A method for preventing and controlling heat damage based on multi-source data analysis in mines

Abstract

This application discloses a method for preventing and controlling heat damage based on multi-source data analysis in mines. The method uses temperature sensors and wind speed sensors to monitor and collect the temperature and wind speed in the mine in real time, and uses data processing and an analysis algorithm based on deep learning to perform a time-series correlation analysis on the temperature and wind speed in the mine. Based on the real-time changes and time-series collaborative correlation relationship of the multi-source data in the mine, it automatically determines whether ventilation needs to be strengthened. In this way, it is possible to promptly detect the situation of rising temperature or insufficient wind speed in the mine, and generate corresponding early warning prompts to strengthen ventilation, so that measures to strengthen ventilation can be taken in a timely manner to improve the ventilation effect in the mine, reduce the temperature in the mine, and thus reduce the risk of heat damage to miners.

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 phenomenon of elevated air temperature and humidity within a mine during mining operations, caused by factors such as excessively high ground temperatures, poor ventilation, and heat dissipation from mechanical equipment. This condition can affect the health and safety of miners. Mine heat damage can reduce miner productivity, increase the risk of accidents, and even cause heat stroke, coma, or death. Therefore, preventing and controlling mine heat damage is a crucial component of mine safety.

[0003] However, traditional prevention and control methods are often based on historical data and empirical experience, and are unable to adapt to real-time mine environmental data. Due to the variability and diversity of mine heat damage, fixed and static prevention and control methods cannot promptly respond to and adapt to changes in the mine environment. This can lead to a lag in prevention and control measures, preventing effective measures from being implemented at critical moments, and thus compromising the effectiveness of mine heat damage prevention and control.

[0004] Therefore, a heat damage prevention and control solution based on multi-source data analysis in mines is desired. Summary of the Invention

[0005] 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 uses temperature sensors and wind speed sensors to monitor and collect the temperature and wind speed in the mine in real time, and uses data processing and deep learning-based analysis algorithms to perform time-series correlation analysis on the temperature in the mine and the wind speed in the mine, so as to automatically judge whether ventilation needs to be strengthened based on the real-time changes and time-series collaborative correlation relationship of multi-source data in the mine. In this way, it is possible to timely discover the situation where the temperature in the mine is rising or the wind speed is insufficient, and generate corresponding early warning prompts for strengthening ventilation, so that measures to strengthen ventilation can be taken in time to improve the ventilation effect in the mine, reduce the temperature in the mine, and thus reduce the risk of heat damage to miners.

[0006] According to one aspect of the present application, a method for preventing and controlling heat damage based on multi-source mine data analysis is provided, which includes:

[0007] Obtaining a time series of the temperature in the mine and a time series of the wind speed in the mine collected by a temperature sensor and a wind speed sensor;

[0008] Arrange the time series of the temperature in the mine and the time series of the wind speed in the mine into a time series input vector of the temperature in the mine and a time series input vector of the wind speed in the mine, respectively, according to the time dimension;

[0009] Calculating a sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector to obtain a mine temperature-wind speed time series correlation matrix;

[0010] Performing time series correlation pattern feature extraction on the mine temperature-wind speed time series correlation matrix to obtain a mine temperature-wind speed time series correlation feature graph;

[0011] The mine temperature-wind speed time series correlation feature map is passed through a local feature salient device based on an adaptive attention layer to obtain an adaptive enhanced mine temperature-wind speed time series correlation feature map;

[0012] Processing the adaptive enhanced mine temperature-wind speed time series correlation feature map using a prototype feature extraction network to obtain a mine temperature-wind speed time series prototype semantic feature;

[0013] Based on the semantic features of the mine temperature-wind speed time series prototype, it is determined whether to generate an early warning prompt for enhanced ventilation.

[0014] Compared with the existing technology, the present application provides a heat damage prevention and control method based on multi-source mine data analysis. It uses temperature sensors and wind speed sensors to monitor and collect the temperature and wind speed in the mine in real time, and uses data processing and deep learning-based analysis algorithms to perform time-series correlation analysis on the temperature and wind speed in the mine. Based on the real-time changes and time-series collaborative correlation relationship of the multi-source data in the mine, it automatically determines whether ventilation needs to be strengthened. In this way, it is possible to promptly detect the situation of rising temperature or insufficient wind speed in the mine, and generate corresponding early warning prompts to strengthen ventilation, so that measures to strengthen ventilation can be taken in a timely manner to improve the ventilation effect in the mine, reduce the temperature in the mine, and thus reduce the risk of heat damage to miners. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended 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 drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 Flowchart of a method for preventing and controlling heat damage based on multi-source mine data analysis according to an embodiment of the present application.

[0017] Figure 2 Schematic diagram of the architecture of a heat damage prevention and control method based on multi-source data analysis in a mine according to an embodiment of the present application.

[0018] Figure 3The present invention provides a flowchart for training the mine temperature-wind speed time series correlation pattern feature extractor based on the convolutional neural network model, the local feature highlighter based on the adaptive attention layer, the prototype feature extraction network and the classifier in the heat damage prevention and control method based on multi-source mine data analysis according to an embodiment of the present application.

[0019] Figure 4 This is a flowchart of a method for preventing and controlling heat damage based on mine multi-source data analysis according to an embodiment of the present application, in which the training mine temperature-wind speed time series prototype semantic feature vector is passed through a classifier to obtain a classification loss function value. DETAILED DESCRIPTION

[0020] The following describes embodiments of the present disclosure in more detail with reference to the accompanying drawings. While the drawings illustrate certain embodiments of the present disclosure, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not intended to limit the scope of protection of the present disclosure.

[0021] It should be understood that the various steps described in the method embodiments of the present disclosure may be performed in a different order and / or in parallel. Furthermore, 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.

[0022] In the description of the embodiments of the present disclosure, the term "including" and similar terms should be understood as open inclusion, i.e., "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.

[0023] It should be noted that the modifications of "one" and "multiple" mentioned in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, they should be understood as "one or more".

[0024] Mine heat damage refers to the phenomenon during mining operations whereby the temperature and humidity within a mine exceed the human comfort range due to factors such as excessively high ground temperatures, poor ventilation, and heat dissipation from mechanical equipment, impacting miners' health and work efficiency. Mine heat damage increases the risk of accidents, reduces mine production capacity, wastes resources, and pollutes the environment. Therefore, preventing and controlling mine heat damage is a crucial task in ensuring safe production and sustainable development in mines.

[0025] However, traditional methods for preventing and controlling heat damage in mines are typically based on historical data and empirical experience, failing to adapt to real-time mine environmental data. Due to the variability and diversity of mine heat damage, fixed, static prevention and control methods are unable to promptly respond to and adapt to changes in the mine environment. This can lead to delayed implementation of prevention and control measures, preventing effective action at critical moments, and thus compromising the effectiveness of mine heat damage prevention and control.

[0026] Therefore, in response to the above technical problems, the technical concept of this application is to monitor and collect the temperature and wind speed in the mine in real time through temperature sensors and wind speed sensors, and use data processing and deep learning-based analysis algorithms to perform time-series correlation analysis on the temperature and wind speed in the mine, so as to automatically determine whether ventilation needs to be strengthened based on the real-time changes and time-series collaborative correlation relationship of multi-source data in the mine. In this way, it is possible to promptly detect the situation of rising temperature or insufficient wind speed in the mine, and generate corresponding early warning prompts for strengthening ventilation, so that measures to strengthen ventilation can be taken in time to improve the ventilation effect in the mine, reduce the temperature in the mine, and thus reduce the risk of heat damage to miners.

[0027] Figure 1 Flowchart of a method for preventing and controlling heat damage based on multi-source mine data analysis 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 mine data analysis according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to the embodiment of the present application, the heat damage prevention and control method based on mine multi-source data analysis includes: S110, obtaining the time series of the temperature in the mine and the time series of the wind speed in the mine collected by the temperature sensor and the wind speed sensor; S120, arranging the time series of the temperature in the mine and the time series of the wind speed in the mine into the time series input vector of the temperature in the mine and the time series input vector of the wind speed in the mine according to the time dimension; S130, calculating the sample covariance correlation matrix of the wind speed time series input vector in the mine relative to the temperature time series input vector in the mine to obtain the mine temperature-wind speed time series correlation matrix; S140, The mine temperature-wind speed time series correlation matrix is subjected to time series correlation pattern feature extraction to obtain a mine temperature-wind speed time series correlation feature map; S150, the mine temperature-wind speed time series correlation feature map is passed through a local feature salient device based on an adaptive attention layer to obtain an adaptive enhanced mine temperature-wind speed time series correlation feature map; S160, the adaptive enhanced mine temperature-wind speed time series correlation feature map is processed using a prototype feature extraction network to obtain a mine temperature-wind speed time series prototype semantic feature; and, S170, based on the mine temperature-wind speed time series prototype semantic feature, it is determined whether to generate an early warning prompt for enhanced ventilation.

[0028] In step S110, a time series of the mine temperature and wind speed, collected by the temperature sensor and wind speed sensor, is obtained. It should be understood that mine temperature and wind speed are important indicators for assessing mine safety. Specifically, the mine temperature can reflect the changing trend of the mine's internal temperature. For example, excessively high temperatures may lead to the risk of mine fires. The mine wind speed can help determine the operational status of the mine's ventilation system and can indicate poor ventilation or insufficient air volume, thereby reducing the occurrence of accidents such as explosions and poisoning. Based on this, in the technical solution of the present application, a time series of the mine temperature and wind speed, collected by the temperature sensor and wind speed sensor, is obtained. By analyzing the time series of the mine temperature and wind speed, an early warning model can be established to detect abnormal changes and fluctuations in the mine temperature and wind speed, thereby providing early warning of potential failures and hazards. For example, when the temperature rises beyond a safe range or the wind speed drops abnormally low, the system can automatically issue an alarm, prompting operators to take appropriate measures, such as increasing ventilation and repairing equipment, to ensure the safe operation of the mine.

[0029] In step S120, the time series of the temperature in the mine and the time series of the wind speed in the mine are arranged according to the time dimension into a time series input vector for the temperature in the mine and a time series input vector for the wind speed in the mine, respectively. Accordingly, considering that the changes in the temperature in the mine and the wind speed in the mine occur over time, that is, the temperature in the mine and the wind speed in the mine have certain characteristic information in the time dimension. Therefore, in order to retain the time series information of the temperature in the mine and the wind speed in the mine data, so as to better understand and analyze the characteristics of the change trend, periodicity and volatility of the temperature and the wind speed in the mine in the time dimension, in the technical solution of the present application, the time series of the temperature in the mine and the time series of the wind speed in the mine are arranged according to the time dimension into a time series input vector for the temperature in the mine and a time series input vector for the wind speed in the mine, respectively. In this way, it is convenient to conduct a time-series correlation analysis on the temperature in the mine and the wind speed in the mine, so as to discover the possible time-series collaborative correlation relationship and change and fluctuation trends between them. For example, an increase in temperature may be accompanied by a decrease in wind speed, or the temperature and wind speed may show similar fluctuation patterns within a specific time period, providing support for subsequent ventilation warning prompts in the mine.

[0030] In step S130, the sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector is calculated to obtain the mine temperature-wind speed time series correlation matrix. It should be understood that, considering that the mine temperature and the mine wind speed may show similar change trends and fluctuations at certain time points or time periods, and that there may also be a time series correlation pattern and correlation relationship between the mine temperature and the mine wind speed. Therefore, in the technical solution of the present application, the sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector is calculated to obtain the mine temperature-wind speed time series correlation matrix. It is worth mentioning that the sample covariance correlation matrix can measure the correlation between two variables. Specifically, the mine temperature and the mine wind speed are two variables, and their time series correlation degree can be evaluated by calculating their sample covariance. Specifically, the positive or negative value of the covariance indicates the direction of the linear relationship between the mine temperature and the mine wind speed, while the absolute value of the covariance indicates the strength of the linear relationship between the two variables. In other words, by calculating the sample covariance, we can discover the correlation, association pattern, and fluctuation pattern between the mine wind speed time series input vector and the mine temperature time series input vector. This helps us gain a deeper understanding of the dynamic processes within the mine, thereby improving mine safety and operational efficiency.

[0031] Specifically, in an embodiment of the present application, calculating the sample covariance correlation matrix of the wind speed time series input vector in the mine relative to the temperature time series input vector in the mine to obtain the mine temperature-wind speed time series correlation matrix includes: calculating the sample covariance correlation matrix of the wind speed time series input vector in the mine relative to the temperature time series input vector in the mine using the following sample covariance formula to obtain the mine temperature-wind speed time series correlation matrix; wherein, the sample covariance formula is:

[0032] M=W T XX T W

[0033] Wherein, W is the time series input vector of the wind speed in the mine, X is the time series input vector of the temperature in the mine, and M is the mine temperature-wind speed time series correlation matrix.

[0034] In step S140, the mine temperature-wind speed time series correlation matrix is subjected to time series correlation pattern feature extraction to obtain a mine temperature-wind speed time series correlation feature graph. Specifically, in an embodiment of the present application, the mine temperature-wind speed time series correlation matrix is subjected to time series correlation pattern feature extraction to obtain a mine temperature-wind speed time series correlation feature graph, comprising: passing the mine temperature-wind speed time series correlation matrix through a mine temperature-wind speed time series correlation pattern feature extractor based on a convolutional neural network model to obtain the mine temperature-wind speed time series correlation feature graph. Accordingly, considering that the mine temperature-wind speed time series correlation matrix has a local implicit correlation feature relationship in the time series, and also considering that the convolutional neural network has good feature extraction capabilities when processing time series data and capturing implicit correlation relationships between time series data. Therefore, in the technical solution of the present application, the mine temperature-wind speed time series correlation matrix is passed through a mine temperature-wind speed time series correlation pattern feature extractor based on a convolutional neural network model to extract the time series correlation feature information between the mine temperature and wind speed in the mine temperature-wind speed time series correlation matrix, thereby obtaining a mine temperature-wind speed time series correlation feature diagram with better representation capabilities.

[0035] In step S150, the mine temperature-wind speed time-series correlation feature map is passed through a local feature saliencer based on an adaptive attention layer to obtain an adaptively enhanced mine temperature-wind speed time-series correlation feature map. Accordingly, considering that each channel in the mine temperature-wind speed time-series correlation feature map may have different feature information, some key and significant local time-series correlation features are crucial for revealing important information and correlations between environmental parameters within the mine, while some irrelevant feature information also exists. Therefore, in order to enhance the key and significant temporal synergistic correlation features in the mine temperature-wind speed time-series correlation feature map, thereby better capturing these local time-series correlation patterns and reducing interference from irrelevant features, in the technical solution of the present application, the mine temperature-wind speed time-series correlation feature map is passed through a local feature saliencer based on an adaptive attention layer to obtain an adaptively enhanced mine temperature-wind speed time-series correlation feature map. It should be understood that in the mine temperature-wind speed time-series correlation feature map, the mine temperature-wind speed time-series correlation features on different channels may have different local time-series synergistic correlation patterns. In particular, the adaptive attention layer automatically adjusts the weight distribution of the features of different channels in the mine temperature-wind speed time series correlation feature map, highlights important local significant feature areas, and captures the local significant time series correlation patterns in the mine temperature-wind speed time series correlation feature map, thereby obtaining the more expressive adaptive enhanced mine temperature-wind speed time series correlation feature map, providing useful information and decision support for mine operation and safety management.

[0036] Specifically, in an embodiment of the present application, the mine temperature-wind speed time series correlation feature map is passed through a local feature saliency device based on an adaptive attention layer to obtain an adaptive enhanced mine temperature-wind speed time series correlation feature map, including: passing the mine temperature-wind speed time series correlation feature map through the local feature saliency device based on an adaptive attention layer using the following adaptive saliency formula to obtain the adaptive enhanced mine temperature-wind speed time series correlation feature map; wherein the adaptive saliency formula is:

[0037] v c =pool(F)

[0038] A=σ(W a *v c +B a )

[0039]

[0040] F'=A'⊙F

[0041] Where F represents the mine temperature-wind speed time series correlation feature map, pool(·) represents the global mean pooling process for each feature matrix along the channel dimension in the feature map, and v c The channel feature vector representing the mine temperature-wind speed time series correlation feature map, W a and B a Represents the weight and bias of the convolution layer, σ represents the activation function, A represents the convolution feature vector of the channel feature vector, A i represents the eigenvalue of the i-th position in the convolution feature vector, A' represents the weighted feature vector, ⊙ represents the point multiplication by position, and F' represents the adaptive enhanced mine temperature-wind speed time series correlation feature map.

[0042] In step S160, the adaptively enhanced mine temperature-wind speed temporal correlation feature map is processed using a prototype feature extraction network to obtain a mine temperature-wind speed temporal prototype semantic feature. It should be understood that, considering that each adaptively enhanced mine temperature-wind speed temporal correlation feature vector in the adaptively enhanced mine temperature-wind speed temporal correlation feature map represents a local temporal correlation pattern and feature between the mine temperature and the mine wind speed in different local time periods, and that each adaptively enhanced mine temperature-wind speed temporal correlation feature vector has mutual correlation and influence, that is, each adaptively enhanced mine temperature-wind speed temporal correlation feature vector may have similar temporal fluctuation patterns or change trends, but may also contain some feature information that is not related to the temporal correlation prototype feature between the mine temperature and the mine wind speed. Therefore, in the technical solution of the present application, the prototype feature extraction network is used to process the adaptively enhanced mine temperature-wind speed temporal correlation feature map to obtain a mine temperature-wind speed temporal prototype semantic feature vector. In particular, the prototype feature extraction network weights each of the adaptively enhanced mine temperature-wind speed time series association feature vectors in the adaptively enhanced mine temperature-wind speed time series association feature vector sequence obtained by expanding the adaptively enhanced mine temperature-wind speed time series association feature graph relative to other adaptively enhanced mine temperature-wind speed time series association features as a weight. In this way, the prototype feature vectors that are more representative and important than other feature vectors can be highlighted, and the key semantic information in the adaptively enhanced mine temperature-wind speed time series association feature graph can be extracted, thereby obtaining the mine temperature-wind speed time series prototype semantic feature vector with more expressiveness and discrimination, providing a more reliable and effective feature representation for subsequent ventilation warning analysis and decision-making.

[0043] Specifically, in an embodiment of the present application, a prototype feature extraction network is used to process the adaptive enhanced mine temperature-wind speed time series association feature map to obtain a mine temperature-wind speed time series prototype semantic feature, including: using the prototype feature extraction network to process the adaptive enhanced mine temperature-wind speed time series association feature map using the following prototype feature formula to obtain a mine temperature-wind speed time series prototype semantic feature vector as the mine temperature-wind speed time series prototype semantic feature; wherein the prototype feature formula is:

[0044]

[0045]

[0046] Among them, v irepresents the i-th adaptive enhanced mine temperature-wind speed time series correlation feature vector in the adaptive enhanced mine temperature-wind speed time series correlation feature vector sequence after the expansion of the adaptive enhanced mine temperature-wind speed time series correlation feature graph, v j represents the jth adaptive enhanced mine temperature-wind speed time series correlation feature vector in the adaptive enhanced mine temperature-wind speed time series correlation feature vector sequence after the expansion of the adaptive enhanced mine temperature-wind speed time series correlation feature map, ‖·‖1 represents the norm of the feature vector, V k represents the sequence of the adaptive enhanced mine temperature-wind speed time series correlation feature vector, M represents the number of the adaptive enhanced mine temperature-wind speed time series correlation feature vector sequences - 1, D i represents the eigenvalue of each position in the semantic difference feature vector of the mine temperature-wind speed local time series, P k represents the mine temperature-wind speed time series prototype semantic feature vector, and exp(·) represents an exponential function with the natural constant e as the base.

[0047] In step S170, based on the mine temperature-wind speed time series prototype semantic features, it is determined whether to generate an early warning prompt for enhanced ventilation. Specifically, in an embodiment of the present application, based on the mine temperature-wind speed time series prototype semantic features, it is determined whether to generate an early warning prompt for enhanced ventilation, including: passing the mine temperature-wind speed time series prototype semantic feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether an early warning prompt for enhanced ventilation is generated. That is to say, the mine temperature-wind speed time series prototype semantic features after prototype feature extraction are used for classification processing, thereby automatically judging whether to generate an early warning prompt for enhanced ventilation based on the prototype semantic features of the multi-source data in the mine in the time dimension. In this way, it is possible to timely discover the situation of increased temperature or insufficient wind speed in the mine, and take measures to enhance ventilation. In this way, the ventilation effect in the mine can be effectively improved, the temperature in the mine can be reduced, and the risk of heat damage to miners can be reduced.

[0048] It is worth mentioning that those skilled in the art should be aware that before applying a deep neural network model for inference, the deep neural network model must first be trained so that the deep neural network can implement specific functional capabilities.

[0049] Specifically, in an embodiment of the present application, a training step is also included: for training the mine temperature-wind speed time series correlation pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network and the classifier.

[0050] Figure 3This is a flowchart for training the mine temperature-wind speed temporal correlation pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network, and the classifier 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 3 As shown, the training step includes: S210, obtaining training data, the training data including the time series of the temperature in the training mine and the time series of the wind speed in the training mine collected by the temperature sensor and the wind speed sensor, as well as the real value of the warning prompt of whether to increase ventilation; S220, arranging the time series of the temperature in the training mine and the time series of the wind speed in the training mine into the training mine temperature time series input vector and the training mine wind speed time series input vector according to the time dimension respectively; S230, calculating the sample covariance correlation matrix of the training mine wind speed time series input vector relative to the training mine temperature time series input vector to obtain the training mine temperature-wind speed time series correlation matrix; S240, passing the training mine temperature-wind speed time series correlation matrix through the mine temperature-wind speed time series correlation pattern feature extractor based on the convolutional neural network model to obtain the training mine temperature-wind speed time series correlation feature map; S250, the training mine temperature-wind speed time series correlation matrix The feature map is passed through the local feature salient device based on the adaptive attention layer to obtain a training adaptive enhanced mine temperature-wind speed time series association feature map; S260, the training adaptive enhanced mine temperature-wind speed time series association feature map is processed using the prototype feature extraction network to obtain a training mine temperature-wind speed time series prototype semantic feature vector; S270, the training mine temperature-wind speed time series prototype semantic feature vector is passed through a classifier to obtain a classification loss function value; and, S280, based on the classification loss function value and through gradient descent back propagation, the mine temperature-wind speed time series association pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network and the classifier are trained, wherein each time the training mine temperature-wind speed time series prototype semantic feature vector passes through the classifier for classification regression iteration, the training mine temperature-wind speed time series prototype semantic feature vector is optimized.

[0051] Figure 4 This is a flow chart of the method for preventing and controlling heat damage based on mine multi-source data analysis according to an embodiment of the present application, wherein the training mine temperature-wind speed time series prototype semantic feature vector is passed through a classifier to obtain a classification loss function value. Specifically, in the embodiment of the present application, if Figure 4As shown, the training mine temperature-wind speed time series prototype semantic feature vector is passed through a classifier to obtain a classification loss function value, including: S310, using the classifier to classify the training mine temperature-wind speed time series prototype semantic feature vector to obtain a training classification result; and, S320, calculating the cross entropy loss function value between the training classification result and the true value of the warning prompt for whether to generate enhanced ventilation as the classification loss function value.

[0052] Specifically, in step S280, the mine temperature-wind speed time series association pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network and the classifier are trained based on the classification loss function value and through back propagation of gradient descent, wherein each time the training mine temperature-wind speed time series prototype semantic feature vector is iterated through classification regression by the classifier, the training mine temperature-wind speed time series prototype semantic feature vector is optimized. It should be understood that in the technical solution of the present application, each feature matrix of the training adaptive enhanced mine temperature-wind speed time series correlation feature map expresses the local high-order correlation features of the specific channel distribution enhancement associated with the full time domain covariance of the temperature in the training mine and the wind speed in the training mine, and when the training adaptive enhanced mine temperature-wind speed time series correlation feature map is processed using the prototype feature extraction network, the prototype features are calculated based on the feature matrices of the training adaptive enhanced mine temperature-wind speed time series correlation feature map as units to obtain the training mine temperature-wind speed time series prototype semantic feature vector.

[0053] However, considering the differences in channel feature distribution of each feature matrix of the training adaptive enhanced mine temperature-wind speed time series correlation feature map, when calculating the prototype features of each feature matrix to obtain the training mine temperature-wind speed time series prototype semantic feature vector, the significance of the feature distribution information of the local high-order correlation features of each feature matrix based on its predetermined channel position will also be affected, making it difficult for the training mine temperature-wind speed time series prototype semantic feature vector to stably focus on the significant local distribution of features during the training process, thereby affecting the expression effect of the training mine temperature-wind speed time series prototype semantic feature vector and the training speed of the model. Based on this, in the technical solution of the present application, each time the training mine temperature-wind speed time series prototype semantic feature vector is iterated through classification regression by the classifier, the training mine temperature-wind speed time series prototype semantic feature vector is optimized.

[0054] More specifically, in an embodiment of the present application, each time the training mine temperature-wind speed time series prototype semantic feature vector is subjected to classification regression iteration by the classifier, the training mine temperature-wind speed time series prototype semantic feature vector is optimized, including: each time the training mine temperature-wind speed time series prototype semantic feature vector is subjected to classification regression iteration by the classifier, the training mine temperature-wind speed time series prototype semantic feature vector is optimized using the following optimization formula; wherein, the optimization formula is:

[0055]

[0056] Among them, v i is the eigenvalue of the i-th position in the training mine temperature-wind speed time series prototype semantic feature vector V, and are the 1-norm and 2-norm squares of the training mine temperature-wind speed time series prototype semantic feature vector V, L is the length of the training mine temperature-wind speed time series prototype semantic feature vector V, and ω is a weight hyperparameter, v' i is the eigenvalue of the i-th position in the optimized semantic feature vector of the training mine temperature-wind speed time series prototype, and log is the logarithmic function value with base 2.

[0057] In particular, by geometrically registering the high-dimensional feature manifold shape based on the scale and structural parameters of the training mine temperature-wind speed time series prototype semantic feature vector V, it is possible to focus on features with rich feature semantic information in the feature set composed of the eigenvalues of the training mine temperature-wind speed time series prototype semantic feature vector V, that is, distinguishable stable features of interest based on the dissimilarity of local context information representation when the classifier is performing classification, thereby achieving significant labeling of the feature information of the training mine temperature-wind speed time series prototype semantic feature vector V during the classification process and improving the training speed of the classifier. In this way, it is possible to promptly detect situations where the temperature in the mine is rising or the wind speed is insufficient, and take measures to strengthen ventilation. In this way, the ventilation effect in the mine can be effectively improved, the temperature in the mine can be reduced, and the risk of heat damage to miners can be reduced.

[0058] In summary, the heat damage prevention and control method based on the analysis of multi-source data in mines according to the embodiment of the present application is explained. It uses temperature sensors and wind speed sensors to monitor and collect the temperature and wind speed in the mine in real time, and uses data processing and deep learning-based analysis algorithms to perform time-series correlation analysis on the temperature and wind speed in the mine. Based on the real-time changes and time-series collaborative correlation relationship of the multi-source data in the mine, it automatically determines whether ventilation needs to be strengthened. In this way, it is possible to promptly detect the situation of rising temperature or insufficient wind speed in the mine, and generate corresponding early warning prompts for strengthening ventilation, so that measures to strengthen ventilation can be taken in time to improve the ventilation effect in the mine, reduce the temperature in the mine, and thus reduce the risk of heat damage to miners.

[0059] The foregoing 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 purposes of illustration and not limitation. The present disclosure may also take many forms other than those explicitly described herein. Therefore, it is emphasized that the present disclosure is not limited to the methods, systems, and apparatus explicitly disclosed, but is intended to encompass variations and modifications within the spirit and scope of the appended claims.

Claims

1. A method for preventing and controlling heat damage based on multi-source data analysis in mines, characterized in that: include: Obtaining a time series of the temperature in the mine and a time series of the wind speed in the mine collected by a temperature sensor and a wind speed sensor; Arrange the time series of the temperature in the mine and the time series of the wind speed in the mine into a time series input vector of the temperature in the mine and a time series input vector of the wind speed in the mine, respectively, according to the time dimension; Calculating a sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector to obtain a mine temperature-wind speed time series correlation matrix; Performing time series correlation pattern feature extraction on the mine temperature-wind speed time series correlation matrix to obtain a mine temperature-wind speed time series correlation feature graph; The mine temperature-wind speed time series correlation feature map is passed through a local feature salient device based on an adaptive attention layer to obtain an adaptive enhanced mine temperature-wind speed time series correlation feature map; Processing the adaptive enhanced mine temperature-wind speed time series correlation feature map using a prototype feature extraction network to obtain a mine temperature-wind speed time series prototype semantic feature; Determining whether to generate an early warning prompt for enhanced ventilation based on the semantic features of the mine temperature-wind speed time series prototype; The method of passing the mine temperature-wind speed time series correlation feature map through a local feature saliency device based on an adaptive attention layer to obtain an adaptive enhanced mine temperature-wind speed time series correlation feature map comprises: passing the mine temperature-wind speed time series correlation feature map through the local feature saliency device based on an adaptive attention layer using the following adaptive saliency formula to obtain the adaptive enhanced mine temperature-wind speed time series correlation feature map; The adaptive saliency formula is: in, represents the mine temperature-wind speed time series correlation characteristic diagram, Indicates that global mean pooling is performed on each feature matrix along the channel dimension in the feature map. The channel feature vector representing the mine temperature-wind speed time series correlation feature map, and represents the weights and biases of the convolutional layer, represents the activation function, represents the convolution feature vector of the channel feature vector, Represents the convolution feature vector The eigenvalues at the positions, represents the weight feature vector, Indicates point multiplication by position, The figure represents the adaptive enhanced mine temperature-wind speed time series correlation characteristic diagram.

2. The heat damage prevention and control method based on mine multi-source data analysis according to claim 1 is characterized in that: Calculating a sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector to obtain a mine temperature-wind speed time series correlation matrix, comprising: calculating a sample covariance correlation matrix of the mine wind speed time series input vector relative to the mine temperature time series input vector using the following sample covariance formula to obtain the mine temperature-wind speed time series correlation matrix; Wherein, the sample covariance formula is: in, is the wind speed time series input vector in the mine, is the temperature time series input vector in the mine, is the mine temperature-wind speed time series correlation matrix.

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 temperature-wind speed time series correlation matrix is subjected to time series correlation pattern feature extraction to obtain a mine temperature-wind speed time series correlation feature diagram, including: passing the mine temperature-wind speed time series correlation matrix through a mine temperature-wind speed time series correlation pattern feature extractor based on a convolutional neural network model to obtain the mine temperature-wind speed time series correlation feature diagram.

4. The heat damage prevention and control method based on mine multi-source data analysis according to claim 3 is characterized in that: The method includes processing the adaptive enhanced mine temperature-wind speed time series correlation feature map using a prototype feature extraction network to obtain a mine temperature-wind speed time series prototype semantic feature, including: processing the adaptive enhanced mine temperature-wind speed time series correlation feature map using the prototype feature extraction network using the following prototype feature formula to obtain a mine temperature-wind speed time series prototype semantic feature vector as the mine temperature-wind speed time series prototype semantic feature; Wherein, the prototype characteristic formula is: in, The first one in the sequence of adaptive enhanced mine temperature-wind speed time series correlation feature vectors after the expansion of the adaptive enhanced mine temperature-wind speed time series correlation feature map is represented. Adaptive enhanced mine temperature-wind speed time series correlation feature vector, The first one in the sequence of adaptive enhanced mine temperature-wind speed time series correlation feature vectors after the expansion of the adaptive enhanced mine temperature-wind speed time series correlation feature map is represented. Adaptive enhanced mine temperature-wind speed time series correlation feature vector, represents the norm of the eigenvector, represents the sequence of the adaptive enhanced mine temperature-wind speed time series correlation feature vectors, represents the number of the adaptive enhanced mine temperature-wind speed time series correlation feature vector sequences -1, Represents the eigenvalues of each position in the semantic difference feature vector of the mine temperature-wind speed local time series, represents the mine temperature-wind speed time series prototype semantic feature vector, Expressed as a natural constant An exponential function with base .

5. The heat damage prevention and control method based on mine multi-source data analysis according to claim 4 is characterized in that: Based on the semantic features of the mine temperature-wind speed time series prototype, determining whether to generate an early warning prompt for enhanced ventilation includes: passing the mine temperature-wind speed time series prototype semantic feature vector through a classifier to obtain a classification result, and the classification result is used to indicate whether to generate an early warning prompt for enhanced ventilation.

6. The heat damage prevention and control method based on mine multi-source data analysis according to claim 5 is characterized in that: It also includes a training step: for training the mine temperature-wind speed time series correlation pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network and the classifier.

7. The heat damage prevention and control method based on mine multi-source data analysis according to claim 6 is characterized in that: The training step comprises: Acquiring training data, the training data including a time series of temperature and wind speed within a training mine collected by a temperature sensor and a wind speed sensor, and a true value of the warning prompt for whether to generate enhanced ventilation; Arrange the time series of the temperature in the training mine and the time series of the wind speed in the training mine according to the time dimension into a training mine temperature time series input vector and a training mine wind speed time series input vector respectively; Calculating a sample covariance correlation matrix of the training mine wind speed time series input vector relative to the training mine temperature time series input vector to obtain a training mine temperature-wind speed time series correlation matrix; Passing the training mine temperature-wind speed time series correlation matrix through the mine temperature-wind speed time series correlation pattern feature extractor based on the convolutional neural network model to obtain a training mine temperature-wind speed time series correlation feature map; Passing the training mine temperature-wind speed time series correlation feature map through the local feature salient device based on the adaptive attention layer to obtain a training adaptive enhanced mine temperature-wind speed time series correlation feature map; Processing the training adaptive enhanced mine temperature-wind speed time series correlation feature map using the prototype feature extraction network to obtain a training mine temperature-wind speed time series prototype semantic feature vector; Passing the training mine temperature-wind speed time series prototype semantic feature vector through a classifier to obtain a classification loss function value; The mine temperature-wind speed time series association pattern feature extractor based on the convolutional neural network model, the local feature salient device based on the adaptive attention layer, the prototype feature extraction network and the classifier are trained based on the classification loss function value and through back propagation of gradient descent, wherein each time the training mine temperature-wind speed time series prototype semantic feature vector is iterated through classification regression by the classifier, the training mine temperature-wind speed time series prototype semantic feature vector is optimized.

8. The heat damage prevention and control method based on mine multi-source data analysis according to claim 7 is characterized in that: The training mine temperature-wind speed time series prototype semantic feature vector is passed through a classifier to obtain a classification loss function value, including: Using the classifier to classify the training mine temperature-wind speed time series prototype semantic feature vector to obtain a training classification result; A cross entropy loss function value between the training classification result and the true value of the early warning prompt of whether to generate enhanced ventilation is calculated as the classification loss function value.

Images