Mine pressure regulation and control optimization method for fully mechanized caving face of extra-thick coal seam

Abstract

The invention relates to the technical field of coal mining, in particular to an ultra-thick coal seam fully mechanized caving face mine pressure regulation and control optimization method. The method is used for solving the problems that in the mine pressure detection process, response to mine pressure fluctuation emergencies is slow, multi-dimensional data cannot be effectively combined for comprehensive analysis, and manual judgment errors are caused by excessive manual intervention. By monitoring the mine pressure of the fully mechanized caving face of the coal seam, displacement of a supporting structure and soil settlement data in real time and combining a Kalman filter and a neural network deep learning algorithm, the stability of a mine pressure system is accurately analyzed, the change trend of the mine pressure is predicted, early warning is conducted in advance, mine pressure regulation measures are optimized, and coal mine operation safety is ensured; a convolutional neural network and a recurrent neural network are adopted to process time domain and frequency domain feature data, automatic signal matching and control decision are achieved, the operation modes of a supporting structure and mechanical equipment are dynamically adjusted, the human decision complexity is reduced, and mine safety is ensured.

Description

Technical Field

[0001] This invention relates to the field of coal mining technology, specifically to an optimization method for controlling mine pressure in a fully mechanized longwall mining face of an extra-thick coal seam. Background Technology

[0002] With the increasing depth and difficulty of coal mining, fully mechanized longwall faces in extra-thick coal seams face increasingly complex mine pressure environments. During coal mining, fluctuations in mine pressure and instability in stress distribution can lead to accidents such as overload of support structures and mine collapses, seriously threatening the lives of coal miners and the normal operation of equipment. Therefore, how to monitor changes in mine pressure in real time and accurately, and make scientific adjustments and emergency responses based on this data, has become a key technical issue in coal mine safety management. Existing mine pressure monitoring methods mostly rely on traditional pressure sensors and mechanical equipment. These methods are often slow to respond to sudden fluctuations in mine pressure and cannot effectively combine multi-dimensional data for comprehensive analysis. Furthermore, excessive human intervention can easily lead to human error or delayed control measures. Therefore, adopting mine pressure control technology based on data processing and artificial intelligence, which can more accurately identify the changing trends of mine pressure fluctuations and stress distribution, predict abnormal mine pressure situations in advance, and automatically adjust relevant control measures, has become an important development direction for current coal mine safety management.

[0003] To address the aforementioned issues, it is necessary to propose an optimization method for controlling mine pressure in fully mechanized longwall mining faces of extra-thick coal seams. Summary of the Invention

[0004] The purpose of this invention is to solve the problems existing in the background art and to propose an optimization method for controlling mine pressure in fully mechanized longwall mining faces of extra-thick coal seams.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A method for optimizing mine pressure control in fully mechanized longwall mining faces of extra-thick coal seams includes the following steps:

[0007] Step 1: Data Acquisition and Real-time Monitoring;

[0008] Real-time monitoring of the mining pressure environment around the fully mechanized coal seam face, acquiring mining pressure data, support structure displacement data, and soil settlement data.

[0009] At preset time intervals, the pressure sensors arranged on the fully mechanized mining face are accessed to obtain real-time mine pressure data;

[0010] At preset time intervals, the displacement sensors arranged on the support structure of the fully mechanized mining face are accessed to obtain real-time displacement data of the support structure.

[0011] At preset time intervals, access GPS devices deployed within the soil of the fully mechanized excavation face to obtain real-time soil settlement data.

[0012] The collected mine pressure data, support structure displacement data, and soil settlement data are sent to step two.

[0013] Step 2: Data Processing and Mine Pressure Analysis;

[0014] The collected mine pressure data, support structure displacement data, and soil settlement data are filtered and noise is eliminated to obtain the filtering results. Then, time domain feature data, frequency domain feature data, and soil stress distribution data are extracted from the filtering results.

[0015] Data filtering and noise cancellation are performed using a Kalman filter to obtain the estimated state of each preset time interval of the Kalman filter output, thus obtaining the filtering result.

[0016] As a preferred embodiment of the present invention, time-domain feature data and frequency-domain feature data are extracted from the feature dataset, and the specific process is as follows:

[0017] Extract time-domain feature data and calculate the mean, variance, maximum, and minimum values ​​of the estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data with respect to the timestamp in the estimated state.

[0018] Specifically, it includes:

[0019] The estimated values ​​of mine pressure data with respect to timestamps are: x11, x12, x13, x14; x15, x16; x17, x18; x19, x20; x20, x21; x22, x23; x24; x25, x36; x37, x38; x38, x39; x30, x31; x32; x33; x34; x35, x36; x37, x38; x39, x30; x31, x30; x31, x32; x33; x34; x35, x36.

[0020] Generate a time-domain feature dataset: X1={x11, x12, ..., x34}.

[0021] Extract frequency domain feature data and use Fourier transform to estimate the state. The estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data are converted to the frequency domain to obtain the frequency domain data of the estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data, and the corresponding characteristic frequency amplitude, spectral energy, maximum amplitude, and minimum amplitude are obtained respectively.

[0022] Specifically, it includes:

[0023] The characteristic frequency amplitude y11, the spectral energy y12, the maximum amplitude y13, and the minimum amplitude y14 of the mine pressure data correspond to the mine pressure data. The characteristic frequency amplitude y21, the spectral energy y22, the maximum amplitude y23, and the minimum amplitude y24 of the support structure displacement data correspond to the support structure displacement data. The characteristic frequency amplitude y31, the spectral energy y32, the maximum amplitude y33, and the minimum amplitude y34 of the soil settlement data correspond to the soil settlement data.

[0024] Generate a frequency domain feature dataset: Y1={y11, y12, ..., y34}.

[0025] As a preferred embodiment of the present invention, the soil stress distribution state data is extracted, and the specific process is as follows:

[0026] By simplifying the empirical formulas of elasticity:

[0027]

[0028] Calculate the distance in three-dimensional space from the coordinates (x, y, z) of the pressure sensor's location. The estimated mineral pressure at the location , where α is the preset mine pressure attenuation coefficient.

[0029] As a preferred embodiment of the present invention, based on the inferred mineral pressure Build a mining pressure field for each individual mining pressure data point. Calculate the estimated mineral pressure for all spatial locations Regarding the inferred mineral pressure The average value is taken to obtain the mineral pressure field, and a spatial mineral pressure map is drawn.

[0030] Step 3: Artificial intelligence analysis and optimization decision-making;

[0031] By inputting time-domain feature data, frequency-domain feature data, and soil stress distribution into a neural network-based clustering algorithm, the algorithm analyzes and predicts the future trends of mine pressure and stress distribution, and obtains the mine pressure stability coefficient.

[0032] The sets (x11, x12, ..., x44) and (y11, y12, ..., y44) consisting of time-domain feature data and frequency-domain feature data are respectively input into the first convolutional neural network and the second convolutional neural network to obtain the first abstract feature data and the second abstract feature data.

[0033] The spatial mineral pressure map is input into a recurrent neural network to obtain the third abstract feature data.

[0034] The weighted average summation of all elements in the first, second, and third abstract feature data yields the mine pressure stability coefficient. The numerical range of the mine pressure stability coefficient is -100 to 100.

[0035] The closer the mine pressure stability coefficient is to 0, the closer the mine pressure system of the coal seam fully mechanized mining face is to equilibrium; the further the mine pressure stability coefficient deviates from 0, the more it indicates mine pressure instability. A positive value indicates that the cause of mine pressure instability is the consolidation pressure inside the soil; a negative value indicates that the cause of mine pressure instability is the vibration of external mechanical equipment and environmental influences.

[0036] Specifically, a mine pressure stability coefficient between 1 and 50 indicates relatively good mine pressure stability, but there may still be some fluctuations or instabilities. Mine pressure fluctuations may increase slightly, and the support structure may bear some additional pressure, requiring continuous monitoring and fine-tuning of control strategies.

[0037] A mine pressure stability coefficient between 51 and 100 indicates poor mine pressure stability, violent fluctuations in mine pressure, and significant stress on the support structure. The coal seam fully mechanized mining face may face a greater mine pressure risk, and there may be situations where the mine pressure exceeds the standard or the support structure is overloaded, requiring emergency control measures.

[0038] Specifically, a mine pressure stability coefficient between -1 and -50 indicates poor mine pressure stability, but the mine pressure system is experiencing a certain degree of adverse fluctuation or stress imbalance, potentially indicating localized stress concentration or deformation of the support structure. Smaller negative values ​​may be due to changes in the ore layer structure, external disturbances, or unstable equipment operation.

[0039] A mine pressure stability coefficient between -51 and -100 indicates that the mine pressure system is highly unstable, with extreme fluctuations in mine pressure. The support structure may be subject to significant deformation or damage risks. In this case, the safety of the mine pressure environment is seriously threatened, and coal seam collapse, equipment damage, and dangerous mine accidents may occur. Immediate control measures, mine pressure early warning, and risk management are required.

[0040] Step 4: Signal matching and execution control;

[0041] The execution signal is matched based on the mine pressure stability coefficient, the control decision is matched based on the execution signal, and then sent to the mechanical equipment and support structure.

[0042] The specific rules for matching execution signals based on the mine pressure stability coefficient are as follows:

[0043] If the value range of the mine pressure stability coefficient is 1 to 50, a fine-tuning signal for the support structure is generated, and the management personnel are reminded to ensure that the soil settlement is within a controllable range and to strengthen the data monitoring and alarm system.

[0044] If the value of the mine pressure stability coefficient is between 51 and 100, a strength enhancement signal for the support structure is generated, and the management personnel are reminded to optimize the mining path and increase the soil support force.

[0045] If the value range of the mine pressure stability coefficient is -50 to -1, a significant reinforcement signal for the support structure is generated, and a shutdown inspection signal for large mechanical equipment is generated.

[0046] If the value range of the mine pressure stability coefficient is -51 to -100, generate shutdown inspection signals for all mechanical equipment, generate personnel evacuation signals and mine safety alarms.

[0047] Step 5: Monitoring and Feedback of Execution Results;

[0048] During the execution of control decisions, changes in mine pressure data are continuously monitored, execution errors are calculated, and nonlinear characteristics are captured to obtain nonlinear feedback coefficients. These nonlinear feedback coefficients are then used to correct the mine pressure stability coefficients, thereby optimizing the execution results of control decisions.

[0049] After obtaining the mine pressure stability coefficient and its matching execution signal, real-time mine pressure data is continuously acquired. And calculate the execution error at each coordinate point (x, y, z) at timestamp k. Among them The safety limit for mine pressure at the preset coordinate point (x, y, z).

[0050] Furthermore, the nonlinear characteristics in the execution error are captured, and the specific process is as follows:

[0051] By preset formula

[0052] Among them For nonlinear feedback coefficients, where The execution error is the coordinate point (x, y, z) at the previous timestamp k-1.

[0053] α and β are preset adjustment parameters used to control the sensitivity of the error change rate to feedback.

[0054] Furthermore, the mine pressure stability coefficient is corrected based on the nonlinear feedback coefficient, and the correction formula is as follows:

[0055]

[0056] Among them This is the corrected mine pressure stability coefficient. This is the initial mine pressure stability coefficient.

[0057] T represents the generation time of the mine pressure stability coefficient, and N represents the number of pressure sensors.

[0058] The corrected mine pressure stability coefficient is sent to step four to re-perform the execution signal matching.

[0059] Compared with the prior art, the beneficial effects of the present invention are:

[0060] 1. This invention, by real-time monitoring of coal seam fully mechanized mining face data on mine pressure, support structure displacement, and soil settlement, combined with advanced data processing and filtering technologies such as Kalman filters and deep learning algorithms based on neural networks, can accurately analyze the stability of the mine pressure system and predict mine pressure change trends in real time. Through the extraction of time-domain and frequency-domain feature data and the calculation of soil stress distribution, it can effectively identify mine pressure fluctuations and stress distribution instabilities, provide early warnings, and optimize mine pressure control measures to ensure safe coal mine operations.

[0061] 2. This invention employs convolutional neural networks and recurrent neural networks to abstract and process time-domain and frequency-domain feature data. Through an automated signal matching and control decision generation system, it can automatically adjust control parameters such as the reinforcement degree of the support structure and the operating mode of mechanical equipment based on changes in the mine pressure stability coefficient. Automated control through signal matching not only reduces the complexity of human decision-making but also allows for dynamic adjustments based on real-time data feedback, ensuring the mine remains in a safe state throughout the mining process.

[0062] 3. This invention uses a mine pressure stability coefficient to precisely monitor the dynamic changes in the mine pressure environment, enabling timely detection of potential risks to the mine pressure system. When the mine pressure stability coefficient deviates from its normal value, the system automatically generates an emergency control signal and initiates corresponding risk response measures, such as reinforcing the support structure, shutting down machinery and equipment, or evacuating personnel. This method not only effectively prevents mine accidents but also minimizes the impact of mine pressure fluctuations on the coal seam longwall face and the surrounding environment, providing a more scientific risk management and emergency response mechanism for coal mine operations. Attached Figure Description

[0063] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings:

[0064] Figure 1 This is a flowchart of the method of the present invention;

[0065] Figure 2 This is a spatial mineral pressure map proposed in the embodiments of the present invention. Detailed Implementation

[0066] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] Please see Figure 1 As shown, a method for optimizing mine pressure control in a fully mechanized longwall mining face of an extra-thick coal seam includes the following steps:

[0068] Step 1: Data Acquisition and Real-time Monitoring;

[0069] Real-time monitoring of the mining pressure environment around the fully mechanized coal seam face, acquiring mining pressure data, support structure displacement data, and soil settlement data.

[0070] At preset time intervals, the pressure sensors located on the longwall face are accessed to obtain real-time mine pressure data. ;

[0071] At preset time intervals, the displacement sensors arranged on the support structure of the longwall face are accessed to obtain real-time displacement data of the support structure. ;

[0072] At preset time intervals, GPS devices deployed within the soil of the fully mechanized trenching face are accessed to obtain real-time soil settlement data. .

[0073] (x, y, z) represents the location coordinates of the pressure sensor, displacement sensor, or GPS.

[0074] Furthermore, each data point of mine pressure, support structure displacement, and soil settlement has a data stamp t, where t = t(0), t(1), ..., t(k). The data stamp represents the time of data collection and is an integer multiple of the preset time interval. t(0) is the initial timestamp with a value of 0; t(1) is one time interval of the preset time interval; and t(k) is k times the preset time interval, where k is the number of preset time intervals elapsed from the start of data collection until the current time.

[0075] The collected mine pressure data, support structure displacement data, and soil settlement data are sent to step two.

[0076] Step 2: Data Processing and Mine Pressure Analysis;

[0077] The collected mine pressure data, support structure displacement data, and soil settlement data are filtered and noise is eliminated to obtain the filtering results. Then, time domain feature data, frequency domain feature data, and soil stress distribution data are extracted from the filtering results.

[0078] Data filtering and noise reduction are performed using a Kalman filter. The Kalman filter formula is as follows:

[0079]

[0080] in, It is the estimated state of the k-th preset time interval; where k is the timestamp, and where... , and These are estimated values ​​for mine pressure data, support structure displacement data, and soil settlement data, respectively.

[0081] in It is the estimated state of the (k-1)th preset time interval;

[0082] in It is the Kalman gain for the k-th preset time interval, used to balance the weights between the predicted and actual measured values, and is obtained through the prediction error covariance matrix. , among them Let H be the prediction error covariance matrix for the (k-1)th preset time interval, representing the uncertainty of the state estimate at the previous time step. H is a preset observation matrix, where the elements are preset constants indicating how the estimated state is transformed into the measurement space. R is a preset penetration noise covariance matrix, representing the uncertainty or noise level of the measured values.

[0083] in It represents the actual measured values ​​of the mine pressure data, support structure displacement data, and soil settlement data at the k-th preset time interval, and... ;

[0084] Obtain the estimated state of each preset time interval of the Kalman filter output. The filtered result is obtained.

[0085] Furthermore, time-domain and frequency-domain feature data are extracted from the feature dataset. The specific process is as follows:

[0086] Extract time-domain feature data and calculate the estimated state. In the data, the estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data are expressed as mean, variance, maximum, and minimum with respect to timestamps.

[0087] Specifically, it includes:

[0088] The estimated values ​​of mine pressure data with respect to timestamps are: x11, x12, x13, x14; x15, x16; x17, x18; x19, x20; x20, x21; x22, x23; x24; x25, x36; x37, x38; x38, x39; x30, x31; x32; x33; x34; x35, x36; x37, x38; x39, x30; x31, x30; x31, x32; x33; x34; x35, x36.

[0089] Generate a time-domain feature dataset: X1={x11, x12, ..., x34}.

[0090] Extract frequency domain feature data and use Fourier transform to estimate the state. The estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data are converted to the frequency domain using the following formula:

[0091]

[0092] in The frequency domain data output by the Fourier transform includes the frequency domain data of the estimated values ​​of mine pressure data, support structure displacement data, and soil settlement data, where f is the data frequency component.

[0093] Among them The input is the estimated values ​​of the Fourier transform data of mine pressure, support structure displacement, and soil settlement, where t is the data timestamp component.

[0094] Where j is the imaginary unit.

[0095] The frequency domain data of the estimated values ​​of mine pressure data, support structure displacement data and soil settlement data are obtained, and the corresponding characteristic frequency amplitude, spectral energy, maximum amplitude and minimum amplitude are obtained respectively.

[0096] Specifically, it includes:

[0097] The characteristic frequency amplitude y11, the spectral energy y12, the maximum amplitude y13, and the minimum amplitude y14 of the mine pressure data correspond to the mine pressure data. The characteristic frequency amplitude y21, the spectral energy y22, the maximum amplitude y23, and the minimum amplitude y24 of the support structure displacement data correspond to the support structure displacement data. The characteristic frequency amplitude y31, the spectral energy y32, the maximum amplitude y33, and the minimum amplitude y34 of the soil settlement data correspond to the soil settlement data.

[0098] Generate a frequency domain feature dataset: Y1={y11, y12, ..., y34}.

[0099] Furthermore, the soil stress distribution data is extracted, and the specific process is as follows:

[0100] By simplifying the empirical formulas of elasticity:

[0101]

[0102] Calculate the distance in three-dimensional space from the coordinates (x, y, z) of the pressure sensor's location. The estimated mineral pressure at the location , where α is the preset mine pressure attenuation coefficient.

[0103] Please see Figure 2 As shown, based on the inferred mine pressure Build a mining pressure field for each individual mining pressure data point. Calculate the estimated mineral pressure for all spatial locations Regarding the inferred mineral pressure The average value is taken to obtain the mineral pressure field, and a spatial mineral pressure map is drawn.

[0104] Step 3: Artificial intelligence analysis and optimization decision-making;

[0105] By inputting time-domain feature data, frequency-domain feature data, and soil stress distribution into a neural network-based clustering algorithm, the algorithm analyzes and predicts the future trends of mine pressure and stress distribution, and obtains the mine pressure stability coefficient.

[0106] The sets (x11, x12, ..., x44) and (y11, y12, ..., y44) consisting of time-domain feature data and frequency-domain feature data are respectively input into the first convolutional neural network and the second convolutional neural network to obtain the first abstract feature data and the second abstract feature data.

[0107] The core convolutional layer operation formulas for the first and second convolutional neural networks are:

[0108]

[0109] in The output of the core convolutional layer of the first convolutional neural network is p1, where p1 is the first index of the first convolutional neural network and p2 is the second index. p1 = 1, 2, 3; p2 = 1, 2, 3, 4. The combination of the first and second indices represents an element in the time-domain feature dataset X, pointing to the mean, variance, maximum, and minimum values ​​of the estimated values ​​of the mine pressure data, support structure displacement data, and soil settlement data with respect to the timestamp, respectively. and These are the convolution weight factors and bias factors of the core convolutional layer of the preset first convolutional neural network, respectively;

[0110] Among them The output of the core convolutional layer of the second convolutional neural network is represented by c1, where c1 is the first index and c2 is the second index. c1 = 1, 2, 3; c2 = 1, 2, 3, 4. The combination of the first and second indices represents an element in the frequency domain feature dataset X2, pointing to the characteristic frequency amplitude, spectral energy, maximum amplitude, and minimum amplitude of the estimated values ​​for the mine pressure data, support structure displacement data, and soil settlement data, respectively. and These are the convolution weight factors and bias factors of the core convolutional layer of the preset second convolutional neural network, respectively;

[0111] The spatial mineral pressure map is input into a recurrent neural network to obtain the third abstract feature data.

[0112] The hidden state update formula in a recurrent neural network is:

[0113]

[0114] Among them Let be the updated value of the hidden state at time t, where and The first and second weights of the hidden state are used to hide the data state and input data, respectively. This is the update value of the hidden state at the previous time step t-1, where... The spatial mineral pressure map input at time t, where... For bias terms;

[0115] The weighted average summation of all elements in the first, second, and third abstract feature data yields the mine pressure stability coefficient.

[0116] It should be noted that the value range of the mine pressure stability coefficient is -100 to 100.

[0117] The closer the mine pressure stability coefficient is to 0, the closer the mine pressure system of the coal seam fully mechanized mining face is to equilibrium; the further the mine pressure stability coefficient deviates from 0, the more it indicates mine pressure instability. A positive value indicates that the cause of mine pressure instability is the consolidation pressure inside the soil; a negative value indicates that the cause of mine pressure instability is the vibration of external mechanical equipment and environmental influences.

[0118] Specifically, a mine pressure stability coefficient between 1 and 50 indicates relatively good mine pressure stability, but there may still be some fluctuations or instabilities. Mine pressure fluctuations may increase slightly, and the support structure may bear some additional pressure, requiring continuous monitoring and fine-tuning of control strategies.

[0119] A mine pressure stability coefficient between 51 and 100 indicates poor mine pressure stability, violent fluctuations in mine pressure, and significant stress on the support structure. The coal seam fully mechanized mining face may face a greater mine pressure risk, and there may be situations where the mine pressure exceeds the standard or the support structure is overloaded, requiring emergency control measures.

[0120] Specifically, a mine pressure stability coefficient between -1 and -50 indicates poor mine pressure stability, but the mine pressure system is experiencing a certain degree of adverse fluctuation or stress imbalance, potentially indicating localized stress concentration or deformation of the support structure. Smaller negative values ​​may be due to changes in the ore layer structure, external disturbances, or unstable equipment operation.

[0121] A mine pressure stability coefficient between -51 and -100 indicates that the mine pressure system is highly unstable, with extreme fluctuations in mine pressure. The support structure may be subject to significant deformation or damage risks. In this case, the safety of the mine pressure environment is seriously threatened, and coal seam collapse, equipment damage, and dangerous mine accidents may occur. Immediate control measures, mine pressure early warning, and risk management are required.

[0122] Step 4: Signal matching and execution control;

[0123] The execution signal is matched based on the mine pressure stability coefficient, the control decision is matched based on the execution signal, and then sent to the mechanical equipment and support structure.

[0124] The specific rules for matching execution signals based on the mine pressure stability coefficient are as follows:

[0125] If the value range of the mine pressure stability coefficient is 1 to 50, a fine-tuning signal for the support structure is generated, and the management personnel are reminded to ensure that the soil settlement is within a controllable range and to strengthen the data monitoring and alarm system.

[0126] If the value of the mine pressure stability coefficient is between 51 and 100, a strength enhancement signal for the support structure is generated, and the management personnel are reminded to optimize the mining path and increase the soil support force.

[0127] If the value range of the mine pressure stability coefficient is -50 to -1, a significant reinforcement signal for the support structure is generated, and a shutdown inspection signal for large mechanical equipment is generated.

[0128] If the value range of the mine pressure stability coefficient is -51 to -100, generate shutdown inspection signals for all mechanical equipment, generate personnel evacuation signals and mine safety alarms.

[0129] Step 5: Monitoring and Feedback of Execution Results;

[0130] During the execution of control decisions, changes in mine pressure data are continuously monitored, execution errors are calculated, and nonlinear characteristics are captured to obtain nonlinear feedback coefficients. These nonlinear feedback coefficients are then used to correct the mine pressure stability coefficients, thereby optimizing the execution results of control decisions.

[0131] After obtaining the mine pressure stability coefficient and its matching execution signal, real-time mine pressure data is continuously acquired. And calculate the execution error at each coordinate point (x, y, z) at timestamp k. Among them The safety limit for mine pressure at the preset coordinate point (x, y, z).

[0132] Furthermore, the nonlinear characteristics in the execution error are captured, and the specific process is as follows:

[0133] By preset formula

[0134] Among them For nonlinear feedback coefficients, where The execution error is the coordinate point (x, y, z) at the previous timestamp k-1.

[0135] α and β are preset adjustment parameters used to control the sensitivity of the error change rate to feedback.

[0136] Furthermore, the mine pressure stability coefficient is corrected based on the nonlinear feedback coefficient, and the correction formula is as follows:

[0137]

[0138] Among them This is the corrected mine pressure stability coefficient. This is the initial mine pressure stability coefficient.

[0139] T represents the generation time of the mine pressure stability coefficient, and N represents the number of pressure sensors.

[0140] The corrected mine pressure stability coefficient is sent to step four to re-perform the execution signal matching.

[0141] It should be noted that the significance of the nonlinear feedback coefficient lies in its ability to enhance the control system's response by amplifying the deviation of the mine pressure stability coefficient from zero if the execution effect is unsatisfactory or the mine pressure regulation effect is poor. For example, when the mine pressure system exhibits an increasing error trend during execution, the nonlinear feedback coefficient will increase the deviation of the corrected mine pressure stability coefficient from zero, further strengthening the regulation of mine pressure fluctuations. In this way, the system can react more quickly to unstable or abnormal states, avoiding a slow response that could lead to potential mine pressure instability or disaster risks.

[0142] It should be understood that the terms “comprising” and “including” used in this disclosure and claims indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0143] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. As used in this disclosure and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this disclosure and claims means any combination and all possible combinations of one or more of the associated listed items, and includes such combinations;

[0144] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for optimizing mine pressure regulation of a fully mechanized caving face in a super-thick coal seam, characterized in that, The method comprises the following steps: Step 1, data acquisition and real-time monitoring: real-time monitoring of the mine pressure environment around the fully mechanized coal mining face to obtain mine pressure data, support structure displacement data and soil settlement data; Step 2, data processing and mine pressure analysis: data filtering and noise elimination are performed on the collected mine pressure data, support structure displacement data and soil settlement data to obtain a filtering result, and time domain feature data, frequency domain feature data and soil stress distribution state data are extracted from the filtering result; Step 3, artificial intelligence analysis and optimization decision: the time domain feature data, frequency domain feature data and soil stress distribution data are input into a clustering algorithm based on a neural network for analysis and prediction to identify the future trend of mine pressure and stress distribution, and a mine pressure stability coefficient is obtained; Step 4, signal matching and execution control: execution signal matching is performed according to the mine pressure stability coefficient, control decision matching is performed according to the execution signal, and the mechanical equipment and support structure are sent; Step 5, execution result monitoring and feedback; During the control decision execution process, the change of the mine pressure data is continuously monitored, the execution error is calculated, the nonlinear characteristics are captured, and a nonlinear feedback coefficient is obtained; The mine pressure stability coefficient is corrected according to the nonlinear feedback coefficient to optimize the execution result of the control decision.

2. The pressure regulation and optimization method for fully mechanized caving face in a super-thick coal seam according to claim 1, characterized in that, The specific process of extracting the time domain feature data, frequency domain feature data and soil stress distribution state data from the filtering result is as follows: Data filtering and noise elimination are performed through a Kalman filter to obtain the estimated state of each preset time interval output by the Kalman filter, and the filtering result is obtained; The time domain feature data, frequency domain feature data and soil stress distribution state data are extracted according to the feature data set.

3. The pressure regulation and optimization method for fully mechanized caving face in super thick seam according to claim 2, characterized in that, The specific process of extracting the time domain feature data is as follows: The average value, variance, maximum value and minimum value of the estimated values of the mine pressure data, support structure displacement data and soil settlement data in the estimated state with respect to the time stamp are calculated, including: the average value x11 of the estimated values of the mine pressure data with respect to the time stamp, the variance x12 of the estimated values of the mine pressure data with respect to the time stamp, the maximum value x13 of the estimated values of the mine pressure data, the minimum value x14 of the estimated values of the mine pressure data, the average value x21 of the estimated values of the support structure displacement data with respect to the time stamp, the variance x22 of the estimated values of the support structure displacement data with respect to the time stamp, the maximum value x23 of the estimated values of the support structure displacement data, the minimum value x24 of the estimated values of the support structure displacement data, the average value x31 of the estimated values of the soil settlement data with respect to the time stamp, the variance x32 of the estimated values of the soil settlement data with respect to the time stamp, the maximum value x33 of the estimated values of the soil settlement data, and the minimum value x34 of the estimated values of the soil settlement data; A time domain feature data set X1 is generated: X1={x11, x12,..., x34}.

4. The pressure regulation and optimization method for fully mechanized caving face in a super thick coal seam according to claim 1, characterized in that, The specific process of extracting the frequency domain feature data is as follows: The estimated values of the mine pressure data, the supporting structure displacement data and the soil body settlement data in the estimated state are converted into the frequency domain by Fourier transform to obtain the frequency domain data of the estimated values of the mine pressure data, the supporting structure displacement data and the soil body settlement data, and the corresponding characteristic frequency amplitude, spectral energy, maximum amplitude and minimum amplitude are obtained respectively; including: the characteristic frequency amplitude y11 corresponding to the mine pressure data, the spectral energy y12 corresponding to the mine pressure data, the maximum amplitude y13 corresponding to the mine pressure data, the minimum amplitude y14 corresponding to the mine pressure data, the characteristic frequency amplitude y21 corresponding to the supporting structure displacement data, the spectral energy y22 corresponding to the supporting structure displacement data, the maximum amplitude y23 corresponding to the supporting structure displacement data, the minimum amplitude y24 corresponding to the supporting structure displacement data, the characteristic frequency amplitude y31 corresponding to the soil body settlement data, the spectral energy y32 corresponding to the soil body settlement data, the maximum amplitude y33 corresponding to the soil body settlement data and the minimum amplitude y34 corresponding to the soil body settlement data; The frequency domain feature data set Y1 is generated: Y1={y11, y12,..., y34}.

5. The pressure regulation and optimization method for fully mechanized caving face in ultra thick seam according to claim 2, characterized in that, The specific process of extracting the soil stress distribution state data is: Through the simplified elastic mechanics empirical formula: ， The distance between the position (x, y, z) of the pressure sensor and the position (x, y, z) of the pressure sensor in the three-dimensional space is calculated as The mine pressure at the position (x, y, z) is predicted as Wherein, α is a preset mine pressure attenuation coefficient; wherein The mine pressure environment at the position (x, y, z) is Constructing the pressure field, for each individual pressure data Constructing the pressure field, for each individual pressure data Calculating the pressure for all spatial positions ; for the pressure Taking the average, resulting in a pressure field, and plotting the spatial pressure map.

6. The pressure regulation and optimization method for fully mechanized caving face in a super thick coal seam according to claim 1, characterized in that, The clustering algorithm based on the neural network is specifically: The set (x11, x12,..., x44) and (y11, y12,..., y44) composed of the time domain feature data and the frequency domain feature data are respectively input into the first convolutional neural network and the second convolutional neural network to obtain the first abstract feature data and the second abstract feature data respectively; The spatial mine pressure graph is input into the recurrent neural network to obtain the third abstract feature data; The weighted average sum of all elements in the first, second and third abstract feature data is obtained to obtain the mine pressure stability coefficient.

7. The pressure regulation and optimization method for fully mechanized caving face in a super thick seam according to claim 1, characterized in that, The specific process of performing signal matching according to the mine pressure stability coefficient is: If the numerical range of the mine pressure stability coefficient is 1 to 50, a fine tuning signal of the supporting structure is generated; If the numerical range of the mine pressure stability coefficient is 51 to 100, a strength improvement signal of the supporting structure is generated; If the numerical range of the mine pressure stability coefficient is -50 to -1, a large-scale reinforcement signal of the supporting structure and a shutdown inspection signal of the large-scale mechanical equipment are generated; If the numerical range of the mine pressure stability coefficient is -51 to -100, a shutdown inspection signal of all mechanical equipment is generated, a personnel evacuation signal and a mine safety alarm signal are generated.

8. The pressure regulation and optimization method for a fully mechanized caving face in a super-thick coal seam according to claim 7, characterized in that, The specific process of calculating the nonlinear feedback coefficient is: After obtaining the mine pressure stability coefficient and the matched execution signal, real-time mine pressure data is continuously acquired , and the execution error of each coordinate point (x, y, z) at the time stamp k is calculated ; wherein is a preset mine pressure safety limit value at the coordinate point (x, y, z). The nonlinear characteristics in the execution error are captured, and the specific process is: by a preset formula , wherein is a non-linear feedback coefficient, wherein is the execution error at coordinate point (x, y, z) at previous timestamp k-1; Wherein, α and β are preset adjustment parameters; The mine pressure stability coefficient is corrected according to the nonlinear feedback coefficient.

9. The pressure regulation and optimization method for a fully mechanized caving face in a super-thick coal seam according to claim 8, characterized in that, The specific process of correcting the mine pressure stability coefficient according to the nonlinear feedback coefficient is: Through the correction formula: ， calculating a corrected rock pressure stability factor , is the initial rock pressure stability factor; T is the generation time of the mine pressure stability coefficient, and N is the number of pressure sensors; The corrected mine pressure stability coefficient is sent to step four to perform signal matching again.

Images