Coal mining operation optimization scheduling system based on Internet of Things

Abstract

The invention discloses a coal mining operation optimization scheduling system based on the Internet of Things. The system comprises a preliminary demand calculation module, a real-time data acquisition module, a feature extraction and abnormal trend analysis module, an abnormal risk prediction module, a dynamic demand adjustment module and a material distribution management module. The preliminary demand calculation module is used for calculating the preliminary support material demand quantity in the early stage of coal mining according to the tunneling speed of mining equipment, the working face scale and the roadway design specification; potential abnormity is predicted by monitoring the mine roof state and machine learning in real time, the support material supply amount is dynamically adjusted in combination with fuzzy logic, it is ensured that the support materials are added in time when the roof is abnormal, and the collapse risk is prevented. The optimized supply plan preferentially responds to a high-risk area, distribution delay is reduced, the supporting efficiency is improved, mine safety and continuous operation are guaranteed, meanwhile, the material use efficiency is maximized, and production stagnation and economic losses caused by insufficient supporting are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of coal mining, and in particular to an Internet of Things-based coal mining operation optimization and scheduling system. Background Art

[0002] The application of the Internet of Things (IoT) in the scheduling of coal mining operations refers to the use of IoT technology to monitor, analyze and optimize the scheduling of various types of equipment, personnel, environment and operation data in the coal mining process in real time to improve operation efficiency and safety. In this process, sensors, smart devices and network communication technologies are widely used to collect data from the coal mine operation site in real time (such as the operating status of mining equipment, temperature and humidity of the coal mine environment, gas concentration, etc.) and transmit these data to the central scheduling platform. The platform uses technologies such as big data analysis and machine learning to carry out intelligent scheduling and resource allocation in combination with the actual production needs of the mine. For example, according to factors such as the working status of the equipment, the location of the personnel, and the progress of mining, the allocation and scheduling strategies of the operation tasks are dynamically adjusted to achieve equipment load balancing, optimization of production progress and safety of miners. In this way, IoT technology can significantly improve the safety, efficiency and sustainability of coal mine production.

[0003] The prior art has the following deficiencies: During coal mining, the supply of support materials is usually predetermined based on the excavation speed, the scale of the working surface, and the tunnel design specifications. However, when an abnormality occurs at the top of the mine and support materials need to be urgently dispatched, the supply of support materials often cannot meet the actual demand due to insufficient dispatch. In such cases, the mine roof may not be adequately supported, which can easily lead to roof collapse. Roof collapse not only directly threatens the lives of miners, but may also cause the mine operation to stagnate or be forced to close, causing serious economic losses and delays in production plans.

[0004] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0005] The purpose of the present invention is to provide an Internet of Things-based coal mining operation optimization scheduling system, which monitors the status of the mine roof in real time, combines machine learning models to predict potential roof anomalies, and dynamically adjusts the supply of support materials according to fuzzy logic to ensure that when the roof is abnormal, support materials can be added in time to avoid the risk of roof collapse. The optimized supply plan is executed through the scheduling system, which can give priority to responding to high-risk areas, reduce material distribution delays, and improve support efficiency, thereby ensuring the safety and continuity of mine operations, which not only improves the safety of the mine, but also maximizes the use efficiency of support materials, reduces production stagnation and economic losses caused by insufficient support, so as to solve the problems in the above-mentioned background technology.

[0006] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: an Internet of Things coal mining operation optimization scheduling system, including a preliminary demand calculation module, a real-time data acquisition module, a feature extraction and abnormal trend analysis module, an abnormal risk prediction module, a dynamic demand adjustment module and a material distribution management module; The initial demand calculation module calculates the initial support material demand in the early stage of coal mining based on the excavation speed of the mining equipment, the scale of the working surface and the tunnel design specifications; The real-time data acquisition module uses a sensor network to collect the status data of the mine roof in real time during the mining process and stores it as an analysis set for subsequent analysis; The feature extraction and abnormal trend analysis module extracts the features reflecting the abnormality of the mine roof by analyzing the data in the collection. Within the set monitoring time window, it analyzes the trend and change of the extracted features to identify whether there are early signs of abnormality of the roof. The abnormal risk prediction module inputs the analyzed features into the machine learning model, predicts the potential abnormalities of the mine roof through the pre-trained machine learning model, and outputs the abnormal risk assessment results to quantify the potential abnormalities of the mine roof; The dynamic demand adjustment module preliminarily determines the risk level of the roof based on the prediction results of the machine learning model, and dynamically adjusts the support material demand plan, giving priority to high-risk areas. It uses fuzzy logic to analyze the abnormal prediction results of the mine roof and dynamically determines the required additional support material supply; The material distribution management module adjusts the actual support material distribution according to the final support material supply output by fuzzy logic to ensure that the mine support needs are met in a timely manner and to ensure operational safety.

[0007] Preferably, the steps for determining the initial support material requirement are as follows: First, calculate the number of preliminary support points. The calculation expression is: ,in: is the number of preliminary support points, indicating the total number of points where support materials need to be laid. L is the excavation speed, which represents the linear distance mined per unit time, is the cross-sectional area of ​​the working surface, is the coverage area of ​​a single support point, is the support efficiency factor, and its value range is , used to represent the efficiency loss of the actual support arrangement; Based on the support point characteristics and material parameters, the support material requirements are calculated. The calculation expression is: ,in, is the total demand for support materials, indicating the total amount of materials in the initial supply plan. The material mass required for each support point is determined by the strength requirements of the support point. is the material utilization rate, which indicates the proportion of materials that can be used in the actual layout process. i is the index of the support point, It is the pressure resistance parameter of the support point material, which is used to indicate the adjustment coefficient of the material under different roof pressure conditions.

[0008] Preferably, by analyzing the data in the set, features reflecting the abnormality of the mine roof are extracted, wherein the extracted features include the vertical and horizontal displacement changes of the roof per unit time and the crack expansion of the mine roof. Within the set monitoring time window, after analyzing the extracted features, displacement reference values ​​and crack expansion reference values ​​are generated respectively. The displacement reference value is used to quantify the degree of displacement change of the mine roof per unit time, and the crack expansion reference value is used to quantify the expansion speed and range of the roof cracks.

[0009] Preferably, after analyzing the characteristics reflecting the abnormality of the mine roof, the displacement reference value and crack extension reference value generated after the analysis are input into the machine learning model to generate a potential risk factor, and the potential abnormality of the mine roof is predicted by the potential risk factor, and the potential abnormality of the mine roof is quantified by the potential risk factor.

[0010] Preferably, within the set monitoring time window, the specific steps of analyzing the extracted vertical and horizontal displacement variation characteristics of the top plate within a unit time and generating the displacement reference value are as follows: First, the vertical and horizontal displacement changes of the mine roof in unit time are calculated, and a weighted displacement transformation matrix is ​​constructed to comprehensively reflect the relative importance and influence of displacements in different directions. The expression of the weighted displacement transformation matrix is: ,in, Dis the weighted displacement transformation matrix, which represents the overall displacement of the top plate within the monitoring time window. is the vertical displacement weight coefficient, is the vertical displacement change per unit time, is the horizontal displacement weight coefficient, is the change in horizontal displacement per unit time; Using the weighted displacement transformation matrix, the vertical and horizontal displacements are comprehensively processed through a nonlinear dynamic fusion function to generate the final displacement reference value to quantify the displacement change degree and potential instability risk of the mine roof. The generation expression of the displacement reference value is: ,in, is the displacement reference value, which indicates the degree of comprehensive displacement change of the mine roof per unit time. is the weighted vertical displacement transformation, , is the weighted horizontal displacement transformation, , is the adjustment coefficient used to control the sensitivity of nonlinear growth, is the displacement interaction function, defined as: ,in, is a minimum value used to prevent the denominator from being zero.

[0011] Preferably, within the set monitoring time window, after analyzing the extracted crack extension characteristics of the mine roof, the specific steps of generating the crack extension reference value are as follows: The crack characteristic function is calculated based on the crack extension speed and range. The calculation expression is: ,in, is the crack characteristic function, which represents the cumulative crack extension characteristic value within time t. is the crack expansion rate, expressed in time x The growth rate of the roof crack is is the area change rate of crack expansion, indicating time x The dynamic changes of the area affected by the crack expansion, is the time decay factor, is a positive parameter that controls the decay rate and is used to reduce the impact of early crack extension on the current assessment. x is the time variable, used for accumulation within the integral range; The crack extension reference value is constructed using the characteristic function, and the constructed expression is: ,in, is the crack expansion reference value, which is used to quantify the expansion speed and range of roof cracks. T To monitor the length of the time window for analysis within a limited time, and is the weight coefficient, is the logarithmic value of the crack characteristic function, is the time derivative of the characteristic function, which indicates the sensitivity of the dynamic change of crack extension. is the maximum ratio of the crack area change rate to the expansion speed, which is used for normalization processing. The start time of the monitoring time window.

[0012] Preferably, according to the prediction results of the machine learning model, fuzzy logic is used to analyze the abnormal prediction results of the mine roof, and the required additional support material supply is dynamically determined. The specific steps are as follows: Predict the status of the mine roof through a pre-trained machine learning model to obtain potential risk factors; After determining the risk level of the mine roof, fuzzy logic is used to dynamically adjust the demand for support materials. The fuzzy logic system fuzzifies the output of potential risk factors and outputs a fuzzy risk value. , which is used to represent the abnormal risk of the roof, and the fuzzification process is calculated through the following fuzzy rules: ,That represents the fuzzy logic mapping function, It is a potential risk factor predicted by a machine learning model. Represents a rule set, a rule set include: If the potential risk factor is greater than or equal to the potential risk factor reference threshold, the fuzzy risk value Displays "high risk" and increases the demand for support materials; If the potential risk factor is less than the potential risk factor reference threshold, the fuzzy risk value Displays “low risk” and the demand for support materials remains normal.

[0013] Preferably, the fuzzy risk value outputted by fuzzy logic , combined with the actual situation of the mine, determine the supply of new support materials. The calculation expression for the supply of new support materials is: ,in, Indicates the additional supply of support materials. is an adjustment factor, which indicates the adjustment ratio of the supply of support materials under different risk levels. S It is the scale of the mine operating surface.

[0014] In the above technical solution, the technical effects and advantages provided by the present invention are: The present invention monitors the status of the mine roof in real time, combines machine learning models to predict potential roof anomalies, and dynamically adjusts the supply of support materials according to fuzzy logic to ensure that when the roof is abnormal, support materials can be added in time to avoid the risk of roof collapse. The optimized supply plan is executed through the scheduling system, which can give priority to responding to high-risk areas, reduce material delivery delays, and improve support efficiency, thereby ensuring the safety and continuity of mine operations. It not only improves the safety of the mine, but also maximizes the use efficiency of support materials and reduces production stagnation and economic losses caused by insufficient support. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0016] Figure 1 This is a module schematic diagram of a coal mining operation optimization scheduling system of the Internet of Things of the present invention. DETAILED DESCRIPTION

[0017] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0018] The present invention provides Figure 1 The coal mining operation optimization scheduling system of the Internet of Things shown includes a preliminary demand calculation module, a real-time data acquisition module, a feature extraction and abnormal trend analysis module, an abnormal risk prediction module, a dynamic demand adjustment module and a material distribution management module; The initial demand calculation module calculates the initial support material demand in the early stage of coal mining based on the excavation speed of the mining equipment, the scale of the working surface and the tunnel design specifications; The steps to determine the initial support material requirements are as follows: First, calculate the number of preliminary support points. The calculation expression is: ,in: is the number of preliminary support points, indicating the total number of points where support materials need to be laid. L is the excavation speed, which represents the linear distance mined per unit time, is the cross-sectional area of ​​the working surface, which is determined by the tunnel design specifications. It is the coverage area of ​​a single support point, which is determined by the support type (such as anchor support, shotcrete support, etc.). is the support efficiency factor, and its value range is , which is used to represent the efficiency loss of the actual support arrangement (e.g., unsupported areas caused by equipment position adjustment); This step calculates the number of preliminary support points based on the relationship between the coverage of the support material layout and the cross-sectional area of ​​the tunnel during the mining operation. The support efficiency factor is introduced into the formula to correct the support efficiency loss caused by equipment deviation, rock formation irregularities, etc. in actual operations.

[0019] For example, when the excavation speed , tunnel cross-sectional area , Single point support area , support efficiency factor When , the number of initial support points is: The calculation in this step can quickly determine the number of points required for support layout, providing a basis for subsequent material demand calculations.

[0020] Based on the support point characteristics and material parameters, the support material requirements are calculated. The calculation expression is: ,in, is the total demand for support materials, indicating the total amount of materials in the initial supply plan. The material mass required for each support point is determined by the strength requirements of the support point. is the material utilization rate, which indicates the proportion of materials that can be used in the actual layout process. i is the index of the support point, It is the pressure resistance parameter of the support point material, which is used to indicate the adjustment coefficient of the material under different roof pressure conditions (such as anchor rod length, shotcrete thickness, etc.); The total demand for support materials is calculated by adding up the material demand at each support point. The formula takes into account material properties and material waste or deviation during construction. In addition, additional reserve materials are added to provide protection for demand in abnormal situations. For example, when , , , When the total demand for support materials is: , the calculated preliminary support material requirement can ensure that the material supply meets the operational needs.

[0021] The initial demand for support materials is dynamic, because the mining progress, working surface status and geological conditions of the mine will change over time. The demand for support materials will be adjusted with the excavation speed and the scale of the working surface. Usually, the supply of support materials will be dynamically updated based on real-time mining data and monitoring results. The supply frequency can be set according to the specific situation of the mine. The common cycle is daily or after each excavation progress reaches a certain level.

[0022] Real-time data acquisition module, which uses a sensor network (such as roof pressure sensors and rock displacement monitoring equipment) to collect real-time status data of the mine roof during the mining process and stores it as an analysis set for subsequent analysis; Information about the mine roof is a core data source for predicting potential anomalies. Sensors can measure parameters such as roof pressure, rock deformation, vibration amplitude, etc., generating large-scale time series data sets. The establishment of an analysis set includes the process of data collection, cleaning, and formatting. Cleaning data can remove sensor noise or erroneous readings to make the data more reliable. For example, a sensor may collect pressure values ​​once a second and store them as a time series to form a "time-pressure" data analysis set.

[0023] The feature extraction and abnormal trend analysis module extracts the features reflecting the abnormality of the mine roof by analyzing the data in the collection, analyzes the trend and change of the extracted features within the set monitoring time window (such as the past 10 minutes or 1 hour), and identifies whether there are early signs of abnormal roof. By analyzing the data in the collection, the features reflecting the abnormalities of the mine roof are extracted, wherein the extracted features include the vertical and horizontal displacement changes of the roof in unit time and the crack expansion of the mine roof. Within the set monitoring time window, after analyzing the extracted features, the displacement reference value and the crack expansion reference value are generated respectively. The displacement reference value is used to quantify the displacement change degree of the mine roof in unit time, reflecting the stability of the roof and the potential risk of instability; the crack expansion reference value is used to quantify the expansion speed and range of the roof cracks, reflecting the development of the cracks in the mine roof.

[0024] During the coal mining process, the vertical and horizontal displacements of the mine roof change greatly per unit time, which usually indicates that there is a potential abnormality in the roof. This abnormal displacement may be caused by uneven settlement of the rock strata, stress redistribution or unstable support structure. If the roof is displaced over a large range, especially rapidly, it indicates that the roof may be close to instability or rupture, and there is a risk of roof collapse. In order to ensure the safety of mine operations, it is necessary to increase support materials in a timely manner according to the displacement changes, strengthen the support of the mine roof, prevent more serious disasters caused by excessive deformation of the roof, and ensure the safety of miners and the stability of the working environment.

[0025] Within the set monitoring time window, the specific steps of analyzing the vertical and horizontal displacement change characteristics of the extracted top plate within unit time and generating the displacement reference value are as follows: First, the vertical and horizontal displacement changes of the mine roof in unit time are calculated, and a weighted displacement transformation matrix is ​​constructed to comprehensively reflect the relative importance and influence of displacements in different directions. The expression of the weighted displacement transformation matrix is: ,in, D is the weighted displacement transformation matrix, which represents the overall displacement of the top plate within the monitoring time window. is the vertical displacement weight coefficient, reflecting the relative importance of vertical displacement to the stability of the roof. is the vertical displacement change per unit time, which is collected in real time by the vertical displacement sensor. is the horizontal displacement weight coefficient, reflecting the relative importance of horizontal displacement to the stability of the roof. is the horizontal displacement change per unit time, which is collected in real time by the horizontal displacement sensor; This step generates a comprehensive weighted displacement transformation matrix by weighted combination of vertical and horizontal displacement changes. and Used to adjust the effects of vertical and horizontal displacements to reflect the different contributions of displacements in different directions to roof stability. For example, in some mines, vertical displacement may have a greater impact on roof stability than horizontal displacement, so you can set Through this weighted method, the weighted displacement transformation matrix can more accurately reflect the displacement changes of the top plate in all directions and its comprehensive impact on stability, providing a basis for subsequent calculations.

[0026] Using the weighted displacement transformation matrix, the vertical and horizontal displacements are comprehensively processed through a nonlinear dynamic fusion function to generate the final displacement reference value to quantify the displacement change degree and potential instability risk of the mine roof. The generation expression of the displacement reference value is: ,in, is the displacement reference value, which indicates the degree of comprehensive displacement change of the mine roof per unit time. is the weighted vertical displacement transformation, , is the weighted horizontal displacement transformation, , is the adjustment coefficient used to control the sensitivity of nonlinear growth, usually determined by optimizing historical data. is the displacement interaction function, defined as: ,in, is a minimum value used to prevent the denominator from being zero; Through the nonlinear dynamic fusion function, the vertical and horizontal displacement transformation quantities are integrated with the weighted displacement transformation matrix to generate a comprehensive displacement reference value. First, the cube root function is used The weighted displacements are combined nonlinearly to enhance sensitivity to large displacement changes. Subsequently, the exponential term Further zooming in on the displacement interaction function The calculated comprehensive displacement effect is Used to adjust the magnitude of exponential growth.

[0027] Displacement interaction function The mutual influence between vertical and horizontal displacements is taken into account, avoiding the information loss or error amplification that may be caused by simple linear combination. In this way, the displacement reference value not only reflects the change in displacement in each direction, but also takes into account the interaction between them, thereby more accurately quantifying the stability and potential instability risk of the roof.

[0028] Ultimately, the displacement reference value can be used as a comprehensive indicator to reflect the overall displacement changes of the mine roof within the monitoring time window, helping mine managers to promptly identify and respond to potential roof anomalies and ensure the safety and stability of mine operations.

[0029] It can be seen from the displacement reference value that within the set monitoring time window, the larger the performance value of the displacement reference value generated after analyzing the vertical and horizontal displacement change characteristics of the extracted roof within a unit time, the more drastic the displacement change of the mine roof within the set monitoring time window, and there may be a greater potential abnormal risk. In this case, the mine roof may be experiencing a large displacement or deformation, suggesting that the roof may be at risk of instability or collapse, so additional support materials are needed to ensure the stability of the roof and the safety of mine operations. On the contrary, when the displacement reference value performance value is small, it means that the displacement change of the roof is relatively stable and the potential abnormal risk is low. At this time, the stability of the mine roof is good, and usually no additional support materials are required.

[0030] During coal mining, if the cracks in the mine roof expand quickly and over a large area, it usually means that the roof has begun to have structural problems. The expansion of the cracks may cause a significant decrease in the strength of the roof, which in turn causes local instability or collapse. This abnormal crack expansion is a clear indication of potential abnormalities in the mine roof, indicating that the original support materials can no longer effectively support the mine roof, posing a serious safety hazard. Therefore, it is necessary to add support materials in a timely manner, especially in the crack expansion area, to reinforce the roof to prevent further damage and mine accidents.

[0031] After analyzing the extracted crack propagation characteristics of the mine roof within the set monitoring time window, the specific steps for generating the crack propagation reference value are as follows: The crack characteristic function is calculated based on the crack extension speed and range. The calculation expression is: ,in, is the crack characteristic function, which represents the cumulative crack extension characteristic value within time t. is the crack expansion rate, expressed in time x The growth rate of the roof crack is is the area change rate of crack expansion, indicating time x The dynamic changes of the area affected by the crack expansion, is the time decay factor, is a positive parameter that controls the decay rate and is used to reduce the impact of early crack extension on the current assessment. x is the time variable, used for accumulation within the integral range; This step takes into account the crack propagation rate and area change rate , and decays exponentially Weigh the contribution of different time points to the characteristic function. For example, the closer the time is to the current moment, the greater the impact of crack extension, while the impact of early extension gradually decreases. The purpose of this step is to construct a dynamic function that captures the comprehensive impact of the speed and range of crack extension on roof stability, providing a core foundation for subsequent calculations.

[0032] The crack extension reference value is constructed using the characteristic function, and the constructed expression is: ,in, is the crack expansion reference value, which is used to quantify the expansion speed and range of roof cracks. T To monitor the length of the time window for analysis within a limited time, and is the weight coefficient, which is used to adjust the relative influence of the crack characteristic value and characteristic change rate. is the logarithmic value of the crack characteristic function, emphasizing the change of crack extension intensity in the higher value area. is the time derivative of the characteristic function, which indicates the sensitivity of the dynamic change of crack extension. is the maximum ratio of the crack area change rate to the expansion speed, which is used for normalization processing. The starting time of the monitoring time window; The fracture extension reference value measures the overall impact and dynamic changes of fracture extension through the logarithm and time derivative of the fracture characteristic function, and uses the normalized ratio to balance the relative impact between the fracture extension speed and range. and A flexible adjustment mechanism is provided to enhance the contribution of certain features according to the actual needs of the mine.

[0033] It can be seen from the crack expansion reference value that within the set monitoring time window, the larger the performance value of the crack expansion reference value generated after analyzing the extracted crack expansion characteristics of the mine roof, the greater the crack expansion speed and range of the mine roof, reflecting the increased risk of uneven force on the roof and structural instability, which may cause the roof to become unstable or collapse, so additional support materials are needed to reinforce the roof to ensure operational safety. On the contrary, if the crack expansion reference value is small, it means that the roof crack expands slowly or has a small range, the stability of the mine roof is better, and the potential abnormal risk is low, so no additional support materials are needed.

[0034] The abnormal risk prediction module inputs the analyzed features into a machine learning model (such as a support vector machine, random forest, or deep neural network), predicts potential abnormalities of the mine roof through a pre-trained machine learning model, and outputs abnormal risk assessment results to quantify potential abnormalities of the mine roof; After analyzing the characteristics reflecting the abnormalities of the mine roof, the displacement reference values ​​and crack extension reference values ​​generated after the analysis are input into the machine learning model to generate potential risk factors. The potential abnormalities of the mine roof are predicted through the potential risk factors, and the potential abnormalities of the mine roof are quantified through the potential risk factors.

[0035] Pre-trained machine learning models refer to models that are trained using labeled mine roof data (including data in normal and abnormal states) in specific historical data sets and experimental environments. These models learn complex patterns and rules in the data, so that they can automatically identify and predict potential abnormal risks when faced with new real-time data. Pre-trained machine learning models have been repeatedly verified and adjusted, and have strong generalization capabilities and reliability, and can adapt to application needs in different mine environments.

[0036] In the process of coal mining, the stability of the mine roof is crucial. Real-time data obtained by sensors and other equipment reflect various dynamic characteristics of the roof (such as displacement, pressure, crack expansion, etc.). These data themselves may be complex and contain noise, making it difficult to use them directly for anomaly detection and risk prediction. In order to improve the accuracy and timeliness of predictions, machine learning technology has been introduced into the prediction of coal mine roof anomalies. By using historical data sets, engineers can select appropriate features (such as displacement reference values, crack expansion reference values, etc.) and use these data to train a machine learning model, such as a support vector machine, random forest, or deep neural network. These algorithms can identify the relationship between different features during the training process and maximize the prediction effect by optimizing model parameters.

[0037] Pre-trained machine learning models usually need to go through a large amount of training data to adjust their internal parameters until they can accurately predict the potential risk factors of the mine roof. The training data includes historical data from actual mine operations, which contain various states of the mine roof, such as displacement changes, crack development, and abnormal roof events that have occurred in history. Each data sample has been labeled whether it is an abnormal state, or the type and severity of the abnormality. By learning from these labeled data, the machine learning model can identify which features have a significant impact on roof anomalies and understand the nonlinear relationship between these features. For example, the displacement reference value and the crack extension reference value may not directly show abnormalities in some cases, but some combined features of the two can effectively predict the risk of roof instability.

[0038] The core goal of training is to optimize the parameters of the machine learning model so that it can predict whether there are potential anomalies in the mine roof through the input real-time monitoring data, and then quantify the risk of anomalies. Taking the support vector machine as an example, it maps the training data to a high-dimensional space and uses the method of maximizing the classification interval to find the optimal classification plane. The random forest determines the final prediction result through the voting results of multiple decision trees, which enhances the robustness of the model. The deep neural network performs feature learning through a multi-level network structure, which can automatically extract more complex feature representations from the original data, thereby better predicting potential risks.

[0039] Once the machine learning model is trained, it can be applied in the actual mine operating environment to process data from monitoring equipment in real time, output potential risk factors, and predict potential anomalies in the mine roof. When new data is input into the model, the model will use the experience gained in training to process the data and output the value of the potential risk factor. This potential risk factor represents the possibility of anomalies in the mine roof. The higher the value, the greater the risk of roof instability, and vice versa. The stability of the roof is better. Through this prediction, operators can understand the safety status of the mine roof in a timely manner, so as to take necessary support measures to avoid potential accidents.

[0040] Compared with traditional rule-based early warning systems, pre-trained machine learning models have higher flexibility and accuracy. They can automatically learn complex relationships from large amounts of data without the need for manual setting of fixed rules or thresholds. In mine operations, the abnormal state of the roof is often a complex, multi-factor intertwined process, and machine learning models can comprehensively analyze data from a multi-dimensional and multi-level perspective to make more accurate abnormality predictions. This not only improves the safety of coal mine operations, but also helps operators better formulate support material supply plans and optimize the scheduling of mine resources.

[0041] The machine learning model is not specifically limited here, and can realize the displacement reference value and crack extension reference value Conduct comprehensive analysis to generate potential risk factors In order to realize the technical solution of the present invention, the present invention provides a specific implementation method; potential risk factors The resulting calculation formula is: , where , Displacement reference values and crack extension reference value The preset scaling factor of , Both are greater than 0.

[0042] It can be seen from the potential risk factor and the displacement reference value that, within the set monitoring time window, the greater the performance value of the displacement reference value generated after analyzing the vertical and horizontal displacement change characteristics of the extracted roof within unit time, the greater the performance value of the crack extension reference value generated after analyzing the extracted crack extension characteristics of the mine roof. That is, the greater the performance value of the potential risk factor generated when predicting the potential abnormality of the mine roof through the pre-trained machine learning model, the greater the potential abnormality risk of the mine roof, and additional support materials are needed to reinforce the roof to ensure operation safety. Otherwise, it indicates that the potential abnormality risk of the mine roof is smaller, and no additional support materials are needed to reinforce the roof.

[0043] The dynamic demand adjustment module preliminarily determines the risk level of the roof based on the prediction results of the machine learning model, and dynamically adjusts the support material demand plan, giving priority to high-risk areas. It uses fuzzy logic to analyze the abnormal prediction results of the mine roof and dynamically determines the required additional support material supply; According to the prediction results of the machine learning model, fuzzy logic is used to analyze the abnormal prediction results of the mine roof and dynamically determine the additional support material supply required. The specific steps are as follows: Predict the status of the mine roof through a pre-trained machine learning model to obtain potential risk factors; The purpose of this step is to preliminarily determine the risk level of the mine roof through a comprehensive analysis of the roof status data, so as to provide a basis for subsequent support material supply decisions.

[0044] After determining the risk level of the mine roof, fuzzy logic is used to dynamically adjust the demand for support materials. The fuzzy logic system fuzzifies the output of potential risk factors and outputs a fuzzy risk value. , which is used to represent the abnormal risk of the roof, and the fuzzification process is calculated through the following fuzzy rules: ,That represents the fuzzy logic mapping function, It is a potential risk factor predicted by a machine learning model. Represents a rule set, a rule set include: If the potential risk factor is greater than or equal to the potential risk factor reference threshold, the fuzzy risk value Displays "high risk" and increases the demand for support materials; If the potential risk factor is less than the potential risk factor reference threshold, the fuzzy risk value Displays “low risk” and the demand for support materials remains normal.

[0045] The use of fuzzy logic can handle the complexity and uncertainty of potential risk factors, dynamically evaluate abnormal risks of the roof and provide flexible adjustment solutions.

[0046] Fuzzy risk value output based on fuzzy logic , combined with the actual situation of the mine, determine the supply of new support materials. The calculation expression for the supply of new support materials is: ,in, Indicates the additional supply of support materials. is an adjustment factor that indicates the proportion of support material supply adjustment under different risk levels (e.g. Indicates that the material increment is half of the basic demand when the risk is low. Indicates that the increment is 1.5 times the basic demand when the risk is high). S It is the size of the mine working surface (i.e. the area of ​​the mine working surface), taking into account the direct relationship between the size of the working surface and the amount of support required.

[0047] The material distribution management module adjusts the actual support material distribution according to the final support material supply output by fuzzy logic to ensure that the mine support needs are met in a timely manner and to ensure operational safety; The final support material supply is the support material demand calculated based on the excavation speed of the mining equipment, the scale of the working surface, and the tunnel design specifications, plus the additional support material supply dynamically determined by analyzing the abnormal prediction results of the mine roof using fuzzy logic. The optimized supply plan is executed through the scheduling system to deliver materials from the warehouse or reserve point to the working surface. During the distribution process, the system will monitor the supply progress in real time to ensure that high-risk areas are given priority and delays are reduced. At the same time, the supply execution results (such as delivery time and arrival volume) will be fed back to the system as reference data for the next adjustment. This closed-loop management model can continuously optimize supply efficiency and maximize the mine support capacity.

[0048] The actual support material distribution is adjusted according to the final support material supply output by fuzzy logic, so as to ensure that the mine can obtain the required support materials in time during the mining process, especially in high-risk areas, so as to ensure the safe support of the roof and prevent accidents such as roof collapse. By dynamically adjusting the supply, the system can flexibly respond to the support needs of different areas according to the real-time risk assessment results of the mine, and give priority to distributing materials to areas with higher potential risks, avoiding safety hazards caused by insufficient or delayed material supply. At the same time, this approach can optimize the overall material distribution process, improve the efficiency and safety of mine support, ensure the continuous and stable operation of mine operations, and reduce the safety risks of miners and the possibility of production interruptions.

[0049] Through the above-mentioned IoT optimization and scheduling system, dynamic regulation of coal mining operations can be achieved, effectively solving the problem of insufficient supply of support materials. The system monitors the status of the mine roof in real time, combines machine learning models to predict potential abnormalities of the roof, and dynamically adjusts the supply of support materials according to fuzzy logic to ensure that when the roof is abnormal, support materials can be added in time to avoid the risk of roof collapse. The optimized supply plan is executed through the scheduling system, which can give priority to responding to high-risk areas, reduce material delivery delays, and improve support efficiency, thereby ensuring the safety and continuity of mine operations. This system not only improves the safety of the mine, but also maximizes the efficiency of the use of support materials and reduces production stagnation and economic losses caused by insufficient support.

[0050] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0051] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0052] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. An Internet of Things coal mining operation optimization scheduling system, characterized in that: It includes preliminary demand calculation module, real-time data collection module, feature extraction and abnormal trend analysis module, abnormal risk prediction module, dynamic demand adjustment module and material distribution management module; The initial demand calculation module calculates the initial support material demand in the early stage of coal mining based on the excavation speed of the mining equipment, the scale of the working surface and the tunnel design specifications; The real-time data acquisition module uses a sensor network to collect the status data of the mine roof in real time during the mining process and stores it as an analysis set for subsequent analysis; The feature extraction and abnormal trend analysis module extracts the features reflecting the abnormality of the mine roof by analyzing the data in the collection. Within the set monitoring time window, it analyzes the trend and change of the extracted features to identify whether there are early signs of abnormality of the roof. The abnormal risk prediction module inputs the analyzed features into the machine learning model, predicts the potential abnormalities of the mine roof through the pre-trained machine learning model, and outputs the abnormal risk assessment results to quantify the potential abnormalities of the mine roof; The dynamic demand adjustment module preliminarily determines the risk level of the roof based on the prediction results of the machine learning model, and dynamically adjusts the support material demand plan, giving priority to high-risk areas. It uses fuzzy logic to analyze the abnormal prediction results of the mine roof and dynamically determines the required additional support material supply; The material distribution management module adjusts the actual support material distribution according to the final support material supply output by fuzzy logic to ensure that the mine support needs are met in a timely manner and to ensure operational safety.

2. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 1 is characterized in that: The steps to determine the initial support material requirements are as follows: First, calculate the number of preliminary support points. The calculation expression is: ， in: is the number of preliminary support points, indicating the total number of points where support materials need to be laid. L is the excavation speed, which represents the linear distance mined per unit time, is the cross-sectional area of ​​the working surface, is the coverage area of ​​a single support point, is the support efficiency factor, and its value range is , used to represent the efficiency loss of the actual support arrangement; Based on the support point characteristics and material parameters, the support material requirements are calculated. The calculation expression is: ， in, is the total demand for support materials, indicating the total amount of materials in the initial supply plan. The material mass required for each support point is determined by the strength requirements of the support point. is the material utilization rate, which indicates the proportion of materials that can be used in the actual layout process. i is the index of the support point, It is the pressure resistance parameter of the support point material, which is used to indicate the adjustment coefficient of the material under different roof pressure conditions.

3. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 1 is characterized in that: By analyzing the data in the collection, the features reflecting the abnormalities of the mine roof are extracted, where the extracted features include the vertical and horizontal displacement changes of the roof per unit time and the crack expansion of the mine roof. Within the set monitoring time window, after analyzing the extracted features, the displacement reference value and the crack expansion reference value are generated respectively. The displacement reference value is used to quantify the degree of displacement change of the mine roof per unit time, and the crack expansion reference value is used to quantify the expansion speed and range of the roof cracks.

4. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 3 is characterized in that: After analyzing the characteristics reflecting the abnormalities of the mine roof, the displacement reference values ​​and crack extension reference values ​​generated after the analysis are input into the machine learning model to generate potential risk factors. The potential abnormalities of the mine roof are predicted through the potential risk factors, and the potential abnormalities of the mine roof are quantified through the potential risk factors.

5. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 3 is characterized in that: Within the set monitoring time window, the specific steps of analyzing the vertical and horizontal displacement change characteristics of the extracted top plate within unit time and generating the displacement reference value are as follows: First, the vertical and horizontal displacement changes of the mine roof in unit time are calculated, and a weighted displacement transformation matrix is ​​constructed to comprehensively reflect the relative importance and influence of displacements in different directions. The expression of the weighted displacement transformation matrix is: ， in, D is the weighted displacement transformation matrix, which represents the overall displacement of the top plate within the monitoring time window. is the vertical displacement weight coefficient, is the vertical displacement change per unit time, is the horizontal displacement weight coefficient, is the change in horizontal displacement per unit time; Using the weighted displacement transformation matrix, the vertical and horizontal displacements are comprehensively processed through a nonlinear dynamic fusion function to generate the final displacement reference value to quantify the displacement change degree and potential instability risk of the mine roof. The generation expression of the displacement reference value is: ， in, is the displacement reference value, which indicates the degree of comprehensive displacement change of the mine roof per unit time. is the weighted vertical displacement transformation, , is the weighted horizontal displacement transformation, , is the adjustment coefficient used to control the sensitivity of nonlinear growth, is the displacement interaction function, defined as: ， in, is a minimum value used to prevent the denominator from being zero.

6. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 3 is characterized in that: After analyzing the extracted crack propagation characteristics of the mine roof within the set monitoring time window, the specific steps for generating the crack propagation reference value are as follows: The crack characteristic function is calculated based on the crack extension speed and range. The calculation expression is: ， in, is the crack characteristic function, which represents the cumulative crack extension characteristic value within time t. is the crack expansion rate, expressed in time x The growth rate of the roof crack is is the area change rate of crack expansion, indicating time x The dynamic changes of the area affected by the crack expansion, is the time decay factor, is a positive parameter that controls the decay rate and is used to reduce the impact of early crack extension on the current assessment. x is the time variable, used for accumulation within the integral range; The crack extension reference value is constructed using the characteristic function, and the constructed expression is: ， in, is the crack expansion reference value, which is used to quantify the expansion speed and range of roof cracks. T To monitor the length of the time window for analysis within a limited time, and is the weight coefficient, is the logarithmic value of the crack characteristic function, is the time derivative of the characteristic function, which indicates the sensitivity of the dynamic change of crack extension. is the maximum ratio of the crack area change rate to the expansion speed, which is used for normalization processing. The start time of the monitoring time window.

7. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 4 is characterized in that: According to the prediction results of the machine learning model, fuzzy logic is used to analyze the abnormal prediction results of the mine roof and dynamically determine the additional support material supply required. The specific steps are as follows: Predict the status of the mine roof through a pre-trained machine learning model to obtain potential risk factors; After determining the risk level of the mine roof, fuzzy logic is used to dynamically adjust the demand for support materials. The fuzzy logic system fuzzifies the output of potential risk factors and outputs a fuzzy risk value. , which is used to represent the abnormal risk of the roof, and the fuzzification process is calculated through the following fuzzy rules: ， That represents the fuzzy logic mapping function, It is a potential risk factor predicted by a machine learning model. Represents a rule set, a rule set include: If the potential risk factor is greater than or equal to the potential risk factor reference threshold, the fuzzy risk value Display "high risk" and increase the demand for support materials; If the potential risk factor is less than the potential risk factor reference threshold, the fuzzy risk value Displays "low risk" and the demand for support materials remains normal.

8. The coal mining operation optimization and scheduling system of the Internet of Things according to claim 7 is characterized in that: Fuzzy risk value output based on fuzzy logic , combined with the actual situation of the mine, determine the supply of new support materials. The calculation expression for the supply of new support materials is: ， in, Indicates the additional supply of support materials. is an adjustment factor, which indicates the adjustment ratio of the supply of support materials under different risk levels. S It is the scale of the mine operating surface.

Images