses a negative pressure extraction system to mix low-concentration methane with high-concentration methane at a 4:6 ratio, maintaining the methane concentration in the pipeline network between 30% and 45%. The mixing chamber is made of 304 stainless steel with a wall thickness of 15 mm and a volume of 5 m³.3 It is equipped with an internal turbine agitator, the speed of which can be adjusted from 0 to 500 rpm. The intelligent mixing and air distribution effect is shown in Table 4:

[0092] Table 4 Comparison of Gas Concentration Before and After Mixing

[0093]

[0094]

[0095] like Figure 6 As shown, a gas emission prediction model based on a long short-term memory neural network was established to predict the gas emission volume and concentration trends of the working face over the next 12 hours by analyzing historical data. The model was trained using 90 days of historical data, including working face advance speed data, borehole outlet gas concentration data, gas concentration data at key nodes of the main pipeline, and negative pressure value data. The performance evaluation results of the prediction model are shown in Table 5.

[0096] Table 5 Performance Evaluation of Predictive Models

[0097] Evaluation indicators training set Validation set test set Root Mean Square Error (RMSE) 1.35 1.62 1.78 Mean Absolute Error (MAE) 0.94 1.15 1.23 <![CDATA[Coefficient of determination (R 2 )]]> 0.921 0.893 0.875 1-hour prediction accuracy (%) 95.8 93.6 92.1 6-hour prediction accuracy (%) 87.5 84.2 82.6 12-hour prediction accuracy (%) 78.3 75.4 73.8

[0098] like Figure 7 As shown, a gas concentration-based utilization system is implemented, classifying gas into three levels for utilization based on concentration. High-concentration gas (>30%) is directly used for power generation; medium-concentration gas (10%–30%) is converted into heat energy through a catalytic oxidation device; and low-concentration gas (5%–10%) is treated using thermal oxidation technology before being released. The utilization effects of the three different gas concentrations are shown in Table 6.

[0099] Table 6. Utilization Effect of Gas at Different Concentrations

[0100]

[0101] like Figure 8 As shown, a smart monitoring platform for gas extraction is deployed, integrating extraction negative pressure data, flow data, borehole outlet gas concentration data, gas concentration data at key nodes of the main pipeline, and system status data to achieve remote monitoring and fault early warning. The platform consists of four parts: a perception layer, a network layer, a platform layer, and an application layer, enabling real-time monitoring and intelligent early warning of the entire system. The key performance indicators of this monitoring platform are shown in Table 7.

[0102] Table 7 Performance Indicators of Intelligent Monitoring Platform

[0103] Indicator Name Indicator value Technical Standards System reliability (MTBF, hours) 12580 >10000 System availability (%) 99.986 >99.98 Response time (ms) 356 <1000 Data storage period (days) 365 >180 Number of early warning rules (items) 128 >100 Historical case database capacity (number of cases) 625 >500

[0104] In the gas drainage effectiveness evaluation phase, residual gas content was determined using borehole sampling to calculate the drainage rate and gas utilization rate. The residual gas content was tested at eight measuring points on the working face, and the calculated drainage effectiveness is shown in Table 8.

[0105] Table 8. Evaluation Table of Gas Drainage Effectiveness

[0106]

[0107]

[0108] Traditional gas drainage methods typically employ single vertical or horizontal boreholes, resulting in uneven borehole distribution and low drainage rates, generally only reaching 45%–55%. Furthermore, these methods lack intelligent prediction and control mechanisms, making it difficult to cope with changes in gas emission levels and posing safety hazards. Unstable gas concentrations also lead to low overall utilization efficiency, with gas utilization typically around 60%.

[0109] The Y-type ventilation gas extraction method of this invention represents a significant improvement over traditional methods: First, the Y-type borehole structure increases the contact area with the coal seam, improving the extraction coverage rate, which is increased from approximately 50% to 77.5%, an improvement of about 55%. Second, the gas concentration regulation and buffer system and intelligent gas mixing device achieve precise control of the gas concentration, maintaining it within the required range of 30% to 45%, significantly improving the gas utilization rate to 89.3%, nearly 50% higher than traditional methods. Third, the predictive control model based on a long short-term memory neural network enables early prediction of gas outbursts, with a 12-hour prediction accuracy of 73.8%, providing a scientific basis for dynamic adjustment of extraction parameters and significantly improving system safety.

[0110] It should be noted that the variables involved in this invention are explained in detail in Tables 9 and 10 below.

[0111] Table 9. Variable Explanation Table (Part 1)

[0112]

[0113]

[0114] Table 10 Variable Explanation Table (Part Two)

[0115]

[0116] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for Y-type ventilation and gas extraction in close-range coal seam groups, characterized in that, include: Deploy a Y-shaped ventilation and extraction borehole group; install a gas concentration monitoring sensor network; construct a gas concentration regulation and buffer system; install an intelligent gas mixing and distribution device; establish a dynamic control system for gas extraction parameters; apply a predictive control model to analyze the gas emission patterns of the working face based on historical data, receive data on the working face advance speed, borehole outlet gas concentration data, key node gas concentration data and negative pressure data of the main pipeline, output predicted values ​​of gas emission and concentration change trends for the next 12 hours, and adjust the extraction negative pressure and flow rate 12 hours in advance; implement a gas concentration-based utilization system. Deploy an intelligent monitoring platform for gas extraction; perform gas extraction effect evaluation; the gas concentration regulation and buffer system is specifically a device composed of a main buffer tank, an auxiliary buffer tank, a flow regulating valve, a pressure sensor, and a concentration sensor. When the extracted gas concentration fluctuates, the gas concentration regulation and buffer system controls the mixing ratio of different concentrations of gas by adjusting the valve opening to keep the output gas concentration stable; the intelligent mixing and gas distribution device is specifically a device composed of a mixing chamber, multiple air inlet pipes, an electric proportional valve, a gas analyzer, and a controller. The multiple air inlet pipes are respectively connected to high-concentration gas sources, medium-concentration gas sources, and low-concentration gas sources; an electric proportional valve is installed on each air inlet pipe, and the controller precisely controls the opening of each valve according to the real-time concentration data measured by the gas analyzer to achieve precise mixing of different concentrations of gas.

2. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 1, characterized in that, The Y-type ventilation and extraction borehole group is a special borehole structure pre-arranged at the front of coal seam mining. It consists of a main borehole and two branch boreholes. The main borehole is drilled perpendicular to the coal seam direction to the depth of the coal seam to be mined. The two branch boreholes extend from the bottom of the main borehole in the upstream and downstream directions respectively, with an inclination angle of 45°, covering the goaf and adjacent coal seams to form a three-dimensional extraction network.

3. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 2, characterized in that, Specifically, the gas concentration monitoring sensor network sets up a monitoring point every 50 meters at the borehole outlet and key nodes of the main pipeline to collect real-time gas concentration data at the borehole outlet, gas concentration data at key nodes of the main pipeline, and negative pressure value data.

4. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 3, characterized in that, The gas extraction parameter dynamic control system automatically adjusts the extraction negative pressure and flow rate based on the working face advance speed data, borehole outlet gas concentration data, main pipeline key node gas concentration data, and negative pressure value data.

5. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 4, characterized in that, The predictive control model is specifically a gas outburst prediction model built on a long short-term memory neural network. It consists of three parts: an input layer, a hidden layer, and an output layer. The input layer receives data on the working face advance speed, borehole outlet gas concentration, key node gas concentration data of the main pipeline, and negative pressure value data. The hidden layer contains four layers of long short-term memory units, with 128 neurons in each layer.

6. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 5, characterized in that, The specific structure of the predictive control model is a multivariate temporal prediction network structure, which consists of a data preprocessing unit, a feature extraction unit, a temporal learning unit, and a prediction unit. The feature extraction unit adopts a bidirectional long short-term memory network structure, and the temporal learning unit adopts an attention mechanism to enhance the model's ability to capture long temporal dependencies.

7. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 6, characterized in that, The steps for establishing the training dataset of the predictive control model specifically include four parts: data acquisition, data cleaning, feature extraction, and dataset partitioning. In the data acquisition stage, historical monitoring data for 90 consecutive days is obtained from the gas concentration monitoring sensor network.

8. The Y-type ventilation and gas extraction method for closely spaced coal seams according to claim 7, characterized in that, The temporal learning unit of the predictive control model is specifically composed of an attention layer consisting of a multi-head self-attention module with 8 heads and an attention dimension of 64. The model stability is improved by combining residual connections and layer normalization. The prediction unit consists of two fully connected layers, with the first layer containing 64 neurons.

Images