A method and system for monitoring ventilation in underground coal mines
Through the random forest model, the ventilation parameter value of coal mine underground tunnels is predicted, combined with mean square error and abnormal probability, and the prompt information is output, which solves the problem of long maintenance time limit of ventilation equipment in the existing technology, and realizes efficient monitoring and early warning of underground tunnels of coal mine underground tunnels. .
Patent Information
- Application Number
- CN202411312029.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-20
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-09-20
AI Technical Summary
The prior art is difficult to achieve effective monitoring of downhole tunnel ventilation in coal mines, resulting in a long maintenance period of ventilation equipment and it is difficult to ensure a good ventilation environment in the tunnel.
By obtaining multiple random forest models, using these models to predict ventilation parameter values, and output prompt information when there is an abnormality in the predicted value. Combining mean square error and abnormal probability calculation, the most accurate model is determined for prediction to achieve ventilation of coal mine tunnels Better monitoring.
Accurate prediction of the ventilation parameter values of coal mine tunnels is achieved, potential abnormalities are discovered in advance, the time limit requirements for ventilation equipment maintenance is reduced, and the monitoring efficiency and accuracy of ventilation environment is improved.
Smart Images

Figure CN118836051B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of safety monitoring, and in particular to a method and system for monitoring ventilation in underground coal mines. Background Art
[0002] A coal mine is an area rich in coal where coal resources are mined. It is generally divided into underground coal mines and open-pit coal mines. Usually, tunnels are excavated underground for coal mining. Coal mines usually include tunnels, shafts, and mining faces. If the ventilation of the coal mine tunnels is abnormal, it will seriously affect the environmental safety of the tunnels, thereby threatening the health of the workers in the tunnels. Therefore, the ventilation of the coal mine tunnels can be monitored to ensure good ventilation of the coal mine tunnels, for example, it helps to deal with abnormal ventilation equipment in a timely manner.
[0003] In order to realize the monitoring of ventilation of coal mine tunnels, in the relevant technology, for example, the Chinese patent application document with publication number CN113738428A, a monitoring method for local ventilation in underground coal mines is provided, in which monitoring devices are arranged at the air outlet of each fan of local ventilation and the air outlet of the upper wind duct, and all monitoring devices are connected to the monitoring station, in which the specified air leakage rate and air supply volume of the working face are set, and the monitoring station has an alarm function; each monitoring device measures the air volume and air pressure at its location, the monitoring station monitors whether the air volume and air pressure of the two fans match, the monitoring station monitors whether the air leakage rate at the air outlet matches the specified air leakage rate, the monitoring station monitors whether the air outlet volume of the wind duct matches the air supply volume of the working face, and if they do not match, a corresponding prompt is issued, and the power mechanism of the working face is cut off when the air supply volume of the working face cannot match.
[0004] The relevant technology can provide a prompt when there are abnormalities in the wind pressure and air volume in the coal mine at the current moment. However, when the wind pressure and air volume are already abnormal at the current moment, the operation and maintenance personnel of the ventilation equipment may find it difficult to maintain the ventilation equipment within the effective time. Therefore, it is difficult to achieve effective monitoring of the ventilation of the coal mine tunnels in the relevant technology. Summary of the invention
[0005] In order to overcome the problem in the related art that it is difficult to effectively monitor the ventilation of the tunnels of a coal mine, the present application provides a method and system for monitoring the ventilation of an underground coal mine.
[0006] According to a first aspect of an embodiment of the present application, a method for monitoring ventilation in underground coal mines is provided, comprising: obtaining a plurality of random forest models, wherein the random forest models are obtained by performing model training according to a training set, and the training sets of different random forest models correspond to different data subsets of a ventilation parameter data set, and the ventilation parameter data set includes ventilation parameter values of a target coal mine tunnel at different times; obtaining a mean square error of the random forest model, and obtaining a first evaluation value of the data subset, wherein the first evaluation value is used to characterize the probability that an abnormality exists in the ventilation parameter value in the data subset; multiplying the mean square error of the random forest model by the first evaluation value to obtain a second evaluation value of the random forest model, and determining a target random forest model with the smallest second evaluation value from a plurality of random forest models; using the target random forest model to predict a ventilation parameter prediction value at a target time, and outputting a prompt message when it is determined that an abnormality exists in the ventilation parameter prediction value.
[0007] In this way, by determining the target random forest model from multiple random forest models, and using the target random forest model to predict the ventilation parameter prediction value at the target time, when it is determined that the ventilation parameter prediction value is abnormal, a prompt message is output, which can realize early warning of the ventilation parameter value at the target time, so as to achieve better monitoring of the ventilation of the target coal mine tunnel.
[0008] At the same time, the first evaluation value is used to characterize the probability that there is an abnormality in the ventilation parameter values in the data subset. The second evaluation value is obtained by multiplying the mean square error of the random forest model by the first evaluation value. The second evaluation value can comprehensively consider the prediction accuracy of the random forest model and the probability that there is an abnormality in the ventilation parameter values in the training set of the random forest model. The target random forest model has the smallest second evaluation value among multiple random forest models. Therefore, the prediction result output by the target random forest model is more accurate, which helps to obtain more accurate prediction values of the ventilation parameters at the target time.
[0009] Furthermore, since the predicted values of the ventilation parameters at the target time can be predicted, possible abnormalities in the ventilation parameter values at the target time can be known in advance, and ventilation equipment that may have abnormalities can be maintained in advance, reducing the time limit requirements for maintenance of the ventilation equipment.
[0010] Optionally, the first evaluation value of the data subset is obtained in the following manner: obtaining a target monitoring parameter value of the target coal mine tunnel at a target historical moment, and obtaining a first mean of multiple monitoring parameter values corresponding to multiple historical moments to determine a first difference value between the target monitoring parameter value and the first mean; obtaining a first ratio of a first quantity to a second quantity, the first quantity being the total number of monitoring parameter values, and the second quantity being the number of monitoring parameter values among the multiple monitoring parameter values that are within a first preset range, and the first preset range is determined based on the target monitoring parameter value; determining a second evaluation value of the target ventilation parameter value at the target historical moment based on a first multiplication value of the first ratio and the first difference value, and obtaining the first evaluation value of the data subset based on a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset.
[0011] In this way, the larger the first ratio of the first number to the second number is, the smaller the number of monitoring parameter values within the first preset range among the multiple monitoring parameter values is, and the greater the probability that the target monitoring parameter value is abnormal; on the contrary, the smaller the first ratio of the first number to the second number is, the larger the number of monitoring parameter values within the first preset range among the multiple monitoring parameter values is, and the smaller the probability that the target monitoring parameter value is abnormal. Therefore, the first ratio between the first number and the second number can effectively characterize the probability that the target monitoring parameter value is abnormal.
[0012] Optionally, the target monitoring parameter value includes multiple monitoring values corresponding to multiple monitoring dimensions, and a second evaluation value of the target ventilation parameter value at the target historical moment is determined according to a first product of the first ratio and the first difference value, including: determining, for the target monitoring dimension among the multiple monitoring dimensions, the first product corresponding to the monitoring value of the target monitoring dimension; and taking a third average of the multiple first product values corresponding to the multiple monitoring values as the second evaluation value of the target ventilation parameter value at the target historical moment.
[0013] In this way, since the first product value of the monitoring value can represent the probability of abnormality in the monitoring value under the monitoring dimension, and there is a correlation between the monitoring value and the ventilation parameter value, the third mean of multiple first product values corresponding to multiple monitoring values is obtained. The third mean can reflect the possible abnormalities in multiple monitoring dimensions as a whole. The third mean is used as the second evaluation value of the target ventilation parameter value at the target historical moment, so that the second evaluation value can better represent the probability of abnormality in the ventilation parameter value of the target coal mine roadway.
[0014] Optionally, a first evaluation value of the data subset is obtained based on a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset, including: obtaining a monitoring parameter value sequence including multiple monitoring parameter values corresponding to multiple historical moments, and obtaining a ventilation parameter value sequence including multiple ventilation parameter values corresponding to multiple historical moments, the monitoring parameter value sequence and the ventilation parameter value sequence corresponding to the same historical time period; determining a correlation coefficient between the monitoring parameter value and the ventilation parameter value based on the monitoring parameter value sequence and the ventilation parameter value sequence; and multiplying the correlation coefficient by the second mean to obtain the first evaluation value of the data subset.
[0015] In this way, the correlation coefficient is multiplied by the second mean to obtain the first evaluation value of the data subset. The correlation coefficient can reflect the correlation between the monitoring parameter value and the ventilation parameter value, and the second mean can reflect the difference between the target monitoring parameter value and other monitoring parameter values. Therefore, the first evaluation value can effectively characterize the probability of abnormal ventilation parameter values.
[0016] Optionally, the first evaluation value of the data subset is obtained based on the second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset, including: obtaining the fourth mean of the ventilation parameter value of the target coal mine tunnel in the target historical time period, the target historical time period being the historical time period in which the target historical moment is located, and determining the second difference value between the target monitoring parameter value of the target coal mine tunnel at the target historical time and the fourth mean; subtracting a preset coefficient from the second difference value to obtain a first value, and using the result of operating the linear rectification function on the first value as the third evaluation value of the target ventilation parameter value at the target historical moment; determining a fifth mean of multiple third evaluation values corresponding to multiple target historical moments, multiplying the fifth mean by the second mean, and obtaining the first evaluation value of the data subset.
[0017] In this way, by determining the fifth mean of multiple third evaluation values corresponding to multiple target historical moments, multiplying the fifth mean by the second mean, and obtaining the first evaluation value of the data subset, the abnormal conditions of the target coal mine tunnel at multiple target historical moments can be reflected through the first evaluation value, thereby more accurately characterizing the probability of abnormalities in the ventilation parameter values through the first evaluation value.
[0018] Optionally, the target historical time period is a historical time period centered on a target historical moment.
[0019] Optionally, the training set of the random forest model is obtained in the following manner: randomly selecting a preset number of ventilation parameter values from a ventilation parameter data set, and using the randomly selected preset number of ventilation parameter values as the training set of the random forest model.
[0020] Optionally, the ventilation parameter data set is acquired in the following manner: receiving ventilation parameter data sent by multiple ventilation parameter acquisition devices, wherein the multiple ventilation parameter acquisition devices are arranged at different spatial positions within a preset spatial range of the target coal mine laneway; generating the ventilation parameter data set corresponding to the target coal mine laneway based on the ventilation parameter data sent by the multiple ventilation parameter acquisition devices.
[0021] According to the second aspect of an embodiment of the present application, a coal mine underground ventilation monitoring system is provided, comprising: a processor and a memory, the memory storing computer program instructions, and the computer program instructions, when executed by the processor, implement the steps of the coal mine underground ventilation monitoring method provided in the first aspect of the present application.
[0022] The technical solution provided by the embodiments of the present application may include the following beneficial effects: by determining a target random forest model from multiple random forest models, and using the target random forest model to predict the ventilation parameter prediction value at the target moment, when it is determined that the ventilation parameter prediction value is abnormal, a prompt message is output, and an early warning of the ventilation parameter value at the target moment can be achieved, so as to achieve better monitoring of the ventilation of the target coal mine tunnel.
[0023] At the same time, the first evaluation value is used to characterize the probability that there is an abnormality in the ventilation parameter values in the data subset. The second evaluation value is obtained by multiplying the mean square error of the random forest model by the first evaluation value. The second evaluation value can comprehensively consider the prediction accuracy of the random forest model and the probability that there is an abnormality in the ventilation parameter values in the training set of the random forest model. The target random forest model has the smallest second evaluation value among multiple random forest models. Therefore, the prediction result output by the target random forest model is more accurate, which helps to obtain a more accurate prediction value of the ventilation parameters at the target time.
[0024] Furthermore, since the predicted values of the ventilation parameters at the target time can be predicted, possible abnormalities in the ventilation parameter values at the target time can be known in advance, and ventilation equipment that may have abnormalities can be maintained in advance, reducing the time limit requirements for maintenance of the ventilation equipment.
[0025] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.
[0027] Figure 1 is a flow chart of a method for monitoring ventilation in underground coal mines according to an exemplary embodiment;
[0028] Figure 2 It is a structural schematic diagram of a coal mine underground ventilation monitoring system according to an exemplary embodiment. DETAILED DESCRIPTION
[0029] Here, exemplary embodiments are described in detail, and examples thereof are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application.
[0030] First, a brief introduction is given to the application scenario of the embodiment of the present application. In the application scenario of the present application, in order to ensure good ventilation of the coal mine tunnels, the ventilation of the coal mine tunnels can be monitored. When the relevant technology monitors the ventilation of the tunnels, it is determined whether there is an abnormality in the ventilation parameter values at the current moment by collecting the ventilation parameter values at the current moment. Therefore, it is difficult for the relevant technology to know whether there may be abnormalities in the ventilation parameter values in the tunnels in the future time period, and it is difficult to achieve effective monitoring of the ventilation parameter values of the coal mine tunnels.
[0031] For example, ventilation in the tunnels of a coal mine is mainly achieved through ventilation equipment. If the ventilation parameter values at the current moment are already abnormal, it means that the ventilation equipment at the current moment may have become abnormal, and the time limit for repairing or replacing the ventilation equipment is high, making it difficult to ensure good ventilation in the coal mine tunnels.
[0032] In response to the above technical problems, an embodiment of the present application provides a method and system for monitoring ventilation in underground coal mines. By determining a target random forest model from multiple random forest models, and using the target random forest model to predict the ventilation parameter prediction values at the target time, when it is determined that the ventilation parameter prediction values are abnormal, a prompt message is output, thereby enabling early warning of the ventilation parameter values at the target time to achieve better monitoring of the ventilation of the target coal mine tunnels.
[0033] At the same time, the first evaluation value is used to characterize the probability that there is an abnormality in the ventilation parameter values in the data subset. The second evaluation value is obtained by multiplying the mean square error of the random forest model by the first evaluation value. The second evaluation value can comprehensively consider the prediction accuracy of the random forest model and the probability that there is an abnormality in the ventilation parameter values in the training set of the random forest model. The target random forest model has the smallest second evaluation value among multiple random forest models. Therefore, the prediction result output by the target random forest model is more accurate, which helps to obtain more accurate prediction values of the ventilation parameters at the target time.
[0034] Furthermore, since the predicted values of the ventilation parameters at the target time can be predicted, possible abnormalities in the ventilation parameter values at the target time can be known in advance, and ventilation equipment that may have abnormalities can be maintained in advance, reducing the time limit requirements for maintenance of the ventilation equipment.
[0035] Figure 1 is a flow chart of a method for monitoring ventilation in underground coal mines according to an exemplary embodiment, which can be applied to terminal equipment. The terminal equipment can be used to monitor ventilation in underground coal mines, such as Figure 1 As shown, the method includes the following steps.
[0036] In step S101, multiple random forest models are obtained.
[0037] The random forest model is obtained by model training based on the training set. The training sets of different random forest models correspond to different data subsets of the ventilation parameter data set. The ventilation parameter data set includes ventilation parameter values of the target coal mine tunnel at different times.
[0038] Multiple ventilation parameter values of the target coal mine tunnel at multiple historical moments before the current moment can be obtained, and the multiple ventilation parameter values obtained are combined into a ventilation parameter data set. By randomly selecting ventilation parameter values from the ventilation parameter data set, different data subsets of the ventilation parameter data set can be obtained respectively.
[0039] In one embodiment, the training set of the random forest model is obtained in the following manner: randomly selecting a preset number of ventilation parameter values from a ventilation parameter data set, and using the randomly selected preset number of ventilation parameter values as the training set of the random forest model.
[0040] For example, the preset number may be between 20 and 50. To avoid errors that may be caused by different numbers of ventilation parameter values in the training sets, the numbers of ventilation parameter values in the training sets of different random forest models may be the same.
[0041] By randomly selecting a preset number of ventilation parameter values from the ventilation parameter data set, it is possible to avoid using exactly the same data subsets in different random forest models, thereby improving the diversity of the training set and thus improving the diversity of the random forest model.
[0042] The pre-built initial random forest model is trained using the training set. When the training is completed, the random forest model corresponding to the training set can be obtained. Different random forest models corresponding to different training sets can be obtained by model training using different training sets.
[0043] In an embodiment of the present application, the training sets of different random forest models correspond to different data subsets of the ventilation parameter data set, and the performance of the ventilation parameter values in different data subsets may be different. Therefore, it is helpful to determine the most accurate random forest model from multiple random forest models.
[0044] The ventilation parameter value refers to the value of the ventilation parameter. The ventilation parameter may refer to the ventilation wind speed or effective ventilation area in the alley. The effective ventilation area may be determined by the ventilation area blocked in the alley and the total ventilation area in the alley.
[0045] For example, by setting up an anemometer at the target coal mine tunnel, the ventilation speed at the target coal mine tunnel can be collected. By setting up an image sensor or multiple distance sensors at the tunnel, the ventilation area blocked in the tunnel and the total ventilation area in the tunnel can be determined.
[0046] In one embodiment, the ventilation parameter data set can be obtained in the following manner: receiving ventilation parameter data sent by multiple ventilation parameter acquisition devices, wherein the multiple ventilation parameter acquisition devices are arranged at different spatial positions within a preset spatial range of the target coal mine tunnel; generating the ventilation parameter data set corresponding to the target coal mine tunnel based on the ventilation parameter data sent by the multiple ventilation parameter acquisition devices.
[0047] For example, the ventilation parameter collection device can be an anemometer or other device with wind speed measurement function, and the ventilation parameter collection device can be communicatively connected with the terminal device; through the communication connection with the terminal device, the ventilation parameter collection device can send the collected ventilation parameter data to the terminal device.
[0048] The terminal equipment for monitoring the underground ventilation of a coal mine can receive the ventilation parameter data sent by multiple ventilation parameter collection devices. By using multiple ventilation parameter collection devices at different spatial positions within a preset spatial range of the target coal mine tunnel, the accuracy of collecting the ventilation parameters of the target coal mine tunnel can be improved. The ventilation parameter data can be the ventilation parameter values at the target tunnel.
[0049] Based on the ventilation parameter data sent by multiple ventilation parameter acquisition devices, the ventilation parameter data set corresponding to the target coal mine tunnel is generated. For example, the ventilation parameter data sent by multiple ventilation parameter acquisition devices can be directly combined to form a ventilation parameter data set corresponding to the target coal mine tunnel; or, the multiple ventilation parameter data corresponding to the same moment sent by multiple ventilation parameter acquisition devices can be averaged or the mode can be taken to obtain the ventilation parameter data corresponding to the same moment, so as to combine the multiple ventilation parameter data corresponding to multiple moments into the ventilation parameter data set corresponding to the target coal mine tunnel.
[0050] In step S102, the mean square error of the random forest model is obtained, and a first evaluation value of the data subset is obtained, where the first evaluation value is used to characterize the probability that the ventilation parameter value in the data subset is abnormal.
[0051] In one embodiment, the first evaluation value of the data subset is obtained in the following manner: obtaining the target monitoring parameter value of the target coal mine tunnel at the target historical moment, and obtaining the first mean of multiple monitoring parameter values corresponding to multiple historical moments to determine a first difference value between the target monitoring parameter value and the first mean; obtaining a first ratio of a first quantity to a second quantity, the first quantity being the total number of monitoring parameter values, and the second quantity being the number of monitoring parameter values among the multiple monitoring parameter values that are within a first preset range, and the first preset range is determined based on the target monitoring parameter value; determining a second evaluation value of the target ventilation parameter value at the target historical moment based on a first multiplication value of the first ratio and the first difference value, and obtaining the first evaluation value of the data subset based on a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset.
[0052] The first difference value between the target monitoring parameter value and the first mean, for example, can be the absolute value of the difference between the target monitoring parameter value and the first mean; or, the absolute value of the difference between the target monitoring parameter value and the first mean can be determined, and the ratio of the absolute value to the first mean is taken as the first difference value between the target monitoring parameter value and the first mean; the first difference value can characterize the degree of difference between the target monitoring parameter value and the first mean.
[0053] The first preset range is determined based on the target monitoring parameter value. The first preset range can be, for example, a range formed by 80% to 120% of the target monitoring parameter value. The first number is the number of monitoring parameter values among the multiple monitoring parameter values that are within the first preset range. The more monitoring parameter values among the multiple monitoring parameter values that are within the first preset range, the more other monitoring parameter values that are close to the target monitoring parameter value, and the higher the probability that the target monitoring parameter value is not abnormal. On the contrary, the fewer monitoring parameter values among the multiple monitoring parameter values that are within the first preset range, the fewer other monitoring parameter values that are close to the target monitoring parameter value, and the higher the probability that the target monitoring parameter value is abnormal.
[0054] In this way, the larger the first ratio of the first number to the second number is, the smaller the number of monitoring parameter values within the first preset range among the multiple monitoring parameter values is, and the greater the probability that the target monitoring parameter value is abnormal; on the contrary, the smaller the first ratio of the first number to the second number is, the larger the number of monitoring parameter values within the first preset range among the multiple monitoring parameter values is, and the smaller the probability that the target monitoring parameter value is abnormal. Therefore, the first ratio between the first number and the second number can effectively characterize the probability that the target monitoring parameter value is abnormal.
[0055] The monitoring parameter value may be the value of the monitoring parameter, and the monitoring parameter may be, for example, parameters such as temperature, humidity, and air pressure; when there are abnormalities in parameters such as temperature, humidity, and air pressure in the tunnel, the ventilation of the coal mine tunnel will be affected, causing the ventilation parameter value of the coal mine tunnel to also be abnormal; or, the abnormalities in parameters such as temperature, humidity, and air pressure in the tunnel may be caused by abnormalities in the ventilation equipment of the coal mine tunnel. When there are abnormalities in the ventilation equipment of the coal mine tunnel, the ventilation parameter value of the coal mine tunnel will also be abnormal. Therefore, through the first product value of the first ratio and the first difference value, the possible abnormalities in the target ventilation parameter value at the target historical moment can be reflected from the side; wherein, the target historical moment may be any one of multiple historical moments.
[0056] The following uses an exemplary calculation formula as an example to explain the process of obtaining the first multiplication value in the embodiment of the present application: , where D is the first multiplication value; N1 is the first number, i.e., the total number of monitoring parameter values; N2 is the second number, i.e., the number of monitoring parameter values in the first preset range among the multiple monitoring parameter values; G is the target monitoring parameter value, and M is the first mean value of the multiple monitoring parameter values corresponding to the multiple historical moments, It means taking the absolute value, and norm is the normalization function.
[0057] A first evaluation value of the data subset is obtained according to a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset, and the first evaluation value can reflect the overall abnormality degree of multiple target ventilation parameter values in the data subset.
[0058] In one embodiment, the target monitoring parameter value includes multiple monitoring values corresponding to multiple monitoring dimensions, and the second evaluation value of the target ventilation parameter value at the target historical moment is determined according to the first product of the first ratio and the first difference value, including: for the target monitoring dimension among the multiple monitoring dimensions, determining the first product corresponding to the monitoring value of the target monitoring dimension; and using the third average of the multiple first product values corresponding to the multiple monitoring values as the second evaluation value of the target ventilation parameter value at the target historical moment.
[0059] Multiple monitoring dimensions may include, for example, temperature, humidity, and air pressure. When the target monitoring dimension is temperature, the monitoring value of the target monitoring dimension is the temperature value at the target coal mine tunnel; when the target monitoring dimension is humidity, the monitoring value of the target monitoring dimension is the humidity value at the target coal mine tunnel; when the target monitoring dimension is air pressure, the monitoring value of the target monitoring dimension is the air pressure value at the target coal mine tunnel.
[0060] By referring to the process of obtaining the first product of the first ratio and the first difference value, the first product of the monitoring values corresponding to different monitoring dimensions can be obtained. Since the first product of the monitoring value can characterize the probability of abnormality in the monitoring value under the monitoring dimension, and there is a correlation between the monitoring value and the ventilation parameter value, the third mean of multiple first product values corresponding to multiple monitoring values is obtained. The third mean can reflect the possible abnormalities in multiple monitoring dimensions as a whole. Using the third mean as the second evaluation value of the target ventilation parameter value at the target historical moment can enable the second evaluation value to better characterize the probability of abnormality in the ventilation parameter value of the target coal mine roadway.
[0061] In one embodiment, a first evaluation value of a data subset is obtained based on a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset, including: obtaining a monitoring parameter value sequence including multiple monitoring parameter values corresponding to multiple historical moments, and obtaining a ventilation parameter value sequence including multiple ventilation parameter values corresponding to multiple historical moments, the monitoring parameter value sequence and the ventilation parameter value sequence corresponding to the same historical time period; determining a correlation coefficient between the monitoring parameter values and the ventilation parameter values based on the monitoring parameter value sequence and the ventilation parameter value sequence; and multiplying the correlation coefficient by the second mean to obtain the first evaluation value of the data subset.
[0062] For example, the Pearson correlation coefficient between the monitoring parameter values and the ventilation parameter values can be determined based on the monitoring parameter value sequence and the ventilation parameter value sequence, and the absolute value of the Pearson correlation coefficient is used as the correlation coefficient between the monitoring parameter value and the ventilation parameter value; there is a connection between the ventilation parameter value and the monitoring parameter value at the target coal mine tunnel, for example, the temperature, humidity and air pressure at the target coal mine tunnel may affect the ventilation parameter value at the target coal mine tunnel; here, the correlation coefficient is obtained based on the monitoring parameter value sequence and the ventilation parameter value sequence corresponding to the same historical time period, and the obtained correlation coefficient can reflect possible anomalies in the ventilation parameter value, or the correlation coefficient can reflect possible anomalies in the monitoring parameter value.
[0063] In this way, the correlation coefficient is multiplied by the second mean to obtain the first evaluation value of the data subset. The correlation coefficient can reflect the correlation between the monitoring parameter value and the ventilation parameter value, and the second mean can reflect the difference between the target monitoring parameter value and other monitoring parameter values. Therefore, the first evaluation value can effectively characterize the probability of abnormal ventilation parameter values.
[0064] In one embodiment, a first evaluation value of a data subset is obtained based on a second mean of multiple second evaluation values of multiple target ventilation parameter values in a data subset, including: obtaining a fourth mean of the ventilation parameter value of a target coal mine tunnel within a target historical time period, wherein the target historical time period is a historical time period in which a target historical moment is located, and determining a second difference value between a target monitoring parameter value of the target coal mine tunnel at a target historical time and the fourth mean value; subtracting a preset coefficient from the second difference value to obtain a first value, and using a result of operating a linear rectification function on the first value as a third evaluation value of the target ventilation parameter value at the target historical moment; determining a fifth mean of multiple third evaluation values corresponding to multiple target historical moments, and multiplying the fifth mean value by the second mean value to obtain the first evaluation value of the data subset.
[0065] The target historical time period may be a historical time period centered on the target historical moment. For example, when the target historical moment is 08:00:00, the target historical time period may be the time period from 07:59:00 to 08:01:00.
[0066] The fourth mean of the ventilation parameter values of the target coal mine tunnels during the target historical time period can reflect the average level of the ventilation parameter values of the target coal mine tunnels during the target historical time period. The second difference value between the target monitoring parameter value and the fourth mean can characterize the probability of abnormality in the target monitoring parameter value.
[0067] When monitoring the ventilation of coal mine tunnels, due to the certain fluidity of the gas in the tunnel, even if the ventilation equipment does not adjust the ventilation, the gas in the tunnel may still have a certain wind speed. The first value is obtained by subtracting the preset coefficient from the second difference value, and the result of the linear rectification function on the first value is used as the third evaluation value of the target ventilation parameter value at the target historical moment. This can take into account the impact of the fluidity of the gas in the tunnel itself on the ventilation parameter value.
[0068] For example, when the first value obtained by subtracting the preset coefficient from the second difference value is less than 0, the result of the linear rectification function on the first value is 0; when the first value obtained by subtracting the preset coefficient from the second difference value is greater than or equal to 0, the result of the linear rectification function on the first value is greater than or equal to 0, for example, the result of the linear rectification function on the first value is equal to the first value.
[0069] In this way, by determining the fifth mean of multiple third evaluation values corresponding to multiple target historical moments, multiplying the fifth mean by the second mean, and obtaining the first evaluation value of the data subset, the abnormal conditions of the target coal mine tunnel at multiple target historical moments can be reflected through the first evaluation value, thereby more accurately characterizing the probability of abnormalities in the ventilation parameter values through the first evaluation value.
[0070] In step S103, the mean square error of the random forest model is multiplied by the first evaluation value to obtain a second evaluation value of the random forest model, and a target random forest model with the smallest second evaluation value is determined from multiple random forest models.
[0071] The mean square error of the random forest model can be determined, for example, by the square of the difference between the predicted value and the actual value of the random forest model. The mean square error of the random forest model can characterize the degree of difference between the predicted value and the actual value of the random forest model. The larger the mean square error of the random forest model, the greater the difference between the predicted value and the actual value of the random forest model. On the contrary, the smaller the mean square error of the random forest model, the smaller the difference between the predicted value and the actual value of the random forest model.
[0072] The following uses an exemplary calculation formula as an example to illustrate the process of obtaining the second evaluation value of the random forest model in the embodiment of the present application. , where E is the second evaluation value of the random forest model, H is the first evaluation value of the data subset corresponding to the random forest model, P is the number of predicted values of the random forest model, and B i is the i-th predicted value of the random forest model, A i is the i-th actual value of the random forest model, where i is a positive integer less than or equal to P.
[0073] Here, the mean square error of the random forest model can characterize the degree of difference between the predicted value and the actual value of the random forest model. The first evaluation value can characterize the probability that the ventilation parameter value in the data subset exists abnormally, and the data subset is the training set of the random forest model. That is, the first evaluation value can characterize the probability that the ventilation parameter value in the training set of the random forest model exists abnormally. Therefore, the mean square error of the random forest model is multiplied by the first evaluation value to obtain the second evaluation value of the random forest model. The second evaluation value can better reflect the probability that the random forest model exists abnormally.
[0074] For example, when the training set includes abnormal ventilation parameter values, the random forest model obtained by using the training set for model training may output the abnormal ventilation parameter values as prediction results when predicting the ventilation parameter values due to the influence of the abnormal ventilation parameter values on the model training process, thereby affecting the reference value of the prediction results output by the random forest model.
[0075] Since the second evaluation value can comprehensively reflect the prediction consistency and the probability of anomalies in the training set of the random forest model, determining the target random forest model with the smallest second evaluation value from multiple random forest models helps to obtain more accurate prediction results for ventilation parameter values through the target random forest model.
[0076] In step S104, the target random forest model is used to predict the ventilation parameter prediction value at the target time, and when it is determined that the ventilation parameter prediction value is abnormal, a prompt message is output.
[0077] For example, multiple ventilation parameter values at multiple historical moments before the current moment can be input into a pre-selected and trained target random forest model to obtain the ventilation parameter prediction value at the target moment output by the target random forest model.
[0078] By comparing the predicted ventilation parameter value with the preset ventilation parameter value range, when the predicted ventilation parameter value is within the preset ventilation parameter value range, it can be determined that there is no abnormality in the predicted ventilation parameter value; on the contrary, when the predicted ventilation parameter value is outside the preset ventilation parameter value range, it can be determined that there is an abnormality in the predicted ventilation parameter value; the preset ventilation parameter value range can be determined based on the number of workers in the coal mine and the volume of the coal mine. For example, the preset ventilation parameter value range can be 0.3 meters per second to 4 meters per second.
[0079] When it is determined that there is an abnormality in the predicted value of the ventilation parameters, in order to effectively remind the maintenance personnel, the output prompt information may be one or more of audio prompt information, light prompt information and vibration prompt information. For example, an audio prompt information with the content "There is an abnormality in the ventilation of the target coal mine tunnel, please check carefully" may be output, or a light of a preset color may be output for a constant or flashing prompt, and the preset color may be, for example, a striking color such as red or orange; or, when it is determined that there is an abnormality in the predicted value of the ventilation parameters, the terminal equipment used to monitor the underground ventilation of the coal mine may provide a vibration reminder.
[0080] Through the coal mine underground ventilation monitoring method provided in the embodiment of the present application, a target random forest model is determined from multiple random forest models, and the target random forest model is used to predict the ventilation parameter prediction value at the target time. When it is determined that the ventilation parameter prediction value is abnormal, a prompt message is output, which can realize early warning of the ventilation parameter value at the target time, so as to achieve better monitoring of the ventilation of the target coal mine tunnel.
[0081] At the same time, the first evaluation value is used to characterize the probability that there is an abnormality in the ventilation parameter values in the data subset. The second evaluation value is obtained by multiplying the mean square error of the random forest model by the first evaluation value. The second evaluation value can comprehensively consider the prediction accuracy of the random forest model and the probability that there is an abnormality in the ventilation parameter values in the training set of the random forest model. The target random forest model has the smallest second evaluation value among multiple random forest models. Therefore, the prediction result output by the target random forest model is more accurate, which helps to obtain more accurate prediction values of the ventilation parameters at the target time.
[0082] Furthermore, since the predicted values of the ventilation parameters at the target time can be predicted, possible abnormalities in the ventilation parameter values at the target time can be known in advance, and ventilation equipment that may have abnormalities can be maintained in advance, reducing the time limit requirements for maintenance of the ventilation equipment.
[0083] Figure 2 FIG. 1 is a schematic diagram of a ventilation monitoring system for underground coal mines according to an exemplary embodiment. Figure 2 The coal mine underground ventilation monitoring system includes: a processor 1100 and a memory 1200, wherein the memory 1200 stores computer program instructions, and when the computer program instructions are executed by the processor 1100, all or part of the steps of the coal mine underground ventilation monitoring method in the present application are implemented.
[0084] It should be understood that the features of some embodiments of the various applications described herein may be combined with each other, unless specifically stated otherwise.
[0085] Although terms such as "first", "second", and "third" may be used herein to describe various components, parts, regions, layers, or sections, these components, parts, regions, layers, or sections are not limited to these terms. Instead, these terms are only used to distinguish one component, part, region, layer, or section from another component, part, region, layer, or section. Therefore, without departing from the teachings of the examples described herein, the first component, part, region, layer, or section mentioned in the examples may also be referred to as the second component, part, region, layer, or section.
[0086] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include at least one of the features. In the description herein, the meaning of "plurality" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined.
[0087] In addition, the word "exemplary" is used herein to mean serving as an example, instance, or diagram. Any aspect or design described herein as "exemplary" is not necessarily to be understood as being advantageous over other aspects or designs. Instead, the use of the word exemplary is intended to present concepts in a concrete manner. As used herein, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or".
[0088] Likewise, although the present application has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. With particular regard to the various functions performed by the components (e.g., elements, resources, etc.) described above, unless otherwise noted, the terms used to describe such components are intended to correspond to any component (functionally equivalent) that performs the specific functions of the described components, even if not structurally equivalent to the disclosed structures.
[0089] Additionally, while particular features of the present application may have been disclosed with respect to only one of several implementations, such features may be combined with one or more other features of other implementations as may be desired and advantageous for any given or particular application.
[0090] Those skilled in the art will readily appreciate other embodiments of the present application after considering the specification and practicing the invention disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary technical means in the art that are not disclosed in the present application, and the specification and embodiments are only considered as exemplary.
[0091] It should be understood that the present application is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.
Claims
1. A method for monitoring ventilation in underground coal mines, characterized in that: include: Acquire multiple random forest models, wherein the random forest models are obtained by performing model training according to a training set, and the training sets of different random forest models correspond to different data subsets of a ventilation parameter data set, wherein the ventilation parameter data set includes ventilation parameter values of a target coal mine tunnel at different times; Obtaining a mean square error of a random forest model, and obtaining a first evaluation value of the data subset, wherein the first evaluation value is used to characterize a probability that a ventilation parameter value in the data subset is abnormal; Multiplying the mean square error of the random forest model by the first evaluation value to obtain a second evaluation value of the random forest model, and determining a target random forest model with the smallest second evaluation value from multiple random forest models; Using the target random forest model to predict the ventilation parameter prediction value at the target time, and when it is determined that the ventilation parameter prediction value is abnormal, outputting prompt information; The first evaluation value of the data subset is obtained by: obtaining a target monitoring parameter value of a target coal mine tunnel at a target historical moment, and obtaining a first mean value of a plurality of monitoring parameter values corresponding to a plurality of historical moments, so as to determine a first difference value between the target monitoring parameter value and the first mean value; Obtaining a first ratio of a first quantity to a second quantity, wherein the first quantity is the total quantity of monitoring parameter values, and the second quantity is the quantity of monitoring parameter values in a plurality of monitoring parameter values that are within a first preset range, wherein the first preset range is determined based on a target monitoring parameter value; determining a second evaluation value of a target ventilation parameter value at a target historical moment based on a first multiplication value of the first ratio and the first difference value, and obtaining a first evaluation value of the data subset based on a second mean of a plurality of second evaluation values of a plurality of target ventilation parameter values in the data subset; The target monitoring parameter value includes multiple monitoring values corresponding to multiple monitoring dimensions, and according to the first multiplication value of the first ratio and the first difference value, a second evaluation value of the target ventilation parameter value at the target historical moment is determined, including: for the target monitoring dimension in the multiple monitoring dimensions, determining the first multiplication value corresponding to the monitoring value of the target monitoring dimension; and using a third mean of the multiple first multiplication values corresponding to the multiple monitoring values as the second evaluation value of the target ventilation parameter value at the target historical moment; Obtaining a first evaluation value of the data subset according to a second mean of multiple second evaluation values of multiple target ventilation parameter values in the data subset, including: obtaining a monitoring parameter value sequence including multiple monitoring parameter values corresponding to multiple historical moments, and obtaining a ventilation parameter value sequence including multiple ventilation parameter values corresponding to multiple historical moments, wherein the monitoring parameter value sequence and the ventilation parameter value sequence correspond to the same historical time period; determining a correlation coefficient between the monitoring parameter value and the ventilation parameter value according to the monitoring parameter value sequence and the ventilation parameter value sequence; and multiplying the correlation coefficient by the second mean to obtain a first evaluation value of the data subset; Or it includes: obtaining a fourth mean of the ventilation parameter value of the target coal mine tunnel within a target historical time period, wherein the target historical time period is the historical time period in which the target historical moment is located, and determining a second difference value between the target monitoring parameter value of the target coal mine tunnel at the target historical time and the fourth mean value; subtracting a preset coefficient from the second difference value to obtain a first value, and using the result of the operation of the linear rectification function on the first value as a third evaluation value of the target ventilation parameter value at the target historical moment; determining a fifth mean of multiple third evaluation values corresponding to multiple target historical moments, multiplying the fifth mean by the second mean, and obtaining a first evaluation value of the data subset.
2. The method for monitoring ventilation in underground coal mines according to claim 1, characterized in that: The target historical time period is a historical time period centered on a target historical moment.
3. The method for monitoring ventilation in underground coal mines according to claim 1, characterized in that: The training set of the random forest model is obtained in the following way: A preset number of ventilation parameter values are randomly selected from the ventilation parameter data set, and the randomly selected preset number of ventilation parameter values are used as a training set for the random forest model.
4. The method for monitoring ventilation in underground coal mines according to any one of claims 1 to 3, characterized in that: The ventilation parameter data set is obtained in the following way: Receiving ventilation parameter data sent by a plurality of ventilation parameter acquisition devices, wherein the plurality of ventilation parameter acquisition devices are arranged at different spatial positions within a preset spatial range of a target coal mine laneway; The ventilation parameter data set corresponding to the target coal mine tunnel is generated according to the ventilation parameter data sent by multiple ventilation parameter acquisition devices.
5. A coal mine underground ventilation monitoring system, characterized in that: include: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the coal mine underground ventilation monitoring method according to any one of claims 1 to 4 is implemented.
