Coal-pillar-free self-formed roadway dynamic pressure zone management and control method based on intelligent prediction and dynamic regulation and control

Abstract

The invention discloses a coal-pillar-free self-formed roadway dynamic pressure zone management and control method based on intelligent prediction and dynamic regulation and control, and belongs to the technical field of coal mining, and the method comprises the steps: building a machine learning database, carrying out the feature correlation judgment, selection and standardization processing of data in the database, and obtaining a data set; and then establishing and optimizing a dynamic pressure area stable distance prediction model according to the data set, obtaining a final model through screening, adding target mine parameters into a final model test set to obtain a dynamic pressure area range prediction result, and performing model selection and design of support equipment, real-time monitoring of the stability of a dynamic pressure area roadway and dynamic adjustment of a control strategy according to a monitoring result. According to the coal-pillar-free self-forming roadway dynamic pressure area management and control method based on intelligent prediction and dynamic regulation and control, a scientific basis can be provided for dynamic pressure area support design and equipment type selection, the safety of workers is guaranteed, meanwhile, the monitoring efficiency is improved, the accuracy of monitoring data is ensured, and effective management and control of a dynamic pressure area roadway are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of coal mining, and in particular relates to a method for controlling dynamic pressure zones in a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation. Background Art

[0002] The pillar-free, self-forming roadway technology uses directional roof slitting to cut stress transfer between the roadway roof and the goaf roof, effectively improving the stress environment in the roadway's surrounding rock. It then leverages mine pressure to automatically form a new roadway for the next working face. This technology significantly improves resource recovery, eliminates the stress concentration in the surrounding rock caused by leaving coal pillars, and significantly promotes the sustainable development of coal mining.

[0003] Based on the temporal and spatial continuity of the pillar-free self-forming roadway, the surrounding rock structure of the roadway can be divided into three zones: the leading coal support zone, the lagging working face dynamic pressure zone, and the roadway stability zone. The dynamic pressure zone is located some distance behind the working face and the hydraulic support. Due to the collapse of the goaf roof and the movement of the overlying rock, the roadway within this area is unstable. Stable control of the roadway in the dynamic pressure zone is crucial to ensuring the success of pillar-free self-forming roadway.

[0004] Traditional techniques for determining the stability of roadways in dynamic pressure zones rely primarily on post-mining on-site monitoring of rock pressure, lacking corresponding pre-mining prediction methods. This results in pre-mining temporary support design and equipment selection relying primarily on engineering experience, lacking theoretical basis and scientific guidance. Because traditional on-site rock pressure monitoring is largely manual, it is significantly limited in terms of safety, efficiency, and accuracy, failing to meet modern requirements for high efficiency and precision. Furthermore, traditional techniques for controlling the surrounding rock of roadways in dynamic pressure zones often employ a single support method, making it difficult to adapt to the complex and changing demands of site environments. Therefore, a new approach is urgently needed. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for controlling the dynamic pressure zone of a self-forming coal-pillar-free tunnel based on intelligent prediction and dynamic regulation. This method can provide a scientific basis for the design of dynamic pressure zone support and equipment selection. While ensuring the safety of workers, it also improves monitoring efficiency, ensures the accuracy of monitoring data, and realizes effective control of dynamic pressure zone tunnels.

[0006] To achieve the above objectives, the present invention provides a method for controlling dynamic pressure zones in self-forming lanes without coal pillars based on intelligent prediction and dynamic regulation, comprising the following steps:

[0007] Step 1: Collect mining data and build a machine learning database;

[0008] Step 2: perform correlation determination and feature selection on the data features in the database in step 1, and perform standardization on the data features to obtain a processed data set;

[0009] Step 3: Using the data in the dataset in step 2, a prediction model for the stable distance of the dynamic pressure zone is established, and the prediction model is optimized to obtain a model to be evaluated;

[0010] Step 4: Screen the models to be evaluated in step 3 to obtain the final model;

[0011] Step 5: Before mining, relevant parameters of the target mine are added as samples to the test set of the final model in Step 4 to obtain the prediction results of the dynamic pressure zone range; support equipment selection and support design of the dynamic pressure zone are carried out based on the prediction results;

[0012] Step 6: During the mining process, the stability of the roadway in the dynamic pressure zone is monitored in real time, and the control strategy is dynamically adjusted according to the monitoring results.

[0013] Preferably, in step 1, a machine learning database is established, and mine-related data is collected through mine surveys and literature summaries, including mining depth, coal seam thickness, coal seam inclination, top cutting height, top cutting angle, roof rock hardness, tunnel height, tunnel width, working face inclination length, working face strike length, and dynamic pressure stability distance.

[0014] Preferably, in step 2, the correlation determination and feature selection of data features include using 3D modeling software FLAC 3D Orthogonal simulation experiments are carried out to obtain parameters with a high correlation with the dynamic pressure stability distance; the data features are standardized, including using the Box-Cox transformation technology to standardize the feature engineering database.

[0015] Preferably, in step 3, the prediction model, including an artificial neural network model, continuously reduces the error through iteration to meet the set learning objectives; the optimization processing of the prediction model includes optimizing the prediction model using a particle swarm optimization algorithm and a genetic optimization algorithm.

[0016] Preferably, in step 4, the model to be evaluated in step 3 is screened, including using correlation coefficient, mean absolute error, mean absolute percentage error and root mean square error evaluation indicators to screen the model.

[0017] Preferably, step 5 also includes designing temporary support for the tunnel in the dynamic pressure zone based on the prediction results.

[0018] Preferably, in step 6, the stability of the dynamic pressure zone tunnel is monitored in real time, including using an optical fiber intelligent monitoring device to monitor the stress and displacement of the dynamic pressure zone tunnel in real time; the dynamic adjustment control strategy includes timely adjusting the temporary support design of the dynamic pressure zone according to the monitoring results.

[0019] A coal pillar-free self-forming roadway dynamic pressure zone management device based on intelligent prediction and dynamic regulation, comprising:

[0020] Data collection module, used to collect mine mining data and build a machine learning database;

[0021] A data feature processing module, connected to the data collection module, is used to determine the relevance and select features of the data features in the database, and to perform standardization on the data features to obtain a processed data set;

[0022] A model training module is connected to the data feature processing module, uses the data in the data set to establish a prediction model for the stable distance of the dynamic pressure zone, and optimizes the prediction model to obtain a model to be evaluated;

[0023] A model screening module, connected to the model training module, is used to screen the model to be evaluated and obtain a final model;

[0024] A result prediction module is connected to the model screening module and is used to add the relevant parameters of the target mine as samples to the test set of the final model before mining to obtain the prediction results of the dynamic pressure zone range; and select the support equipment and support design for the dynamic pressure zone according to the prediction results;

[0025] The real-time monitoring module is connected to the result prediction module and is used to monitor the stability of the dynamic pressure zone roadway in real time during the mining process and dynamically adjust the control strategy according to the monitoring results.

[0026] Preferably, the result prediction module further includes a temporary support design submodule for designing temporary support for the dynamic pressure zone tunnel according to the prediction results.

[0027] Preferably, the computer program implements the steps of any one of the above methods when executed by a processor.

[0028] Therefore, the present invention adopts the above-mentioned method for controlling the dynamic pressure zone of the self-forming lane without coal pillars based on intelligent prediction and dynamic regulation. Compared with the existing technology, the present invention has the following significant beneficial effects:

[0029] (1) The method used in the present invention can predict the stable distance of the dynamic pressure zone of the self-forming roadway without coal pillars before mining, providing scientific guidance for the support design and equipment selection of the dynamic pressure zone;

[0030] (2) The method adopted by the present invention eliminates manual monitoring underground through optical fiber intelligent monitoring, which can ensure worker safety, improve efficiency, and ensure monitoring accuracy;

[0031] (3) The method adopted by the present invention can realize dynamic control of the tunnel in the dynamic pressure zone, and realize safe and efficient mining of the tunnel without coal pillars.

[0032] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A schematic diagram of the zoning of a non-pillar self-forming lane according to the method for controlling dynamic pressure zones of a non-pillar self-forming lane based on intelligent prediction and dynamic regulation of the present invention;

[0034] Figure 2 This is a flow chart of an embodiment of a method for controlling dynamic pressure zones in a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation according to the present invention;

[0035] Figure 3 Parameter diagram of the orthogonal simulation experimental method of the method for controlling the dynamic pressure zone of the self-forming lane without coal pillars based on intelligent prediction and dynamic regulation of the present invention;

[0036] Figure 4 Schematic diagram of an artificial neural network model of the method for controlling dynamic pressure zone of a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation of the present invention;

[0037] Figure 5 This is a flow chart of the artificial neural network learning process of the method for controlling the dynamic pressure zone of the self-forming coal-pillar-free lane based on intelligent prediction and dynamic regulation of the present invention;

[0038] Figure 6 The PSO-ANN dynamic pressure stability distance prediction model of the present invention's method for controlling dynamic pressure zones in a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation;

[0039] Figure 7 The GO-ANN dynamic pressure stability distance prediction model of the present invention's method for controlling dynamic pressure zones in a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation;

[0040] Figure 8 This is a schematic diagram of the temporary support of a single pillar and π-shaped beam in the method for controlling the dynamic pressure zone of a self-forming lane without coal pillars based on intelligent prediction and dynamic regulation of the present invention, wherein Figure 8 (a) in the figure shows the temporary support cross section of a single pillar and a π-beam. Figure 8 (b) shows the axial cross-section of the temporary support of the single pillar and π-beam. Figure 8 (c) shows the axial plane view of the temporary support of the single pillar and π-beam;

[0041] Figure 9 This is a schematic diagram of temporary support for unit supports in the method for controlling dynamic pressure zones of self-forming lanes without coal pillars based on intelligent prediction and dynamic regulation of the present invention, wherein Figure 9 (a) in the figure shows the temporary support cross section of the unit support. Figure 9 (b) in the figure shows the axial cross-section of the temporary support of the unit support. Figure 9 (c) in the figure shows the axial plane view of the temporary support of the unit bracket.

[0042] Reference numerals

[0043] 1. Mining working face; 2. Hydraulic support on the working face; 3. Advance coal body support area; 4. Dynamic pressure area on the lagging working face; 5. Stable zone for tunnelling; 6. Cutting line; 7. Goaf; 8. Single pillar; 9. π-shaped beam; 10. Unit support; 11. Continuous working face; 12. Constant resistance anchor cable; 13. Gangue retaining support. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. Unless otherwise defined, the technical terms or scientific terms used in the present invention should be the common meanings understood by people with ordinary skills in the field to which the present invention belongs.

[0045] Example 1

[0046] like Figure 1 As shown, the mining face 1 is the working area where coal mining is ongoing. A working face hydraulic support 2 is installed on one side of the mining face 1 to support the roof and protect the safety of workers and equipment. Based on the temporal and spatial continuity of the pillarless self-forming tunneling process, the surrounding rock structure of the tunnel can be divided into three areas: the advanced coal body support area 3, the lagging working face dynamic pressure area 4, and the tunneling stability area 5. The advanced coal body support area 3 is located in the unmined coal body area in front of the mining face 1. The lagging working face dynamic pressure area 4 is located behind the mining face 1 and the working face hydraulic support 2, and the tunneling stability area 5 is located behind the lagging working face dynamic pressure area 4. The roof cutting line 6 marks the location of the directional cut in the roof. The goaf 7 is located behind the mining face 1 and is the area where coal mining has been completed.

[0047] like Figure 2 As shown, the embodiment process of the method for controlling the dynamic pressure zone of the self-forming lane without coal pillars based on intelligent prediction and dynamic regulation of the present invention includes the following steps:

[0048] Step 1: Through mine research and literature review, collect relevant data from multiple mines, including mining depth, coal seam thickness, coal seam inclination, top cutting height, top cutting angle, roof rock hardness, roadway height, roadway width, working face inclination length, working face strike length, and dynamic pressure stability distance. Organize the collected data, filter out abnormal data, and build a machine learning database.

[0049] Step 2: Use 3D modeling software FLAC 3D Orthogonal simulation tests were carried out to analyze the effects of mining depth, coal seam thickness, coal seam inclination, top cutting height, top cutting angle, roof rock hardness, roadway height, and roadway width on the dynamic pressure stability distance. Figure 2 As shown in the figure, factors related to the dynamic pressure stability distance are selected to conduct multi-factor (m) multi-level (n) orthogonal simulation experiments. m Through simulation experiments, we obtained parameters that have a strong correlation with the dynamic pressure stability distance.

[0050] After the simulation determines the relevant parameters, they are used as input parameters for data standardization. The input variable database is preprocessed using the Box-Cox transformation technique. The Box-Cox transformation normalizes samples of the same type, concentrating scattered data in a small area, greatly improving classification accuracy. Since the transformation is performed on samples of the same type, the independence of the data can be guaranteed. After the transformation, the distribution of each type of data is closer to normality. The general form of the Box-Cox transformation is:

[0051]

[0052] Among them, y is a continuous variable, λ is a transformation parameter, different λ corresponds to different transformation methods, and the specific transformation method can be determined by solving the λ value. The estimation method of λ value can be used to use maximum likelihood estimation, y (λ) is the new data value obtained after applying the Box-Cox transformation.

[0053] Step 3: Establish a prediction model for the stable distance of the dynamic pressure zone using an artificial neural network model (ANN), which includes an input layer, a hidden layer, and an output layer. Figure 3 As shown, ξ k represents the input unit, C j represents the hidden unit, O i Represents the output unit. jk and W ijis the connection weight. w = {W, w} represents all connection permissions. Output unit subscript i = 1, 2…, I; hidden unit subscript j = 1, 2…, J; input unit subscript k = 1, 2…, K. In addition, u represents different input modes; P represents the number of input modes, u = 1, 2…, P; g1 and g2 correspond to the activation functions of the appropriately selected hidden layer and output layer, respectively. Under a given input mode u, the input of hidden unit j is:

[0054]

[0055] The output is:

[0056]

[0057] in, is the output value of the jth hidden unit for the uth input pattern, w jk is the connection weight from the input layer to the hidden layer unit j, is the value of the k-th input feature under the u-th input mode, and the input of output unit i is:

[0058]

[0059] in, is the input value of the i-th hidden unit for the u-th input pattern, W ij is the connection weight from hidden layer unit j to output layer unit i, and the final output result for:

[0060]

[0061] In the data processing process, the forward propagation of the signal and the back propagation of the error are iterative. In this process, the current weight w is modified to obtain a new weight △w. The error is continuously reduced through iteration to meet the set learning goal. The learning process flow chart of the BP neural network is as follows Figure 4 shown.

[0062] On this basis, the particle swarm optimization (PSO) and genetic optimization (GO) algorithms are introduced to optimize the model. The PSO algorithm dynamically adjusts the optimal solution of each particle and the global optimal solution to improve the candidate solution. In the PSO algorithm, the solution to the problem is a particle. The set of n particles in D-dimensional space is X. In each iteration, the update formula is:

[0063]

[0064] Where z is the inertia weight; 1≤d≤D; 1≤h≤n; t is the number of iterations; v hd is the velocity of the particle, xhd is the position of the particle; c1 and c2 are learning factors, c1=c2=2; r1 and r2 are random numbers uniformly distributed in [0,1].

[0065] like Figure 5 As shown in Figure 2, the establishment process of the PSO-ANN model is as follows:

[0066] Step 301, determining the neural network structure;

[0067] Step 302, initializing the weights and thresholds of the ANN;

[0068] Step 303: Establish the relationship between PSO, weights, and thresholds, and initialize PSO;

[0069] Step 304, calculate the particle fitness value and update the position and velocity of each particle;

[0070] Step 305: Calculate the error of the PSO-ANN. If the error meets the requirements, proceed to the next step. If not, repeat step 304.

[0071] Step 306: Output the optimization results, train the ANN model, and obtain the PSO-ANN structure model.

[0072] The GO algorithm has multiple search paths, which are easy to parallelize and further improve the efficiency of the algorithm. During the iteration process, an adaptive crossover probability is used. During the calculation process, the crossover probability is automatically adjusted according to the value of the fitness function. The calculation formula is as follows:

[0073]

[0074] Among them, Y s is the adaptive crossover probability; f max is the maximum fitness function value of individuals in the population; f avg is the average value of the fitness function of each generation of the population; f is the value of the fitness function that produces two individuals crossing, one of which is larger; Q1, Q2 are constants in the interval from 0 to 1.

[0075] Adaptive mutation probability Y m The calculation is as follows:

[0076]

[0077] like Figure 6 As shown in the figure, the process of establishing the GO-ANN structure is as follows:

[0078] Step 307, establishing the ANN structure;

[0079] Step 308, using ANN to train the data to generate an initial population;

[0080] Step 309, initial encoding and preprocessing of the GO algorithm;

[0081] Step 310, performing fitness calculation using different initial values;

[0082] Step 311, gene manipulation processing;

[0083] Step 312: Obtain the optimal weight and threshold. If the optimal value is not reached, repeat steps 309 and 310.

[0084] Step 313: Assign values to the neural network, find the global optimal solution, and establish a GO-ANN structure model.

[0085] Step 4: Evaluate the two optimization models using the correlation coefficient (R 2 ), mean absolute error (MAE), mean absolute percentage error (MAPE) and root mean square error (RMSE) were used to evaluate the reliability and accuracy of the model and to screen the optimal prediction model.

[0086] Correlation coefficient (R 2 ) represents the proportion of the variance explained by the model to the total variance, and its value range is between 0 and 1. 2 The closer it is to 1, the better the model. The calculation formula is as follows:

[0087]

[0088] Where n is the number of samples; y i is the true value; is the predicted value, is the average of the true values.

[0089] The mean absolute error (MAE) is the average of the squared prediction errors. A smaller MAE indicates a better model. The calculation formula is as follows:

[0090]

[0091] The mean absolute percentage error (MAPE) is the absolute error between the predicted value and the actual value relative to the actual value. The smaller the MAPE, the better the model. The calculation formula is as follows:

[0092]

[0093] The root mean square error (RMSE) is the square root of the mean square error. The smaller the RMSE, the better. The calculation formula is as follows:

[0094]

[0095] Step 5: Before mining, the relevant parameters of the target mine are added as samples to the test set of the prediction model to obtain the predicted stable distance L of the dynamic pressure zone. The target mine is then designed for temporary support based on the dynamic pressure zone stable distance L output by the prediction model.

[0096] like Figure 8 As shown, one side of the single pillar 8 is the connecting working surface 11. The other side of the single pillar 8 is provided with a slag retaining support 13. A π-shaped beam 9 and a unit support 10 are provided above the single pillar 8. When the dynamic pressure zone is predicted to be stable within 200m of the lagging working surface, that is, when 0≤L≤200m, the single pillar 8 and the π-shaped beam 9 are used together for temporary support.

[0097] like Figure 9 As shown, when the dynamic pressure zone is predicted to be more than 200m behind the working surface and stable, that is, L>200m, a unit support 10 is used for temporary support. A constant resistance anchor cable 12 is set above the unit support 10.

[0098] The roof of the roadway in the dynamic pressure zone is reinforced with a constant resistance anchor cable 12. According to the stability distance of the dynamic pressure zone, a single pillar 8 or a unit bracket 10 is selected for temporary support. U-shaped steel is arranged on the side of the goaf 7 of the roadway in the dynamic pressure zone for gangue support 13. The gangue support 13 can effectively prevent the gangue from the goaf 7 from rushing into the roadway, and form a stable gravel roadway wall structure with the gangue from the goaf 7.

[0099] Step 6: Use a fiber-optic intelligent monitoring system to conduct real-time monitoring and analysis of changes in roof separation in the dynamic pressure zone, the amount of movement of the roof and floor plates, the stress on the constant resistance anchor cables 12, and the stress on the support equipment. Based on the analysis results, timely adjustments are made to the temporary support design for the roadway in the dynamic pressure zone. If the monitoring data in the dynamic pressure zone consistently exceeds the predicted value, the support strength is increased to ensure safety. If the monitoring data in the dynamic pressure zone consistently falls below the predicted value, the support strength is gradually reduced to conserve resources.

[0100] The optical fiber intelligent monitoring system monitors the roof separation. If the roof separation continues to increase, the number of anchor cables and grouting reinforcement can be added at the location with large roof separation.

[0101] The optical fiber intelligent monitoring system monitors the amount of movement of the roadway roof and floor. When the amount of movement of the roof and floor continues to increase, the number of anchor cables can be increased and single pillars 8 can be erected at the location where the roof and floor are moving closer.

[0102] The optical fiber intelligent monitoring system monitors the stress on the constant resistance anchor cable 12. When it is detected that the stress on the constant resistance anchor cable 12 exceeds the preset stress threshold of the constant resistance anchor cable 12, the spacing between the constant resistance anchor cables 12 can be reduced and additional anchor cables can be installed.

[0103] The optical fiber intelligent monitoring system monitors the lateral pressure of the gravel side. When the lateral pressure of the gravel side exceeds the preset threshold, the spacing between the U-shaped steel rows can be reduced.

[0104] The optical fiber intelligent monitoring system monitors the stress on the single pillar 8. When it is detected that the stress on the single pillar 8 exceeds the preset stress threshold, the spacing between the single pillars 8 can be reduced.

[0105] The optical fiber intelligent monitoring system monitors the stress on the unit bracket 10 , and when it is detected that the stress on the unit bracket 10 exceeds a preset stress threshold, the row spacing of the unit bracket 10 can be reduced.

[0106] Furthermore, when single-strut support 8 is used for temporary support, and severe mine pressure in the roadway is observed, the surrounding rock is severely deformed, and the constant-resistance anchor cable 12 fails, the unit support 10 is used to replace the single-strut support 8 for temporary support. When switching between single-strut support and unit support 10, the single-strut support is replaced in sections along the roadway direction, with each section less than 10m long. Single-strut support 8 on the side away from the working face where the roof has not been cut is removed first to avoid secondary disturbance of the surrounding rock. During the replacement process, temporary anchor cables are added to assist in stabilizing the roof.

[0107] Therefore, the present invention adopts the above-mentioned method for controlling the dynamic pressure zone of the self-forming coal-pillar-free tunnel based on intelligent prediction and dynamic regulation. This method can provide a scientific basis for the support design and equipment selection of the dynamic pressure zone. While ensuring the safety of workers, it also improves the monitoring efficiency, ensures the accuracy of the monitoring data, and realizes the effective control of the dynamic pressure zone tunnel.

[0108] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit the same. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical solutions of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for controlling dynamic pressure zones in self-forming lanes without coal pillars based on intelligent prediction and dynamic regulation, characterized in that: The following steps are involved: Step 1: Collect mining data and build a machine learning database; Step 2: perform correlation determination and feature selection on the data features in the database in step 1, and perform standardization on the data features to obtain a processed data set; Step 3: Using the data in the dataset in step 2, a prediction model for the stable distance of the dynamic pressure zone is established, and the prediction model is optimized to obtain a model to be evaluated; Step 4: Screen the models to be evaluated in step 3 to obtain the final model; Step 5: Before mining, relevant parameters of the target mine are added as samples to the test set of the final model in Step 4 to obtain the prediction results of the dynamic pressure zone range; support equipment selection and support design of the dynamic pressure zone are carried out based on the prediction results; Step 6: During the mining process, the stability of the roadway in the dynamic pressure zone is monitored in real time, and the control strategy is dynamically adjusted according to the monitoring results.

2. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: In step 1, a machine learning database is established, and mine-related data is collected through mine surveys and literature summaries, including mining depth, coal seam thickness, coal seam inclination, top cutting height, top cutting angle, roof rock hardness, tunnel height, tunnel width, working face inclination length, working face strike length, and dynamic pressure stability distance.

3. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: In step 2, the data features are correlated and selected, including the use of 3D modeling software FLAC 3D Orthogonal simulation experiments are carried out to obtain parameters with a high correlation with the dynamic pressure stability distance; the data features are standardized, including using the Box-Cox transformation technology to standardize the feature engineering database.

4. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: In step 3, the prediction model, including the artificial neural network model, continuously reduces the error through iteration to meet the set learning objectives; the optimization processing of the prediction model includes optimizing the prediction model using a particle swarm optimization algorithm and a genetic optimization algorithm.

5. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: In step 4, the model to be evaluated in step 3 is screened, including using correlation coefficient, mean absolute error, mean absolute percentage error and root mean square error evaluation indicators to screen the model.

6. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: Step 5 also includes designing temporary support for the roadway in the dynamic pressure zone based on the prediction results.

7. The method for controlling dynamic pressure zone of coal pillar-free self-forming roadway based on intelligent prediction and dynamic regulation according to claim 1 is characterized in that: In step 6, the stability of the dynamic pressure zone tunnel is monitored in real time, including using an optical fiber intelligent monitoring device to monitor the stress and displacement of the dynamic pressure zone tunnel in real time; the dynamic adjustment control strategy includes timely adjusting the temporary support design of the dynamic pressure zone according to the monitoring results.

8. A coal pillar-free self-forming lane dynamic pressure zone management device based on intelligent prediction and dynamic control, characterized in that: include: Data collection module, used to collect mine mining data and build a machine learning database; A data feature processing module, connected to the data collection module, is used to perform correlation determination and feature selection on the data features in the machine learning database, and to perform standardization on the data features to obtain a processed data set; A model training module is connected to the data feature processing module, uses the data in the data set to establish a prediction model for the stable distance of the dynamic pressure zone, and optimizes the prediction model to obtain a model to be evaluated; A model screening module, connected to the model training module, is used to screen the model to be evaluated and obtain a final model; A result prediction module is connected to the model screening module and is used to add the relevant parameters of the target mine as samples to the test set of the final model before mining to obtain the prediction results of the dynamic pressure zone range; and select the support equipment and support design for the dynamic pressure zone according to the prediction results; The real-time monitoring module is connected to the result prediction module and is used to monitor the stability of the dynamic pressure zone roadway in real time during the mining process and dynamically adjust the control strategy according to the monitoring results.

9. The coal pillar-free self-forming lane dynamic pressure zone management device based on intelligent prediction and dynamic control according to claim 8 is characterized in that: The result prediction module also includes a temporary support design submodule, which is used to design temporary support for the dynamic pressure zone tunnel according to the prediction results.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images