Heat damage prevention and control method based on mine multi-source data analysis

Through deep learning technology and data processing, combined with the characteristics of the mine environment and ventilation system, the angle of the mine ventilation fan blades is adjusted in real time, which solves the poor ventilation effect and energy waste caused by the fixation of the traditional ventilation fan angle, and improves the ventilation efficiency of the mine environment and Energy utilization efficiency.

CN118965067AInactive Publication Date: 2024-11-15LIAOBO INFORMATION TECH (ZIBO) CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411021370.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-29
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The blade angle of traditional mine ventilation fans is fixed, and cannot be adaptively adjusted according to the actual mine environment changes, resulting in poor ventilation effect or waste of energy, and cannot effectively prevent and control heat damage.

Method used

By obtaining the temperature, humidity, gas concentration and ventilation fan blade operation angle data at multiple predetermined time points of the mine, deep learning technology and data processing methods are used to extract the mine environment and blade angle characteristics, fuse it into global correlation feature vectors, and adjust the blades in real time Running angle.

Benefits of technology

It realizes the adjustment of the angle of the fan blades according to the real-time status of the mine environment, improves ventilation effect, saves energy, and reduces heat damage risks and energy waste.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118965067A_ABST
    Figure CN118965067A_ABST
Patent Text Reader

Abstract

The invention relates to the field of intelligent analysis, and particularly discloses a mine multi-source data analysis-based heat damage prevention and control method, which comprises the following steps of: acquiring temperature, humidity and gas concentration of a plurality of preset time points in a mine preset time and blade operation angles of a ventilator at a plurality of preset time points in a preset time; and a deep learning technology and a data processing method are adopted, and whether the blade operation angle of the ventilator at the current time point needs to be increased or decreased or not is determined based on the real-time state of the mine environment. By means of the method, multiple factors of the mine environment and the ventilation system can be comprehensively considered, the operation angle of the fan blade is adjusted according to actual requirements, energy can be saved, and the energy utilization efficiency is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent analysis, and more specifically, to a method for preventing and controlling heat damage based on multi-source data analysis in mines. Background Art

[0002] Mine heat damage refers to the harm to the health and safety of miners caused by factors such as high temperature and high humidity in the mine working environment.

[0003] Mine heat damage is a common occupational disease in mining work. Heat damage factors such as high temperature and high humidity can affect the miners' work efficiency and physical activity ability, leading to fatigue, weakness, etc. Therefore, through effective heat damage prevention and control measures, the time and degree of miners' exposure to harmful environments such as high temperature and high humidity can be reduced, the risk of miners suffering from heat stress and related diseases can be reduced, the physical burden of miners can be reduced, the health of miners can be protected, the risk of accidents can be reduced, and the working environment of miners can be improved, comfortable working conditions can be provided, so as to improve work efficiency and production benefits.

[0004] However, since the blade angle of the ventilator in traditional heat damage prevention is usually fixed, it cannot be adaptively adjusted according to changes in the actual mine environment. This may lead to poor ventilation and fail to meet the needs of miners. For example, when a local high temperature area appears in the mine, the traditional ventilator cannot adjust the angle in time to increase the ventilation volume in this area, resulting in the aggravation of the heat damage problem. In addition, the angle of the traditional ventilator is fixed, and the ventilation volume cannot be adjusted according to actual needs. This may cause the ventilator to over-ventilate at certain times, wasting energy. For example, when the temperature in the mine is low or the miners' work intensity is low, the ventilator still provides the maximum ventilation volume at a fixed angle, resulting in a waste of energy.

[0005] Therefore, an optimized heat damage prevention and control solution based on multi-source data analysis in mines is expected. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a method for preventing and controlling heat damage based on multi-source data analysis in mines, which obtains the temperature, humidity and gas concentration at multiple predetermined time points within a predetermined time in the mine and the blade operating angle of the ventilator at multiple predetermined time points within a predetermined time, and adopts deep learning technology and data processing methods to determine whether the blade operating angle of the ventilator at the current time point should be increased, decreased or unchanged based on the real-time state of the mine environment. This method can comprehensively consider multiple factors of the mine environment and ventilation system, and adjust the blade operating angle of the ventilator according to actual needs, which can save energy and improve energy utilization efficiency.

[0007] According to one aspect of the present application, a method for preventing and controlling heat damage based on multi-source data analysis in a mine is provided, which comprises:

[0008] Obtain the temperature, humidity and gas concentration at multiple predetermined time points within a predetermined time period of the mine and the blade running angles at multiple predetermined time points within a predetermined time period of the ventilator;

[0009] Arranging the temperature, humidity and gas concentration at the plurality of predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension, and then extracting the mine environment characteristics to obtain a mine environment characteristic vector;

[0010] Arranging the blade operating angles of the ventilator at a plurality of predetermined time points within a predetermined time according to the time dimension as a blade angle time series feature input vector, and then extracting the blade angle feature to obtain a blade angle feature vector;

[0011] Fusion of the mine environment feature vector and the blade angle feature vector to obtain a mine environment global correlation feature vector;

[0012] Based on the global correlation feature vector of the mine environment, it is determined whether the blade operating angle of the ventilator at the current time point should be increased, decreased, or kept unchanged.

[0013] Compared with the prior art, the present application provides a method for preventing and controlling heat damage based on multi-source data analysis in mines. It obtains the temperature, humidity and gas concentration at multiple predetermined time points in the mine within a predetermined time and the blade running angle of the ventilator at multiple predetermined time points within a predetermined time, and uses deep learning technology and data processing methods to determine whether the blade running angle of the ventilator at the current time point should be increased, decreased or unchanged based on the real-time state of the mine environment. This method can comprehensively consider multiple factors of the mine environment and ventilation system, and adjust the blade running angle of the ventilator according to actual needs, which can save energy and improve energy utilization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0015] Figure 1 The present invention is a flowchart of a method for preventing and controlling heat damage based on multi-source data analysis in a mine according to an embodiment of the present application.

[0016] Figure 2 Schematic diagram of the architecture of a method for preventing and controlling heat damage based on multi-source data analysis in a mine according to an embodiment of the present application.

[0017] Figure 3This is a flowchart of a method for preventing and controlling heat damage based on multi-source data analysis in mines according to an embodiment of the present application, in which the temperature, humidity and gas concentration at multiple predetermined time points are arranged according to the time dimension and the sample dimension as a mine environment input matrix, and then the mine environment features are extracted to obtain a mine environment feature vector.

[0018] Figure 4 This is a flowchart of a method for preventing and controlling heat damage based on multi-source data analysis in mines according to an embodiment of the present application, in which the blade operating angles of the ventilator at multiple predetermined time points within a predetermined time are arranged according to the time dimension as a blade angle timing feature input vector, and then blade angle features are extracted to obtain a blade angle feature vector. DETAILED DESCRIPTION

[0019] The embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as being limited to the embodiments described herein, but rather these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes and are not intended to limit the scope of protection of the present disclosure.

[0020] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in different orders and / or in parallel. In addition, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present disclosure is not limited in this respect.

[0021] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0022] It should be noted that the modifications of "one" and "plurality" mentioned in the present disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0023] Mine heat damage refers to a series of health problems and safety risks caused by high temperature environments in mines. When a mine goes deeper underground, the underground temperature will gradually rise, and the work activities and operation of mechanical equipment inside the mine will generate a lot of heat. If the mine is poorly ventilated and the heat cannot be effectively dissipated, the temperature inside the mine will rise, forming mine heat damage. Mine heat damage has a serious impact on the health and safety of miners. High temperature environments can cause temperature regulation disorders in miners, causing heat stress diseases such as heat stroke and heat exhaustion. Long-term exposure to high temperature environments can also cause diseases of multiple organs in miners, such as the cardiovascular system, respiratory system, and metabolic system. In addition, high temperature environments will increase miners' fatigue, reduce work efficiency, and increase the risk of work accidents.

[0024] High temperature environment will have certain negative effects on workers' physical and mental health, such as fatigue and lack of concentration. Therefore, heat damage prevention and control measures can effectively reduce the health risks of workers in high temperature environments. Specifically, high temperature environments are prone to cause heat stress diseases such as heat stroke and heat exhaustion, which may even be life-threatening in severe cases. By taking appropriate protective measures, the incidence of heat stress diseases among workers can be reduced, the degree of fatigue of workers in high temperature environments can be reduced, their physical health can be protected, and their work efficiency and production capacity can be improved, thereby improving production efficiency and reducing the occurrence of accidents.

[0025] The problem of fixed blade angle of the ventilator used in traditional heat damage prevention and control does exist, which may lead to poor ventilation effect and fail to meet the needs of the actual mine environment. Specifically, the blade angle of the traditional ventilator is usually fixed and cannot be adaptively adjusted according to the changes in the actual mine environment. When a local high temperature area appears in the mine, the traditional ventilator cannot adjust the angle in time to increase the ventilation volume in this area, resulting in the aggravation of the heat damage problem. This may cause the working environment of miners in high temperature areas to deteriorate and increase health risks. In addition, the angle of the traditional ventilator is fixed, and the ventilation volume cannot be adjusted according to actual needs. When the temperature in the mine is low or the miners' work intensity is low, the ventilator still provides the maximum ventilation volume at a fixed angle, resulting in energy waste. This not only increases energy costs, but may also have unnecessary impacts on the environment. Therefore, an optimized heat damage prevention and control method based on multi-source data analysis in mines is desired.

[0026] Figure 1 The present invention is a flowchart of a method for preventing and controlling heat damage based on multi-source data analysis in a mine according to an embodiment of the present application. Figure 2 FIG. 1 is a schematic diagram of the architecture of a heat damage prevention method based on multi-source data analysis of a mine according to an embodiment of the present application. Figure 1 and Figure 2As shown, according to the embodiment of the present application, the method for preventing and controlling heat damage based on multi-source data analysis of mines includes: S110, obtaining the temperature, humidity and gas concentration at multiple predetermined time points within the predetermined time of the mine and the blade operating angles at multiple predetermined time points within the predetermined time of the ventilator; S120, arranging the temperature, humidity and gas concentration at the multiple predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension, and then performing mine environment feature extraction to obtain a mine environment feature vector; S130, arranging the blade operating angles at multiple predetermined time points within the predetermined time of the ventilator into a blade angle time series feature input vector according to the time dimension, and then performing blade angle feature extraction to obtain a blade angle feature vector; S140, fusing the mine environment feature vector and the blade angle feature vector to obtain a mine environment global correlation feature vector; and, S150, based on the mine environment global correlation feature vector, determining whether the blade operating angle of the ventilator at the current time point should be increased, decreased or unchanged.

[0027] In step S110, the temperature, humidity and gas concentration at multiple predetermined time points in the mine and the blade running angle at multiple predetermined time points in the fan are obtained. It should be understood that the high temperature environment in the mine is one of the main causes of heat damage. By obtaining the temperature and humidity data at multiple predetermined time points, the temperature and humidity distribution in different areas of the mine can be understood. This helps to determine the high temperature areas that may exist in the mine and take corresponding measures, such as strengthening ventilation, local ventilation, etc., to reduce the temperature and humidity and reduce the risk of heat stress disease for miners. In addition, there may be a risk of gas accumulation in the mine, and high concentration of gas is one of the main factors causing mine accidents. By obtaining the gas concentration data at multiple predetermined time points, the change of gas concentration in the mine can be monitored in time to determine whether there is a danger of gas accumulation. In this way, corresponding safety measures can be taken, such as increasing the ventilation volume, removing gas in advance, etc., to ensure that the gas concentration in the mine is within a safe range. In addition, ventilation is one of the important means of preventing and controlling heat damage in mines. By obtaining the data of the fan blade operating angle at multiple predetermined time points, we can understand the actual operation of the fan and evaluate the working status and ventilation effect of the fan. If the fan blade angle is fixed or improperly adjusted, it may lead to poor ventilation and fail to meet the needs of miners. Therefore, by understanding the data of the fan blade operating angle, we can adjust the fan's working parameters in time to ensure the best ventilation effect.

[0028] In particular, in a specific embodiment of the present application, the temperature, humidity, gas concentration and fan blade running angle parameters in the mine can be monitored in real time by installing temperature sensors, humidity sensors, gas concentration sensors and blade running angle sensors in the mine. It should be understood that the temperature sensor is used to measure the temperature of the mine environment. It can provide real-time temperature values ​​to help determine the temperature conditions inside the mine so as to carry out reasonable ventilation management and temperature control. The humidity sensor is used to measure the humidity level of the mine environment. Humidity refers to the concentration of moisture content in the air, which is very important for the comfort of the mine environment and the health of miners. The humidity sensor can provide humidity values ​​to help evaluate the humidity conditions inside the mine so as to carry out corresponding ventilation adjustment and humidity control. The gas concentration sensor is used to measure the concentration of gas (such as methane) in the mine environment. Gas is a common mine hazardous gas that is flammable and explosive. The gas concentration sensor can monitor the changes in gas concentration inside the mine and promptly detect gas exceeding the standard so as to take corresponding safety measures and ventilation adjustment to ensure the safety of the mine. The blade running angle sensor is used to measure the running angle of the fan blade. The fan is the core equipment of the mine ventilation system. The air volume and air pressure of the fan can be controlled by adjusting the operating angle of the blades. The blade operating angle sensor can monitor the operating status of the blades in real time and provide blade angle data to evaluate the operating effect of the fan and make corresponding adjustments.

[0029] In step S120, the temperature, humidity and gas concentration at the plurality of predetermined time points are arranged into a mine environment input matrix according to the time dimension and the sample dimension, and then the mine environment feature extraction is performed to obtain a mine environment feature vector.

[0030] Figure 3 The present invention is a flowchart of arranging the temperature, humidity and gas concentration of the plurality of predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension in the heat damage prevention method based on mine multi-source data analysis according to the embodiment of the present application, and then extracting the mine environment feature to obtain the mine environment feature vector. Specifically, in the embodiment of the present application, if Figure 3 As shown, the temperature, humidity and gas concentration of the multiple predetermined time points are arranged according to the time dimension and the sample dimension as a mine environment input matrix, and then the mine environment feature extraction is performed to obtain the mine environment feature vector, including: S210, arranging the temperature, humidity and gas concentration of the multiple predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension; S220, performing mixed feature extraction on the mine environment input matrix to obtain a mine environment feature matrix; and, S230, passing the mine environment feature matrix through a linear embedding layer to obtain the mine environment feature vector.

[0031] Specifically, in step S210, the temperature, humidity and gas concentration at the multiple predetermined time points are arranged as a mine environment input matrix according to the time dimension and the sample dimension. Accordingly, considering that in the mine environment, parameters such as temperature, humidity and gas concentration have characteristics in the time series dimension, that is, temperature, humidity and gas concentration will change over time. Based on this, in the technical solution of the present application, the temperature, humidity and gas concentration at the multiple predetermined time points are arranged as a mine environment input matrix according to the time dimension and the sample dimension, which can retain the information of the time series, so that the model can capture the dynamic characteristics of the mine environment changing over time, so as to better understand and predict the evolution trend of the mine environment, thereby providing more accurate prediction and decision support. By arranging according to the sample dimension, different data can be integrated together, and different characteristics can be considered at the same time, so as to more comprehensively understand and predict the state of the entire mine environment.

[0032] Specifically, in step S220, the mine environment input matrix is ​​subjected to hybrid feature extraction to obtain a mine environment feature matrix. Specifically, in an embodiment of the present application, the mine environment input matrix is ​​subjected to hybrid feature extraction to obtain a mine environment feature matrix, including: passing the mine environment input matrix through a deep convolutional neural network model including a hybrid convolutional layer to obtain a mine environment feature matrix. It should be understood that, considering that in a mine environment, parameters such as temperature, humidity, and gas concentration usually have spatial and temporal correlations, and features at different locations and sample points have different spatial scales. Based on this, in the technical solution of the present application, the mine environment input matrix is ​​further subjected to feature extraction through a deep convolutional neural network model including a hybrid convolutional layer, so as to obtain a mine environment feature matrix with better representation capabilities. It is worth mentioning that deep convolutional neural networks (DCNNs) perform well in image and sequence data processing, can effectively extract features from input data, and can use hybrid convolutional layers to automatically learn and extract spatial and temporal features in mine environments. By using hybrid convolutional layers, features of different scales can be considered simultaneously, thereby better capturing spatial information in mine environments and enhancing the model's ability to represent features at different time points and sample points.

[0033] More specifically, in an embodiment of the present application, the mine environment input matrix is ​​passed through a deep convolutional neural network model including a hybrid convolutional layer to obtain a mine environment feature matrix, including: each hybrid convolutional layer of the deep convolutional neural network model including a hybrid convolutional layer includes a parallel first convolution branch structure, a second convolution branch structure, a third convolution branch structure and a fourth convolution branch structure, and a multi-scale fusion structure connected to the first to fourth convolution branch structures, wherein the first convolution branch uses a first convolution kernel having a first size, the second convolution branch uses a second convolution kernel having a first size and a first void ratio, the third convolution branch uses a third convolution kernel having a first size and a second void ratio, and the fourth convolution branch uses a fourth convolution kernel having a first size and a third void ratio.

[0034] In particular, in another achievable embodiment, the temperature, humidity and gas concentration at the plurality of predetermined time points are arranged according to the time dimension and the sample dimension into a mine environment input matrix, and then the mine environment feature extraction is performed to obtain a mine environment feature vector, which can be achieved by the following steps. First, it is necessary to collect relevant data of the mine environment. This includes real-time monitoring data of parameters such as temperature, humidity and gas concentration at a plurality of predetermined time points. Then, the collected data is arranged according to the time dimension and the sample dimension to construct a mine environment input matrix. The time dimension represents different time points, and the sample dimension represents different samples. For example, assuming that there are 3 time points and 4 samples, the mine environment input matrix can be represented as a 3x4 matrix, in which each element represents the value of a parameter. Then, for the data at each time point, some statistical features, such as the average value, the maximum value, the minimum value, the standard deviation, etc., can be calculated. These features can reflect the overall trend and range of variation of the mine environment. Then, some time series features, such as differential features, sliding window statistical features, etc., are calculated on the sample dimension. These features can reflect the dynamic changes and trends of the mine environment between different samples. Finally, the statistical features and time domain features are combined to obtain the final mine environment feature matrix. Feature selection methods, such as principal component analysis (PCA) or correlation analysis, can be used to screen and reduce the features to reduce feature redundancy and improve the effect of the model. Through the above steps, the temperature, humidity and gas concentration at multiple predetermined time points can be arranged according to the time dimension and sample dimension as a mine environment input matrix, and then the mine environment feature extraction is performed to obtain the mine environment feature vector. Such a feature vector can provide key information for mine environment analysis, prediction and decision-making.

[0035] Specifically, in step S230, the mine environment feature matrix is ​​passed through a linear embedding layer to obtain the mine environment feature vector. Accordingly, considering that the mine environment feature matrix usually has a high dimension and contains a large amount of feature information, including key feature information and background redundant information, each mine environment feature vector has mutual correlation and influence. Based on this, in the technical solution of the present application, by passing the mine environment feature matrix through a linear embedding layer to obtain the mine environment feature vector, the dimension of the data can be reduced, the cost of storage and calculation can be reduced, and the redundant information can be reduced while retaining the key features, and the correlation and interaction between them can be captured, so that the feature vector can better represent the comprehensive characteristics of the mine environment. In addition, linear embedding can convert the mine environment feature matrix into a feature vector, in which each dimension corresponds to a feature. In this way, the characteristics of the mine environment can be represented as a vector. That is, each dimension in the feature vector can represent different features, such as temperature, humidity, and gas concentration. By further analyzing and processing the feature vector, the mine environment can be better understood and predicted.

[0036] In step S130, the blade operating angles at a plurality of predetermined time points within a predetermined time of the ventilator are arranged according to the time dimension as a blade angle time series feature input vector, and then blade angle feature extraction is performed to obtain a blade angle feature vector. Figure 4 The present invention is a flow chart of arranging the blade operating angles of the ventilator at multiple predetermined time points within a predetermined time according to the time dimension as a blade angle time series feature input vector, and then extracting the blade angle feature to obtain a blade angle feature vector in the heat damage prevention method based on mine multi-source data analysis according to an embodiment of the present application. Specifically, in the embodiment of the present application, Figure 4 As shown, the blade operating angles of a plurality of predetermined time points within a predetermined time of the ventilator are arranged according to the time dimension as a blade angle timing feature input vector, and then blade angle feature extraction is performed to obtain a blade angle feature vector, including: S310, arranging the blade operating angles of a plurality of predetermined time points within a predetermined time of the ventilator according to the time dimension as a blade angle timing feature input vector; and, S320, passing the blade angle timing feature input vector through a blade angle feature extractor based on a fully connected layer to obtain the blade angle feature vector.

[0037] Specifically, in step S310, the blade operating angles at multiple predetermined time points within the predetermined time of the ventilator are arranged as a blade angle time series feature input vector according to the time dimension. It should be understood that, considering that the blade operating angle of the ventilator changes over time, that is, it has time series in the time dimension. Based on this, in the technical solution of the present application, the blade operating angles at multiple predetermined time points within the predetermined time of the ventilator are arranged as a blade angle time series feature input vector according to the time dimension, and the information of the time series can be retained, so as to better capture the changing trend and periodicity of the ventilator blade operating angle, and provide more accurate input for subsequent model analysis. At the same time, arranging the blade angles according to the time dimension can enable the model to learn the dynamic characteristics of the ventilator blade operating angle. By considering the changes in the blade angles at different time points, the model can better understand the operating state and behavior of the ventilator. This helps to improve the model's ability to predict and control the operation of the ventilator, thereby improving the reliability and efficiency of the ventilation system.

[0038] Specifically, in step S320, the blade angle timing feature input vector is passed through a blade angle feature extractor based on a fully connected layer to obtain the blade angle feature vector. Furthermore, considering that the blade angle timing feature input vector may contain complex timing patterns and correlations, for example, blade angles at different time points may have certain correlations and interactions. Therefore, in the technical solution of the present application, in order to better capture the information in the blade angle timing features so as to provide support for the subsequent control of the blade angle, in the technical solution of the present application, the blade angle timing feature input vector is passed through a blade angle feature extractor based on a fully connected layer to obtain the blade angle feature vector, so that the complex interactive relationship between these blade angle features can be learned, thereby improving the expressive power of the blade angle features.

[0039] It is worth mentioning that the blade angle feature extractor based on the fully connected layer refers to a structure or component that uses a fully connected layer in a neural network to extract blade angle features. It should be understood that the fully connected layer is one of the simplest and most commonly used layers in a neural network. Each of its neurons is connected to all neurons in the previous layer to form a fully connected structure. In the fully connected layer, each neuron has a set of weight parameters for linearly combining and weighted summing the input features, and performing nonlinear transformations through activation functions. In the blade angle feature extractor, the fully connected layer is used to convert the blade angle time series feature input vector into a blade angle feature vector. Specifically, the blade angle time series feature input vector is used as the input of the fully connected layer, and each neuron corresponds to a feature dimension. By learning appropriate weight parameters, the fully connected layer can combine, map and nonlinearly transform each dimensional feature of the input vector to obtain a blade angle feature vector. Usually, after the fully connected layer, an activation function such as ReLU, Sigma ID, etc. can be added to introduce nonlinear transformations. This helps to extract more complex nonlinear features and enhance the expressiveness of the blade angle feature. Therefore, the blade angle feature extractor based on the fully connected layer utilizes the linear combination and nonlinear transformation capabilities of the fully connected layer to convert the blade angle time series feature input vector into a blade angle feature vector to improve the model's ability to express blade angle features.

[0040] In step S140, the mine environment feature vector and the blade angle feature vector are fused to obtain a mine environment global correlation feature vector. It should be understood that the mine environment feature vector and the blade angle feature vector capture different aspects of information respectively. Specifically, the mine environment feature vector contains various parameters and indicators related to the mine environment, such as temperature, humidity, gas concentration, etc. The blade angle feature vector reflects the operating state and effect of the ventilator. Based on this, in the technical solution of the present application, the mine environment feature vector and the blade angle feature vector are fused to obtain a mine environment global correlation feature vector, which can comprehensively consider multiple factors of the mine environment and the ventilation system and provide a more comprehensive and comprehensive feature representation. Furthermore, the mine environment and the ventilation system are interrelated, and there is a complex global correlation relationship between them. Changes in the mine environment will affect the operating effect of the ventilation system, and the operating state of the ventilation system will also affect the mine environment. Therefore, by fusing the mine environment feature vector and the blade angle feature vector, this global correlation can be better captured and modeled, and the ability to understand the complex relationship between the mine environment and the ventilation system can be improved. In addition, the fused feature vectors can also be used for feature interaction and impact analysis, revealing the relationship and mode of action between different features. Specifically, by analyzing the interactive impact between mine environment characteristics and blade angle characteristics, we can gain a deep understanding of the regulation mechanism of the ventilation system on the mine environment, and provide guidance and decision support for the optimization and control of the ventilation system.

[0041] In step S150, based on the global correlation feature vector of the mine environment, it is determined whether the blade operating angle of the ventilator at the current time point should be increased, decreased or unchanged. It should be understood that considering the changes in the mine environment will directly affect the operation effect of the ventilator and the safety of the miners. For example, when the temperature, humidity or gas concentration in the mine environment exceeds a certain limit, it may cause the ventilator to operate insufficiently to meet the needs of the miners, or even cause danger. The blade operating angle of the ventilator is an important parameter for adjusting the ventilation system. Based on this, in the technical solution of the present application, based on the global correlation feature vector of the mine environment, it is determined whether the blade operating angle of the ventilator at the current time point should be increased, decreased or unchanged. By analyzing the global correlation feature vector of the mine environment, the overall trend and change of the mine environment can be understood, so as to determine whether it is necessary to adjust the blade operating angle of the ventilator, and based on adjusting the blade operating angle, the air volume and air pressure of the ventilator can be changed to adapt to the changes in the mine environment and the needs of the miners.

[0042] Specifically, in an embodiment of the present application, based on the global correlation feature vector of the mine environment, it is determined whether the blade operating angle of the ventilator at the current time point should be increased, decreased, or remained unchanged, including: performing feature optimization on the global correlation feature vector of the mine environment to obtain an optimized global correlation feature vector of the mine environment; passing the optimized global correlation feature vector of the mine environment through a classifier to obtain a classification result, and the classification result is used to indicate whether the blade operating angle of the ventilator at the current time point should be increased, decreased, or remained unchanged.

[0043] More specifically, the feature optimization of the global correlation feature vector of the mine environment to obtain the optimized global correlation feature vector of the mine environment includes: performing category harmony matching based on local similarity on the global correlation feature vector of the mine environment to obtain the optimized global correlation feature vector of the mine environment. In particular, in the technical solution of the present application, considering that the temperature, humidity and gas concentration in the mine may show large spatial and temporal differences due to factors such as geographical location, ventilation conditions, and work activities, this diversity may lead to inconsistent distribution of feature matrices in local areas. Environmental data collected at different time points may have inconsistent data quality due to factors such as sensor accuracy, installation location, and data transmission, affecting the accuracy of feature extraction. The deep convolutional neural network model may not be able to fully capture all feature information related to the mine environment, or the sensitivity to certain features is different, resulting in differences in the distribution of feature vectors in local areas. When the mine environment feature vector and the blade angle feature vector are fused, the importance and weights of different features may not be properly processed, resulting in inconsistency in some local areas of the fused global correlation feature vector of the mine environment. However, the global correlation feature vector of the mine environment has local distribution differences, which leads to the problem that when the classifier is used for classification judgment, there is a class harmony offset relative to the predetermined class label of the classifier, which affects the classification iteration effect. Based on this, the global correlation feature vector of the mine environment is matched with the class harmony based on local similarity to obtain the optimized global correlation feature vector of the mine environment.

[0044] Specifically, in an embodiment of the present application, the global correlation feature vector of the mine environment is matched with categories based on local similarity to obtain the optimized global correlation feature vector of the mine environment, including: calculating the global autocorrelation association matrix of the mine environment between the global correlation feature vector of the mine environment and its transposed vector; matrix multiplying the weight parameter matrix of the classifier with the global autocorrelation association matrix of the mine environment to obtain a weight parameter modulated classification autocorrelation association matrix; calculating the correlation perception factor between any two row vectors in the weight parameter modulated classification autocorrelation association matrix to obtain a metric correlation perception matrix; multiplying the metric correlation perception matrix with the global correlation feature vector of the mine environment to obtain the optimized global correlation feature vector of the mine environment.

[0045] In the embodiment of the present application, specifically, performing category harmony matching based on local similarity on the global correlation feature vector of the mine environment to obtain the optimized global correlation feature vector of the mine environment includes: processing the global correlation feature vector of the mine environment with the following formula to obtain the optimized global correlation feature vector of the mine environment;

[0046]

[0047] Among them, V c is the global correlation feature vector of the mine environment, V c T is the transposed vector of the global correlation feature vector of the mine environment, M c is the global autocorrelation matrix of the mine environment, M p is the weight parameter matrix of the classifier, M t is the weight parameter modulated classification autocorrelation matrix, V i is the row vector of the i-th row in the classification autocorrelation matrix modulated by the weight parameter, V j is the row vector of the jth row in the autocorrelation matrix of the weight parameter modulation classification, V j T is the transposed vector of the row vector of the jth row in the weight parameter modulation classification autocorrelation association matrix, L is the total number of rows in the weight parameter modulation classification autocorrelation association matrix, m i,j is the row vector V i and the row vector V j The correlation perception coefficient between 2 is the bi-norm of the eigenvector, represents matrix multiplication, M′ is the metric correlation perception matrix, R is a set of real numbers, and V′ is the global correlation feature vector of the optimized mine environment.

[0048] That is, in the technical solution of the present application, it is considered that there are local distribution differences in the global correlation feature vector of the mine environment, which leads to the problem that when the classifier performs classification judgment, there will be a class harmony offset relative to the predetermined class label of the classifier, which affects the classification iteration effect. Based on this, in the technical solution of the present application, based on the autocorrelation dimension of the global correlation feature vector of the mine environment, the spatial sub-dimensional complexity of the high-dimensional feature space of the feature distribution is used to perform correlation expansion to obtain the global autocorrelation association matrix of the mine environment. Furthermore, the global autocorrelation association matrix of the mine environment is subjected to metric correlation perception based on weight learning parameters. By introducing weight learning parameters, it can effectively quantify and adjust the local distribution differences of the global correlation feature vectors of different mine environments, thereby reducing the class harmony offset caused by these local distribution differences when the classifier performs classification judgment.

[0049] More specifically, in an embodiment of the present application, the optimized mine environment global correlation feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the blade operating angle of the ventilator at the current time point should be increased, decreased, or unchanged, including: using the fully connected layer of the classifier to fully connect encode the optimized mine environment global correlation feature vector to obtain a mine environment global correlation encoded classification feature vector; and, passing the mine environment global correlation encoded classification feature vector through the Softmax classification function of the classifier to obtain the classification result.

[0050] In summary, the heat damage prevention and control method based on the mine multi-source data analysis based on the embodiment of the present application is explained, which obtains the temperature, humidity and gas concentration at multiple predetermined time points within the predetermined time of the mine and the blade operation angle of the ventilator at multiple predetermined time points within the predetermined time, and adopts deep learning technology and data processing methods to determine whether the blade operation angle of the ventilator at the current time point should be increased, decreased or unchanged based on the real-time state of the mine environment. This method can comprehensively consider multiple factors of the mine environment and ventilation system, and adjust the ventilator blade operation angle according to actual needs, which can save energy and improve energy utilization efficiency.

[0051] The above is merely an example of the principles of the present disclosure, and various modifications may be made by those skilled in the art without departing from the scope of the present disclosure. The above embodiments are presented for the purpose of illustration rather than limitation. The present disclosure may also take many forms other than those explicitly described herein.

Claims

1. A method for preventing and controlling heat damage based on multi-source data analysis in mines, characterized in that: include: Obtain the temperature, humidity and gas concentration at multiple predetermined time points within a predetermined time period of the mine and the blade running angles at multiple predetermined time points within a predetermined time period of the ventilator; Arranging the temperature, humidity and gas concentration at the plurality of predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension, and then extracting the mine environment characteristics to obtain a mine environment characteristic vector; Arranging the blade operating angles of the ventilator at a plurality of predetermined time points within a predetermined time according to the time dimension as a blade angle time series feature input vector, and then extracting the blade angle feature to obtain a blade angle feature vector; Fusion of the mine environment feature vector and the blade angle feature vector to obtain a mine environment global correlation feature vector; Based on the global correlation feature vector of the mine environment, it is determined whether the blade operating angle of the ventilator at the current time point should be increased, decreased, or kept unchanged.

2. The heat damage prevention and control method based on mine multi-source data analysis according to claim 1 is characterized in that: Arranging the temperature, humidity and gas concentration at the plurality of predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension, and then extracting the mine environment feature to obtain a mine environment feature vector, including: Arranging the temperature, humidity and gas concentration at the plurality of predetermined time points into a mine environment input matrix according to the time dimension and the sample dimension; Performing mixed feature extraction on the mine environment input matrix to obtain a mine environment feature matrix; The mine environment feature matrix is ​​passed through a linear embedding layer to obtain the mine environment feature vector.

3. The heat damage prevention and control method based on mine multi-source data analysis according to claim 2 is characterized in that: The mine environment input matrix is ​​subjected to mixed feature extraction to obtain a mine environment feature matrix, including: the mine environment input matrix is ​​passed through a deep convolutional neural network model including a mixed convolutional layer to obtain a mine environment feature matrix.

4. The heat damage prevention and control method based on mine multi-source data analysis according to claim 3 is characterized in that: The deep convolutional neural network model containing hybrid convolutional layers includes: each hybrid convolutional layer of the deep convolutional neural network model containing hybrid convolutional layers includes a parallel first convolution branch structure, a second convolution branch structure, a third convolution branch structure and a fourth convolution branch structure, and a multi-scale fusion structure connected to the first to fourth convolution branch structures, wherein the first convolution branch uses a first convolution kernel with a first size, the second convolution branch uses a second convolution kernel with a first size and a first dilation rate, the third convolution branch uses a third convolution kernel with a first size and a second dilation rate, and the fourth convolution branch uses a fourth convolution kernel with a first size and a third dilation rate.

5. The heat damage prevention and control method based on mine multi-source data analysis according to claim 4 is characterized in that: Arranging the blade operating angles of the ventilator at multiple predetermined time points within a predetermined time according to the time dimension as a blade angle time series feature input vector, and then extracting the blade angle feature to obtain a blade angle feature vector, including: Arranging the blade operating angles of the ventilator at a plurality of predetermined time points within a predetermined time according to the time dimension into a blade angle time series feature input vector; The blade angle time series feature input vector is passed through a blade angle feature extractor based on a fully connected layer to obtain the blade angle feature vector.

6. The method for preventing and controlling heat damage based on multi-source data analysis in mines according to claim 5, characterized in that: Based on the global correlation feature vector of the mine environment, determining whether the blade operation angle of the ventilator at the current time point should be increased, decreased, or kept unchanged includes: Performing feature optimization on the mine environment global correlation feature vector to obtain an optimized mine environment global correlation feature vector; The optimized mine environment global correlation feature vector is passed through a classifier to obtain a classification result, and the classification result is used to indicate whether the blade operation angle of the ventilator at the current time point should be increased, decreased, or kept unchanged.

7. The method for preventing and controlling heat damage based on multi-source data analysis in mines according to claim 6, characterized in that: Performing feature optimization on the mine environment global correlation feature vector to obtain an optimized mine environment global correlation feature vector, including: The global correlation feature vector of the mine environment is matched for category harmony based on local similarity to obtain the optimized global correlation feature vector of the mine environment.

8. The method for preventing and controlling heat damage based on multi-source data analysis in mines according to claim 7 is characterized in that: Performing category harmony matching based on local similarity on the global correlation feature vector of the mine environment to obtain the optimized global correlation feature vector of the mine environment includes: Calculating a mine environment global autocorrelation matrix between the mine environment global correlation feature vector and its transposed vector; Performing matrix multiplication of the weight parameter matrix of the classifier and the global autocorrelation matrix of the mine environment to obtain a weight parameter modulated classification autocorrelation matrix; Calculating the correlation perception factor between any two row vectors in the weight parameter modulation classification autocorrelation association matrix to obtain a metric correlation perception matrix; The metric correlation perception matrix is ​​multiplied by the mine environment global correlation feature vector to obtain the optimized mine environment global correlation feature vector.

9. The method for preventing and controlling heat damage based on multi-source data analysis in mines according to claim 8, characterized in that: Calculating the correlation perception factor between any two row vectors in the weight parameter modulation classification autocorrelation association matrix to obtain a metric correlation perception matrix includes: The correlation perception factor between any two row vectors in the weight parameter modulation classification autocorrelation association matrix is ​​calculated using the following correlation perception formula to obtain the metric correlation perception matrix; Wherein, the correlation perception formula is: Among them, V i is the row vector of the i-th row in the classification autocorrelation matrix modulated by the weight parameter, V j is the row vector of the jth row in the autocorrelation matrix of the weight parameter modulation classification, V j T is the transposed vector of the row vector of the jth row in the weight parameter modulation classification autocorrelation association matrix, L is the total number of rows in the weight parameter modulation classification autocorrelation association matrix, m i,j is the row vector V i and the row vector V j The correlation perception coefficient between them, ‖·‖2 is the bi-norm of the feature vector, denotes matrix multiplication, and M′ is the metric correlation-aware matrix.

Citation Information

Cited By

  • Intelligent cooling device for mine heat damage treatment

    CN120367636A