A coal mine multi-parameter linkage intelligent ventilation regulation system and a regulation method

Abstract

This invention relates to a multi-parameter linkage intelligent ventilation control system and method for coal mines, belonging to the field of coal mine ventilation control technology. It includes: a multi-source sensor array, a data transmission gateway, an edge computing core controller, and an airflow adjustment execution terminal. The edge computing core controller collects real-time data on methane concentration, carbon monoxide concentration, dust concentration, and wind speed and pressure from multiple working faces underground, constructs a multi-parameter coupling matrix, calculates ventilation demand assessment indicators, and generates airflow control commands based on a fuzzy neural network. This commands control the stepper motor of the regulating damper and the frequency converter of the local ventilation fan in the airflow adjustment execution terminal, achieving precise on-demand air supply. This application solves the problems of lagging ventilation control in existing mines and the inability to achieve dynamic and precise on-demand control based on changes in multiple parameters through multi-parameter linkage calculation and edge intelligent control, thereby improving the safety and energy efficiency of mine ventilation.

Description

Technical Field

[0001] This invention relates to the field of coal mine ventilation control technology, and in particular to a coal mine multi-parameter linkage intelligent ventilation control system and control method. Background Technology

[0002] The underground mining environment in coal mines is complex and variable, with the emission of harmful substances such as gas, carbon monoxide, and dust occurring suddenly and unevenly. To ensure the safety of underground workers and the normal operation of equipment, stable and sufficient ventilation must be maintained. Traditional mine ventilation control mainly relies on manual timed detection or threshold alarms for single parameters, followed by manual or semi-automatic adjustment of damper openings and ventilation fan frequencies.

[0003] However, in actual mining operations, gas emission, carbon monoxide concentration, and dust concentration are often interrelated and dynamically change with mining intensity. For example, during coal cutting, dust concentration and gas emission usually increase simultaneously. Most existing ventilation control systems adopt a single-parameter independent control mode, that is, they adjust only based on gas concentration or only based on wind speed, lacking a multi-parameter linkage analysis mechanism. This easily leads to lag in ventilation regulation, and even makes it impossible to make timely comprehensive control decisions when multiple parameters exceed the limits.

[0004] Furthermore, due to the lack of real-time assessment of the dynamic resistance of underground ventilation networks, existing adjustment methods often suffer from "over-adjustment" or "under-adjustment," failing to achieve precise on-demand air supply and easily leading to significant energy waste in the ventilation system. Therefore, designing a ventilation control system capable of real-time linkage with multiple environmental parameters, intelligent decision-making, and precise control is a pressing issue for safe and intelligent mining in current mines. Summary of the Invention

[0005] To address the aforementioned issues, this application provides a multi-parameter linkage intelligent ventilation control system and control method for coal mines.

[0006] In a first aspect, this application provides a multi-parameter linkage intelligent ventilation control system for coal mines, comprising: A multi-source sensor array is deployed in various monitoring areas underground in the coal mine to collect environmental parameter data and ventilation status parameter data in real time; A data transmission gateway, connected to the multi-source sensor array, is used to perform protocol conversion and packet transmission of the collected data; The edge computing core controller is connected to the data transmission gateway and is used to receive packaged data, construct a multi-parameter coupling matrix, calculate the ventilation demand assessment index of each monitoring area based on the multi-parameter coupling matrix, and generate air volume control instructions according to the ventilation demand assessment index. The air volume adjustment execution terminal is connected to the edge computing core controller and is used to receive the air volume adjustment command and drive the adjustment mechanism to perform air volume adjustment actions.

[0007] Preferably, the multi-source sensor array includes a gas sensor, a carbon monoxide sensor, a dust sensor, a wind speed sensor, and a wind pressure sensor.

[0008] By adopting the above technical solutions, it is possible to obtain air quality parameters and aerodynamic states of the downhole working face in a comprehensive and multi-dimensional manner, providing a comprehensive and accurate data foundation for multi-parameter linkage control.

[0009] Preferably, the air volume regulation execution terminal includes a damper adjustment driver, a local ventilation fan frequency converter, and an air curtain control box.

[0010] By adopting the above technical solution, it is possible to achieve coordinated control of airflow branch resistance and air source power output, thereby improving the flexibility and response accuracy of airflow regulation.

[0011] Preferably, the edge computing core controller includes a data preprocessing module, a coupling matrix construction module, a ventilation demand assessment module, a control command decision module, and a communication interface module.

[0012] By adopting the above technical solution, a complete closed loop from data input, feature extraction, solution decision to control output is realized at the edge, which reduces the latency of data uploading to the cloud and improves the real-time response speed of the control system.

[0013] Secondly, this application provides a multi-parameter linkage intelligent ventilation control method for coal mines, comprising the following steps: Step S1: Real-time data on methane concentration, carbon monoxide concentration, dust concentration, wind speed, and wind pressure in various monitoring areas underground in the coal mine are collected using a multi-source sensor array; Step S2: Transmit the collected data to the edge computing core controller via the data transmission gateway; Step S3: The edge computing core controller normalizes the received data and constructs a multi-parameter coupling matrix; Step S4: The edge computing core controller calculates the ventilation demand assessment index of each monitoring area based on the multi-parameter coupling matrix and uses a multi-physical quantity coupling correlation algorithm. Step S5: The edge computing core controller inputs the ventilation demand assessment index into the preset fuzzy neural network control model, outputs the target air volume for each monitoring area, and generates the corresponding air volume control command. Step S6: The air volume adjustment execution terminal receives the air volume control command and drives the control mechanism to adjust the ventilation volume of the corresponding area.

[0014] By adopting the above technical solution, complex downhole multi-parameters are fused into a coupling matrix, and fuzzy neural networks are used for intelligent decision-making, realizing automated closed-loop control from multi-dimensional environmental perception to precise airflow regulation.

[0015] Preferably, the specific process of constructing the multi-parameter coupling matrix in step S3 is as follows: Collect gas concentration at the same time t carbon monoxide concentration Dust concentration and actual wind speed Construct state vector ; Introducing a sliding time window T, we calculate the rate of change of each parameter within the window and construct a rate of change vector. ; Multiplying the state vector and the rate of change vector by a matrix yields the multi-parameter coupling matrix. .

[0016] By adopting the above technical solution, not only are the absolute values ​​of the parameters at the current instantaneous moment considered, but the changing trends of each parameter over time are also introduced, enabling the control system to have a forward-looking prediction and response capability for gas outbursts and dust accumulation.

[0017] Preferably, the calculation formula for the ventilation demand assessment index of each monitoring area in step S4 is as follows: ; In the formula, As an indicator for assessing ventilation demand, The preset weighting coefficients satisfy the following conditions: , The safe threshold for gas concentration, The safe threshold for carbon monoxide concentration. This is the safe threshold for dust concentration.

[0018] By adopting the above technical solution, multi-dimensional physical quantities are uniformly transformed into dimensionless comprehensive evaluation indicators, which can intuitively and scientifically reflect the current safety risk level and ventilation urgency of the working face.

[0019] Preferably, the training process of the fuzzy neural network control model in step S5 includes: Acquire historical ventilation control data, including historical multi-parameter coupling matrices and corresponding actual controlled air volumes; The membership function parameters and connection weights of the fuzzy neural network are adjusted using the backpropagation algorithm. The mean square error between the output air volume and the actual controlled air volume is minimized as the optimization objective. Iterative training is then performed to obtain the trained fuzzy neural network control model.

[0020] By adopting the above technical solution, the control model can adaptively learn complex nonlinear mapping relationships downhole, thereby improving the accuracy of air volume prediction.

[0021] Preferably, the driving control mechanism action in step S6 includes: When the target air volume is greater than the current air volume, the air volume regulation execution terminal controls the local ventilation fan frequency converter to increase the output frequency and controls the damper regulating driver to increase the damper opening. When the target air volume is less than the current air volume, the air volume regulation execution terminal first determines whether the current gas concentration and carbon monoxide concentration are lower than 50% of the safety threshold. If they are lower, it controls the damper regulating driver to reduce the damper opening; otherwise, it keeps the current air volume unchanged.

[0022] By adopting the above technical solutions, energy-saving regulation is carried out under the premise of ensuring safety, preventing the sudden accumulation of harmful gases due to blindly reducing the air volume, and ensuring the safety of the regulation process.

[0023] Preferably, after step S6, the method further includes: Step S7: Continuously collect wind speed and wind pressure data, calculate the actual air volume adjustment deviation, and if the deviation continues to exceed the set threshold within a preset time period, trigger the system's closed-loop feedback correction mechanism to fine-tune the output weights of the fuzzy neural network control model.

[0024] By adopting the above technical solution, a closed-loop feedback circuit was constructed, which can effectively eliminate the interference of time-varying factors such as damper wear and changes in roadway air resistance on the adjustment accuracy.

[0025] In summary, this application includes at least one of the following beneficial technical effects: 1. A scheme that uses a multi-source sensor array combined with an edge computing core controller to construct a multi-parameter coupling matrix has been adopted, which has obtained a comprehensive evaluation capability that can correlate instantaneous concentration and time-series change trend, avoiding the lag and limitations of single parameter adjustment; 2. By adopting a fuzzy neural network control model and a closed-loop feedback correction mechanism, which can adaptively learn and correct control rules, the system achieves high-precision and on-demand control of air volume in complex and ever-changing downhole environments, significantly improving ventilation safety and reducing fan energy consumption. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of a multi-parameter linkage intelligent ventilation control system for coal mines.

[0027] Figure 2 This is a schematic diagram of the modules within the edge computing core controller.

[0028] Figure 3This is a flowchart of a multi-parameter linkage intelligent ventilation control method for coal mines. Detailed Implementation

[0029] The present application will be further described in detail below with reference to all the accompanying drawings.

[0030] This application discloses a multi-parameter linkage intelligent ventilation control system for coal mines.

[0031] Reference Figure 1 and Figure 2 A coal mine multi-parameter linkage intelligent ventilation control system includes a multi-source sensor array, a data transmission gateway, an edge computing core controller, and an air volume adjustment execution terminal.

[0032] Multi-source sensor arrays are deployed in various monitoring areas underground in coal mines (e.g., fully mechanized mining faces, tunneling faces, return airways, etc.) to collect environmental parameter data and ventilation status parameter data in real time. Specifically, the multi-source sensor array includes gas sensors (for monitoring gas concentration, such as laser gas sensors or infrared gas sensors), carbon monoxide sensors (for monitoring carbon monoxide concentration, such as electrochemical carbon monoxide sensors), dust sensors (for monitoring the mass concentration of suspended dust in the air, such as light scattering dust sensors), wind speed sensors (for monitoring wind speed in the roadway, such as ultrasonic wind speed sensors), and wind pressure sensors (for monitoring static pressure and dynamic pressure difference in the roadway, such as piezoresistive wind pressure sensors).

[0033] The data transmission gateway connects to the multi-source sensor array via industrial Ethernet or mining wireless mesh network to perform protocol conversion (such as Modbus RTU to Modbus TCP) and packet transmission of the acquired analog or digital signals.

[0034] The edge computing core controller connects to the data transmission gateway and is implemented using an explosion-proof embedded industrial control computer. Internally, the edge computing core controller integrates a data preprocessing module, a coupling matrix construction module, a ventilation demand assessment module, a control command decision module, and a communication interface module. The edge computing core controller receives packaged data, constructs a multi-parameter coupling matrix, calculates the ventilation demand assessment indicators for each monitoring area based on the multi-parameter coupling matrix, and generates airflow control commands based on the ventilation demand assessment indicators.

[0035] The airflow regulation execution terminal is connected to the edge computing core controller via a control bus or wireless link. It receives airflow regulation commands and drives the regulation mechanism to perform airflow regulation actions. Specifically, the airflow regulation execution terminal includes a damper adjustment driver (used to control the blade opening of the automatic damper; the driver can be an explosion-proof stepper motor driver or a hydraulic push rod driver), a local ventilation fan frequency converter (used to control the speed of the local ventilation fan to change the air supply volume), and an air curtain control box (used to control the opening and closing of the air curtain in the roadway to block or guide airflow).

[0036] The working principle of a multi-parameter linkage intelligent ventilation control system for coal mines, as described in this application embodiment, is as follows: When mining operations are carried out underground, a multi-source sensor array continuously monitors the concentrations of methane, carbon monoxide, and dust in various areas, as well as the current wind speed and wind pressure. The collected physical quantity data is converted into unified Ethernet data packets by a data transmission gateway and uploaded to the edge computing core controller. The edge computing core controller uses a built-in algorithm to denoise and normalize the data, and then calculates the parameter change rate by combining a time sliding window, thereby constructing a coupling matrix that comprehensively reflects the dynamic interaction of multiple parameters. The ventilation demand assessment index for the current area is calculated through this matrix. Based on this index, the control command decision module determines the target air volume through a preset intelligent control algorithm and sends control commands to the air volume adjustment execution terminal. The driver of the air volume adjustment execution terminal adjusts the damper opening and fan frequency according to the commands, thereby achieving on-demand, precise, and real-time intelligent ventilation control.

[0037] This embodiment also discloses a multi-parameter linkage intelligent ventilation control method for coal mines, which is applied to a multi-parameter linkage intelligent ventilation control system for coal mines.

[0038] Reference Figure 3 A multi-parameter linkage intelligent ventilation control method for coal mines includes the following steps: Step S1: Real-time data on methane concentration, carbon monoxide concentration, dust concentration, wind speed, and wind pressure in various monitoring areas underground in the coal mine are collected using a multi-source sensor array.

[0039] Step S2: The collected data is formatted and encapsulated through the data transmission gateway and transmitted to the edge computing core controller deployed in the explosion-proof cavity underground via the mining Ethernet.

[0040] Step S3: The edge computing core controller normalizes the received data and constructs a multi-parameter coupling matrix. Normalization maps physical quantities of different dimensions to the [0,1] interval. The specific process of constructing the multi-parameter coupling matrix is ​​as follows: Collect gas concentration at the same time t carbon monoxide concentration Dust concentration and actual wind speed Construct state vector ; Introducing a sliding time window T, we calculate the rate of change of each parameter within the window and construct a rate of change vector. ; Multiplying the state vector and the rate of change vector by a matrix yields the multi-parameter coupling matrix. This matrix reflects the cross-coupling relationship between the current state value and its changing trend.

[0041] Step S4: The edge computing core controller, based on a multi-parameter coupling matrix, uses a multi-physical quantity coupling correlation algorithm to calculate the ventilation demand assessment index for each monitoring area. The calculation formula is: In the formula, This is a ventilation demand assessment indicator (dimensionless; the larger the value, the more urgent the ventilation demand). The preset weighting coefficients satisfy the following conditions: ; For example, in a working face prone to gas outbursts, the weight is set as follows: =0.4, =0.2, =0.2, =0.2, but in working faces with high coal dust, it can be adjusted to =0.2, =0.2, =0.4, =0.2; Set the safe threshold for gas concentration (e.g., 1.0%). The safe threshold for carbon monoxide concentration (e.g., set to 0.0024%). Set the dust concentration safety threshold (e.g., 10). ).

[0042] Step S5: The edge computing core controller inputs the ventilation demand assessment index into the preset fuzzy neural network control model, outputs the target air volume for each monitoring area, and generates the corresponding air volume control command.

[0043] Specifically, in order to make the fuzzy neural network (FNN) of this application clearly feasible, this embodiment adopts a five-layer topology network, including: an input layer, a fuzzification layer, a fuzzy rule inference layer, a normalization layer, and an output layer.

[0044] The input layer has two input variables: the ventilation demand assessment index calculated in step S4 at the current moment. And the rate of change of the index calculated using the differencer. .

[0045] The fuzzification layer uses a Gaussian membership function to convert the input variables into fuzzy activations. and The universe of discourse is defined as [0,1], and within this universe, it is divided into five fuzzy language sets: {NB (negative large), NS (negative small), ZO (zero), PS (positive small), PB (positive large)}. The mathematical expression for the Gaussian membership function is: ; In the formula, x is the input variable. and Let be the center value and width of the i-th input in the j-th fuzzy subset, respectively, and let their initial values ​​be uniformly discretized according to the historical ventilation expert experience distribution.

[0046] The fuzzy rule inference layer contains N=25 expert control rules, and the logical relationship of the rule base is shown in Table 1: Table 1: Fuzzy Neural Network Control Rule Base Table

[0047] The applicability of each rule output is obtained through algebraic product operation: The normalization applicability is calculated through the fourth normalization layer. Finally, the output layer of the fifth layer combines the consequent weights. Perform defuzzing and output the target air volume. By publicly disclosing the aforementioned specific pre-process architecture and rule base, those skilled in the art can fully reproduce this intelligent decision-making process. The training process of the fuzzy neural network control model includes: acquiring historical ventilation control data, including historical multi-parameter coupling matrices and corresponding actual control air volumes; using the backpropagation algorithm to adjust the membership function parameters and connection weights of the fuzzy neural network, and iteratively training with the goal of minimizing the mean square error between the output air volume and the actual control air volume, to obtain the trained fuzzy neural network control model.

[0048] Step S6: The airflow adjustment execution terminal receives the airflow control command and drives the control mechanism to adjust the ventilation volume of the corresponding area. Specifically, driving the control mechanism includes: When the target air volume is greater than the current air volume, it indicates that the environmental parameters are deteriorating or the air volume is insufficient. The air volume adjustment execution terminal controls the local ventilation fan frequency converter to increase the output frequency (e.g., from 35Hz to 45Hz) and controls the damper adjustment driver to increase the damper opening (e.g., increase the opening by 15%). To achieve the target air volume To address the "one cause, multiple effects" mapping and strong coupling conflict problem between local ventilation fans and automatic dampers, this application integrates a feedforward decoupling allocation algorithm into the edge computing core controller. Its core logic is to decompose the total airflow demand into power supply (fan frequency) and branch resistance adjustment (damper opening). The specific allocation formula is as follows: First, based on the kinetic energy mapping of the fan characteristic curve, the target output frequency of the local ventilation fan inverter is... The calculation formula is: ; In the formula, and These are the inverter's rated maximum frequency (50Hz) and minimum operating frequency (25Hz), respectively. This represents the maximum rated exhaust volume of the fan at the current frequency.

[0049] Secondly, in order to match the power output of the blower and eliminate the air resistance coupling in the roadway, the target opening area of ​​the automatic damper is... Decoupling calculations are performed using a pressure feedforward compensation formula: ; In the formula, The dynamic pressure difference is collected in real time by air pressure sensors deployed before and after the damper; Let be the downhole air density (taken as a constant 1.2). ); The flow coefficient of the damper cross section (range: 0.62~0.65).

[0050] During the dynamic execution phase, the controller adopts a coordinated control sequence of "fan first, damper later": when the air volume needs to be increased, the controller prioritizes increasing the frequency of the fan inverter. After a 3-5 second delay to allow the airflow and pressure to stabilize, adjust the pressure based on the latest differential pressure. Calculate and drive the damper regulating actuator to open the dynamic opening degree. When the airflow needs to be reduced, the operation is reversed. This achieves fully decoupled control of wind power and wind resistance, preventing airflow oscillations in the system. When the target air volume is less than the current air volume, in order to save energy, the system considers reducing the air volume. However, the air volume adjustment execution terminal first determines whether the current gas concentration and carbon monoxide concentration are lower than 50% of the safety threshold (i.e., gas concentration < 0.5% and carbon monoxide concentration < 0.0012%). If both are lower, the damper adjustment driver is controlled to reduce the damper opening. Otherwise, the current air volume is kept unchanged to prevent the accumulation of harmful gases.

[0051] Step S7: Continuously collect wind speed and wind pressure data, calculate the actual air volume adjustment deviation. If the deviation exceeds the set deviation threshold (e.g., the set threshold is 10% of the theoretical adjustment value) within a preset time period (e.g., 120 seconds), the system's closed-loop feedback correction mechanism is triggered to fine-tune the output weight of the fuzzy neural network control model to compensate for changes in wind resistance or mechanical wear of the actuator.

[0052] In step S7 above, the implementation steps and principles of the closed-loop feedback correction mechanism are as follows: Actual air volume calculation and deviation extraction: After the airflow adjustment is performed in step S6, the system enters a dynamic transition period. This occurs during the preset observation period. (In this embodiment, 90 to 120 seconds is preferred to avoid mechanical response lag caused by airflow inertia.) Within this timeframe, the edge computing core controller acquires the actual wind speed in real time through wind speed sensors deployed on the return air side of the working face. The actual air volume was calculated based on the standard cross-sectional area A of the tunnel at the test point. The calculation formula is: In the formula, The angle deviation between the wind speed sensor installation direction and the roadway axis (in this embodiment, the sensor is deployed perpendicular or parallel to the airflow). (Default value is 0).

[0053] Calculate the airflow deviation at the current time t. The mechanism is as follows: ; In the formula, The target air volume is the output of the fuzzy neural network in step S5.

[0054] Determining the feedback trigger condition: To avoid frequent system adjustments and oscillations caused by transient disturbances such as underground vehicle passage and temporary ventilation doors, the system is equipped with dual threshold triggering logic: Instantaneous deviation threshold: judgment Is it greater than the set maximum allowable deviation? (In this embodiment, it is set to 10% of the target air volume); Integral deviation threshold: calculated over the sliding time window Cumulative absolute deviation integral value : ; If and only if and ( The preset cumulative deviation energy threshold, in this embodiment, is set to a value of Simultaneously satisfying, or the duration of this state exceeds When the system determines that the resistance characteristics of the current underground ventilation network have undergone structural changes (such as local roof collapse, severe air leakage at the ventilation doors, or aging of the ventilation fans), it formally triggers the online correction procedure for the weights of the fuzzy neural network.

[0055] Implementation of an online weight correction algorithm for fuzzy neural networks: The output of the fuzzy neural network in this application It is the output weight of the consequent of each fuzzy rule. The forward propagation calculation formula for the linear combination weighted sum is as follows: ; In the formula, N is the total number of fuzzy rules. Let be the normalized applicability of the j-th fuzzy rule (i.e., the degree of matching between the input vector and the membership function of the antecedent of the rule).

[0056] When the correction procedure is triggered, the edge computing core controller uses gradient descent and adaptive error feedback to adjust the weights of the consequents. Real-time online correction is performed; the correction formula is as follows: ; In the formula, To adaptively adjust the learning rate, keeping the step size constant can easily lead to system overshoot or slow convergence. Therefore, this application designs an adaptive variable step size factor: ; In the formula, The base learning rate (in this embodiment, the value is 0.01~0.05); This is the sensitivity coefficient (in this embodiment, the value is taken as 0.5~1.2).

[0057] The control logic of this formula lies in: when the air volume deviation... When it is large, The system automatically increases the weights, quickly adjusting them to bring the airflow close to the target value; when the deviation... As it gradually decreases, Approaching 0 prevents the system from generating limit cycle oscillations near the equilibrium point, thus ensuring the asymptotic stability of the control system.

[0058] Security protection and convergence limitations: During the weight correction process, to prevent the weights from increasing indefinitely due to sensor malfunctions (such as a wind speed sensor being clogged with dust causing a constant reading of zero), the system sets weight boundary protection conditions: ; In the formula, and These are the weighted limits calculated based on the rated minimum and maximum air volume of the local ventilation fan. Once the correction value exceeds this boundary, the system immediately stops automatic correction and sends an "Actuator / Sensor Abnormality" alarm signal to the ground control center, switching to a safe alternative control mode (maintaining the current maximum safe air volume operation).

[0059] Through the aforementioned closed-loop feedback correction mechanism, this application can not only complete the conventional deviation elimination, but also enable the system to have the ability to learn and adapt to the time-varying characteristics of roadway wind resistance, effectively solving the interference of complex underground mining dynamic changes on ventilation control accuracy.

[0060] It should be noted that, in order to prevent a single monitoring area from being subject to the aforementioned consequent weighting... Online corrections can cause wind regulation conflicts between adjacent branch wind networks. Therefore, the system and method in this application impose the following limitations on the multi-regional collaborative boundary: The closed-loop feedback correction mechanism in this embodiment primarily regulates independent tunneling headways supplied with local ventilation fans or coal mining faces with independent return air systems. These areas belong to the terminal independent branches in the ventilation network topology.

[0061] When the edge computing core controller detects two or more monitoring areas with series or parallel wind paths simultaneously triggering the weight correction procedure in step S7, the controller will initiate the global wind network impedance matrix reconstruction mechanism. The controller will read the data streams from the wind pressure sensors in each area, calculate the coupling sensitivity matrix H, and incorporate coupling attenuation constraints into the actual adaptive correction step size for each monitoring area. ;in Let be the wind pressure interference coefficient generated by the adjustment of region k in region i. By introducing coupling attenuation constraints, the total system gain when multiple regions are simultaneously corrected is reduced, eliminating the risk of topological oscillation and runaway from the boundaries of system topology and global control during multi-region joint commissioning.

[0062] The working principle of the intelligent multi-parameter linkage ventilation control method for coal mines, as described in this application, is as follows: By constructing a multi-parameter coupling matrix containing instantaneous values ​​and rates of change from the collected environmental parameters, the system can predict not only the current concentration but also the rate of concentration change. A weighted adjustment formula is used to calculate the ventilation demand assessment index, which is then substituted into a fuzzy neural network control model to calculate the target air volume. During adjustment, the actuator follows the principle of "rapid response when increasing air velocity and safe verification when decreasing air velocity" to ensure the safety of the control process. Closed-loop feedback is introduced to adaptively correct the model parameters in real time based on actual air volume deviations, ensuring long-term stability and accuracy of the control.

[0063] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A multi-parameter linkage intelligent ventilation control system for coal mines, characterized in that, include: A multi-source sensor array is deployed in various monitoring areas underground in the coal mine to collect environmental parameter data and ventilation status parameter data in real time; A data transmission gateway, connected to the multi-source sensor array, is used to perform protocol conversion and packet transmission of the collected data; The edge computing core controller is connected to the data transmission gateway and is used to receive packaged data, construct a multi-parameter coupling matrix, calculate the ventilation demand assessment index of each monitoring area based on the multi-parameter coupling matrix, and generate air volume control instructions according to the ventilation demand assessment index. The air volume adjustment execution terminal is connected to the edge computing core controller and is used to receive the air volume adjustment command and drive the adjustment mechanism to perform air volume adjustment actions.

2. The intelligent ventilation control system for coal mines with multi-parameter linkage according to claim 1, characterized in that, The multi-source sensor array includes a gas sensor, a carbon monoxide sensor, a dust sensor, a wind speed sensor, and a wind pressure sensor.

3. The intelligent ventilation control system for coal mines with multi-parameter linkage according to claim 1, characterized in that, The air volume regulation execution terminal includes a damper adjustment driver, a local ventilation fan frequency converter, and an air curtain control box.

4. The intelligent ventilation control system for coal mines with multi-parameter linkage according to claim 1, characterized in that, The edge computing core controller includes a data preprocessing module, a coupling matrix construction module, a ventilation demand assessment module, a control command decision module, and a communication interface module.

5. A method for multi-parameter linkage intelligent ventilation control in coal mines based on the multi-parameter linkage intelligent ventilation control system described in claim 1, characterized in that, Includes the following steps: Step S1: Real-time data on methane concentration, carbon monoxide concentration, dust concentration, wind speed, and wind pressure in various monitoring areas underground in the coal mine are collected using a multi-source sensor array; Step S2: Transmit the collected data to the edge computing core controller via the data transmission gateway; Step S3: The edge computing core controller normalizes the received data and constructs a multi-parameter coupling matrix; Step S4: The edge computing core controller calculates the ventilation demand assessment index of each monitoring area based on the multi-parameter coupling matrix and uses a multi-physical quantity coupling correlation algorithm. Step S5: The edge computing core controller inputs the ventilation demand assessment index into the preset fuzzy neural network control model, outputs the target air volume for each monitoring area, and generates the corresponding air volume control command. Step S6: The air volume adjustment execution terminal receives the air volume control command and drives the control mechanism to adjust the ventilation volume of the corresponding area.

6. The intelligent ventilation control method for multi-parameter linkage in coal mines according to claim 5, characterized in that, The specific process of constructing the multi-parameter coupling matrix in step S3 is as follows: Collect gas concentration at the same time t carbon monoxide concentration Dust concentration and actual wind speed Construct state vector ; Introducing a sliding time window T, we calculate the rate of change of each parameter within the window and construct a rate of change vector. ; Multiplying the state vector and the rate of change vector by a matrix yields the multi-parameter coupling matrix. .

7. The intelligent ventilation control method for multi-parameter linkage in coal mines according to claim 6, characterized in that, The calculation formula for the ventilation demand assessment index of each monitoring area in step S4 is as follows: ； In the formula, As an indicator for assessing ventilation demand, The preset weighting coefficients satisfy the following conditions: , The safe threshold for gas concentration, The safe threshold for carbon monoxide concentration. This is the safe threshold for dust concentration.

8. The intelligent ventilation control method for multi-parameter linkage in coal mines according to claim 5, characterized in that, The training process of the fuzzy neural network control model in step S5 includes: Acquire historical ventilation control data, including historical multi-parameter coupling matrices and corresponding actual controlled air volumes; The membership function parameters and connection weights of the fuzzy neural network are adjusted using the backpropagation algorithm. The mean square error between the output air volume and the actual controlled air volume is minimized as the optimization objective. Iterative training is then performed to obtain the trained fuzzy neural network control model.

9. The intelligent ventilation control method for multi-parameter linkage in coal mines according to claim 5, characterized in that, The actions of the driving control mechanism in step S6 include: When the target air volume is greater than the current air volume, the air volume regulation execution terminal controls the local ventilation fan frequency converter to increase the output frequency and controls the damper regulating driver to increase the damper opening. When the target air volume is less than the current air volume, the air volume regulation execution terminal first determines whether the current gas concentration and carbon monoxide concentration are lower than 50% of the safety threshold. If they are lower, it controls the damper regulating driver to reduce the damper opening; otherwise, it keeps the current air volume unchanged.

10. The intelligent ventilation control method for multi-parameter linkage in coal mines according to claim 5, characterized in that, Following step S6, the method further includes: Step S7: Continuously collect wind speed and wind pressure data, calculate the actual air volume adjustment deviation, and if the deviation continues to exceed the set threshold within a preset time period, trigger the system's closed-loop feedback correction mechanism to fine-tune the output weights of the fuzzy neural network control model.

Images