Dustproof control method and system for underground mining of coal mine

Abstract

The invention relates to the technical field of dustproof control, and discloses a dustproof control method and system for underground mining of a coal mine, and the system comprises a data analysis module which constructs all mine dust information into a mine dust data chain, outputs a dust action value based on a dust action model, and stores the dust action value; the dustproof analysis module determines an association result of the splashing dust image and the whole dust drifting image based on the first dust data chain and an association rule algorithm, and determines a spraying instruction value based on the association result and a Bayesian model, and the dustproof processing module compares the spraying instruction value in a historical spraying database; the dustproof control module determines whether to adjust the dustproof precipitation amount according to the dustproof spray load, determines a precipitation adjustment factor based on a relationship between the dustproof spray load and the dustproof spray use load, and adjusts the dustproof precipitation amount based on the precipitation adjustment factor. The dustproof stability and reliability are improved, and the risk that the dust concentration threatens underground operation safety is effectively reduced.

Description

Technical Field

[0001] This invention relates to the field of dust control technology, and more specifically, to a dust control method and system for underground coal mining. Background Technology

[0002] Underground coal mining is a crucial step in coal resource acquisition. Processes such as cutting, drilling and blasting at the tunnel face, and coal transfer and transportation generate mixed dust, primarily coal dust. The generation of this dust is unavoidable, and failure to effectively control underground dust can lead to multiple risks: for workers, long-term inhalation of coal dust can easily induce pneumoconiosis, and high concentrations of coal dust reaching their explosive limits can trigger a coal dust explosion upon contact with an ignition source, directly threatening underground operational safety; for mining equipment, coal dust adhering to the hydraulic system of the fully mechanized mining machine and the sensor surfaces of the tunneling machine can lead to decreased equipment accuracy and increased malfunction frequency.

[0003] Chinese patent CN119353029A discloses a method and control system for dust suppression using spraying in mining based on image recognition. The method includes: when an actual dust diffusion map contains shaded areas, determining a corresponding simulated mining area model based on the mining area parameters to be controlled; obtaining a simulated dust diffusion map corresponding to the mining area to be controlled based on the construction parameters of the mining area to be controlled and the simulated mining area model; determining a target dust diffusion map based on the actual dust diffusion map and the simulated dust diffusion map; performing edge detection and particle size analysis on the target dust diffusion map to determine the dust information corresponding to the mining area to be controlled from the target dust diffusion map; and generating spray control commands based on the dust information, which are used to control the spraying equipment to perform dust suppression operations based on the spray control commands. However, the target dust diffusion map ignores the dust diffusion process between areas, leading to dust diffusion superposition in the target dust diffusion map. Furthermore, dust suppression operations based on spray control commands cannot be adjusted according to the load of the spraying equipment, resulting in malfunctions such as pipe ruptures and pump overheating, which in turn causes dust concentration rebound and threatens the safety of underground operations.

[0004] Therefore, it is necessary to design a dust control method and system for underground coal mining to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a dust control method and system for underground coal mining, which aims to solve the problem of ignoring the dust diffusion process between areas, resulting in the superposition of dust diffusion, and the inability to make corresponding adjustments according to the load of the spraying equipment, leading to malfunctions such as pipeline rupture and pump overheating of the spraying equipment, thereby causing the dust concentration to rebound and threatening the safety of underground operations.

[0006] In one aspect, the present invention proposes a dust control system for underground coal mining, comprising: The data analysis module is configured to identify several dust control zones and collect mine dust information for each dust control zone, construct a mine dust data chain from all the mine dust information, output dust action values ​​based on the dust action model, and decompose the mine dust data chain according to the dust action values ​​to determine the first dust data chain and the second dust data chain. The dust control analysis module is configured to acquire images of splashed dust from mining equipment and dust dispersion images from several dust control zones, determine the association results between the splashed dust images and all dust dispersion images based on a first dust data chain and an association rule algorithm, and determine the spray command value based on the association results and a Bayesian model. The dust control module is configured to compare the spray command value with the historical spray database, determine the dust control precipitation of the dust control spray equipment based on the comparison result, obtain the dust control spray load of the dust control spray equipment during the dust control period, and determine whether to adjust the dust control precipitation based on the dust control spray load. The dust control module is configured to determine a precipitation adjustment factor based on the relationship between the dust spray load and the dust spray usage load when it is determined that the dust precipitation amount needs to be adjusted, and to adjust the dust precipitation amount based on the precipitation adjustment factor.

[0007] Furthermore, when constructing a mine dust data chain from all the mine dust information, outputting dust action values ​​based on the dust action model, and decomposing the mine dust data chain according to the dust action values, the process includes: The data analysis module acquires the first dust storage chain and the second dust storage chain, substitutes the dust information of each mine on the mine dust data chain into the dust action model, and outputs the dust action value of each mine dust information. The dust effect value of each mine's dust information is compared with the dust effect threshold. All mine dust information with dust effect values ​​greater than or equal to the dust effect threshold is transferred to the first dust shelving chain, and the first dust data chain is determined. All mine dust information with dust effect values ​​less than the dust effect threshold is transferred to the second dust shelving chain, and the second dust data chain is determined.

[0008] Furthermore, when constructing a mine dust data chain from all the mine dust information, outputting dust action values ​​based on the dust action model, and decomposing the mine dust data chain according to the dust action values, the process also includes: The data analysis module determines model parameters and establishes a random forest model based on grid search. The random forest model has 200 decision trees, a maximum depth of 10, a minimum number of samples for node splits of 8, a minimum number of samples for leaf nodes of 5, and stratified sampling based on feature categories during each split. The feature categories are dust properties, environmental factors, and operational parameters. Obtain a dust dataset and divide it into a training set and a test set. Train the random forest model using the training set and test the trained random forest model using the test set to determine the dust effect model.

[0009] Furthermore, when training the random forest model based on the training set and testing the trained random forest model based on the test set to determine the dust action model, the process includes: The data analysis module extracts a subsample set from the training set based on bootstrap sampling and assigns the subsample set to each decision tree. Each decision tree splits its nodes based on the subsample set and stratified sampling and completes the current training. The test set is then substituted into the currently trained random forest model. If the test value of the currently trained random forest model is less than the test value of the previously trained random forest model, then adjust the learning rate of the currently trained random forest model and continue training until the test value of the trained random forest model is greater than or equal to the test value of the previously trained random forest model. If the test value of the currently trained random forest model is greater than or equal to the test value of the previously trained random forest model, then training is stopped, and the currently trained random forest model is determined as the dust action model.

[0010] Furthermore, when determining the association result between the splashed dust image and all dust dispersion images based on the first dust data chain and association rule algorithm, and determining the spray command value based on the association result and Bayesian model, the process includes: The dust control analysis module generates several candidate item sets based on the mine dust information on the first dust data chain and the Eclat algorithm, and determines the frequent itemsets based on the support of the candidate item sets, and determines the association result based on the frequent itemsets. The dust control analysis module pre-trains the Bayesian model and substitutes the correlation results into the Bayesian model to determine the spray command value.

[0011] Furthermore, when comparing the spray command value with the historical spray database and determining the dust suppression precipitation amount of the dust suppression spray equipment based on the comparison results, the process includes: The historical spray database includes several historical spray command values ​​and several historical dust suppression precipitation values, and each historical spray command value corresponds to a historical dust suppression precipitation value. When a historical spray command value with the same spray command value exists in the historical spray database, the dust control module determines a historical precipitation reuse strategy. When there is no historical spray command value in the historical spray database that is the same as the spray command value, the dust control module determines the historical precipitation clustering strategy; When the historical precipitation reuse strategy is determined, if there is a unique historical spray command value in the historical spray database that is the same as the spray command value, then the historical dust suppression precipitation corresponding to the historical spray command value is determined as the dust suppression precipitation. If there are multiple historical spray command values ​​in the historical spray database that are the same as the spray command value, then the average of the historical dust control precipitation corresponding to each historical spray command value shall be determined as the dust control precipitation.

[0012] Furthermore, when comparing the spray command value with the historical spray database and determining the dust suppression precipitation amount of the dust suppression spray equipment based on the comparison results, the method further includes: When the historical precipitation clustering strategy is determined, the dust control module uses the spray command value and the historical spray database as the dust control dataset to be clustered, extracts the historical dust control precipitation corresponding to each historical spray command value in the dust control dataset to be clustered, determines the expected number of clusters k as 2, initializes the parameters of the Gaussian distribution, determines the probability that each data in the dust control dataset to be clustered belongs to each Gaussian distribution to determine the responsibility value, and determines the precipitation dataset corresponding to the spray command value based on the responsibility value. The average historical dust-prevention precipitation in the precipitation data is determined as the dust-prevention precipitation.

[0013] Furthermore, when obtaining the dust suppression spray load of the dust suppression spray equipment during the dust suppression period, and determining whether to adjust the dust suppression precipitation based on the dust suppression spray load, the process includes: The dust prevention period is directly proportional to the speed of the mine tunneling equipment; Obtain the dust suppression spray load and dust suppression spray usage load of the dust suppression spray equipment during the dust suppression period; When the dust suppression spray usage load exceeds the dust suppression spray capacity, the dust suppression treatment module determines to adjust the dust suppression precipitation amount; When the dust suppression spray usage load is less than or equal to the dust suppression spray load, the dust suppression treatment module determines not to adjust the dust suppression precipitation amount and uses the dust suppression precipitation amount for dust suppression.

[0014] Furthermore, when determining the precipitation adjustment factor based on the relationship between the dust suppression spray load and the dust suppression spray usage load, and adjusting the dust suppression precipitation amount based on the precipitation adjustment factor, the process includes: Obtain the ratio of the dust suppression spray load to the dust suppression spray usage load; The dustproof processing module determines a first ratio and a second ratio, wherein the first ratio is greater than the second ratio; The dust control module determines a first precipitation adjustment factor, a second precipitation adjustment factor, and a third precipitation adjustment factor, wherein the first precipitation adjustment factor is greater than the second precipitation adjustment factor, and the second precipitation adjustment factor is greater than the third precipitation adjustment factor; When the ratio is greater than the first ratio, the first precipitation adjustment factor is determined as the precipitation adjustment factor. When the ratio is less than or equal to the first ratio and greater than or equal to the second ratio, the second precipitation adjustment factor is determined as the precipitation adjustment factor. When the ratio is less than the second ratio, the third precipitation adjustment factor is determined as the precipitation adjustment factor. The amount of dust-prevention precipitation is directly proportional to the precipitation adjustment factor.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: The data analysis module divides dust control zones, constructs a mine dust data chain, and decomposes the first and second dust data chains by combining a dust action model. It also focuses on analyzing dust formation factors, avoiding analytical biases caused by neglecting the diffusion process between regions. The dust analysis module associates the images of splashed dust from mine tunneling equipment with the dust dispersion images of each dust control zone. It mines the inherent relationship between the two through association rule algorithms and outputs spray command values ​​by combining a Bayesian model. This avoids one-size-fits-all spray control and the waste of water resources caused by ineffective spraying. It ensures that the dust suppression effect always adapts to the dust generation rhythm, realizes the operational requirements of on-demand dust suppression, and improves the efficiency of dust control. The dust control module and the dust control module form a closed loop of load monitoring and intelligent adjustment. By comparing the historical spray database, the dust control precipitation is determined, avoiding pipeline overload caused by blindly setting the dust control precipitation. On the other hand, the load data of the spray equipment is acquired in real time, and the dust control module determines the precipitation adjustment factor based on the load relationship, dynamically controlling the dust control precipitation, ensuring the dust control stability of underground coal mining.

[0016] On the other hand, this application also provides a dust control method for underground coal mining, for applying the above-mentioned dust control system for underground coal mining, comprising: Several dust control zones are identified and mine dust information is collected for each zone. All the mine dust information is constructed into a mine dust data chain. Based on the dust action model, the dust action value is output. The mine dust data chain is then decomposed according to the dust action value to determine the first dust data chain and the second dust data chain. Acquire images of splashed dust from mining equipment and dust dispersion images from several dust control zones. Determine the association results between the splashed dust images and all dust dispersion images based on the first dust data chain and association rule algorithm. Determine the spray command value based on the association results and Bayesian model. The spray command value is compared with the historical spray database. The dust suppression precipitation of the dust suppression spray equipment is determined based on the comparison result. The dust suppression spray load of the dust suppression spray equipment during the dust suppression period is obtained. The dust suppression precipitation is adjusted based on the dust suppression spray load. When it is determined that the dust suppression precipitation should be adjusted, a precipitation adjustment factor is determined based on the relationship between the dust suppression spray load and the dust suppression spray usage load, and the dust suppression precipitation is adjusted based on the precipitation adjustment factor.

[0017] It is understandable that the dust control method and system described above for underground coal mining have the same beneficial effects, and will not be elaborated further here. Attached Figure Description

[0018] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a functional block diagram of a dust control system for underground coal mining, provided as an embodiment of the present invention.

[0019] Figure 2 A flowchart of a dust control method for underground coal mining provided in an embodiment of the present invention. Detailed Implementation

[0020] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] In some embodiments of this application, see Figure 1 As shown, a dust control system for underground coal mining includes: The data analysis module is configured to identify several dust control zones and collect mine dust information for each zone, construct a mine dust data chain from all the mine dust information, output dust action values ​​based on the dust action model, and decompose the mine dust data chain according to the dust action values ​​to determine the first dust data chain and the second dust data chain.

[0022] The dust control analysis module is configured to acquire images of splashed dust from mining equipment and dust dispersion images from several dust control zones, determine the association results between the splashed dust images and all dust dispersion images based on the first dust data chain and association rule algorithm, and determine the spray command value based on the association results and Bayesian model.

[0023] The dust control module is configured to compare the spray command value with the historical spray database, determine the dust control precipitation of the dust control spray equipment based on the comparison result, obtain the dust control spray load of the dust control spray equipment during the dust control period, and determine whether to adjust the dust control precipitation based on the dust control spray load.

[0024] The dust control module is configured to determine a precipitation adjustment factor based on the relationship between the dust spray load and the dust spray usage load when it is determined that the dust precipitation amount should be adjusted, and to adjust the dust precipitation amount based on the precipitation adjustment factor.

[0025] Specifically, the data analysis module identifies several dust control zones. The number of dust control zones is dynamically determined based on the area requiring dust suppression during underground coal mining. The module collects mine dust information from each dust control zone and constructs a mine dust data chain from all the dust information. This is because the distribution of mixed dust generated by processes such as cutting, drilling and blasting at the tunnel face, and coal transfer and transportation in underground coal mines is regional, with differences in dust concentration and sources in different areas. Constructing a mine dust data chain integrates the dust information from various dispersed areas, forming a complete dust data system that provides comprehensive data support for subsequent analysis. The data analysis module outputs dust action values ​​based on a dust action model and decomposes the mine dust data chain according to these values, determining the first and second dust data chains. The dust action model quantifies which factors are influencing dust generation and which are irrelevant. By decomposing the mine dust data chain based on the dust action values, subsequent modules can focus on key information regarding dust generation, thereby improving the targeting of dust control efforts. The dust control analysis module acquires images of splashed dust from the mine tunneling equipment and dust dispersion images from several dust control zones. The mine tunneling equipment is the immediate source of dust generation. Splashed dust images capture the initial state of dust generation, while dust dispersion images reflect the diffusion of dust in each control zone. Combining these two images allows for a complete understanding of the entire dust generation and diffusion process, avoiding the neglect of dust diffusion between areas and resulting in deviations in dust control. The dust control analysis module determines the correlation results between the splashed dust images and all dust dispersion images based on the first dust data chain and association rule algorithms. The first dust data chain integrates the direct influencing factors of dust generation, and the association rule algorithm can uncover the causal relationship between splashed dust (initial source) and dust dispersion (after diffusion) in each area. For example, when the drum cutting speed of the fully mechanized mining machine reaches 3-5 m / s, the impact on coal... The strong fracturing force of the layer leads to a correlation between the amount of splashed dust and the concentration of drifting dust in a certain dust control area. The association rule algorithm effectively avoids misjudgment of dust diffusion superposition. The spray command value is determined based on the association results and the Bayesian model. Since the drifting of underground dust is affected by uncertain factors such as airflow and mining progress, the association results of the splashed dust image and the drifting dust image obtained by the association rule algorithm alone cannot completely and accurately determine the spray command value. For example, although a certain dust control area is highly correlated with the mine tunneling equipment, if the current underground airflow suddenly weakens and the dust diffusion speed slows down, it may cause water waste if the spray command value is determined solely by the association rule algorithm. However, the Bayesian model can transform the influence of uncertain factors into prior probabilities under the same correlation in the past, avoiding the deviation of the spray command value caused by the complexity of dust conditions.

[0026] Understandably, the dust control module is responsible for comparing spray command values ​​with historical spray databases. These databases store the correspondence between spray command values ​​and dust suppression precipitation under different dust scenarios. By comparing these values, the module can quickly identify dust suppression precipitation that has been verified in practice, while reducing insufficient dust suppression or water waste caused by unreasonable parameters. Based on the comparison results, the module determines the dust suppression precipitation of the dust suppression spray equipment and obtains the dust suppression spray load (the current operating load of the dust suppression spray equipment, such as water pump pressure and motor power) within the dust suppression period. Since the dust suppression spray equipment has a maximum load limit, exceeding the load limit can lead to malfunctions such as water pump damage and pipeline rupture, which will not only interrupt dust suppression work but may also cause safety hazards. When determining the adjustment of dust suppression precipitation, the dust control module determines the precipitation adjustment factor based on the relationship between the dust suppression spray load and the dust suppression spray usage load (the safe operating load range designed for the dust suppression spray equipment). This ensures that the adjustment factor is neither too large, leading to insufficient dust suppression, nor too small, leading to equipment overload, thus ensuring the stability of the dust suppression effect. By dividing the dust control zone and disassembling the dust data chain, the source and influencing factors of dust are accurately located. Combined with association rule algorithms and Bayesian models, the association results of splash dust images and dust dispersion images are established, avoiding the limitations of "one-size-fits-all" dust control. Furthermore, the dust control module determines whether to adjust the dust precipitation by monitoring the dust spray load, while the dust control module determines the precipitation adjustment factor based on the load relationship. This ensures that the dust spray equipment always operates within the safe operating load, reduces failures caused by overload, guarantees the stability and reliability of dust control, and avoids the risk of a sudden increase in dust concentration caused by the shutdown of the dust spray equipment.

[0027] In some embodiments of this application, when constructing a mine dust data chain from all mine dust information, outputting dust action values ​​based on a dust action model, and disassembling the mine dust data chain according to the dust action values, the process includes: a data analysis module acquiring a first dust shelving chain and a second dust shelving chain, substituting each piece of mine dust information on the mine dust data chain into the dust action model, outputting the dust action value of each piece of mine dust information, comparing the dust action value of each piece of mine dust information with a dust action threshold, transferring all mine dust information with dust action values ​​greater than or equal to the dust action threshold to the first dust shelving chain, and determining the first dust data chain; transferring all mine dust information with dust action values ​​less than the dust action threshold to the second dust shelving chain, and determining the second dust data chain.

[0028] Specifically, mine dust information includes blasting operation parameters, coal seam hardness, coal seam moisture content, rock strata composition and structure, the height and angle of the transfer points of belt conveyors and scraper conveyors, air temperature, humidity, sunlight, number of workers, mine surface elevation, equipment type and operating parameters of mine tunneling equipment, etc. The data analysis module obtains the first and second dust storage chains in advance. Because the mine dust data chain integrates dust information from multiple regions and of multiple types, direct analysis and processing within the mine dust data chain can easily lead to a mixture of dust information with high and low impact factors. Both the first and second dust storage chains are equipped with multiple data chain nodes and connecting sections. The connecting sections are used to connect the two data chain nodes, providing a dedicated container for the classification of mine dust information. Each piece of mine dust information in the mine dust data chain is substituted into the dust action model to output the corresponding dust action value. The dust action model can calculate the impact of mine dust information one by one, avoiding deviations in the overall calculation. The dust effect value is compared with the dust effect threshold, which is dynamically determined based on coal mine safety regulations and health protection requirements. The dust effect threshold is a critical value in the 0-10 quantitative range output by the dust effect model. The dust effect threshold is the dividing value that distinguishes between direct influencing factors and irrelevant factors of dust generation. Each dust effect value represents its corresponding mine dust information, that is, the factor that causes the dust concentration to exceed the safe range or poses a risk to the health of workers. In this embodiment, the value is preferably 7. For example, in the "Coal Mine Safety Regulations", the free SiO2 in coal dust is <10%, and the respirable dust PC-TWA is 2.5 mg / m³. 3 Total dust PC-TWA is 4 mg / m³ 3 The requirement is 2.5 mg / m². 3 With 4mg / m 3The ratio is 0.625, meaning that when the respirable particulate matter (PC-TWA) in coal dust is greater than 2.5, it directly affects the dust concentration in the environment. Without dust control measures, the dust concentration poses a threat to the health of workers. Considering the influence of airflow and other conditions on the dispersion of respirable dust, the dust effect threshold is set to 7 when the dust effect model maps the dust effect value (range 0-10). All mine dust information with dust effect values ​​greater than or equal to the dust effect threshold is transferred to the first dust shelving chain, and the first dust data chain is determined. The first dust data chain is a data chain directly related to dust formation. All mine dust information with dust effect values ​​less than the dust effect threshold is transferred to the second dust shelving chain, and the second dust data chain is determined. The second dust data chain is a data chain with almost no direct relationship to dust formation. This includes blasting operation parameters, coal seam hardness, coal seam moisture content, rock strata composition and structure, belt conveyor and scraper... The height and angle of the transfer point of the conveyor belt, the type of equipment and operating parameters of the mine tunneling equipment have a direct relationship with the formation of dust. The higher the corresponding dust effect value, the more likely the first dust data chain is generated. On the other hand, the influence of temperature, humidity, light, number of workers and well surface elevation on dust formation is relatively limited and has almost no direct relationship. The lower the corresponding dust effect value, the more likely the second dust data chain is generated. Through dedicated carriers and individual evaluation, accurate classification of mine dust information is achieved, thereby improving the reliability and stability of dust prevention.

[0029] In some embodiments of this application, when constructing all mine dust information into a mine dust data chain, outputting dust effect values ​​based on a dust effect model, and decomposing the mine dust data chain according to the dust effect values, the method further includes: a data analysis module determining model parameters based on grid search and establishing a random forest model. The random forest model has 200 decision trees, a maximum depth of 10, a minimum number of samples for node splits of 8, a minimum number of samples for leaf nodes of 5, and performs stratified sampling based on feature categories during each split. The feature categories are dust attributes, environmental factors, and operational parameters. The dust dataset is obtained and divided into a training set and a test set. The random forest model is trained based on the training set and tested based on the test set to determine the dust effect model.

[0030] Specifically, the data analysis module determines the model parameters of the random forest model through grid search. Random forest models excel at handling multi-feature, non-linear data, and the ensemble prediction of multiple decision trees reduces the overall model error, improving the stability of dust impact value calculation. Grid search exhaustively searches the parameter space for optimal parameter combinations, such as the number of decision trees and maximum depth, to build the random forest model. The final determined parameters are: 200 decision trees, a maximum depth of 10, a minimum number of samples for node splits of 8, and a minimum number of samples for leaf nodes of 5. Generally, too few decision trees reduce model stability, while too many increase computational costs. 200 decision trees strike a balance between accuracy and efficiency. A maximum depth limit of 10 prevents the model from overfitting the training data, avoiding reduced generalization ability due to excessive learning of details. Limits on the number of samples for node splits and leaf nodes ensure that each decision tree can split based on sufficient sample data, preventing bias in dust impact values ​​caused by accidental sample data. The random forest model uses stratified sampling according to feature categories (dust properties, environmental factors, and operational parameters). In mine dust information, different information features have varying impacts on dust concentration values. Dust properties refer to the inherent parameters of the material that directly determine dust generation characteristics, including coal seam hardness, coal seam moisture content, and rock strata composition and structure. These parameters are the physicochemical properties of the coal and rock strata themselves, directly influencing the material basis for dust formation. Environmental factors refer to underground environmental conditions that affect the dust's existence but do not directly participate in dust generation, including temperature, humidity, light, and wellhead elevation. These parameters are characteristics of the underground natural environment or spatial location. Operational parameters refer to the operational and equipment parameters directly related to human mining operations, including blasting operation parameters, the height and angle of the transfer points of belt conveyors and scraper conveyors, the number of workers, and the equipment type and operating parameters of mine tunneling equipment. Stratified sampling ensures that each type of feature is representative during training, preventing any type of feature from being marginalized, and ensuring that the model can comprehensively capture the correlation between various factors and dust concentration values. A dust dataset is obtained and divided into training and testing sets. The dust dataset includes information on mine dust from different periods and corresponding safety indicators for underground dust concentration. The dust dataset is divided into training and testing sets, usually in a 7:3 ratio, to ensure the generalization ability of the model. The training set is used to train the random forest model, while the testing set is used to test the performance of the trained model and verify its predictive ability on new data. The random forest model confirmed by training and testing is used as the dust effect model.

[0031] Understandably, the parameter optimization and stratified sampling of grid search ensure that the model can accurately capture the impact of various factors on dust, improve the accuracy of dust effect value determination, and provide a reliable basis for the decomposition of the mine dust data chain. The overfitting ability and generalization performance of random forest enable the model to output dust effect value stably when the underground environment changes, ensuring the consistency of classification logic, thereby improving the reliability and stability of the dust control system.

[0032] In some embodiments of this application, when training a random forest model based on a training set, testing the trained random forest model based on a test set, and determining the dust action model, the process includes: a data analysis module extracting a subsample set from the training set based on bootstrap sampling and assigning a subsample set to each decision tree; each decision tree splitting nodes based on the subsample set and stratified sampling and completing the current training; substituting the test set into the currently trained random forest model; if the test value of the currently trained random forest model is less than the test value of the previously trained random forest model, adjusting the learning rate of the currently trained random forest model and continuing training until the test value of the trained random forest model is greater than or equal to the test value of the previously trained random forest model; if the test value of the currently trained random forest model is greater than or equal to the test value of the previously trained random forest model, stopping training and determining the currently trained random forest model as the dust action model.

[0033] Specifically, the data analysis module extracts a subset of samples from the training set based on bootstrap sampling and assigns a subset of samples to each decision tree. Bootstrap sampling randomly samples with replacement from the training set, making the training data of each decision tree different, thereby enhancing the diversity between decision trees. If all decision trees are trained using the same data, the model will become overly dependent on specific samples, reducing its generalization ability. Differentiated subsets of samples allow different decision trees to learn different features of the data, improving the model's adaptability to complex dusty underground environments. Each decision tree splits nodes based on a subset of samples and stratified sampling, completing the current training. During node splitting, the model selects the optimal split point based on the influence of features on dust impact values ​​(such as the correlation between blasting charge and dust impact values). It learns patterns and relationships in the data during each training iteration to improve its prediction or classification capabilities, avoiding the bias caused by a single feature dominating the split. The test set is substituted into the trained model for testing, with metrics including accuracy, loss function value, and recall to measure model performance. If the current test value is lower than the previous test value, it indicates a performance decline, requiring adjustment of the learning rate of the current random forest model before continuing training. This helps the model more stably approach the global optimum. If the current test value is greater than or equal to the previous test value, it indicates stable model performance and improved generalization ability, allowing training to stop as the model has reached a satisfactory performance level. By training the random forest model to obtain the dust impact model, the dust impact values ​​for each mine dust information can be accurately output, ensuring the reliability and stability of dust control.

[0034] Understandably, bootstrap sampling and hierarchical node splitting ensure the accuracy of the model in determining dust impact values. Iterative training optimization through dynamic adjustment of the learning rate effectively avoids the risk of overfitting, enabling the model to stably output dust impact values ​​even when the underground environment changes. This ensures the reliability of the mine dust data chain decomposition. The integrated prediction of multiple decision trees reduces the overall error of the model and provides a basis for the division of the first and second dust data chains, thereby improving the stability and intelligence of the dust control system at the model level.

[0035] In some embodiments of this application, when determining the association results of the splashed dust image and all dust dispersion images based on the first dust data chain and the association rule algorithm, and determining the spray command value based on the association results and the Bayesian model, the process includes: the dust control analysis module generating several candidate item sets based on the mine dust information on the first dust data chain and the Eclat algorithm on the image features of the splashed dust image and all dust dispersion images; determining frequent itemsets based on the support of the candidate item sets; determining the association results based on the frequent itemsets; the dust control analysis module pre-training the Bayesian model; and substituting the association results into the Bayesian model to determine the spray command value.

[0036] Specifically, the underground mining environment in coal mines is complex, harsh, and highly variable, and underground video surveillance suffers from image degradation due to factors such as low lighting conditions. Based on the U-Net neural network for predicting future video frames, which integrates multi-scale appearance feature extraction and temporal information extraction, it can specifically identify complex dynamic patterns between video frames by analyzing the temporal information of consecutive video frames containing flying dust and dust dispersion. Training a G network to predict future video frames improves the image quality of the predicted frames. First, MSTP-Net is used as the G network, with the sequence of video frames (I1, I2, I3, ..., It) before the current frame It+1 as the input tensor, and the predicted frame I*t+1 as the output tensor. A PatchGAN structure is used to distinguish the differences between the real and predicted frames, prompting the G network to generate images consistent with the real frames, making the predicted frame I*t+1 more similar to the real frame It+1. The MSTP-Net model architecture is mainly designed based on the U-Net structure, and the overall network architecture of MSTP-Net includes encoding and decoding parts. The network's input and output sizes are both 256×256×3. All convolutional and deconvolutional kernels are set to 3×3, and the max-pooling layer's kernel size is set to 2×2. Based on the U-Net architecture, the model, when detecting splashed dust and dust dispersion in videos, not only considers the extracted multi-scale spatial features but also compensates for the loss of some image detail information caused by downsampling operations. This ensures the reliability and stability of both splashed dust images and all dust dispersion images.

[0037] Understandably, the dust control analysis module uses the mine dust information (such as tunneling equipment parameters, blasting parameters, etc.) from the first dust data chain, combined with the Eclat algorithm to process the image features of splashed dust images (direct dust generation images from tunneling equipment) and dust dispersion images (dust diffusion images from each control area), such as dust concentration distribution and dispersion trajectory, to generate several candidate sets. Candidate itemsets are combinations of image features and dust data, such as images of splashing dust from high-power tunneling equipment and images of high-concentration dust dispersion in transport tunnels. These are used to initially capture potential associations. Next, the support of the candidate itemsets is calculated, which is the frequency of the combination in the data. Frequent itemsets with support greater than the support threshold are selected. The specific size of the support threshold is adjusted according to the tunneling speed and dust control sensitivity underground. This embodiment does not impose a specific limitation. The support ensures that the association results are statistically significant (not accidental). Frequent itemsets are the results of stable association between splashing dust images and dust dispersion images. For example, the splashing intensity of the tunneling equipment is highly correlated with the dispersion concentration in the No. 3 dust control zone. The dust control analysis module substitutes the association results into a pre-trained Bayesian model to determine the spray command value. The training process of the Bayesian model is consistent with that of the dust action model, and will not be repeated here. The association rule algorithm accurately locates the correlation between dust generation conditions and corresponding dust diffusion, while the Bayesian model further mines potential correlation features based on real-time correlation results and outputs spray command values. This ensures that dust control can effectively adapt to the complex environment downhole, reduce the risk of over-spraying or under-spraying, reduce resource consumption while ensuring dust control effect, and enhance the reliability of the system.

[0038] In some embodiments of this application, when comparing spray command values ​​in a historical spray database and determining the dust suppression precipitation amount of the dust suppression spray device based on the comparison results, the following steps are taken: the historical spray database includes several historical spray command values ​​and several historical dust suppression precipitation amounts, and each historical spray command value corresponds to a historical dust suppression precipitation amount. When there is a historical spray command value in the historical spray database that is the same as the spray command value, the dust suppression processing module determines a historical precipitation reuse strategy. When there is no historical spray command value in the historical spray database that is the same as the spray command value, the dust suppression processing module determines a historical precipitation clustering strategy. When the historical precipitation reuse strategy is determined, if there is only one historical spray command value in the historical spray database that is the same as the spray command value, then the historical dust suppression precipitation amount corresponding to that historical spray command value is determined as the dust suppression precipitation amount. If there are not only one historical spray command value in the historical spray database that is the same as the spray command value, then the average of the historical dust suppression precipitation amounts corresponding to each historical spray command value is determined as the dust suppression precipitation amount.

[0039] In some embodiments of this application, when comparing the spray command value with the historical spray database and determining the dust suppression precipitation amount of the dust suppression spray equipment based on the comparison result, the method further includes: when the historical precipitation clustering strategy is determined, the dust suppression processing module uses the spray command value and the historical spray database as the dust suppression dataset to be clustered, extracts the historical dust suppression precipitation amount corresponding to each historical spray command value in the dust suppression dataset to be clustered, determines the expected number of clusters k as 2, initializes the parameters of the Gaussian distribution, determines the probability that each data in the dust suppression dataset to be clustered belongs to each Gaussian distribution to determine the responsibility value, determines the precipitation dataset corresponding to the spray command value based on the responsibility value, and determines the mean of the historical dust suppression precipitation amount in the precipitation dataset as the dust suppression precipitation amount.

[0040] Specifically, the historical spray database stores several historical spray command values ​​and several historical dust suppression precipitation amounts. The one-to-one correspondence between each historical spray command value and a historical dust suppression precipitation amount provides historical experience for current dust suppression decisions. When a spray command value is generated, the dust suppression processing module searches the historical spray database to see if there is a historical spray command value that is the same as the spray command value. If there is a historical spray command value that is the same, it means that corresponding historical dust suppression precipitation amounts have been used for dust suppression operations at different times. Therefore, the dust suppression processing module executes a historical precipitation reuse strategy. If the same historical spray command value is unique, its corresponding historical dust suppression precipitation amount can be directly reused, thereby ensuring the consistency of dust suppression operations. If the same historical spray command value is not unique, such as if the historical spray command value corresponds to different historical dust suppression precipitation amounts on different dates, the average of the historical precipitation amounts is taken as the current dust suppression precipitation amount. This avoids the random deviation of a single historical data point and thus balances the subtle differences between different historical scenarios. If no identical historical spray command value exists, it means there is no direct reference in the historical data. In this case, a historical precipitation clustering strategy is executed. By matching clusters, the historical data closest to the spray command value is found, ensuring the flexibility and reliability of dust suppression precipitation. Reusing historical data makes dust suppression precipitation more closely match the actual situation downhole, reducing the risk of insufficient or excessive spraying, while avoiding water waste and improving the stability and reliability of the dust suppression system.

[0041] Understandably, by analyzing historical spray databases using clustering algorithms, the precipitation dataset closest to the current spray command value is found, improving the accuracy of dust suppression precipitation determination. Under different downhole conditions, the dust suppression precipitation can be adaptively determined, thus maintaining the stability and efficiency of dust suppression operations. At the same time, it reduces reliance on human experience and judgment, avoids deviations in manually setting dust suppression precipitation, improves the automation level and reliability of the overall system, avoids excessive or insufficient dust suppression precipitation, thereby optimizing resource utilization and reducing the waste of dust suppression energy.

[0042] In some embodiments of this application, when obtaining the dust suppression spray load of the dust suppression spray equipment during the dust suppression period and determining whether to adjust the dust suppression precipitation based on the dust suppression spray load, the process includes: the dust suppression period is directly proportional to the speed of the mine tunneling equipment; obtaining the dust suppression spray load and the dust suppression spray usage load of the dust suppression spray equipment during the dust suppression period; when the dust suppression spray usage load is greater than the dust suppression spray load, the dust suppression processing module determines to adjust the dust suppression precipitation; when the dust suppression spray usage load is less than or equal to the dust suppression spray load, the dust suppression processing module determines not to adjust the dust suppression precipitation and uses the dust suppression precipitation for dust suppression.

[0043] In some embodiments of this application, when determining a precipitation adjustment factor based on the relationship between dust suppression spray load and dust suppression spray usage load, and adjusting the dust suppression precipitation based on the precipitation adjustment factor, the process includes: obtaining the ratio of dust suppression spray load to dust suppression spray usage load; a dust suppression treatment module determining a first ratio and a second ratio, wherein the first ratio is greater than the second ratio; a dust suppression treatment module determining a first precipitation adjustment factor, a second precipitation adjustment factor, and a third precipitation adjustment factor, wherein the first precipitation adjustment factor is greater than the second precipitation adjustment factor, and the second precipitation adjustment factor is greater than the third precipitation adjustment factor; when the ratio is greater than the first ratio, the first precipitation adjustment factor is determined as the precipitation adjustment factor; when the ratio is less than or equal to the first ratio and greater than or equal to the second ratio, the second precipitation adjustment factor is determined as the precipitation adjustment factor; when the ratio is less than the second ratio, the third precipitation adjustment factor is determined as the precipitation adjustment factor; and the dust suppression precipitation is directly proportional to the precipitation adjustment factor.

[0044] Specifically, the dust control module determines the dust control period, which is directly proportional to the speed of the mine tunneling equipment. The faster the tunneling equipment travels, the more coal and rock is broken per unit time, and the longer the dust generation lasts. Furthermore, the mine tunneling equipment is a major factor directly causing dust generation. Therefore, the dust control period is determined according to a direct proportional relationship. The faster the tunneling equipment travels, the longer the corresponding dust control period will be. The dust control module obtains the dust control spray load and dust control spray usage load within the dust control period. The dust suppression spray load is the rated safe load of the spray equipment (such as the maximum allowable pressure of the water pump and the rated power of the motor), representing the maximum operating limit that the equipment can withstand. The dust suppression spray usage load is the current actual operating load of the spray equipment (such as the real-time pressure of the water pump and the working power of the motor), reflecting the current operating intensity of the spray equipment. If the dust suppression spray usage load (current load) is less than or equal to the dust suppression spray load (safety limit), it indicates that the spray equipment has a certain operating redundancy or is just balanced with the operating limit. For example, if the rated pressure is 1.2MPa and the current pressure is only 0.8MPa, it means that the spray equipment can meet the current dust suppression precipitation, so no adjustment is made to the dust suppression precipitation, and dust suppression operation is carried out based on the dust suppression precipitation. If the dust suppression spray usage load is greater than the dust suppression spray load, it means that the spray equipment has exceeded the safety limit. Continuing to execute the dust suppression precipitation will cause the water pump to overload and be damaged, the pipeline to rupture, and thus interrupt the dust suppression work. By comparing the dust suppression spray load and the dust suppression spray usage load, maintenance costs and the risk of dust suppression interruption can be reduced, ensuring the stability and reliability of the dust suppression system.

[0045] Understandably, the dust control module obtains the ratio of the dust suppression spray load (the equipment's rated safe load) to the dust suppression spray usage load (the equipment's current actual load). If the ratio is greater than or equal to 1, it indicates that the spraying equipment has a certain operational redundancy or is just balanced with the operating limit. If the ratio is less than 1, it indicates that the spraying equipment has exceeded the safe limit and needs to be adjusted. The preferred first ratio determined by the dust control module is 0.8, the preferred second ratio is 0.5, the preferred first precipitation adjustment factor is 0.7, the preferred second precipitation adjustment factor is 0.5, and the preferred third precipitation adjustment factor is 0.3. The lower the ratio, the more the dust suppression load needs to be reduced. The more water used, the less risk of overloading the spraying equipment. When the ratio is greater than the first ratio but less than 1, it indicates that the spraying equipment has exceeded its safety limit, but the extent of the exceedance is limited. A larger first precipitation adjustment factor is needed for adjustment. When the ratio is less than or equal to the first ratio but greater than or equal to the second ratio, it indicates that the spraying equipment has exceeded its safety limit, but the extent of the exceedance is relatively moderate. In this case, the second precipitation adjustment factor is determined as the precipitation adjustment factor to adjust the load balance of the spraying equipment. When the ratio is less than the second ratio, it indicates that the spraying equipment has severely exceeded its safety limit, requiring a significant reduction in dust suppression precipitation to avoid pump overload damage and pipeline rupture. Assuming the dust suppression precipitation is V and the precipitation adjustment factor is M, the adjusted dust suppression precipitation is determined to be V*M. When it is necessary to reduce the amount of precipitation for dust suppression, the precipitation adjustment factor will be reduced accordingly. By establishing a direct proportional relationship between the amount of precipitation for dust suppression and the precipitation adjustment factor, precise control of the amount of precipitation for dust suppression is achieved during the dust suppression process. This ensures the safety of the spraying equipment, avoids the risk of pump damage and pipeline rupture due to overload, and the segmented precipitation adjustment factor makes the change in the amount of precipitation for dust suppression relatively gradual, avoiding the impact of sudden changes in the amount of precipitation for dust suppression on the continuity of dust suppression or the spraying equipment, and ensuring the stable operation of the dust suppression system.

[0046] In summary, the beneficial effects of this invention are as follows: The data analysis module divides dust control zones, constructs a mine dust data chain, and decomposes the first and second dust data chains using a dust action model. It also focuses on analyzing dust formation factors, avoiding analytical biases caused by neglecting the diffusion process between regions. The dust analysis module correlates images of splashing dust from mine tunneling equipment with dust dispersion images from each dust control zone. It uses association rule algorithms to uncover the inherent relationship between the two, and then combines this with a Bayesian model to output spray command values. This avoids a one-size-fits-all approach to spray control and prevents water waste caused by ineffective spraying. It ensures that the dust suppression effect always matches the dust generation rhythm, achieving on-demand dust control and improving dust control efficiency. The dust control module and the dust control module form a closed loop of load monitoring and intelligent adjustment. By comparing the historical spray database, the dust control precipitation is determined, avoiding pipeline overload caused by blindly setting the dust control precipitation. On the other hand, the load data of the spray equipment is acquired in real time, and the dust control module determines the precipitation adjustment factor based on the load relationship, dynamically controlling the dust control precipitation, ensuring the dust control stability of underground coal mining.

[0047] In another preferred embodiment based on the above embodiments, see [reference] Figure 2 As shown, this embodiment provides a dust control method for underground coal mining, which applies the aforementioned dust control system for underground coal mining, including: S100: Determine several dust control zones and collect mine dust information for each zone. Construct a mine dust data chain from all the mine dust information. Output dust action values ​​based on the dust action model. Decompose the mine dust data chain according to the dust action values ​​to determine the first dust data chain and the second dust data chain.

[0048] S200: Acquire images of splashed dust from mining equipment and dust dispersion images from several dust control zones; determine the association results between the splashed dust images and all dust dispersion images based on the first dust data chain and association rule algorithm; and determine the spray command value based on the association results and Bayesian model.

[0049] S300: Compare the spray command value with the historical spray database, determine the dust suppression precipitation amount of the dust suppression spray equipment based on the comparison results, obtain the dust suppression spray load of the dust suppression spray equipment during the dust suppression period, and determine whether to adjust the dust suppression precipitation amount based on the dust suppression spray load.

[0050] S400: When it is determined that the dust suppression precipitation should be adjusted, a precipitation adjustment factor is determined based on the relationship between the dust suppression spray load and the dust suppression spray usage load, and the dust suppression precipitation is adjusted based on the precipitation adjustment factor.

[0051] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0052] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0053] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0054] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A dust control system for use in underground coal mining, characterised in that, Comprise: a data analysis module configured to determine a plurality of dust control zones and collect mine dust information of each dust control zone, construct all mine dust information into a mine dust data chain, output a dust effect value based on a dust effect model, and disassemble the mine dust data chain based on the dust effect value to determine a first dust data chain and a second dust data chain; a dust analysis module configured to obtain a splashing dust image of a mine tunneling equipment and a dust drift image of a plurality of dust control zones, determine an association result of the splashing dust image and all dust drift images based on the first dust data chain and an association rule algorithm, and determine a spraying instruction value based on the association result and a Bayesian model; a dust treatment module configured to compare the spraying instruction value in a historical spraying database, determine a dust precipitation amount of a dust spraying equipment according to the comparison result, obtain a dust spraying load of the dust spraying equipment in a dust control period, and determine whether to adjust the dust precipitation amount according to the dust spraying load; a dust control module configured to, when it is determined to adjust the dust precipitation amount, determine a precipitation adjustment factor based on a relationship between the dust spraying load and a dust spraying usage load, and adjust the dust precipitation amount based on the precipitation adjustment factor.

2. A dust control system for use in underground coal mining according to claim 1 wherein, When all mine dust information is constructed into a mine dust data chain, a dust effect value is output based on a dust effect model, and the mine dust data chain is disassembled based on the dust effect value, comprising: The data analysis module obtains a first dust data chain and a second dust data chain, and substitutes each mine dust information on the mine dust data chain into the dust effect model, and outputs a dust effect value of each mine dust information; Compare the dust effect value of each mine dust information with a dust effect threshold value; Transfer all mine dust information with a dust effect value greater than or equal to the dust effect threshold value to the first dust data chain, and determine the first dust data chain; Transfer all mine dust information with a dust effect value less than the dust effect threshold value to the second dust data chain, and determine the second dust data chain.

3. A dust control system for use in underground coal mining according to claim 2, characterised in that, When all mine dust information is constructed into a mine dust data chain, a dust effect value is output based on a dust effect model, and the mine dust data chain is disassembled based on the dust effect value, further comprising: The data analysis module determines model parameters based on grid search and establishes a random forest model; The number of decision trees of the random forest model is 200, the maximum depth of the decision tree is 10, the minimum sample number of node splitting is 8, the minimum sample number of leaf node is 5, and stratified sampling is performed according to feature categories during each splitting, the feature categories are dust attributes, environmental factors and operation parameters; Obtain a dust data set and divide the dust data set into a training set and a test set, train the random forest model according to the training set, test the trained random forest model according to the test set, and determine the dust effect model.

4. A dust control system for use in underground coal mining according to claim 3, characterised in that, In the process of training the random forest model according to the training set, testing the trained random forest model according to the test set, and determining the dust action model, the method comprises: The data analysis module extracts a sub-sample set from the training set based on bootstrap sampling, and assigns the sub-sample set to each decision tree. Each decision tree splits nodes and completes current training based on the sub-sample set and stratified sampling. The test set is substituted into the trained random forest model; If the test value of the current trained random forest model is less than the test value of the previous trained random forest model, the learning rate of the current trained random forest model is adjusted, and the training is continued until the test value of the trained random forest model is greater than or equal to the test value of the previous trained random forest model; If the test value of the current trained random forest model is greater than or equal to the test value of the previous trained random forest model, the training is stopped, and the current trained random forest model is determined as the dust action model.

5. A dust control system for use in underground coal mining according to claim 4, characterised in that, In the process of determining the association result of the splashing dust image and the total dust dispersion image based on the first dust data chain and the association rule algorithm, and determining the spraying instruction value based on the association result and the Bayesian model, the method comprises: The dust analysis module generates a number of candidate item sets based on the image features of the splashing dust image and the total dust dispersion image and the Eclat algorithm on the mine dust information on the first dust data chain, determines the frequent item set according to the support degree of the candidate item set, and determines the association result based on the frequent item set; The dust analysis module pre-trains the Bayesian model, and substitutes the association result into the Bayesian model to determine the spraying instruction value.

6. A dust control system for use in underground coal mining according to claim 5, characterised in that, In the process of comparing the spraying instruction value in the historical spraying database, and determining the dust precipitation amount of the dust spraying device according to the comparison result, the method comprises: The historical spraying database comprises a plurality of historical spraying instruction values and a plurality of historical dust precipitation amounts, and each historical spraying instruction value corresponds to a historical dust precipitation amount; When there is a historical spraying instruction value identical to the spraying instruction value in the historical spraying database, the dust treatment module determines a historical precipitation reuse strategy; When there is no historical spraying instruction value identical to the spraying instruction value in the historical spraying database, the dust treatment module determines a historical precipitation clustering strategy; When the historical precipitation reuse strategy is determined, if there is only one historical spraying instruction value identical to the spraying instruction value in the historical spraying database, the historical dust precipitation amount corresponding to the historical spraying instruction value is determined as the dust precipitation amount; If there are not only one historical spraying instruction value identical to the spraying instruction value in the historical spraying database, the average of the historical dust precipitation amounts corresponding to each historical spraying instruction value is determined as the dust precipitation amount.

7. A dust control system for use in underground coal mining according to claim 6, characterised in that, In the process of comparing the spraying instruction value in the historical spraying database, and determining the dust precipitation amount of the dust spraying device according to the comparison result, the method further comprises: When the historical precipitation clustering strategy is determined, the dust control module takes the spray instruction value and the historical spray database as a clustering dust control data set, extracts the historical dust precipitation corresponding to each historical spray instruction value in the clustering dust control data set, determines the expected cluster number k as 2, initializes the parameters of the Gaussian distribution, determines the responsibility value of each data in the clustering dust control data set belonging to each Gaussian distribution, and determines the precipitation data set corresponding to the spray instruction value according to the responsibility value; The mean value of the historical dust precipitation in the precipitation data set is determined as the dust precipitation.

8. A dust control system for use in underground coal mining according to claim 7, characterised in that, When the dust control time period is proportional to the speed of the mine tunneling equipment, the dust control spray load of the dust control spray equipment in the dust control time period is obtained, and whether the dust precipitation is adjusted is determined according to the dust control spray load, comprising: The dust control time period is proportional to the speed of the mine tunneling equipment; The dust control spray load and the dust control spray use load of the dust control spray equipment in the dust control time period are obtained; When the dust control spray use load is greater than the dust control spray load, the dust control module determines to adjust the dust precipitation; When the dust control spray use load is less than or equal to the dust control spray load, the dust control module determines not to adjust the dust precipitation, and the dust control is performed with the dust precipitation.

9. A dust control system for use in underground coal mining according to claim 8, characterised in that, When the relationship between the dust control spray load and the dust control spray use load is determined, the dust control module determines the precipitation adjustment factor based on the relationship between the dust control spray load and the dust control spray use load, and adjusts the dust precipitation based on the precipitation adjustment factor, comprising: The ratio of the dust control spray load to the dust control spray use load is obtained; The dust control module determines a first ratio and a second ratio, wherein the first ratio is greater than the second ratio; The dust control module determines a first precipitation adjustment factor, a second precipitation adjustment factor and a third precipitation adjustment factor, and the first precipitation adjustment factor is greater than the second precipitation adjustment factor, and the second precipitation adjustment factor is greater than the third precipitation adjustment factor; When the ratio is greater than the first ratio, the first precipitation adjustment factor is determined as the precipitation adjustment factor; When the ratio is less than or equal to the first ratio and greater than or equal to the second ratio, the second precipitation adjustment factor is determined as the precipitation adjustment factor; When the ratio is less than the second ratio, the third precipitation adjustment factor is determined as the precipitation adjustment factor; The dust precipitation is proportional to the precipitation adjustment factor.

10. A dust control method for underground coal mining, for use with a dust control system for underground coal mining as claimed in any one of claims 1 to 9, characterised in that, Comprising: A plurality of dust control zones are determined, and mine dust information of each dust control zone is collected, all mine dust information is constructed as a mine dust data chain, a dust effect value is output based on a dust effect model, and the mine dust data chain is disassembled based on the dust effect value, a first dust data chain and a second dust data chain are determined; The dust control module obtains a splashing dust image of the mine tunneling equipment and a dust floating image of a plurality of dust control zones, determines the association result of the splashing dust image and all dust floating images based on the first dust data chain and the association rule algorithm, and determines the spray instruction value based on the association result and the Bayesian model; The spray instruction value is compared in a historical spray database, a dustproof precipitation of the dustproof spray equipment is determined according to a comparison result, a dustproof spray load of the dustproof spray equipment in a dustproof time period is obtained, and it is determined whether to adjust the dustproof precipitation according to the dustproof spray load; When it is determined to adjust the dustproof precipitation, a precipitation adjustment factor is determined based on a relationship between the dustproof spray load and a dustproof spray use load, and the dustproof precipitation is adjusted based on the precipitation adjustment factor.

Images