Regulation and control method for monitoring and improving oxygen concentration of coal mine working face

Abstract

The invention discloses a regulation and control method for monitoring and improving the oxygen concentration of a coal mine working face, and relates to the technical field of regulation and control of the oxygen concentration of the coal mine working face. Obtaining original environment monitoring data; performing dynamic denoising and error correction on the original environment monitoring data by adopting a Kalman filtering algorithm to obtain a corrected environment parameter sequence; based on the corrected environmental parameter sequence, constructing a time sequence input vector, inputting the time sequence input vector into a trained long-short-term memory network model, and outputting an oxygen concentration change trend predicted value in a future preset time period; according to the oxygen concentration change trend prediction value, in combination with a preset safety threshold range, whether the current ventilation strategy meets the oxygen concentration regulation and control requirement or not is judged, and a preliminary regulation and control instruction is generated;

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of oxygen concentration regulation of coal mine working face, and in particular to a method for monitoring and improving oxygen concentration regulation of coal mine working face. BACKGROUND

[0002] The oxygen concentration regulation technology of coal mine working face refers to a solution for monitoring and maintaining the oxygen concentration in the air of coal mine underground working face within a safe range through a series of technical means. The core purpose is to protect the life safety of workers, improve production efficiency and reduce accident risks. Therefore, how to use advanced technical means to improve the intelligent level and safety of oxygen concentration regulation of coal mine working face has become one of the problems to be solved.

[0003] In the field of oxygen concentration regulation of coal mine working face, the existing monitoring system can only provide data at the current time, lacks the ability to predict the trend of oxygen concentration change in the future period of time, cannot take preventive measures in advance to deal with potential risks, and most of the ventilation systems adjust the ventilation volume by fixed parameter setting or simple threshold triggering mode, which cannot dynamically adjust according to the changes of actual environmental conditions, resulting in increased energy consumption or reduced safety factor. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a method for monitoring and improving oxygen concentration regulation of coal mine working face to solve the problems that the existing monitoring system can only provide data at the current time, lacks the ability to predict the trend of oxygen concentration change in the future period of time, cannot take preventive measures in advance to deal with potential risks, and most of the ventilation systems adjust the ventilation volume by fixed parameter setting or simple threshold triggering mode, which cannot dynamically adjust according to the changes of actual environmental conditions, resulting in increased energy consumption or reduced safety factor.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In the first aspect, the present application provides a method for monitoring and improving oxygen concentration regulation of coal mine working face, which comprises:

[0008] A multi-source sensor cooperative collection technology is used to collect the oxygen concentration, wind speed, gas concentration and environmental temperature and humidity parameters in the coal mine working face to obtain original environmental monitoring data;

[0009] A Kalman filter algorithm is used to dynamically denoise and correct errors of the original environmental monitoring data to obtain a corrected environmental parameter sequence;

[0010] Based on the corrected environmental parameter sequence, a time series input vector is constructed and input into the trained long short-term memory network model to output the predicted value of oxygen concentration change trend within a preset time period.

[0011] Based on the predicted oxygen concentration change trend and the preset safety threshold range, determine whether the current ventilation strategy meets the oxygen concentration control requirements and generate a preliminary control instruction.

[0012] A nonlinear regression modeling method is used to establish a mathematical mapping relationship between air volume input and oxygen concentration response, and an air volume-oxygen concentration coupled control model is constructed.

[0013] Based on the air volume-oxygen concentration coupled control model, the preliminary control command is converted into the corresponding air volume adjustment target value;

[0014] A closed-loop feedback control mechanism is adopted to dynamically adjust the operating frequency of the main ventilation fan or the opening of local dampers based on the target air volume adjustment value, so as to achieve precise control of oxygen concentration.

[0015] Edge computing nodes are deployed in key areas of the coal mine working face. Each node receives local sensor data and executes local control strategies, while maintaining communication and synchronization with the central control system, forming a distributed collaborative control network.

[0016] The distributed collaborative control network integrates the control results and feedback information of each node, updates the global control strategy, and optimizes the overall ventilation efficiency.

[0017] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, wherein: the Kalman filter algorithm is used to dynamically denoise and correct errors in the original environmental monitoring data to obtain a corrected environmental parameter sequence, specifically including:

[0018] The Kalman filter algorithm was used to perform state estimation on the oxygen concentration, wind speed, gas concentration and environmental temperature and humidity parameters in the original environmental monitoring data.

[0019] Construct the state equations and observation equations for each parameter, where the state equations are expressed as follows:

[0020] X k =AX k-1 +Bu k +w k ;

[0021] The observation equation is expressed as:

[0022] Z k =HX k +v k ;

[0023] Among them, X k Let A represent the state vector at time k, containing oxygen concentration, wind speed, gas concentration, and ambient temperature and humidity. Let A be the state transition matrix, B be the control input matrix, and u be the state vector. k To control the input, w k The noise is a process noise that follows a Gaussian distribution, Z. k Let v be the vector of actual observations at time k, H be the observation matrix, and v k To observe noise;

[0024] Based on the above equations, the prediction and update steps are executed sequentially:

[0025] Predict the current state:

[0026]

[0027] P k|k-1 =AP k-1|k-1 A T +Q;

[0028] Update current state estimate:

[0029] K k =P k|k-1 H T HP k|k-1 H T +R) -1 ;

[0030]

[0031] P k|k =(IK k H)P k|k-1 ;

[0032] in, For prior estimation, For posterior estimation, P is the covariance matrix, Q is the process noise covariance, R is the observation noise covariance, and K... k Kalman gain;

[0033] The obtained posterior estimates are output as a corrected sequence of environmental parameters.

[0034] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, the step of constructing a time series input vector based on the corrected environmental parameter sequence and inputting it into a trained long short-term memory network model to output a predicted value of oxygen concentration change trend within a preset time period specifically includes:

[0035] The corrected environmental parameter sequence is segmented using the sliding window method to construct a time series input vector x of length T. t =[x t-T+1 ,...,x t ];

[0036] Multiple input samples are sequentially fed into a pre-trained long short-term memory network model;

[0037] The Long Short-Term Memory network model has completed weight initialization, forward propagation calculation, loss function optimization, and backpropagation update during the training phase.

[0038] The model consists of multiple LSTM units, and each LSTM unit performs the following operations in sequence:

[0039] Input vector x at the current time t Hidden state from the previous moment t-1 The splicing serves as the input for the gating mechanism;

[0040] The forget gate selectively retains or forgets the cell state from the previous moment;

[0041] The input gate determines the current information writing ratio;

[0042] New cell state increments are generated from candidate states;

[0043] After updating the unit state, the output gate controls the output of the currently hidden state;

[0044] Finally, output the hidden state at the current moment. t ;

[0045] The model output layer provides the final hidden state. t Perform a linear transformation to generate predicted oxygen concentration values ​​from time t+1 to t+n.

[0046] During the model training phase, supervised learning is performed using historical monitoring datasets. The input is a sequence of historical environmental parameters, and the output is the actual oxygen concentration value within the target time period.

[0047] During training, the mean squared error loss function is used to measure the deviation between the predicted value and the true value, and the model parameters are continuously adjusted through the gradient descent algorithm until the preset accuracy threshold is reached.

[0048] After training, the model enters the deployment phase, where it receives real-time environmental parameter sequences and outputs predicted values ​​of oxygen concentration change trends.

[0049] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, the step of determining whether the current ventilation strategy meets the oxygen concentration control requirements based on the predicted value of oxygen concentration change trend and a preset safety threshold range, and generating a preliminary control instruction, specifically includes:

[0050] Predicted values ​​based on oxygen concentration change trends Compare with a preset safety threshold range;

[0051] The safety threshold range is set to [O]. low O ig ], where O low The lower limit of oxygen concentration, O ig The upper limit;

[0052] If any predicted value If the current ventilation strategy is deemed insufficient to maintain the required oxygen concentration, a ventilation enhancement command will be generated.

[0053] If any predicted value The system determines that the current ventilation strategy has resulted in excessively high oxygen concentration and generates a ventilation reduction command.

[0054] If all predicted values ​​are in [O low O ig If the area is within the specified range, the current ventilation strategy will remain unchanged.

[0055] The generated preliminary control instructions include the ventilation enhancement instruction flag D. up Or ventilation reduction instruction flag D down This is for use in subsequent steps.

[0056] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, the method employs a nonlinear regression modeling method to establish a mathematical mapping relationship between air volume input and oxygen concentration response, and constructs an air volume-oxygen concentration coupled control model, specifically including:

[0057] By adopting the air volume adjustment direction corresponding to the initial control command and combining historical operating data, a nonlinear functional relationship between air volume V and oxygen concentration response O is established.

[0058] Select the historical dataset {(V1,O1),(V2,O2),...,(V N O N )};

[0059] Assume the function is a power-law function:

[0060] O = a·V b +c;

[0061] Where O is the oxygen concentration response value, V is the air volume input value, and a, b, and c are the parameters to be fitted.

[0062] The least squares method is used to fit the model parameters to minimize the objective function.

[0063]

[0064] Solving for the optimal parameter a * ,b * ,c * The air volume-oxygen concentration coupled control model is obtained:

[0065] The model is used to convert preliminary control commands into specific airflow adjustment target values.

[0066] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces described in this invention, the method employs a closed-loop feedback control mechanism to dynamically adjust the operating frequency of the main ventilation fan or the opening of local dampers based on the target air volume adjustment value, thereby achieving precise control of oxygen concentration. Specifically, this includes:

[0067] Using the target value V for air volume adjustment set As the setpoint for the closed-loop feedback control system;

[0068] By real-time data collection of the main ventilation fan outlet air volume V act With the set value V set Compare and calculate the deviation e = V set -V act ;

[0069] The control output u is calculated using the proportional-integral-derivative (PID) control algorithm, and its expression is:

[0070]

[0071] Among them, K p K i K d These are the proportional, integral, and derivative coefficients of the PID controller, respectively.

[0072] The control output u is used to adjust the frequency of the main ventilation fan inverter or the opening of the local damper actuator, so that the actual air volume approaches the set value.

[0073] The actual air volume V is re-collected after each adjustment. act Repeat the above control process until the deviation is less than the preset error tolerance.

[0074] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, wherein: the deployment of edge computing nodes in key areas of the coal mine working face, each node receiving local sensor data and executing local control strategies, while maintaining communication and synchronization with the central control system, forming a distributed collaborative control network, specifically includes:

[0075] Edge computing nodes are deployed in key areas of the coal mine working face, and each node is equipped with local oxygen concentration, wind speed and gas concentration sensors.

[0076] Each node periodically collects local sensor data, executes local airflow adjustment strategies, and records the adjustment results;

[0077] Each node uploads the local adjustment results to the central control system, including the adjustment timestamp, air volume setpoint, and oxygen concentration change value;

[0078] The central control system receives data uploaded by each node, performs consistency checks and conflict detection, and identifies abnormal nodes.

[0079] When a conflict in regulation between nodes is detected, a coordination mechanism is activated, and the regulation result of the node with higher priority is selected as a reference.

[0080] The coordinated control strategy is then distributed to each edge node to form a unified distributed collaborative control network.

[0081] As a preferred embodiment of the method for monitoring and improving oxygen concentration control in coal mine working faces according to the present invention, the step of integrating the control results and feedback information of each node through a distributed collaborative control network, updating the global control strategy, and optimizing the overall ventilation efficiency specifically includes:

[0082] A global control database is constructed by using the control results and feedback information uploaded by each node in the distributed collaborative control network;

[0083] Extract the control effect indicators for each node, including the increase in oxygen concentration, the adjustment range of air volume, and the change in energy consumption;

[0084] The weighted average method is used to comprehensively evaluate the regulation effect of each node, and the regional regulation weight coefficient w is calculated. i , where i is the node number;

[0085] Update the global control strategy based on the weighting coefficients and the control deviation:

[0086]

[0087] Where M is the total number of edge nodes, V i (t) represents the airflow setting value of the i-th node at time t;

[0088] Update the global airflow setting value V global (t+1) is sent to each edge computing node to guide the local control operations in the next cycle.

[0089] In a second aspect, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the method for monitoring and improving oxygen concentration control in coal mine working faces as described in the first aspect of the present invention.

[0090] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the method for monitoring and improving oxygen concentration control in coal mine working faces as described in the first aspect of the present invention.

[0091] The beneficial effects of this invention are as follows: By employing multi-source sensor collaborative acquisition, Kalman filter data correction, LSTM neural network trend prediction, air volume-oxygen concentration coupling modeling, closed-loop feedback control, and edge computing distributed collaborative regulation, a complete technology chain from environmental perception and intelligent analysis to dynamic regulation is constructed. This enables high-precision real-time monitoring and forward-looking regulation of oxygen concentration in coal mine working faces, improves the response speed and intelligent regulation level of the ventilation system, effectively protects the safety and health of underground workers, and optimizes ventilation energy consumption, resulting in significant safety benefits and energy-saving advantages. Attached Figure Description

[0092] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0093] Figure 1 This is a flowchart of the method for monitoring and improving oxygen concentration control in coal mine working faces in Example 1. Detailed Implementation

[0094] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0095] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0096] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0097] Example, refer to Figure 1 This embodiment of the invention provides a method for monitoring and improving oxygen concentration control in coal mine working faces, comprising the following steps:

[0098] S1. Multi-source sensor collaborative acquisition technology is used to collect oxygen concentration, wind speed, gas concentration and ambient temperature and humidity parameters in the coal mine working face to obtain raw environmental monitoring data;

[0099] Furthermore, multiple sensor nodes are deployed in key areas of the coal mine working face. Each node includes a first sensor for detecting oxygen concentration, a second sensor for detecting wind speed, a third sensor for detecting methane concentration, and a fourth sensor for detecting ambient temperature and humidity.

[0100] Key areas include, but are not limited to, the coal mining machine operating area, return air roadway, intake air roadway, and densely populated work areas;

[0101] Each sensor periodically collects environmental parameters at its location according to a preset sampling frequency, generating raw environmental monitoring data in the form of a time series.

[0102] The raw environmental monitoring data includes oxygen concentration data series, wind speed data series, gas concentration data series, temperature data series, and humidity data series;

[0103] Each sensor transmits the collected data to the local data acquisition terminal via wired or wireless communication, where the terminal performs preliminary encapsulation and timestamp marking on the data.

[0104] The local data acquisition terminal uploads the packaged data to the central control system, forming a raw environmental monitoring dataset in a unified format;

[0105] The data encapsulation process includes adding sensor numbers, acquisition time, and geographic location information, and employing a checksum mechanism to ensure data integrity.

[0106] It should be noted that the sensor nodes are deployed in key areas of the coal mine working face, enabling multi-dimensional and high-frequency acquisition of parameters such as oxygen concentration, wind speed, gas concentration, and temperature and humidity, providing complete and accurate basic data support for subsequent data processing and control decisions; each sensor synchronizes its data through a unified time reference, ensuring the consistency and fusion of data from different sources on the time axis.

[0107] S2. The Kalman filter algorithm is used to dynamically denoise and correct errors in the original environmental monitoring data to obtain the corrected environmental parameter sequence.

[0108] Furthermore, the Kalman filter algorithm is used to perform state estimation on the oxygen concentration, wind speed, gas concentration and environmental temperature and humidity parameters in the original environmental monitoring data.

[0109] Construct the state equations and observation equations for each parameter, where the state equations are expressed as follows:

[0110] X k =AX k-1 +Bu k +w k ;

[0111] The observation equation is expressed as:

[0112] Z k =HX k +v k ;

[0113] Among them, X k Let A represent the state vector at time k, containing oxygen concentration, wind speed, gas concentration, and ambient temperature and humidity. Let A be the state transition matrix, B be the control input matrix, and u be the state vector. k To control the input, w k The noise is a process noise that follows a Gaussian distribution, Z. k Let v be the vector of actual observations at time k, H be the observation matrix, and v k To observe noise;

[0114] Based on the above equations, the prediction and update steps are executed sequentially:

[0115] Predict the current state:

[0116]

[0117]

[0118] Update current state estimate:

[0119] K k =P k|k-1 H T HP k|k-1 H T +R) -1 ;

[0120]

[0121] P k|k =(IK kH)P k|k-1 ;

[0122] in, For prior estimation, For posterior estimation, P is the covariance matrix, Q is the process noise covariance, R is the observation noise covariance, and K... k Kalman gain;

[0123] The obtained posterior estimates are output as a corrected sequence of environmental parameters;

[0124] It should be noted that the Kalman filter algorithm is suitable for state estimation problems of dynamic systems. In this invention, it is used to eliminate random noise and measurement errors in sensor-acquired data and improve the reliability of environmental parameter data. The method is based on the system model and observation data to recursively update the state estimate, which has the characteristics of strong real-time performance and high computational efficiency, and is suitable for the complex and ever-changing environmental monitoring needs in coal mines.

[0125] S3. Based on the corrected environmental parameter sequence, construct a time series input vector and input it into the trained long short-term memory network model to output the predicted value of oxygen concentration change trend within a preset time period.

[0126] Furthermore, a sliding window method is used to segment the corrected environmental parameter sequence, constructing a time series input vector x of length T. t =[x t-T+1 ,...,x t ];

[0127] Multiple input samples are sequentially fed into a pre-trained long short-term memory network model;

[0128] The Long Short-Term Memory (LSTM) network model has completed weight initialization, forward propagation calculation, loss function optimization, and backpropagation update during the training phase.

[0129] The model consists of multiple LSTM units, and each LSTM unit performs the following operations in sequence:

[0130] Input vector x at the current time t Hidden state from the previous moment t-1 The splicing serves as the input for the gating mechanism;

[0131] The forget gate selectively retains or forgets the cell state from the previous moment;

[0132] The input gate determines the current information writing ratio;

[0133] New cell state increments are generated from candidate states;

[0134] After updating the unit state, the output gate controls the output of the currently hidden state;

[0135] Finally, output the hidden state at the current moment. t ;

[0136] The model output layer's response to the final hidden state. t Perform a linear transformation to generate predicted oxygen concentration values ​​from time t+1 to t+n.

[0137] During the model training phase, supervised learning is performed using historical monitoring datasets. The input is a sequence of historical environmental parameters, and the output is the actual oxygen concentration value within the target time period.

[0138] During training, the mean squared error loss function is used to measure the deviation between the predicted value and the true value, and the model parameters are continuously adjusted through the gradient descent algorithm until the preset accuracy threshold is reached.

[0139] After training is completed, the model enters the deployment phase, which is used to receive real-time collected environmental parameter sequences and output predicted values ​​of oxygen concentration change trends.

[0140] It should be noted that Long Short-Term Memory (LSTM) networks are a deep learning architecture suitable for time series modeling. They can effectively capture the changing trends of historical environmental parameters and predict future oxygen concentration trends. By constructing input vectors through a sliding window approach, historical data is transformed into a form that can be used for model inference, thereby enabling a forward-looking judgment on the trend of oxygen concentration changes and providing a scientific basis for adjusting ventilation strategies.

[0141] S4. Based on the predicted trend of oxygen concentration change and the preset safety threshold range, determine whether the current ventilation strategy meets the oxygen concentration control requirements and generate preliminary control instructions.

[0142] Furthermore, using predicted values ​​of oxygen concentration change trends Compare with a preset safety threshold range;

[0143] The safety threshold range is set to [O] low O ig ], where O low The lower limit of oxygen concentration, O ig The upper limit;

[0144] If any predicted value If the current ventilation strategy is deemed insufficient to maintain the required oxygen concentration, a ventilation enhancement command will be generated.

[0145] If any predicted value The system determines that the current ventilation strategy has resulted in excessively high oxygen concentration and generates a ventilation reduction command.

[0146] If all predicted values ​​are in [O low O ig If the area is within the specified range, the current ventilation strategy will remain unchanged.

[0147] The generated preliminary control instructions include the ventilation enhancement instruction flag D. up Or ventilation reduction instruction flag D down This is for use in subsequent steps;

[0148] It should be noted that setting a safety threshold range is to ensure that the oxygen concentration in the coal mine working environment is always within a safe range, avoiding the risk of personnel suffocation due to excessively low oxygen concentration or the risk of fire due to excessively high oxygen concentration. By comparing the predicted value of oxygen concentration trend with the safety threshold, potential risks can be identified in advance and corresponding control instructions can be generated, realizing the transformation from passive response to proactive prevention.

[0149] S5. Using a nonlinear regression modeling method, establish a mathematical mapping relationship between air volume input and oxygen concentration response, construct an air volume-oxygen concentration coupled control model, and based on the air volume-oxygen concentration coupled control model, convert the initial control command into the corresponding air volume adjustment target value.

[0150] Furthermore, by adopting the air volume adjustment direction corresponding to the initial control command and combining it with historical operating data, a nonlinear functional relationship between air volume V and oxygen concentration response O is established.

[0151] Select the historical dataset {(V1,O1),(V2,O2),...,(V N O N )};

[0152] Assume the function is a power-law function:

[0153] O = a·V b +c;

[0154] Where O is the oxygen concentration response value, V is the air volume input value, and a, b, and c are the parameters to be fitted.

[0155] The least squares method is used to fit the model parameters to minimize the objective function.

[0156]

[0157] Solving for the optimal parameter a * ,b * ,c * The air volume-oxygen concentration coupled control model is obtained:

[0158] The model is used to convert preliminary control commands into specific airflow adjustment target values;

[0159] It should be noted that the air volume-oxygen concentration coupled control model is an empirical model based on a large amount of historical operating data. It can reflect the actual impact of air volume regulation on oxygen concentration distribution. The use of nonlinear regression modeling method is more in line with the complex physical characteristics of mine ventilation system, which improves the control accuracy and adaptability, and makes air volume regulation more scientific and reasonable.

[0160] S6. It adopts a closed-loop feedback control mechanism to dynamically adjust the operating frequency of the main ventilation fan or the opening of local dampers based on the target value of air volume adjustment, so as to achieve precise control of oxygen concentration.

[0161] Furthermore, the target value V for airflow regulation is adopted. set As the setpoint for the closed-loop feedback control system;

[0162] By real-time data collection of the main ventilation fan outlet air volume V act With the set value V set Compare and calculate the deviation e = V set -V act ;

[0163] The control output u is calculated using the proportional-integral-derivative (PID) control algorithm, and its expression is:

[0164]

[0165] Among them, K p K i K d These are the proportional, integral, and derivative coefficients of the PID controller, respectively.

[0166] The control output u is used to adjust the frequency of the main ventilation fan inverter or the opening of the local damper actuator, so that the actual air volume approaches the set value.

[0167] The actual air volume V is re-collected after each adjustment. act Repeat the above control process until the deviation is less than the preset error tolerance.

[0168] It should be noted that the closed-loop feedback control mechanism can continuously adjust the output of the actuator according to the deviation between the actual air volume and the target air volume, ensuring the stable operation of the ventilation system; the PID control algorithm has a simple structure, fast response speed, and high adjustment accuracy, and is suitable for application scenarios in coal mines where oxygen concentration control requires rapid and precise control.

[0169] S7. Deploy edge computing nodes in key areas of the coal mine working face. Each node receives local sensor data and executes local control strategies, while maintaining communication and synchronization with the central control system to form a distributed collaborative control network.

[0170] Furthermore, edge computing nodes are deployed in key areas of the coal mine working face, with each node equipped with local oxygen concentration, wind speed, and gas concentration sensors;

[0171] Each node periodically collects local sensor data, executes local airflow adjustment strategies, and records the adjustment results;

[0172] Each node uploads the local adjustment results to the central control system, including the adjustment timestamp, air volume setpoint, and oxygen concentration change value;

[0173] The central control system receives data uploaded by each node, performs consistency checks and conflict detection, and identifies abnormal nodes.

[0174] When a conflict in regulation between nodes is detected, a coordination mechanism is activated, and the regulation result of the node with higher priority is selected as a reference.

[0175] The coordinated control strategy is then distributed to each edge node to form a unified distributed collaborative control network.

[0176] It should be noted that edge computing nodes have local data processing and decision-making capabilities, enabling them to complete local control tasks without relying on the central control system. This reduces system communication latency and the load pressure on the central node. At the same time, edge nodes maintain collaborative communication with the central control system, ensuring the consistency and optimization effect of the global control strategy.

[0177] S8. By integrating the control results and feedback information of each node through a distributed collaborative control network, the global control strategy is updated to optimize the overall ventilation efficiency.

[0178] Furthermore, a global control database is constructed by using the control results and feedback information uploaded by each node in the distributed collaborative control network;

[0179] Extract the control effect indicators for each node, including the increase in oxygen concentration, the adjustment range of air volume, and the change in energy consumption;

[0180] The weighted average method is used to comprehensively evaluate the regulation effect of each node, and the regional regulation weight coefficient w is calculated. i , where i is the node number;

[0181] Update the global control strategy based on the weighting coefficients and the control deviation:

[0182]

[0183] Where M is the total number of edge nodes, V i (t) represents the airflow setting value of the i-th node at time t;

[0184] Update the global airflow setting value V global(t+1) is sent to each edge computing node to guide the local control operations in the next cycle;

[0185] It should be noted that the update process of the global control strategy takes into account the control effect and weight allocation of each edge node, so as to avoid the abnormal behavior of a single node from having too great an impact on the overall system. By optimizing the air volume setpoint through the weighted average method, the coordination and energy efficiency of the entire mine ventilation system are improved, which helps to achieve intelligent and refined ventilation management.

[0186] This embodiment also provides a computer device suitable for monitoring and improving the oxygen concentration control method at a coal mine working face, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method for monitoring and improving the oxygen concentration control method at a coal mine working face as proposed in the above embodiment.

[0187] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0188] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the method for monitoring and improving oxygen concentration control in coal mine working faces as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0189] In summary, this invention constructs a complete technology chain from environmental perception and intelligent analysis to dynamic regulation through multi-source sensor collaborative acquisition, Kalman filter data correction, LSTM neural network trend prediction, airflow-oxygen concentration coupled modeling, closed-loop feedback control, and edge computing distributed collaborative regulation. This enables high-precision real-time monitoring and forward-looking regulation of oxygen concentration in coal mine working faces, improves the response speed and intelligent regulation level of ventilation systems, effectively protects the safety and health of underground workers, and optimizes ventilation energy consumption, resulting in significant safety benefits and energy-saving advantages.

[0190] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for monitoring and improving oxygen concentration control in coal mine working faces, characterized in that: include: Multi-source sensor collaborative acquisition technology is used to collect oxygen concentration, wind speed, gas concentration and ambient temperature and humidity parameters in the coal mine working face to obtain raw environmental monitoring data; The Kalman filter algorithm is used to dynamically denoise and correct errors in the original environmental monitoring data to obtain a corrected environmental parameter sequence. Based on the corrected environmental parameter sequence, a time series input vector is constructed and input into the trained long short-term memory network model to output the predicted value of oxygen concentration change trend within a preset time period. Based on the predicted oxygen concentration change trend and the preset safety threshold range, determine whether the current ventilation strategy meets the oxygen concentration control requirements and generate a preliminary control instruction. A nonlinear regression modeling method is used to establish a mathematical mapping relationship between air volume input and oxygen concentration response, and an air volume-oxygen concentration coupled control model is constructed. Based on the air volume-oxygen concentration coupled control model, the preliminary control command is converted into the corresponding air volume adjustment target value; A closed-loop feedback control mechanism is adopted to dynamically adjust the operating frequency of the main ventilation fan or the opening of local dampers based on the target air volume adjustment value, so as to achieve precise control of oxygen concentration. Edge computing nodes are deployed in key areas of the coal mine working face. Each node receives local sensor data and executes local control strategies, while maintaining communication and synchronization with the central control system, forming a distributed collaborative control network. The distributed collaborative control network integrates the control results and feedback information of each node, updates the global control strategy, and optimizes the overall ventilation efficiency.

2. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 1, characterized in that: The process of using the Kalman filter algorithm to dynamically denoise and correct errors in the original environmental monitoring data to obtain a corrected environmental parameter sequence specifically includes: The Kalman filter algorithm was used to perform state estimation on the oxygen concentration, wind speed, gas concentration and environmental temperature and humidity parameters in the original environmental monitoring data. Construct the state equations and observation equations for each parameter, where the state equations are expressed as follows: X k =AX k-1 +Bu k +w k ; The observation equation is expressed as: Z k =HX k +v k ; Among them, X k Let A represent the state vector at time k, containing oxygen concentration, wind speed, gas concentration, and ambient temperature and humidity. Let A be the state transition matrix, B be the control input matrix, and u be the state vector. k To control the input, w k The noise is a process noise that follows a Gaussian distribution, Z. k Let v be the vector of actual observations at time k, H be the observation matrix, and v k To observe noise; Based on the above equations, the prediction and update steps are executed sequentially: Predict the current state: P k|k-1 =AP k-1|k-1 From T +Q; Update current state estimate: K k =P k|k-1 H T (HP k|k-1 H T +R) -1 ; P k|k =(I-K k H)P k|k-1 ; in, For prior estimation, For posterior estimation, P is the covariance matrix, Q is the process noise covariance, R is the observation noise covariance, and k is the k-th percentile. k Kalman gain; The obtained posterior estimates are output as a corrected sequence of environmental parameters.

3. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 2, characterized in that: The process involves constructing a time-series input vector based on the corrected environmental parameter sequence and inputting it into a trained long short-term memory network model. The model then outputs a predicted trend of oxygen concentration changes over a predetermined time period. Specifically, this includes: The corrected environmental parameter sequence is segmented using the sliding window method to construct a time series input vector x of length T. t =[x t-T+1 ,...,x t ]; Multiple input samples are sequentially fed into a pre-trained long short-term memory network model; The Long Short-Term Memory network model has completed weight initialization, forward propagation calculation, loss function optimization, and backpropagation update during the training phase. The model consists of multiple LSTM units, and each LSTM unit performs the following operations in sequence: Input vector x at the current time t Hidden state from the previous moment t-1 The splicing serves as the input for the gating mechanism; The forget gate selectively retains or forgets the cell state from the previous moment; The input gate determines the current information writing ratio; New cell state increments are generated from candidate states; After updating the unit state, the output gate controls the output of the currently hidden state; Finally, output the hidden state at the current moment. t ; The model output layer provides the final hidden state. t Perform a linear transformation to generate predicted oxygen concentration values ​​from time t+1 to t+n. During the model training phase, supervised learning is performed using historical monitoring datasets. The input is a sequence of historical environmental parameters, and the output is the actual oxygen concentration value within the target time period. During training, the mean squared error loss function is used to measure the deviation between the predicted value and the true value, and the model parameters are continuously adjusted through the gradient descent algorithm until the preset accuracy threshold is reached. After training, the model enters the deployment phase, where it receives real-time environmental parameter sequences and outputs predicted values ​​of oxygen concentration change trends.

4. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 3, characterized in that: The process involves determining whether the current ventilation strategy meets the oxygen concentration control requirements based on the predicted oxygen concentration change trend and a preset safety threshold range, and generating preliminary control instructions. Specifically, this includes: Predicted values ​​based on oxygen concentration change trends Compare with a preset safety threshold range; The safety threshold range is set to [O]. low O ig ], where O low The lower limit of oxygen concentration, O ig The upper limit; If any predicted value If the current ventilation strategy is deemed insufficient to maintain the required oxygen concentration, a ventilation enhancement command will be generated. If any predicted value The system determines that the current ventilation strategy has resulted in excessively high oxygen concentration and generates a ventilation reduction command. If all predicted values ​​are in [O low O ig If the area is within the specified range, the current ventilation strategy will remain unchanged. The generated preliminary control instructions include the ventilation enhancement instruction flag D. up Or ventilation reduction instruction flag D down This is for use in subsequent steps.

5. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 4, characterized in that: The nonlinear regression modeling method is used to establish a mathematical mapping relationship between air volume input and oxygen concentration response, and to construct an air volume-oxygen concentration coupled control model, specifically including: By adopting the air volume adjustment direction corresponding to the initial control command and combining historical operating data, a nonlinear functional relationship between air volume V and oxygen concentration response O is established. Select the historical dataset {(V1,O1),(V2,O2),...,(V N O N )}; Assume the function is a power-law function: O=a·V b +c; Where O is the oxygen concentration response value, V is the air volume input value, and a, b, and c are the parameters to be fitted. The least squares method is used to fit the model parameters to minimize the objective function. Solving for the optimal parameter a * ,b * ,c * The air volume-oxygen concentration coupled control model is obtained: The model is used to convert preliminary control commands into specific airflow adjustment target values.

6. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 5, characterized in that: The aforementioned closed-loop feedback control mechanism dynamically adjusts the operating frequency of the main ventilation fan or the opening of local dampers based on the target air volume adjustment value, thereby achieving precise control of oxygen concentration. Specifically, this includes: Using the target value V for air volume adjustment set As the setpoint for the closed-loop feedback control system; By real-time data collection of the main ventilation fan outlet air volume V act With the set value V set Compare and calculate the deviation e = V set -V act ; The control output u is calculated using the proportional-integral-derivative (PID) control algorithm, and its expression is: Among them, K p K i K d These are the proportional, integral, and derivative coefficients of the PID controller, respectively. The control output u is used to adjust the frequency of the main ventilation fan inverter or the opening of the local damper actuator, so that the actual air volume approaches the set value. The actual air volume V is re-collected after each adjustment. act Repeat the above control process until the deviation is less than the preset error tolerance.

7. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 6, characterized in that: The deployment of edge computing nodes in key areas of the coal mine working face, where each node receives local sensor data and executes local control strategies while maintaining communication and synchronization with the central control system, forms a distributed collaborative control network, specifically including: Edge computing nodes are deployed in key areas of the coal mine working face, and each node is equipped with local oxygen concentration, wind speed and gas concentration sensors. Each node periodically collects local sensor data, executes local airflow adjustment strategies, and records the adjustment results; Each node uploads the local adjustment results to the central control system, including the adjustment timestamp, air volume setpoint, and oxygen concentration change value; The central control system receives data uploaded by each node, performs consistency checks and conflict detection, and identifies abnormal nodes. When a conflict in regulation between nodes is detected, a coordination mechanism is activated, and the regulation result of the node with higher priority is selected as a reference. The coordinated control strategy is then distributed to each edge node to form a unified distributed collaborative control network.

8. The method for monitoring and improving oxygen concentration control in coal mine working faces as described in claim 7, characterized in that: The process of integrating the control results and feedback information of each node through a distributed collaborative control network, updating the global control strategy, and optimizing overall ventilation efficiency specifically includes: A global control database is constructed by using the control results and feedback information uploaded by each node in the distributed collaborative control network; Extract the control effect indicators for each node, including the increase in oxygen concentration, the adjustment range of air volume, and the change in energy consumption; The weighted average method is used to comprehensively evaluate the regulation effect of each node, and the regional regulation weight coefficient w is calculated. i , where i is the node number; Update the global control strategy based on the weighting coefficients and the control deviation: Where M is the total number of edge nodes, V i (t) represents the airflow setting value of the i-th node at time t; Update the global airflow setting value V global (t+1) is sent to each edge computing node to guide the local control operations in the next cycle.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method for monitoring and improving oxygen concentration control in coal mine working faces as described in any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the method for monitoring and improving the oxygen concentration control of a coal mine working face as described in any one of claims 1 to 8.

Images