A Coal Mine Directional Grouting Control Method and System Based on Geological Features

Abstract

The present invention discloses a coal mine directional grouting control method and system based on geological features, which relates to the technical field of coal mine grouting; the coal mine directional grouting control method includes: obtaining the geological features of the target coal mine; evaluating the target coal mine to obtain the coal mine integrity index; constructing a pre-grouting model of the target coal mine based on the geological features of the target coal mine and the coal mine integrity index; performing directional grouting design to obtain a coal mine directional grouting model; performing two-stage analysis to obtain an optimized coal mine directional grouting control strategy for adjusting the initial coal mine directional grouting control parameters in the coal mine directional grouting model; the coal mine directional grouting control system includes a pre-grouting analysis module and a directional grouting regulation module; the pre-grouting model of the target coal mine constructed by the present invention is highly matched with the actual coal mine situation, and through the integrity index, the pre-grouting model can fully consider the overall situation of the coal mine; through directional grouting design, the grouting efficiency is improved, and the grouting time and cost are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mine grouting, and particularly to a method and system for controlling directional grouting in coal mines based on geological characteristics. Background Art

[0002] Currently, common problems in coal mine grouting technology include poor geological adaptability and extensive grouting path planning; traditional methods rely on empirical hole layout design, making it difficult to accurately match complex geological characteristics, resulting in uneven distribution of grouting materials and residues in local weak areas, thus easily leading to safety hazards such as water inrush and gas leakage; at the same time, traditional static grouting models do not consider the dynamic changes in geological conditions during construction, often resulting in excessive injection of grout into dense rock formations or failure to effectively cover high-risk areas, thereby causing waste of resources or project failures.

[0003] Therefore, it is urgent to integrate geological feature analysis and holistic assessment methods, construct a dynamic pre-grouting model based on geological changes, and optimize the grouting path in combination with bionic structures to achieve precise targeted delivery of grout; through a two-stage feedback control mechanism, dynamically adapt to changes in geological conditions, thereby significantly improving grouting efficiency and reliability, and ensuring the safety of coal mines and the efficient utilization of resources. Summary of the Invention

[0004] The present invention aims to provide a method and system for controlling directional grouting in coal mines based on geological characteristics, so as to improve the grouting efficiency and reliability of coal mines.

[0005] A method for controlling directional grouting in coal mines based on geological characteristics includes the following steps:

[0006] Obtain the geological characteristics of the target coal mine; evaluate the target coal mine based on the geological characteristics of the target coal mine to obtain the overall coal mine index; at the same time, construct a pre-grouting model for the target coal mine based on the geological characteristics of the target coal mine and the overall coal mine index;

[0007] Carry out directional grouting design on the pre-grouting model of the target coal mine in combination with a grouting bionic structure to obtain a directional grouting model for the coal mine; the directional grouting model for the coal mine contains initial control parameters for directional grouting in the coal mine;

[0008] Conduct two-stage analysis based on the directional grouting model for the coal mine and actual grouting reaction parameters to obtain an optimized control strategy for directional grouting in the coal mine; the two-stage analysis includes real-time grouting feedback analysis and predictive grouting feedback analysis;

[0009] The optimized control strategy for directional grouting in the coal mine is used to adjust the initial control parameters for directional grouting in the coal mine in the directional grouting model for the coal mine.

[0010] As a preferred technical solution of the present invention, the target coal mine is evaluated based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; meanwhile, the specific steps for constructing the pre-grouting model of the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index include:

[0011] The geological characteristics of the target coal mine include geological structure stability characteristics, geological safety coupling characteristics, geological environment impact degree characteristics, and geological three-dimensional characteristics;

[0012] Based on the geological structure stability characteristics, the structure stability assessment is carried out to obtain the grouting stability factor;

[0013] Based on the geological safety coupling characteristics, the grouting safety assessment is carried out to obtain the grouting safety factor;

[0014] Based on the geological environment impact degree characteristics, the environmental impact assessment is carried out to obtain the grouting impact degree factor;

[0015] The grouting stability factor, the grouting safety factor, and the grouting impact degree factor are subjected to feature fusion through the secondary weight formula to obtain the coal mine integrity index;

[0016] Based on the coal mine integrity index, the three-dimensional details of the geological three-dimensional characteristics are optimized to obtain the pre-grouting model of the target coal mine.

[0017] As a preferred technical solution of the present invention, the pre-grouting model of the target coal mine is subjected to directional grouting design in combination with the grouting bionic structure to obtain the coal mine directional grouting model; the specific steps for including the initial coal mine directional grouting control parameters in the coal mine directional grouting model include:

[0018] The grouting bionic structure includes N grouting simulation roots G n , n = 1, 2,..., N; based on the grouting simulation root G n Set the root growth vector S n , the environmental response parameter H n And the root random error parameter C n ; among them, the initial value of the root growth vector S n Is S n0 ; set the total growth time of the grouting simulation roots in the grouting bionic structure to T, t = 1, 2,..., T; t represents the growth time step of the grouting simulation root G n ;

[0019] Based on the grouting bionic structure and the pre-grouting model of the target coal mine, the root growth is simulated, and several grouting bionic structures are set to simulate the root growth synchronously;

[0020] Within the total time, according to the grouting simulation root G nThe growth time step is used to perform grouting evolution on several grouting bionic structures. In each grouting bionic structure, the random error parameter C of the grouting root system n fluctuates randomly based on the grouting spatio-temporal evolution equation;

[0021] When the growth time step reaches the total time, the grouting directional environment evolution result P is obtained n ; Based on the grouting directional environment evolution result P n Select the grouting bionic structure with the best growth state and construct a coal mine directional grouting model;

[0022] Set the initial coal mine directional grouting control parameters according to the coal mine directional grouting model.

[0023] As a preferred technical solution of the present invention, the specific steps for performing real-time grouting feedback analysis include:

[0024] The actual grouting reaction parameters include the slurry real-time solidification index, the slurry diffusion acoustic signal, and the grouting orifice pressure signal;

[0025] Perform heterogeneous reconstruction on the actual grouting reaction parameters to obtain the current slurry diffusion data;

[0026] Analyze based on the current slurry diffusion data and the pre-trained grouting effect evaluation model to obtain the current coal mine grouting result.

[0027] As a preferred technical solution of the present invention, the specific steps for performing predictive grouting feedback analysis include:

[0028] Analyze based on the current slurry diffusion data in combination with the coal mine slurry diffusion model. The coal mine slurry diffusion model includes a reinforcement learning prediction layer and a dynamic plasticity strategy layer;

[0029] The reinforcement learning prediction layer is used to make predictions based on the current slurry diffusion data to obtain the coal mine grouting prediction result;

[0030] The dynamic plasticity strategy layer is used to perform reverse parameter adjustment according to the coal mine grouting prediction result and the current coal mine grouting result to obtain an optimized coal mine directional grouting control strategy.

[0031] As a preferred technical solution of the present invention, the coal mine slurry diffusion model is constructed based on the random forest model.

[0032] A coal mine directional grouting control system based on geological characteristics includes:

[0033] The pre-grouting analysis module includes a model construction unit and an orientation design unit; the model construction unit is used to obtain the geological characteristics of the target coal mine; evaluate the target coal mine based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; at the same time, construct a pre-grouting model of the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index; the orientation design unit is used to perform directional grouting design on the pre-grouting model of the target coal mine in combination with the grouting bionic structure to obtain a coal mine directional grouting model; the coal mine directional grouting model contains initial coal mine directional grouting control parameters.

[0034] The directional grouting control module includes a double-layer control unit, which is used to perform two-stage analysis based on the coal mine directional grouting model and the actual grouting reaction parameters to obtain an optimized coal mine directional grouting control strategy; the two-stage analysis includes real-time grouting feedback analysis and predictive grouting feedback analysis; the optimized coal mine directional grouting control strategy is used to adjust the initial coal mine directional grouting control parameters in the coal mine directional grouting model.

[0035] The present invention has the following advantages:

[0036] 1. By constructing a pre-grouting model of the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index, the present invention can ensure a high degree of matching between the model and the actual coal mine situation. The integrity index comprehensively reflects various factors such as the geological stability and mining difficulty of the coal mine. Incorporating it into the model construction process enables the pre-grouting model to fully consider the overall situation of the coal mine; by performing directional grouting design on the pre-grouting model of the target coal mine in combination with the grouting bionic structure, the introduction of the grouting bionic structure can simulate the structural advantages of organisms, avoiding problems such as uneven slurry distribution and waste of slurry in areas that do not require reinforcement in traditional grouting methods, greatly improving the grouting efficiency and reducing the grouting time and cost.

[0037] 2. Through the real-time grouting feedback analysis in the two-stage analysis, the present invention can timely monitor various parameter changes during the grouting process, such as slurry pressure, flow rate, diffusion range, etc.; once it is found that the actual grouting reaction parameters deviate from the expected values, the grouting control parameters can be immediately adjusted according to the feedback information. This real-time feedback mechanism makes the grouting process more flexible and controllable, and can effectively respond to various emergencies. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a schematic structural diagram of a coal mine directional grouting control system based on geological characteristics adopted in Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0039] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention.

[0040] Example 1, a method for controlling directional grouting in coal mines based on geological characteristics, includes the following steps:

[0041] Obtain the geological characteristics of the target coal mine;

[0042] Evaluate the target coal mine based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; at the same time, construct a pre-grouting model for the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index, specifically including:

[0043] The geological characteristics of the target coal mine include geological structure stability characteristics, geological safety coupling characteristics, geological environment influence degree characteristics, and geological three-dimensional characteristics;

[0044] Conduct structural stability evaluation based on the geological structure stability characteristics to obtain the grouting stability factor;

[0045] Collect the geological exploration data of the coal mine, including coal seam thickness, rock strength, structural characteristics (such as faults, folds, etc.), fracture development degree, groundwater distribution, etc.; based on the geological exploration data, use stability analysis methods, such as geological mechanics models or numerical simulation methods, to evaluate the rock layer stability of the coal mine to obtain the geological structure stability characteristics; use the stability evaluation results to calculate the grouting stability factor, and conduct quantitative analysis considering the influence of structural stability on the grouting effect to obtain the grouting stability factor;

[0046] Conduct grouting safety evaluation based on the geological safety coupling characteristics to obtain the grouting safety factor;

[0047] Analyze the geological safety coupling characteristics of the coal mine, that is, the interaction between mining activities and geological conditions, mainly focusing on the possible impacts on coal seams, mine shafts, and surrounding environments during the mining process; according to the impacts of mining activities, evaluate the geological safety of the coal mine, using safety factor analysis methods or hazard assessment models to obtain the geological safety coupling characteristics; according to the geological safety coupling characteristics, calculate the grouting safety factor. Generally speaking, lower geological safety risks will result in higher grouting safety factors;

[0048] Conduct environmental impact assessment based on the geological environment influence degree characteristics to obtain the grouting influence degree factor;

[0049] Evaluate the impacts of coal mine mining activities on the surrounding environment, including factors such as groundwater changes, land subsidence, and environmental pollution; according to the coal mine environmental impact assessment results, use the multi-factor evaluation method to obtain the geological environment influence degree characteristics; use the environmental impact degree analysis model to evaluate the environmental impacts of grouting activities through data such as environmental pollutant concentrations and groundwater flow to obtain the grouting influence degree factor.

[0050] The grouting stability factor, grouting safety factor, and grouting influence factor are subjected to feature fusion through a secondary weight formula to obtain the coal mine integrity index, which specifically includes:

[0051] The secondary weight formula is ; i = 1, 2, 3, used to number the grouting stability factor, grouting safety factor, and grouting influence factor;

[0052] αi is the primary weight index, and β is the secondary weight adjustment index; Y1 represents the grouting stability factor, Y2 represents the grouting safety factor, Y3 represents the grouting influence factor, and Q represents the coal mine integrity index;

[0053] The method of setting the primary weight index is to subjectively evaluate the importance of these three factors by professional and technical personnel to give preliminary weight values. The scale method can be used to evaluate the importance of each factor in the overall model. At the same time, based on sufficient historical data, the relative importance of each factor is calculated through data analysis methods, so as to derive a comprehensive primary weight index;

[0054] The method of setting the secondary weight adjustment index is determined according to the coal mine's mining environment, geological conditions, mine safety requirements, and environmental protection requirements. If the geological structure stability requirements of the coal mine are relatively high, the value of the secondary weight adjustment index can be appropriately increased so that the weighted sum of the three occupies an important position in the final integrity index.

[0055] Based on the coal mine integrity index, three-dimensional detail optimization is carried out on the three-dimensional geological features to obtain the pre-grouting model of the target coal mine.

[0056] The three-dimensional geological features are a three-dimensional model composed of the coal mine geological structure, coal seam and ore body distribution, groundwater, mining area pressure, and other features of the target coal mine; according to the value of the coal mine integrity index, the geological stability, safety, and environmental impact of the coal mine are evaluated, and then the geological features in the model are adjusted; for example, in the coal mine area, areas with relatively poor integrity may require enhanced grouting and obtain a larger weight when deploying grouting holes; while areas with better integrity may optimize resource extraction and deploy fewer grouting holes to balance the grouting requirements; according to the Q value, the grouting volume, grouting method, and other mine stability parameters in specific areas of the three-dimensional model are adjusted. These adjustments help to enhance the safety during coal mine grouting and select the areas that most need grouting for three-dimensional detail optimization; for example, areas with relatively poor stability may require more grouting treatment, adjusting the grouting depth and coverage area; by locally optimizing the three-dimensional geological model, the stability of fault and fracture development areas can be improved, and the grouting effect can be enhanced; after the above optimizations are completed, the final pre-grouting model of the target coal mine is generated to ensure that the model can effectively guide grouting operations during actual mining and meet the requirements of geological safety, environmental protection, and economic benefits.

[0057] Set the overall coal mine index to enable more reasonable allocation of grouting resources, avoid wasting grouting resources in unnecessary areas, and at the same time strengthen the grouting work in high-risk areas to improve the safety of coal mine mining; it can more accurately reflect the actual geological conditions of the coal mine, optimize the actual effect of the 3D model, and improve the effectiveness of the pre-grouting plan.

[0058] Combine the grouting bionic structure to conduct directional grouting design on the pre-grouting model of the target coal mine to obtain the coal mine directional grouting model; the coal mine directional grouting model contains initial coal mine directional grouting control parameters.

[0059] Combine the grouting bionic structure to conduct directional grouting design on the pre-grouting model of the target coal mine to obtain the coal mine directional grouting model; the specific steps for the coal mine directional grouting model to contain initial coal mine directional grouting control parameters include:

[0060] The grouting bionic structure contains N grouting simulation roots G n , n = 1, 2,..., N; based on the grouting simulation root G n Set the root growth vector S n , the environmental response parameter H n and the root random error parameter C n ; among them, the initial value of the root growth vector S n is S n0 ; set the total growth time of the grouting simulation roots in the grouting bionic structure to T, t = 1, 2,..., T; t represents the growth time step of the grouting simulation root G n ; the total time and the growth time step are the core parameters for controlling the dynamic growth process of the simulated roots. Its essence is to transform the continuous grouting process into computable iterative steps through mathematical discretization methods, so as to achieve refined prediction and control of the slurry diffusion path; the total time represents the expected cycle from the start of grouting to the complete filling of the target area, which is the abstract mapping of the total duration of the grouting project from start to end in the bionic model; the growth time step is the discretized time unit in the grouting bionic structure simulation. Each time step corresponds to a certain small time period in the actual grouting process. Within each time step, calculate the expansion direction, branch generation, and random perturbation of the grouting simulation roots, and interact with the geological conditions in real time;

[0061] In the grouting bionic structure, each grouting simulation root G n represents an independent grouting path branch, that is, the setting of the grouting hole, simulating the growth logic of plant roots, and its behavior is controlled by the following parameters:

[0062] S n0 is usually set to be along the direction of the maximum principal stress or the dominant orientation of the fracture, such as determined according to the ground penetrating radar data; the root growth vector Sn Its actual function is to guide the initial extension direction of the grouting holes, avoid blind hole layout, improve the coverage efficiency of the grout for key areas, and the environmental response parameter H n reflects the dynamic correlation between the grouting effect and geological conditions. Its actual function is to adjust the branch growth priority according to the real-time grouting effect, and prioritize the reinforcement of high-risk areas; the root random error parameter C n introduces controllable random perturbations through the time-space evolution equation to simulate the adaptive exploration ability of the roots to complex environments. Its actual function is to simulate the strain ability when encountering complex geological environments;

[0063] Based on the grouting bionic structure and the pre-grouting model of the target coal mine, simulate the root growth, and set several grouting bionic structures to simulate the root growth synchronously;

[0064] According to the grouting simulation root system G within the total time n Perform grouting evolution on several grouting bionic structures according to the growth time step of the root system. In each grouting bionic structure, the grouting root random error parameter C n Performs random fluctuations based on the grouting time-space evolution equation;

[0065] The equation of the grouting time-space evolution equation is ;

[0066] Among them, C0 represents the basic error parameter, tanh() is the hyperbolic tangent function, which is used for smoothing and limiting, reducing random perturbations, ensuring the stable extension of the main path in the high-pressure area; enhancing randomness in the low-pressure area to explore potential channels; ▽R(t) represents the pressure of the grouting simulation root system G at the current growth time step t n ; R0 is the initial simulation pressure; D(t) represents the depth of the grouting simulation root system G at the current growth time step t n ; e ─σD(t) represents the exponential decay amount, which is used to suppress the random fluctuations of deep branches; levy(t) represents the random noise obeying the levy distribution, which represents the uncertainty of root growth, and is used to simulate the accidental large deviation of the roots, such as breaking through local rock strata barriers in coal mines; γ, ε, and σ are adjustment parameters, which are set by professional technicians;

[0067] When the growth time step reaches the total time, the grouting directional environment evolution result P is obtained n ; Based on the grouting directional environment evolution result P n Select the grouting bionic structure with the best growth state to construct the coal mine directional grouting model;

[0068] Set the initial coal mine directional grouting control parameters according to the coal mine directional grouting model.

[0069] The process of directional grouting design for the target coal mine pre-grouting model based on the grouting bionic structure involves multi-parameter dynamic optimization of simulating root growth. As the growth time step progresses, the grouting simulated roots continuously expand in space and undergo spatio-temporal evolution. Finally, by evaluating the results of different grouting paths, several grouting directional environment evolution results are obtained, and each grouting directional environment evolution result corresponds to a specific grouting effect. At this time, the grouting path can be described as a highly adaptable and flexible network that can automatically adjust the pressure, depth, and path during the grouting process. By evaluating the results of different grouting paths, a grouting bionic structure with the best growth state is selected. The best structure usually refers to the one that can efficiently cover the key areas of the mining area while ensuring the grouting effect, especially those areas with high-risk, complex geological environments or concentrated fractures, so as to achieve the effect of directional grouting in the coal mine. Based on the selected grouting bionic structure with the best growth state, a directional grouting model for the coal mine is constructed. This model can optimize the grouting path and grouting volume according to geological conditions, mining progress, and safety requirements, thereby improving the safety and production efficiency of the coal mine. By simulating root growth through the bionic structure, blind hole layout is avoided, ensuring that the grouting holes can cover the key areas of the coal mine and improving the grouting effect. Simulating the adaptive exploration ability of the roots enables the grouting path to cope with complex geological environments and flexibly adjust the grouting direction and depth. By adjusting the grouting strategy in real time and strengthening the grouting density in high-risk areas, the geological stability and mining safety of the coal mine are ensured. Directional grouting can reduce unnecessary resource waste and improve the resource utilization efficiency of the coal mine on the premise of ensuring safety. This model can perform real-time simulation according to the actual situation, predict the grouting effect, and make adaptive adjustments in a dynamic environment, providing decision-making support for the long-term stable mining of the coal mine.

[0070] Based on the coal mine directional grouting model and actual grouting reaction parameters, a two-stage analysis is carried out to obtain an optimized control strategy for coal mine directional grouting. The two-stage analysis includes real-time grouting feedback analysis and predictive grouting feedback analysis.

[0071] The specific steps for carrying out real-time grouting feedback analysis include:

[0072] The actual grouting reaction parameters include the slurry real-time solidification index, slurry diffusion acoustic signal, and grouting hole mouth pressure signal.

[0073] The actual grouting reaction parameters are obtained based on sensors. The solidification index sensor is embedded in the slurry pipeline or drill hole to monitor chemical curing parameters such as slurry viscosity and temperature in real time. High-frequency acoustic receivers are arranged along the drill hole to capture the vibration signals generated by the slurry flow. At the same time, pressure sensors are installed at the outlet of the grouting pump and the drill hole mouth to record the grouting pressure fluctuation data.

[0074] The actual grouting reaction parameters are heterogeneously reconstructed to obtain the current slurry diffusion data.

[0075] The specific steps for heterogeneous reconstruction include:

[0076] The real-time setting index of the slurry reflects the setting speed and degree of the slurry, and can show the setting state of the slurry in the geological environment; the diffusion acoustic wave signal of the slurry reflects the diffusion behavior of the slurry during the injection process and the morphology of the injection channel; the pressure signal at the grouting orifice records the pressure change at the orifice during the grouting process, and is used to judge the flow state and flow resistance of the slurry;

[0077] Convert the data from different sources into the same format to ensure that they can be unified into the same data structure or standard for subsequent analysis; convert the acoustic wave signal, pressure signal and setting index into unified time series data, and combine the data from different sources through time alignment and spatial matching to fuse multi-dimensional data features, reconstruct the diffusion state of the slurry, reflect the expansion path and effect of the slurry at different positions and depths, and obtain the current diffusion data of the slurry; monitor the data change in real time during this process, and verify and adjust the reconstruction result according to the on-site feedback and known models to ensure the accuracy and real-time performance of the model, achieve accurate data output of grouting diffusion, and provide real-time support for the subsequent evaluation of grouting effect;

[0078] Analyze based on the current diffusion data of the slurry and the pre-trained grouting effect evaluation model to obtain the current result of coal mine grouting;

[0079] The training process of the pre-trained grouting effect evaluation model includes:

[0080] The training process of the pre-trained grouting effect evaluation model mainly includes steps such as data preparation, model selection, training and optimization, verification and adjustment, etc. First, a large amount of data during the historical grouting process needs to be collected, including grouting hole positions, grouting pressures, grouting times, slurry diffusion effects, mine geological characteristics, etc. These data will serve as the training set to help the model learn the relationship between grouting effects and geological conditions. Provide grouting effect labels for each historical sample, such as slurry coverage rate, diffusion depth, fracture filling degree, etc. These labels will serve as supervision signals to guide the model to learn how to evaluate grouting effects. Process the collected data through standardization, noise removal, missing value handling, etc. to ensure the quality of the data. The data can be normalized and adjusted to a unified scale to avoid the influence of differences in different data sources on the model effect. Select features that have an important impact on grouting effect evaluation from the historical grouting data, such as geological conditions, grouting pressure, flow rate, time step, etc. Based on the original data, generate some new features, such as calculating the velocity of slurry diffusion, spatial distribution characteristics of grouting holes, change trends of geological conditions, etc. These derived features help improve the accuracy of the model. Select the most important features through methods such as correlation analysis and principal component analysis to reduce redundant information and improve the model training efficiency. According to the nature of the task, select a suitable supervised learning algorithm for model training. Use the training set to train the model. During the training process, the model learns the relationship between features and grouting effects by minimizing the loss function. The training process requires continuous adjustment of model parameters to optimize the model performance. Evaluate the generalization ability of the model through cross-validation to avoid overfitting. Apply the trained grouting effect evaluation model. The model can predict and evaluate the current grouting effect based on real-time grouting response parameters.

[0081] The specific steps for conducting predictive grouting feedback analysis include:

[0082] Based on the current slurry diffusion data, analyze it in combination with the coal mine slurry diffusion model. The coal mine slurry diffusion model contains a reinforcement learning prediction layer and a dynamically plastic strategy layer. The coal mine slurry diffusion model is constructed based on the random forest model.

[0083] The reinforcement learning prediction layer is used to make predictions based on the current slurry diffusion data to obtain the coal mine grouting prediction result.

[0084] The dynamically plastic strategy layer is used to perform reverse parameter adjustment according to the coal mine grouting prediction result and the current coal mine grouting result to obtain an optimized coal mine directional grouting control strategy.

[0085] Introduce a multi-branch feature extractor in the reinforcement learning prediction layer to split the current diffusion data of the slurry. Use a graph attention network to model the topological relationship of coal mine directional grouting, and analyze the temporal change trend of the pressure field through a temporal convolutional network for prediction analysis; in the dynamic plastic policy layer, introduce a model-agnostic meta-learning framework to enable the policy layer to quickly adapt to new tasks. At the same time, design a dual-objective optimization function to minimize both the current error and future potential risks; based on the generation of basic parameters by the random forest model, optimize and adjust the parameters to output the optimal control strategy for optimizing coal mine directional grouting.

[0086] The coal mine slurry diffusion model is constructed based on the random forest model. This model enhances its prediction ability and adaptability by introducing a reinforcement learning prediction layer and a dynamic plastic policy layer; in the reinforcement learning prediction layer, the model uses a multi-branch feature extractor to split the current diffusion data of the slurry, models the topological relationship of coal mine directional grouting through a graph attention network, and analyzes the temporal change trend of the pressure field with the help of a temporal convolutional network to achieve accurate prediction of the slurry diffusion process; the dynamic plastic policy layer adjusts the parameters backward according to the prediction results and the current grouting results to optimize the grouting control strategy; in addition, the model also introduces a model-agnostic meta-learning framework to enable the policy layer to quickly adapt to new tasks and designs a dual-objective optimization function aimed at minimizing both the current error and future potential risks.

[0087] When designing a dual-objective optimization function in the coal mine directional grouting control strategy, the common design ideas of multi-objective optimization problems can be referred to. Specifically, the dual-objective optimization function needs to consider two objectives simultaneously: minimizing the current error and minimizing future potential risks; the current error can be measured by the difference between the actual measurement value and the predicted value during the grouting process, and the future potential risk can be evaluated by the long-term impact on the grouting process. Using the method of weighted sum, a dual-objective optimization function is defined.

[0088] When training a coal mine slurry diffusion model, basic parameters are first generated based on a random forest model. The random forest conducts classification or regression prediction by constructing multiple decision trees that introduce randomness to improve the accuracy and generalization ability of the model. During the training process, the Bayesian optimization method is used to tune the hyperparameters of the random forest model. Bayesian optimization constructs a prior distribution of the objective function and continuously updates the posterior distribution based on the existing observed data, thereby efficiently searching for the optimal combination of hyperparameters. In the reinforcement learning prediction layer, the model predicts and analyzes by monitoring the data during the slurry diffusion process in real time and combining a graph attention network and a temporal convolutional network. The dynamic plasticity strategy layer adjusts the parameters backward according to the prediction results and the current grouting results to optimize the grouting control strategy. Finally, based on the basic parameters generated by the random forest model, combined with the optimization results of the reinforcement learning prediction layer and the dynamic plasticity strategy layer, the parameters are further adjusted to output the optimal optimized coal mine directional grouting control strategy.

[0089] By introducing a multi-branch feature extractor to split the current slurry diffusion data, the influence of different features on slurry diffusion can be analyzed more meticulously. Combining the graph attention network to model the topological relationship of coal mine directional grouting can capture the propagation path and interaction of the slurry in complex geological structures, thereby improving the accuracy and reliability of the prediction. Using the temporal convolutional network to analyze the temporal variation trend of the pressure field can conduct real-time prediction of the dynamic changes during the slurry diffusion process. This time series analysis method can effectively capture the short-term fluctuations and long-term trends during the slurry diffusion process, providing more accurate decision-making support in the time dimension for grouting operations. Introducing a model-agnostic meta-learning framework into the dynamic plasticity strategy layer enables the strategy layer to quickly adapt to new tasks, which means that even when the geological conditions change or the grouting target is adjusted, the model can quickly adjust the strategy without having to train from scratch, greatly improving the adaptability and flexibility of the model. Designing a dual-objective optimization function to simultaneously minimize the current error and future potential risks. This optimization strategy not only focuses on the accuracy of the current grouting effect but also considers the potential risks that may occur in the future, making the grouting control strategy more robust and reducing the potential losses caused by geological condition changes or operation errors. Based on the basic parameters generated by the random forest model, combined with the optimization results of the reinforcement learning prediction layer and the dynamic plasticity strategy layer, the parameters are further adjusted to output the optimal optimized coal mine directional grouting control strategy. This hierarchical optimization method can make full use of the advantages of each layer and finally obtain a grouting control strategy with the optimal comprehensive performance.

[0090] In this embodiment, for example, a target coal mine is located in a complex geological area with multiple rock layers, faults, and fissures. To ensure the safety and stability of coal mine mining, directional grouting reinforcement of the coal mine is required. The geological characteristics of this coal mine are complex, and traditional grouting methods are difficult to achieve the ideal reinforcement effect. Therefore, a coal mine directional grouting control method based on geological characteristics is decided to be adopted. Through geological exploration and borehole sampling, the rock layer stability of the coal mine is analyzed, including the compressive strength, elastic modulus, etc. of the rock. The interaction between the geological structure and the grouting process during coal mine mining is evaluated, and the influence of faults and fissures on grouting safety is analyzed. The potential impact of the grouting process on the surrounding ecological environment is evaluated, including the risk of groundwater pollution, surface subsidence, etc. Using 3D geological modeling software, a 3D geological model of the coal mine is constructed to detail the rock layer distribution, fault positions, and fissure trends. According to the overall integrity index of the coal mine, the 3D geological model is optimized in detail to obtain a pre-grouting model of the target coal mine. A bionic structure containing multiple grouting simulation roots is designed for model optimization to determine the coal mine directional grouting model. During the grouting process, the real-time solidification index of the grout, the grout diffusion acoustic signal, and the grouting hole pressure signal are collected in real time. The actual grouting reaction parameters collected are isomerically reconstructed to obtain the current grout diffusion data. Using a pre-trained grouting effect evaluation model and combining the current grout diffusion data, the current coal mine grouting result is analyzed. In the reinforcement learning prediction layer, a multi-branch feature extractor is used to split the current grout diffusion data, the topological relationship is modeled through a graph attention network, and the temporal change trend of the pressure field is analyzed through a temporal convolutional network to predict the future diffusion of the grout. In the dynamic plasticity strategy layer, reverse parameter adjustment is performed according to the prediction result and the current grouting result to optimize the grouting control strategy. According to the optimized grouting control strategy, the initial grouting control parameters in the coal mine directional grouting model are adjusted, such as grouting pressure, grouting time, grouting hole spacing, etc. Grouting operations are carried out according to the optimized grouting control strategy to ensure the efficiency and safety of the grouting process. The grouting process of this coal mine realizes precise control, the grout diffuses evenly, has a wide coverage range, and effectively reinforces the geological structure of the coal mine. Through the real-time feedback and prediction analysis mechanism during the grouting process, the grouting parameters can be adjusted in time, avoiding grout waste and the risk of geological disasters, improving the mining safety of the coal mine, and at the same time improving the mining efficiency.

[0091] Embodiment 2, a coal mine directional grouting control system based on geological characteristics, see Figure 1 shown, including:

[0092] The pre-grouting analysis module includes a model construction unit and an orientation design unit; the model construction unit is used to obtain the geological characteristics of the target coal mine; evaluate the target coal mine based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; at the same time, construct a pre-grouting model of the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index; the orientation design unit is used to combine the grouting bionic structure to carry out orientation grouting design on the pre-grouting model of the target coal mine to obtain a coal mine orientation grouting model; the coal mine orientation grouting model contains initial coal mine orientation grouting control parameters.

[0093] The orientation grouting regulation module includes a double-layer regulation unit, which is used to perform two-stage analysis based on the coal mine orientation grouting model and the actual grouting reaction parameters to obtain an optimized coal mine orientation grouting control strategy; the two-stage analysis includes real-time grouting feedback analysis and predicted grouting feedback analysis; the optimized coal mine orientation grouting control strategy is used to adjust the initial coal mine orientation grouting control parameters in the coal mine orientation grouting model.

[0094] It should be understood that for those of ordinary skill in the art, improvements or transformations can be made according to the above description, and all such improvements and transformations should fall within the protection scope of the appended claims of the present invention. The parts not described in detail in this specification belong to the prior art well-known to those of ordinary skill in the art.

Claims

1. A method for controlling directional grouting in coal mines based on geological features, characterized in that, It includes the following steps: Obtain the geological characteristics of the target coal mine; Evaluate the target coal mine based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; at the same time, construct a pre-grouting model for the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index; Carry out directional grouting design on the pre-grouting model of the target coal mine in combination with the grouting bionic structure to obtain the directional grouting model of the coal mine; The directional grouting model of the coal mine contains initial coal mine directional grouting control parameters; Conduct a two-stage analysis based on the directional grouting model of the coal mine and the actual grouting reaction parameters to obtain an optimized coal mine directional grouting control strategy; the two-stage analysis includes real-time grouting feedback analysis and predictive grouting feedback analysis; The optimized coal mine directional grouting control strategy is used to adjust the initial coal mine directional grouting control parameters in the directional grouting model of the coal mine; The specific steps of evaluating the target coal mine based on the geological characteristics of the target coal mine to obtain the coal mine integrity index; at the same time, constructing a pre-grouting model for the target coal mine based on the geological characteristics of the target coal mine and the coal mine integrity index include: The geological characteristics of the target coal mine include geological structure stability characteristics, geological safety coupling characteristics, geological environment impact degree characteristics, and geological three-dimensional characteristics; Conduct structural stability assessment based on the geological structure stability characteristics to obtain the grouting stability factor; Conduct grouting safety assessment based on the geological safety coupling characteristics to obtain the grouting safety factor; Conduct environmental impact assessment based on the geological environment impact degree characteristics to obtain the grouting impact degree factor; Carry out feature fusion on the grouting stability factor, grouting safety factor, and grouting impact degree factor through a two-level weight formula to obtain the coal mine integrity index; Conduct three-dimensional detail optimization on the geological three-dimensional characteristics based on the coal mine integrity index to obtain the pre-grouting model of the target coal mine; Carry out directional grouting design on the pre-grouting model of the target coal mine in combination with the grouting bionic structure to obtain the directional grouting model of the coal mine; the specific steps of the directional grouting model of the coal mine containing the initial coal mine directional grouting control parameters include: The grouting bionic structure contains N grouting simulation roots G n , where n = 1, 2, …, N; based on the grouting simulation root G n the root growth vector S is set n , the environmental response parameter H n and the root random error parameter C n ; among them, the initial value of the root growth vector S n is S n0 ; the total time for the growth of the grouting simulation roots in the grouting bionic structure is set to T, where t = 1, 2, …, T; t represents the growth time step of the grouting simulation root G n . Simulate root growth based on the grouting bionic structure and the pre-grouting model of the target coal mine, and set several grouting bionic structures to simulate root growth synchronously; Perform grouting evolution on a number of grouting bionic structures according to the growth time steps of the grouting simulated root system G within the total time. In each grouting bionic structure, the random error parameter C of the grouting root system n fluctuates randomly based on the grouting spatio-temporal evolution equation; n ​ When the growth time step reaches the total time, the evolution result P of the grouting directional environment is obtained n ; Based on the evolution result P of the grouting directional environment n Select the grouting bionic structure with the best growth state and construct a coal mine directional grouting model; Set the initial coal mine directional grouting control parameters according to the directional grouting model of the coal mine.

2. The coal mine directional grouting control method based on geological features according to claim 1, wherein, The specific steps of conducting real-time grouting feedback analysis include: The actual grouting reaction parameters include the slurry real-time solidification index, slurry diffusion acoustic signal, and grouting orifice pressure signal; Perform heterogeneous reconstruction on the actual grouting reaction parameters to obtain the current slurry diffusion data; Conduct analysis based on the current slurry diffusion data and the pre-trained grouting effect evaluation model to obtain the current coal mine grouting result.

3. The coal mine directional grouting control method based on geological features according to claim 2, characterized in that, The specific steps of conducting predictive grouting feedback analysis include: Conduct analysis based on the current slurry diffusion data in combination with the coal mine slurry diffusion model, and the coal mine slurry diffusion model contains a reinforcement learning prediction layer and a dynamic plastic strategy layer; The reinforcement learning prediction layer is used to make predictions based on the current slurry diffusion data to obtain the coal mine grouting prediction result; The dynamic plastic strategy layer is used to perform reverse parameter adjustment according to the coal mine grouting prediction result and the current coal mine grouting result to obtain the optimized coal mine directional grouting control strategy.

4. The method for controlling directional grouting in coal mines based on geological features according to claim 3, wherein The coal mine slurry diffusion model is constructed based on the random forest model.

5. A coal mine directional grouting control system based on geological features, characterized in that, The system is used to implement the coal mine directional grouting control method based on geological features as described in any one of the above claims 1-4, and includes: A pre-grouting analysis module, including a model construction unit and a directional design unit; the model construction unit is used to obtain the geological features of the target coal mine; evaluate the target coal mine based on the geological features of the target coal mine to obtain the coal mine integrity index; at the same time, construct a pre-grouting model of the target coal mine based on the geological features of the target coal mine and the coal mine integrity index; the directional design unit is used to carry out directional grouting design on the pre-grouting model of the target coal mine in combination with the grouting bionic structure to obtain a coal mine directional grouting model; the coal mine directional grouting model contains initial coal mine directional grouting control parameters; A directional grouting regulation module, including a double-layer regulation unit, which is used to perform two-stage analysis based on the coal mine directional grouting model and the actual grouting reaction parameters to obtain an optimized coal mine directional grouting control strategy; the two-stage analysis includes real-time grouting feedback analysis and predictive grouting feedback analysis; the optimized coal mine directional grouting control strategy is used to adjust the initial coal mine directional grouting control parameters in the coal mine directional grouting model.

Images