Coal production optimization method and system based on data mining

Abstract

The invention discloses a coal production optimization method and system based on data mining, and relates to the technical field of data mining, and the method comprises the steps: collecting multi-source safety monitoring data and historical accident data of a target coal production scene, and correspondingly constructing a dynamic graph neural network model according to the historical accident data; according to the multi-source safety monitoring data and the dynamic graph neural network model, a production control scheme is randomly selected in combination with the design working condition domain of the target scene to carry out iterative accident prediction, safety boundary analysis is carried out based on a fitting method, and a real-time production control safety boundary is obtained; and carrying out production parameter control on the target coal production scene by taking the real-time production control safety boundary as a constraint. The technical problem of poor safety of coal production parameter control in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of data mining, and in particular to a coal production optimization method and system based on data mining. Background Art

[0002] Coal production safety monitoring has long relied on static early warning systems based on fixed thresholds and manual decision-making based on experience. Existing technologies typically deploy sensor networks to collect single-dimensional data in real time and set static safety thresholds to trigger alarms. However, this approach has significant drawbacks: First, static thresholds cannot adapt to the complex and dynamic underground environment, resulting in high false alarm rates or omissions of critical risks; second, siloed analysis of multi-source data leads to inaccurate results. Furthermore, existing technologies cannot dynamically adapt to changing operating conditions, making it difficult to balance production efficiency and safety assurance. Summary of the Invention

[0003] The present application provides a coal production optimization method and system based on data mining, which is used to solve the technical problem of poor safety of coal production parameter control in the existing technology.

[0004] In view of the above problems, the present application provides a coal production optimization method and system based on data mining.

[0005] In a first aspect, the present application provides a coal production optimization method based on data mining, the method comprising: Collect multi-source safety monitoring data and historical accident data of the target coal production scene, and construct a dynamic graph neural network model based on the historical accident data; In combination with the multi-source safety monitoring data, an iterative prediction analysis is performed on the production control plan through the dynamic graph neural network model to obtain an iterative prediction analysis result, wherein the production control plan is determined based on a design operating condition domain of a target production scenario; Based on the iterative prediction analysis results and the preset safety boundary constraints, a production control scheme that satisfies the safety boundary constraints is selected as a safe production control scheme set, wherein the safety boundary constraints are determined based on the gas emission volume of the working face, and the gas emission volume of the working face includes an absolute emission volume and a relative emission volume; Performing fitting analysis on the production safety control scheme set to generate a real-time production control safety boundary; The production parameter control of the target coal production scenario is performed with the real-time production control safety boundary as a constraint.

[0006] In a second aspect, the present application provides a coal production optimization system based on data mining, comprising: A model building module is used to collect multi-source safety monitoring data and historical accident data of the target coal production scenario, and to build a dynamic graph neural network model based on the historical accident data; an iterative prediction module, configured to combine the multi-source safety monitoring data and perform iterative prediction analysis on the production control plan through the dynamic graph neural network model to obtain iterative prediction analysis results, wherein the production control plan is determined based on a design operating condition domain of a target production scenario; a scheme screening module, configured to screen, based on the iterative prediction analysis results and a preset safety boundary constraint, a production control scheme that satisfies the safety boundary constraint as a safe production control scheme set, wherein the safety boundary constraint is determined based on the gas emission rate of the working face, and the gas emission rate of the working face includes an absolute emission rate and a relative emission rate; A fitting analysis module, configured to perform fitting analysis on the production safety control scheme set to generate a real-time production control safety boundary; The production parameter control module is used to control the production parameters of the target coal production scenario based on the real-time production control safety boundary.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application proposes a data mining-based coal production optimization method and system. By integrating dynamic graph neural network modeling with collaborative analysis of multi-source safety data, it significantly improves the dynamic adaptability and safety of production control parameters. Compared with traditional methods, the technical solution provided by this application significantly overcomes the limitations of static threshold control, enabling the mining of spatiotemporal correlations between multi-dimensional safety indicators and the real-time collaborative optimization of production control parameters, achieving the technical effect of improving the safety of coal production control parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0009] Figure 1 A flow chart of a coal production optimization method based on data mining provided in an embodiment of the present application.

[0010] Figure 2 A structural diagram of a coal production optimization system based on data mining provided in an embodiment of the present application.

[0011] In the accompanying drawings, the components represented by the reference numerals are described as follows: Model building module 100, iterative prediction module 200, solution screening module 300, fitting analysis module 400, production parameter control module 500. DETAILED DESCRIPTION

[0012] This application provides a coal production optimization method and system based on data mining, which is used to solve the technical problem of poor safety of coal production parameter control in the existing technology.

[0013] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0014] It should be noted that the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0015] Example 1, as Figure 1 As shown, the present application provides a coal production optimization method based on data mining, wherein the method includes: S10: Collect multi-source safety monitoring data and historical accident data of the target coal production scene, and construct a dynamic graph neural network model according to the historical accident data.

[0016] Traditional methods for coal production safety monitoring rely on static models or independent analysis of single-source data, failing to effectively connect the dynamic coupling relationships among multiple factors underground. Existing technologies often store collected data in isolation, with historical accident records used only for retrospective analysis. This leads to overlooking the correlations between monitoring points, failing to capture accident transmission mechanisms such as equipment failure chains and stress migration chains, and causing risk prediction to lag behind actual production dynamics.

[0017] Step S10 in the method provided in the embodiment of the present application includes: Collecting multiple types of safety monitoring data during coal production in real time to generate the multi-source safety monitoring data; Synchronously obtain the accident type, accident location, and accident-related monitoring data recorded during the historical production cycle to form the historical accident data; The multi-type safety monitoring data includes at least gas concentration data, dust concentration data, surrounding rock stress data, equipment operation status data and geological environment monitoring data; According to the historical accident data, a dynamic graph neural network model is constructed, including: Map the safety monitoring points of the target coal production scenario into nodes in the graph structure; Based on the physical dependency, geological distribution, and historical accident correlation between the safety monitoring points, directed connection edges between nodes are constructed and corresponding edge weight matrices are generated to form the graph structure; Combining the graph structure with time series monitoring data and historical production control data in the historical accident data to perform model training to obtain the dynamic graph neural network model; Wherein, according to the historical accident data, a dynamic graph neural network model is constructed, which includes: Performing data enhancement processing on the historical accident data, wherein the data enhancement processing includes data expansion, data smoothing, and outlier correction; Based on the historical accident data after data enhancement processing, constructing a generative adversarial network model for accident data expansion, and generating synthetic accident data through the generative adversarial network model; The synthetic accident data is added to the historical accident data for training the dynamic graph neural network model.

[0018] In an embodiment of the present application, multiple types of safety monitoring data during the coal production process are collected in real time, such as generating multi-source safety monitoring data. The multiple types of safety monitoring data include at least: gas concentration data, such as oxygen concentration data collected using a gas sensor, in units of %; dust concentration data, such as dust particle concentration collected using a dust sensor, in units of mg / m³; surrounding rock stress data, such as surrounding rock stress data collected using a vibrating wire stress gauge, in units of MPa; equipment operating status data, such as current parameters collected from motor equipment, in units of A; and geological environment monitoring data, such as rock layer thickness, in units of cm.

[0019] Based on the production logs during the coal production process, the accident types recorded in the historical production cycle, such as roof collapse accidents, are obtained synchronously; the accident location is obtained by collecting the specific accident coordinates and longitude through GPS; accident-related monitoring data is obtained by collecting multiple types of safety monitoring data within 1 hour before the accident; the accident type, accident location and accident-related monitoring data are integrated to form historical accident data.

[0020] Data enhancement processing is performed on historical accident data, including data expansion, data smoothing, and outlier correction. Data expansion can be supplemented using the linear interpolation method to ensure that there are data points at each time point. Data smoothing can be performed using the moving average method, with the sliding window size set to 5 data points to calculate the average and eliminate sensor noise. Outlier correction can be performed using the interquartile range method to correct outliers and eliminate outliers that may be caused by sensor failures and other issues.

[0021] Based on historical accident data after data augmentation, a generative adversarial network model is constructed for accident data expansion to address the problem of insufficient accident data. Synthetic accident data is generated using the generative adversarial network model. For example, the generator input layer receives historical accident data, the fully connected layer uses 64 nodes activated by the ReLU function, and the output layer outputs the synthetic accident data. The discriminator input layer receives synthetic accident data, the fully connected layer uses 64 nodes activated by the LeakyReLU function, and the output layer uses the Sigmoid function to output the accuracy of the discriminant synthetic accident data. The generative adversarial network model is trained until convergence. For example, if the output synthetic accident data has an accuracy rate of over 90%, the generative adversarial network training is complete.

[0022] Synthetic accident data is added to historical accident data for training dynamic graph neural network models.

[0023] According to historical accident data, a dynamic graph neural network model is constructed accordingly.

[0024] Specifically, the safety monitoring points of the target coal production scenario are mapped as nodes in the graph structure; Based on the physical dependencies, geological distribution, and historical accident correlations between safety monitoring points, directed edges are constructed between nodes and a corresponding edge weight matrix is ​​generated to form a graph structure. For example, bidirectional edges are established between safety monitoring points in connected tunnels, undirected edges are established between safety monitoring points within 50 meters of the same rock layer, and directed edges are established between safety monitoring points that have triggered the same accident alarm simultaneously.

[0025] Model training is performed by combining the graph structure with time series monitoring data and historical production control data from historical accident data to obtain a dynamic graph neural network model. The time series monitoring data in historical accident data is a time series integration of accident data that occurred within a historical period. For example, the historical period can be set to the past 360 days, including multiple safety monitoring records from various safety monitoring points. Historical production control data refers to control data from coal mine production within a historical period, such as the power of mining machinery (in W) and the ventilation volume of coal mine passages (in m³ / s). A three-layer graph convolutional network structure is used to construct a dynamic graph neural network model. The first layer receives the time series monitoring data and historical production control data from historical accident data. The neighbor aggregation method is average, and the activation function is the ReLU function. The second layer uses 128 nodes, the neighbor aggregation method is average, and the activation function is the ReLU function. The third layer outputs risk probabilities and potential accident propagation paths, and the activation function is Softmax. The model is trained repeatedly until convergence conditions are met, such as inputting graph structure nodes, time series monitoring data in historical accident data, historical production control data, and outputting the risk probability corresponding to the graph structure node. The risk probability error is within the range of ±5%, and the accuracy of the potential accident propagation path is above 90%. This means that the dynamic graph neural network is constructed.

[0026] By mapping safety monitoring points into graph nodes and dynamically constructing an edge weight matrix based on physical dependencies, geological distribution, and accident correlations, we achieve deep coupling of multi-source data, break through the limitations of single-point data, and accurately capture the chain transmission laws of multi-dimensional data. By training dynamic graph neural networks with time series monitoring data, we significantly improve the model's predictive sensitivity to complex accidents, providing a high-precision spatiotemporal evolution basis for subsequent safety boundary analysis.

[0027] S20: In combination with the multi-source safety monitoring data, the production control plan is iteratively predicted and analyzed through the dynamic graph neural network model to obtain iterative prediction and analysis results, wherein the production control plan is determined based on the design operating condition domain of the target production scenario.

[0028] There are limitations in the formulation of existing production control strategies. Safety boundaries based on fixed rules often ignore real-time changes in operating conditions, resulting in conservative strategies reducing production capacity or aggressive strategies triggering accidents. There is a lack of a quantitative mapping mechanism between multi-dimensional safety indicators and production control parameters, which causes a disconnect between safety management and production optimization. Manual experience adjustments are subject to the risk of response lag and subjective bias.

[0029] Among them, also include: Collect security boundary analysis results and generate a security boundary analysis result set; Extracting and constructing a historical multi-source security monitoring dataset based on the security boundary analysis result set, and combining the security boundary analysis result set with the historical multi-source security monitoring dataset to construct an enhanced sample dataset; Constructing a lightweight safety margin rapid prediction model, and performing supervised training of the safety margin rapid prediction model based on the enhanced sample dataset; Evaluate and obtain the analytical computing power indicators and analytical timeliness indicators of the anchoring and mining equipment management nodes of multiple production working faces in the target coal production scenario, and deploy and verify the safety margin rapid prediction model based on the analytical computing power indicators and the analytical timeliness indicators; If the deployment verification result shows that the analysis timeliness index is not met, iterative model compression is performed on the safety boundary rapid prediction model; If the deployment verification result shows that the analysis timeliness index is met, the safety margin rapid prediction model is deployed to the target coal production scenario to perform safety margin analysis.

[0030] In an embodiment of the present application, a dynamic graph neural network model is used to iteratively predict and analyze the production control plan in combination with multi-source safety monitoring data to obtain iterative prediction and analysis results, wherein the production control plan is determined based on the design operating condition domain of the target production scenario. For example, the production control plan includes multiple production control parameters, such as the power of the mining machinery, in W, and the ventilation volume of the coal mine passage, in m³ / s. Multiple production control plans and multi-source safety monitoring data are input into the dynamic graph neural network model, and the risk probabilities and potential accident propagation paths of the multiple production control plans are output.

[0031] S30: Based on the iterative prediction analysis result and the preset safety boundary constraint, selecting a production control scheme that satisfies the safety boundary constraint as a safe production control scheme set, wherein the safety boundary constraint is determined based on the gas emission rate of the working face, and the gas emission rate of the working face includes an absolute emission rate and a relative emission rate; Based on the iterative prediction analysis results and the preset safety boundary constraints, production control solutions that meet the safety boundary constraints are selected as the safety production control solution set. The safety boundary constraints can be, for example, risk probability constraints. For example, if the risk probability of a production control solution is greater than or equal to 30%, or if the potential accident propagation path exceeds four nodes, the safety boundary constraints are not met. Multiple production control solutions that meet the safety boundary production constraints are selected as the safety production control solution set.

[0032] Optionally, the safety boundary constraint can also be determined based on the working face gas emission rate. The working face gas emission rate is the volume of gas released from the coal seam or rock stratum into the working space in underground projects such as coal mines. It is divided into absolute emission rate (the volume of gas released from the entire working face per unit time, in m³ / min) and relative emission rate (the volume of gas released per unit coal volume, in m³ / t). S40: performing fitting analysis on the production safety control solution set to generate a real-time production control safety boundary; A fitting analysis is performed on the set of safety production control schemes to generate a real-time production control safety margin. Specifically, the union of each safety production control parameter in the safety production control scheme is calculated. For example, the union of the mining machinery power is calculated to obtain the maximum and minimum power values ​​of the mining machinery, which serve as the real-time production control safety margin for the collected machinery power.

[0033] The method provided in the embodiment of the present application also includes: Collect security boundary analysis results and generate a security boundary analysis result set.

[0034] According to the security boundary analysis result set, the historical multi-source security monitoring dataset is extracted and constructed accordingly, and the security boundary analysis result set and the historical multi-source security monitoring dataset are combined to construct an enhanced sample dataset.

[0035] Build a lightweight, rapid safety margin prediction model, such as a fully connected neural network. The input layer receives multi-source safety monitoring datasets, the hidden layer uses 32 nodes, activated using the ReLU function, and the output layer outputs the predicted safety margin. Supervised training of the rapid safety margin prediction model is performed based on the enhanced sample dataset until the model converges. For example, if the output predicts the safety margin within an error range of ±5% based on the multi-source safety monitoring dataset, the rapid safety margin prediction model is considered trained.

[0036] Evaluate and obtain the analytical computing power and analysis timeliness indicators of the anchoring and mining equipment management nodes at multiple production faces in the target coal production scenario. Verify the deployment of a rapid safety margin prediction model based on these indicators. For example, deploy a rapid safety margin prediction model based on the analytical computing power of the onboard edge computing node (for safety management) corresponding to the anchoring and mining equipment in the target coal production scenario. The analysis timeliness indicator indicates the time required to complete the analysis for the target coal production scenario, measured in seconds.

[0037] When the analysis completion time of the safety margin rapid prediction model is greater than the analysis completion time target required by the target coal production scenario, it means that the current safety margin rapid prediction model does not meet the analysis timeliness index of the target coal production scenario.

[0038] If the deployment verification result shows that the analysis timeliness index is not met, the fast prediction model of the safety boundary is iteratively compressed. For example, the fast prediction model of the safety boundary can be iteratively compressed using a model compression method based on knowledge distillation.

[0039] If the deployment verification results show that the analysis timeliness indicators are met, the safety margin rapid prediction model will be deployed to the target coal production scenario to perform safety margin analysis.

[0040] The accident prediction results based on dynamic graph neural networks are collaboratively analyzed with multi-dimensional safety indicators to achieve real-time quantitative generation of safety boundaries. By iteratively predicting the accident probability under different production control schemes, the parameter combination that meets all safety indicator constraints is identified, breaking through the limitations of empirical thresholds, and converting abstract prediction values ​​into specific boundaries, such as the specific range of mining power, to directly guide production decisions. At the same time, the prediction model is iteratively compressed and optimized according to the corresponding timeliness, making it more adaptable to the needs of limited computing power and the need for rapid response in actual production scenarios.

[0041] S50: Controlling the production parameters of the target coal production scenario with the real-time production control safety boundary as a constraint.

[0042] The current control of coal production parameters has a passive response defect. When the sensor alarms, the system usually shuts down urgently or calls a fixed emergency plan, causing production interruption or excessive handling.

[0043] Step S50 in the method provided in the embodiment of the present application includes: Extract real-time production control solutions for target coal production scenarios; When any production control parameter of the real-time production control plan exceeds the real-time production control safety boundary, an emergency response plan is generated based on the historical case library and the expert knowledge base; In combination with a preset degradation step, executing adaptive production control degradation according to the emergency response plan; A production risk warning signal is generated according to the adaptive production control degradation result, and the production risk warning signal is fed back to the monitoring terminal of the target coal production scene.

[0044] In the embodiment of the present application, a variety of production control parameters of the target coal production scenario are collected, and a real-time production control plan for the target coal production scenario is extracted.

[0045] When any production control parameter of the real-time production control plan exceeds the real-time production control safety boundary, an emergency response plan is generated based on the historical case library and expert knowledge base. For example, the stress response plan may include reducing the power of mining machinery, increasing ventilation volume, etc.

[0046] Combined with the preset degradation step, adaptive production control degradation is performed according to the emergency response plan, where the preset degradation step refers to the pre-set parameters for degrading production control when there is a risk, such as reducing the power of mining equipment by 20%.

[0047] Based on the results of adaptive production control degradation, a production risk warning signal is generated and fed back to the monitoring terminal of the target coal production scene. For example, a text message "There are safety risks in coal mine production, please confirm in time" is generated and sent to the monitoring terminal of the target coal production scene for early warning.

[0048] Using real-time safety boundaries as hard constraints, a dual guarantee system for adaptive control and risk response was established. First, production parameters were continuously compared with dynamic boundaries, triggering warnings and initiating case-based emergency response plans before they crossed the boundary, avoiding unnecessary downtime. Second, parameters were adjusted in stages using preset degradation steps to prevent sudden changes in control instructions that could cause other risks, maximizing production continuity while ensuring safety. Ultimately, this achieved precise and safe control of coal mine production parameters.

[0049] Example 2, as Figure 2 As shown, based on the same inventive concept as the coal production optimization method based on data mining provided in the first embodiment, the embodiment of the present invention further provides a coal production optimization system based on data mining, comprising: The model building module 100 is used to collect multi-source safety monitoring data and historical accident data of the target coal production scene, and to build a dynamic graph neural network model according to the historical accident data; An iterative prediction module 200 is configured to perform iterative prediction analysis on a production control plan using the dynamic graph neural network model in combination with the multi-source safety monitoring data to obtain iterative prediction analysis results, wherein the production control plan is determined based on a design operating condition domain of a target production scenario; a scheme screening module 300 for screening, based on the iterative prediction analysis results and a preset safety boundary constraint, a production control scheme that satisfies the safety boundary constraint as a safe production control scheme set, wherein the safety boundary constraint is determined based on the gas emission rate of the working face, and the gas emission rate of the working face includes an absolute emission rate and a relative emission rate; A fitting analysis module 400 is used to perform fitting analysis on the production safety control solution set to generate a real-time production control safety boundary; The production parameter control module 500 is used to control the production parameters of the target coal production scenario based on the real-time production control safety boundary.

[0050] In one embodiment, the model building module 100 is further configured to: Collecting multiple types of safety monitoring data during coal production in real time to generate the multi-source safety monitoring data; Synchronously obtain the accident type, accident location, and accident-related monitoring data recorded during the historical production cycle to form the historical accident data; The multi-type safety monitoring data includes at least gas concentration data, dust concentration data, surrounding rock stress data, equipment operation status data and geological environment monitoring data; According to the historical accident data, a dynamic graph neural network model is constructed, including: Map the safety monitoring points of the target coal production scenario into nodes in the graph structure; Based on the physical dependency, geological distribution, and historical accident correlation between the safety monitoring points, directed connection edges between nodes are constructed and corresponding edge weight matrices are generated to form the graph structure; Combining the graph structure with time series monitoring data and historical production control data in the historical accident data to perform model training to obtain the dynamic graph neural network model; Wherein, according to the historical accident data, a dynamic graph neural network model is constructed, which includes: Performing data enhancement processing on the historical accident data, wherein the data enhancement processing includes data expansion, data smoothing, and outlier correction; Based on the historical accident data after data enhancement processing, constructing a generative adversarial network model for accident data expansion, and generating synthetic accident data through the generative adversarial network model; The synthetic accident data is added to the historical accident data for training the dynamic graph neural network model.

[0051] In one embodiment, it further includes: Collect security boundary analysis results and generate a security boundary analysis result set; Extracting and constructing a historical multi-source security monitoring dataset based on the security boundary analysis result set, and combining the security boundary analysis result set with the historical multi-source security monitoring dataset to construct an enhanced sample dataset; Constructing a lightweight safety margin rapid prediction model, and performing supervised training of the safety margin rapid prediction model based on the enhanced sample dataset; Evaluate and obtain the analytical computing power indicators and analytical timeliness indicators of the anchoring and mining equipment management nodes of multiple production working faces in the target coal production scenario, and deploy and verify the safety margin rapid prediction model based on the analytical computing power indicators and the analytical timeliness indicators; If the deployment verification result shows that the analysis timeliness index is not met, iterative model compression is performed on the safety boundary rapid prediction model; If the deployment verification result shows that the analysis timeliness index is met, the safety margin rapid prediction model is deployed to the target coal production scenario to perform safety margin analysis.

[0052] In one embodiment, the production parameter control module 500 is further configured to: Extract real-time production control solutions for target coal production scenarios; When any production control parameter of the real-time production control plan exceeds the real-time production control safety boundary, an emergency response plan is generated based on the historical case library and the expert knowledge base; In combination with a preset degradation step, executing adaptive production control degradation according to the emergency response plan; A production risk warning signal is generated according to the adaptive production control degradation result, and the production risk warning signal is fed back to the monitoring terminal of the target coal production scene.

[0053] In summary, the embodiments of the present application have at least the following technical effects: This application proposes a coal production optimization method and system based on data mining. By integrating dynamic graph neural network modeling with collaborative analysis of multi-source safety data, the dynamic adaptability and safety of production control parameters are significantly improved. First, by mapping safety monitoring points into dynamic graph nodes and constructing an edge weight matrix of physical dependence, geological distribution and accident association, the nonlinear coupling relationship between heterogeneous data such as gas concentration and surrounding rock stress is effectively captured, solving the problem of missing accident evolution laws caused by traditional single-dimensional analysis; secondly, based on the accident prediction results, the safety boundary of production parameters is reversed, so that the control strategies such as mining machinery power and ventilation volume are always adaptively adjusted within the dynamic safety threshold, avoiding the conservative defects or blind risks of manual experience decision-making; thirdly, by utilizing lightweight model deployment mechanism and enhanced sample training, the safety boundary can still respond quickly in the limited computing power environment underground, ensuring the real-time closed-loop linkage of analysis timeliness and production control; finally, by integrating the adaptive degradation mechanism of the emergency response plan with the historical case library, the risk of system collapse caused by parameter out-of-bounds is greatly reduced. Compared with traditional methods, the technical solution provided by this application significantly breaks through the limitations of static threshold control, realizes the spatiotemporal correlation mining of multi-dimensional safety indicators and real-time collaborative optimization of production control parameters, and achieves the technical effect of improving the safety of coal production control parameters.

[0054] It should be noted that the order in which the embodiments of the present application are presented is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. Furthermore, the foregoing descriptions of specific embodiments of this specification are provided. Furthermore, the processes depicted in the accompanying drawings do not necessarily require the specific order or sequential sequence shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0055] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should be included in the scope of protection of the present application.

[0056] This specification and drawings are merely illustrative of the present application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Obviously, those skilled in the art may make various modifications and variations to this application without departing from the scope of this application. Thus, this application is intended to include such modifications and variations as fall within the scope of this application and its equivalents.

Claims

1. A coal production optimization method based on data mining, characterized in that: include: Collect multi-source safety monitoring data and historical accident data of the target coal production scene, and construct a dynamic graph neural network model based on the historical accident data; In combination with the multi-source safety monitoring data, an iterative prediction analysis is performed on the production control plan through the dynamic graph neural network model to obtain an iterative prediction analysis result, wherein the production control plan is determined based on a design operating condition domain of a target production scenario; Based on the iterative prediction analysis results and the preset safety boundary constraints, a production control scheme that satisfies the safety boundary constraints is selected as a safe production control scheme set, wherein the safety boundary constraints are determined based on the gas emission volume of the working face, and the gas emission volume of the working face includes an absolute emission volume and a relative emission volume; Performing fitting analysis on the production safety control scheme set to generate a real-time production control safety boundary; The production parameter control of the target coal production scenario is performed with the real-time production control safety boundary as a constraint.

2. The method for optimizing coal production based on data mining according to claim 1, wherein: Collect multi-source safety monitoring data and historical accident data for target coal production scenarios, including: Collecting multiple types of safety monitoring data during coal production in real time to generate the multi-source safety monitoring data; Synchronously obtain the accident type, accident location, and accident-related monitoring data recorded during the historical production cycle to form the historical accident data; Among them, the multi-type safety monitoring data at least includes gas concentration data, dust concentration data, surrounding rock stress data, equipment operation status data and geological environment monitoring data.

3. A coal production optimization method based on data mining according to claim 2, characterized in that: Based on the historical accident data, a dynamic graph neural network model is constructed, including: Map the safety monitoring points of the target coal production scenario into nodes in the graph structure; Based on the physical dependency, geological distribution, and historical accident correlation between the safety monitoring points, directed connection edges between nodes are constructed and corresponding edge weight matrices are generated to form the graph structure; The graph structure is combined with the time series monitoring data and historical production control data in the historical accident data to perform model training to obtain the dynamic graph neural network model.

4. The method for optimizing coal production based on data mining according to claim 1, wherein: Also includes: Collect security boundary analysis results and generate a security boundary analysis result set; Extracting and constructing a historical multi-source security monitoring dataset based on the security boundary analysis result set, and combining the security boundary analysis result set with the historical multi-source security monitoring dataset to construct an enhanced sample dataset; Constructing a lightweight safety margin rapid prediction model, and performing supervised training of the safety margin rapid prediction model based on the enhanced sample dataset; Evaluate and obtain the analytical computing power indicators and analytical timeliness indicators of the anchoring and mining equipment management nodes of multiple production working faces in the target coal production scenario, and deploy and verify the safety margin rapid prediction model based on the analytical computing power indicators and the analytical timeliness indicators; If the deployment verification result shows that the analysis timeliness index is not met, iterative model compression is performed on the safety boundary rapid prediction model; If the deployment verification result shows that the analysis timeliness index is met, the safety margin rapid prediction model is deployed to the target coal production scenario to perform safety margin analysis.

5. The method for optimizing coal production based on data mining according to claim 1, wherein: Based on the historical accident data, a dynamic graph neural network model is constructed, which includes: Performing data enhancement processing on the historical accident data, wherein the data enhancement processing includes data expansion, data smoothing, and outlier correction; Based on the historical accident data after data enhancement processing, constructing a generative adversarial network model for accident data expansion, and generating synthetic accident data through the generative adversarial network model; The synthetic accident data is added to the historical accident data for training the dynamic graph neural network model.

6. The method for optimizing coal production based on data mining according to claim 1, wherein: The method further includes: controlling the production parameters of the target coal production scenario using the real-time production control safety boundary as a constraint; and Extract real-time production control solutions for target coal production scenarios; When any production control parameter of the real-time production control plan exceeds the real-time production control safety boundary, an emergency response plan is generated based on the historical case library and the expert knowledge base; In combination with a preset degradation step, executing adaptive production control degradation according to the emergency response plan; A production risk warning signal is generated according to the adaptive production control degradation result, and the production risk warning signal is fed back to the monitoring terminal of the target coal production scene.

7. A coal production optimization system based on data mining, characterized in that: A method for optimizing coal production based on data mining according to any one of claims 1 to 6, the system comprising: A model building module is used to collect multi-source safety monitoring data and historical accident data of the target coal production scenario, and to build a dynamic graph neural network model based on the historical accident data; an iterative prediction module, configured to combine the multi-source safety monitoring data and perform iterative prediction analysis on the production control plan through the dynamic graph neural network model to obtain iterative prediction analysis results, wherein the production control plan is determined based on a design operating condition domain of a target production scenario; a scheme screening module, configured to screen, based on the iterative prediction analysis results and a preset safety boundary constraint, a production control scheme that satisfies the safety boundary constraint as a safe production control scheme set, wherein the safety boundary constraint is determined based on the gas emission rate of the working face, and the gas emission rate of the working face includes an absolute emission rate and a relative emission rate; A fitting analysis module, configured to perform fitting analysis on the production safety control scheme set to generate a real-time production control safety boundary; The production parameter control module is used to control the production parameters of the target coal production scenario based on the real-time production control safety boundary.

Images