Multi-modal data-based rock burst prediction large model construction method

Abstract

The invention discloses a rock burst prediction large model construction method based on multi-modal data, and the method comprises the steps: constructing a multi-modal data set through collecting data of different modals; preprocessing the multi-modal data set, and constructing a precursor mode sequence of model training; according to the characteristics of different mining areas, the precursor mode sequence is converted into a corresponding grade form, and corresponding rock burst danger grade labels are divided; transform is adopted as a core framework, the precursor mode sequence after grade division is processed, and finally prediction of the rock burst grade occurrence probability is achieved. A comprehensive index method is used for independently evaluating the danger degree of the mining information data and the geological structure data, and the overall danger level of the rock burst is comprehensively evaluated in combination with a danger probability result output by the rock burst prediction module; according to the method, the applicability and prediction precision of the model under different mining area conditions can be improved, and accurate prediction of the rock burst risk is realized.

Description

Technical Field

[0001] The invention relates to a method for constructing a rock burst model, in particular to a method for constructing a large rock burst prediction model based on multimodal data, and belongs to the technical field of coal mine monitoring and early warning. Background Art

[0002] Rock burst is a typical high-energy dynamic disaster in the process of coal mining. It is characterized by strong suddenness and great destructiveness, and it is very easy to cause serious consequences such as damage to mine equipment and personal injury. The occurrence mechanism of rock burst is complex and is affected by multiple factors such as mine geological structure, rock mass stress state, and mining depth. In recent years, with the continuous increase in the mining depth of mine resources, shallow resources have gradually been exhausted, and the focus of underground mining activities has gradually shifted to deep layers. Complex geological conditions have aggravated the frequency of rock burst, and the intensity of disasters has also shown an upward trend. Therefore, how to achieve accurate monitoring and early warning of rock burst has become one of the core research issues in the field of mine safety.

[0003] Rock burst prediction involves multidisciplinary knowledge such as geology, rock mechanics, and data science. Its prediction accuracy and response timeliness are crucial to mine safety prevention and control. However, the existing monitoring and early warning system still has deficiencies in the identification and prediction of impact risk sources. There are problems such as "inaccurate location of disaster sources and low early warning efficiency", making it difficult to accurately predict impact risks. In addition, the generalization of existing rock burst prediction models is low, and the hazard level standards of different coal mines are different, which makes it difficult to directly apply the constructed models to different mining areas. Traditional prediction methods mostly rely on single modal data or expert experience, or are often limited to a specific physical indicator. They are not adaptable enough when dealing with complex geological conditions, which seriously restricts the actual prevention and control effect of rock burst prediction models. Summary of the invention

[0004] The purpose of the present invention is to provide a method for constructing a large model for rock burst prediction based on multimodal data. By fusing multimodal data of rock burst, the level and probability of large-energy events that may occur in the future are predicted in the time dimension. Combined with the information entropy dynamic weight calculation method based on time window design, the weights of multimodal data are comprehensively evaluated to construct a basic large model for rock burst prediction, improve the applicability and prediction accuracy of the model under different mining conditions, and achieve accurate prediction of rock burst risks.

[0005] In order to achieve the above object, the present invention provides a method for constructing a large rock burst prediction model based on multimodal data, including a multimodal data acquisition and preprocessing module, a rock burst prediction module and a hazard level determination module, and the specific steps are as follows:

[0006] S1. In the multimodal data acquisition and preprocessing module, first, a multimodal data set is constructed by collecting data of different modes; then, the multimodal data set is preprocessed to construct a precursor pattern sequence for model training; finally, according to the characteristics of different mining areas, the precursor pattern sequence is converted into a corresponding grade form and the corresponding rock burst hazard grade label is divided;

[0007] S2. In the rock burst prediction module, Transformer is used as the core framework to process the precursor pattern sequence after level division. The rock burst prediction module mainly includes input embedding and position encoding layer, Transformer encoder and fully connected layer. Each module works together to finally predict the probability of rock burst level.

[0008] S3. In the hazard level determination module, the comprehensive index method is used to independently evaluate the hazard level of mining information data and geological structure data, and the overall hazard level of rock burst is comprehensively evaluated in combination with the hazard probability results output by the rock burst prediction module; specifically, for the three parts of data, namely mining information, geological structure and prediction results, their contribution ratios in the comprehensive index are allocated through the weight division method, thereby achieving a comprehensive prediction of the hazard level.

[0009] The multimodal data set in step S1 of the present invention includes a fusion of dynamic data consisting of sensor system data and mining information data and static data consisting of geological structure data;

[0010] The sensor system data is collected in real time through high-precision sensors reasonably arranged in the mine, mainly including microseismic monitoring waveform data, geoacoustic waveform data, rock stress waveform data and electromagnetic signal waveform data. The microseismic monitoring waveform data captures the vibration signal caused by the stress change of the rock mass through the microseismic sensor array arranged in the mine. The geoacoustic waveform data captures the tiny sound fluctuations of the rock mass through the geoacoustic sensors arranged in the mine, reflecting the dynamic changes of the stress of the rock formation. The rock stress waveform data collects the dynamic changes of stress in the rock formation in the mine through the stress sensor. The electromagnetic signal waveform data monitors the changes of electromagnetic signals in the process of rock formation stress in real time through the electromagnetic sensors installed in the mine. Through the joint application of multiple sensor systems, high-frequency collection of multi-dimensional data is realized, providing multi-angle information for rock burst prediction.

[0011] Mining information data is used to describe the mining status of the mine. As the mining process continues to change, it is crucial for the risk assessment and prediction of rock burst. Mining information data includes the shortest distance W between the mining location and the irregular working surface such as the “knife handle” shape or the opening of multiple working surfaces and the misalignment of the stop line. e 1, the shortest distance W between the mining position and the “square” area of ​​the goaf of the working face e 2 , the shortest distance W between the mining location and the intersection area of ​​the "triangle" tunnel e 3 , mining speed W e 4 And the closest distance between the mining location and the structural features around the mine, such as the closest distance W between the mining location and the fault (with a drop greater than 3m) e 5 , the closest distance to the fold (inclination greater than 15°) W e 6 and the shortest distance to the goaf W e 7 , the change rate of coal seam thickness (relative to the average coal thickness) at the mining location W e 8 ;

[0012] Geological structure data is used to describe the geological factors of mining mines and evaluate the overall rock burst hazard level of coal mines before mining; geological structure data includes geological data and mining data; among them, geological data includes the historical number of rock burst occurrences W1 1 , Mining depth W1 2 , the distance between the hard and thick rock layer in the overlying fracture zone and the coal seam W1 3 , roof rock thickness characteristic parameter W1 4 , the concentration degree of structural stress in the mining area W1 5 , uniaxial compressive strength of coal W1 6 The elastic energy index of coal W1 7 ; Mining data includes the degree of decompression of the protective layer W2 1 , the horizontal distance W2 between the working face and the coal pillar left over from the upper protective layer mining 2 , Relationship with adjacent goaf W2 3 , working surface strength W2 4 , Section coal pillar width W2 5 , thickness of bottom coal W2 6 , the distance from the goaf when excavating into the goaf W2 7 , the distance from the goaf when advancing to the goaf W2 8 , distance from the fault W2 9 , distance from the fold W2 10 and the distance from the coal seam phase change zone W2 11 ;

[0013] The specific method of step S1 of the present invention is as follows:

[0014] S1.1: First, the raw data of the sensor system data is preprocessed. For the microseismic monitoring waveform data and ground sound waveform data, the low-frequency or high-frequency background noise is removed by using the bandpass filtering method;

[0015] For rock stress waveform data, outlier detection methods are used to remove data deviations caused by sensor errors or environmental interference;

[0016] For electromagnetic signal data, wavelet transform is used to perform denoising and extract effective electromagnetic signal components;

[0017] S1.2: Secondly, the denoised raw sensor data is converted into a format. Specifically, the microseismic monitoring waveform data and the ground sound waveform data are converted into time-energy format data;

[0018] Convert rock mass stress waveform data into time-stress format data;

[0019] Convert electromagnetic signal waveform data into time-magnetic field data;

[0020] Through this step, high-quality multimodal data that meets the needs of model training and prediction can be generated, providing reliable input support for subsequent modeling and analysis. i Can be recorded as Then the j-th data of the i-th sensor can be expressed as:

[0021]

[0022] Where: d i represents the i-th sensor system data set;

[0023] T i j It is represented as the time corresponding to the jth data of the i-th sensor;

[0024] It is represented as the energy / stress / magnetic field corresponding to the jth data of the i-th sensor;

[0025] Using k fixed time windows to count sensor system data, the number of data is n, then the i-th sensor system time window sequence data set is It can be expressed as:

[0026]

[0027] Statistical analysis of time window series data sets The obtained sensor data set is recorded as U, and the k-th time window data record of the i-th sensor is It can be expressed as:

[0028]

[0029] Where: is the number of the kth time window of the i-th sensor;

[0030] is the maximum energy / stress / magnetic field in the kth time window;

[0031] is the average energy / stress / magnetic field in the kth time window;

[0032] f i k is the frequency of energy / stress / magnetic field in the kth time window;

[0033] According to the sensor data set U, the precursor pattern sequence w is constructed, and the precursor pattern sequence of the e-th sensor of the i-th sensor can be expressed as

[0034]

[0035] Where: g is the sampling step length;

[0036] p is the length of the precursor pattern sequence, and the precursor pattern sequence set W of the constructed i-th sensor i It is expressed as:

[0037]

[0038] Where: q is the total number of precursor pattern sequences;

[0039] S1.3: The sensor data in the precursory pattern sequence are standardized. By converting numerical data such as microseismic energy / magnetic field / stress and frequency into graded information, the grade is used instead of the specific numerical value as the model input, thereby effectively improving the adaptability and prediction performance of the model under different mining conditions.

[0040] The input embedding and position encoding layer of step S2 of the present invention is specifically:

[0041] Input embedding maps the input sequence (precursor pattern sequence) to a high-dimensional space, forming a fixed length suitable for model processing. For each input segment x i , the embedding vector e is obtained by linear transformation i :

[0042] e i =W e x i +b e (6)

[0043] Where: W eand b e are the weights and biases of the embedding layer;

[0044] Since the Transformer itself does not have the ability to process position information, the timing information is introduced through the position encoding layer, which uses sine and cosine functions to generate position information:

[0045]

[0046]

[0047] Where: pos represents the sequence position;

[0048] a is the dimension index;

[0049] d model is the embedding dimension;

[0050] The sequence obtained after the input embedding plus the position encoding layer can be expressed as:

[0051] z0=[e1+PE1,e2+PE2,...,e N +PE N ] (9)

[0052] The Transformer encoder is specifically:

[0053] The Transformer encoder is the core part of the entire network, which is used to extract the global features of the precursor pattern sequence. The module is composed of multiple encoders stacked inside. Each encoder includes a multi-head self-attention mechanism, a feedforward neural network, and a residual connection and normalization. The output of the encoder is a high-dimensional feature representation containing complex relationships between different time segments, as follows:

[0054] The multi-head self-attention mechanism is used to calculate the display weights between sequence segments, so as to dynamically capture the temporal dependency and cross-modal correlation of rock burst precursor patterns and effectively mine potential feature patterns. The multi-head attention mechanism includes the self-attention mechanism and the multi-head mechanism.

[0055] The self-attention mechanism generates a query vector, a key vector, and a value vector for each input sequence z0, and calculates the similarity weight through a dot product operation (MatMul):

[0056] Q=z0W Q ,K=z0W K ,V=z0W V (10)

[0057] Where: W Q , W K , W V is a learnable weight matrix;

[0058] The similarity weights are scaled by the dot product operation and then normalized by the softmax activation function:

[0059]

[0060] Where: d k is the dimension of the key vector;

[0061] In order to enhance the feature extraction capability of the model, multiple heads are used to calculate attention in parallel, and each head has an independent W Q , W K , W V :

[0062] MultiHead(Q,K,V)=Concat(head1,...,head h )W O (12)

[0063] Where:

[0064]

[0065] h is the number of attention heads, W O is the output linear transformation matrix;

[0066] The feedforward neural network performs a nonlinear transformation on the features output by the self-attention mechanism. After the multi-head attention mechanism, the feature vector of each position passes through two layers of fully connected networks and adds the ReLU activation function:

[0067] FFN(x)=W2(ReLU(W1x+b1))+b2 (14)

[0068] Where: W1 and b1 are the weights and biases of the first fully connected network layer;

[0069] W2 and b2 are the weights and biases of the second fully connected network layer;

[0070] Residual connections and normalization

[0071] To avoid the problem of gradient disappearance and gradient explosion, residual connections and layer normalization are added after each sublayer:

[0072] Output=LayerNorm(x+SubLayer(x)) (15)

[0073] Where: LayerNorm(·) is the layer normalization calculation;

[0074] SubLayer(x) is the output of a multi-head attention or feed-forward neural network;

[0075] The fully connected layer is specifically:

[0076] The high-dimensional feature representation Z generated by the Transformer encoder L As the input of the fully connected layer, after one or more layers of linear transformation, the probability distribution of the danger level of rock burst is finally output, and the Softmax activation function can be used to output the level with the highest probability as the prediction result:

[0077] p c =softmax(W d (x)+b d ) (16)

[0078] Where: W d and b d are the weights and biases of the first fully connected layer;

[0079] p c is the predicted probability that the rock burst category is predicted to be c.

[0080] The details of step S3 of the present invention are as follows:

[0081] First, the rock burst hazard level RL is normalized to the interval [0,1], and the mining information, geological structure, and prediction results are also divided into the interval [0,1] to facilitate the final assessment of the rock burst hazard.

[0082] The mining information data is divided into criteria using the comprehensive index method, and the specific criteria for classification of some factors can be modified accordingly according to actual conditions;

[0083] Among them, the mining information data impact factor W e The calculation is as follows:

[0084]

[0085] The geological structure data is divided into rules using the comprehensive index method, and the geological structure analysis of the impact of the comprehensive index method on the above geological data and mining data is as follows:

[0086] Geological data: Mining data:

[0087] Influencing factor of geological structure data W g Select the maximum value of geological data and mining comprehensive index, that is:

[0088] W g =max{W g1 , W g2} (19)

[0089] Deep learning data impact factor W m To output the danger level value, first determine the danger level (none, weak, medium, strong) through the maximum probability value output by the model, then divide the maximum probability value into 5 levels, and divide the 4 danger level ranges into 5 levels of maximum probability value ranges, which are used as the deep learning data influencing factor W. m , further improving the accuracy and practicality of the prediction.

[0090] Analysis shows that the maximum value of the probability of the model outputting the danger level (p c ) max The range is (0.25, 1]. In order to fully reflect the different levels of danger, the present invention constructs different deep learning data impact factors W according to the distribution characteristics of the model output probability. m ,Through this classification method, each hazard level can not only reflect the ,probabilistic output results of the model, but can also effectively ,improve the classification accuracy of the hazard degree, thereby achieving a more ,reliable risk assessment of rock burst.

[0091] Compared with the prior art, the present invention uses a multimodal data acquisition and preprocessing module and a multimodal data fusion technology to convert the raw data collected by the sensor system into a precursor pattern sequence. Compared with the method of directly using the raw data in the prior art, the present invention innovatively uses a hierarchical form to standardize the data, which can significantly improve the adaptability and prediction accuracy of the model under different mining conditions. In the rock burst prediction module, the model architecture based on Transformer is used. Different from the mode of directly outputting fixed results in the traditional deep learning method, the present invention uses a probability distribution form to refine the prediction of the possibility of the occurrence of the rock burst danger level. In the danger level determination module, a dynamic weight calculation method based on time window information entropy is proposed to achieve multi-source information fusion of mining information, geological structure data and prediction results, and comprehensively evaluate the danger level of rock burst. The present invention provides a method for constructing a large model for rock burst prediction based on multimodal data. After training the basic large model on the historical data of other working faces, it can be migrated and applied to the new working face, providing a reference for the time series prediction and prevention of rock burst, improving the applicability and prediction accuracy of the model under different mining conditions, and achieving accurate prediction of rock burst risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0092] Figure 1 It is a schematic diagram of the overall structure of the present invention;

[0093] Figure 2 Schematic diagram of the self-attention mechanism of the present invention;

[0094] Figure 3 Schematic diagram of the multi-head attention mechanism of the present invention;

[0095] Figure 4 Schematic diagram of residual connection and normalization of the present invention;

[0096] Figure 5 Schematic diagram of a feedforward neural network of the present invention;

[0097] Figure 6 Schematic diagram of residual connection and layer normalization of the present invention. DETAILED DESCRIPTION

[0098] The present invention will be further described below in conjunction with the accompanying drawings.

[0099] like Figure 1 As shown, a method for constructing a large rock burst prediction model based on multimodal data includes a multimodal data acquisition and preprocessing module, a rock burst prediction module and a hazard level determination module. The specific steps are as follows:

[0100] S1. In the multimodal data acquisition and preprocessing module, first, a multimodal data set is constructed by collecting data of different modes; then, the multimodal data set is preprocessed to construct a precursor pattern sequence for model training; finally, according to the characteristics of different mining areas, the precursor pattern sequence is converted into a corresponding grade form and the corresponding rock burst hazard grade label is divided;

[0101] The multimodal data set includes dynamic data consisting of sensor system data and mining information data and static data consisting of geological structure data.

[0102] The sensor system data is collected in real time through high-precision sensors reasonably arranged in the mine, mainly including microseismic monitoring waveform data, geoacoustic waveform data, rock stress waveform data, and electromagnetic signal waveform data; the microseismic monitoring waveform data captures the vibration signal caused by the rock stress change through the microseismic sensor array arranged in the mine; the geoacoustic waveform data captures the tiny sound fluctuations of the rock mass through the geoacoustic sensors arranged in the mine, reflecting the dynamic changes of the rock formation stress; the rock stress waveform data collects the dynamic changes of the stress in the rock formation in the mine through the stress sensor; the electromagnetic signal waveform data monitors the electromagnetic signal changes in the rock formation stress process in real time through the electromagnetic sensors placed in the mine; through the joint application of multiple sensor systems, high-frequency collection of multi-dimensional data is realized, providing multi-angle information for rock burst prediction;

[0103] Mining information data is used to describe the mining status of the mine. As the mining process continues to change, it is crucial for the risk assessment and prediction of rock burst. Mining information data includes the shortest distance W between the mining location and the irregular working surface such as the “knife handle” shape or the opening of multiple working surfaces and the misalignment of the stop line. e1 , the shortest distance W between the mining position and the “square” area of ​​the goaf of the working face e 2 , the shortest distance W between the mining location and the intersection area of ​​the "triangle" tunnel e 3 , mining speed W e 4 And the closest distance between the mining location and the structural features around the mine, such as the closest distance W between the mining location and the fault (with a drop greater than 3m) e 5 , the closest distance to the fold (inclination greater than 15°) W e 6 and the shortest distance to the goaf W e 7 , the change rate of coal seam thickness (relative to the average coal thickness) at the mining location W e 8 Mining information data can provide the basis for the changes in the overall stress field of the mine and is the key part of the multi-modal data input for building the rock burst prediction model;

[0104] Geological structure data is used to describe the geological factors of mining mines and evaluate the overall rock burst hazard level of coal mines before mining; geological structure data includes geological data and mining data; among them, geological data includes the historical number of rock burst occurrences W1 1 , Mining depth W1 2 , the distance between the hard and thick rock layer in the overlying fracture zone and the coal seam W1 3 , roof rock thickness characteristic parameter W1 4 , the concentration degree of structural stress in the mining area W1 5 , uniaxial compressive strength of coal W1 6 The elastic energy index of coal W1 7 ; Mining data includes the degree of decompression of the protective layer W2 1 , the horizontal distance W2 between the working face and the coal pillar left over from the upper protective layer mining 2 , Relationship with adjacent goaf W2 3 , working surface strength W2 4 , Section coal pillar width W2 5 , thickness of bottom coal W2 6 , the distance from the goaf when excavating into the goaf W2 7 , the distance from the goaf when advancing to the goaf W2 8 , distance from the fault W2 9 , distance from the fold W2 10 and the distance from the coal seam phase change zone W2 11 ;

[0105] The specific method of step S1 is as follows:

[0106] S1.1: First, due to the large noise interference in the mine environment, in order to ensure that the multimodal data has high quality when input into the model, it is necessary to preprocess the raw data of the sensor system data. For the microseismic monitoring waveform data and ground sound waveform data, the bandpass filtering method is used to remove the low-frequency or high-frequency background noise;

[0107] For rock stress waveform data, outlier detection methods are used to remove data deviations caused by sensor errors or environmental interference;

[0108] For electromagnetic signal data, wavelet transform is used to perform denoising and extract effective electromagnetic signal components;

[0109] S1.2: Secondly, in order to make the denoised sensor system data consistent and suitable for the training and prediction process of the rock burst prediction model, it is necessary to convert the format of the denoised sensor raw data. Specifically, the microseismic monitoring waveform data and ground sound waveform data are converted into time-energy format data.

[0110] Convert rock mass stress waveform data into time-stress format data;

[0111] Convert electromagnetic signal waveform data into time-magnetic field data;

[0112] Through this step, high-quality multimodal data that meets the needs of model training and prediction can be generated, providing reliable input support for subsequent modeling and analysis. i Can be recorded as Then the j-th data of the i-th sensor can be expressed as:

[0113]

[0114] Where: d i represents the i-th sensor system data set;

[0115] T i j It is represented as the time corresponding to the jth data of the i-th sensor;

[0116] It is represented as the energy / stress / magnetic field corresponding to the jth data of the i-th sensor;

[0117] Using k fixed time windows to count sensor system data, the number of data is n, then the i-th sensor system time window sequence data set is It can be expressed as:

[0118]

[0119] Statistical analysis of time window series data sets The obtained sensor data set is recorded as U, and the k-th time window data record of the i-th sensor is It can be expressed as:

[0120]

[0121] Where: is the number of the kth time window of the i-th sensor;

[0122] is the maximum energy / stress / magnetic field in the kth time window;

[0123] is the average energy / stress / magnetic field in the kth time window;

[0124] f i k is the frequency of energy / stress / magnetic field in the kth time window;

[0125] According to the sensor data set U, the precursor pattern sequence w is constructed, and the precursor pattern sequence of the e-th sensor of the i-th sensor can be expressed as

[0126]

[0127] Where: g is the sampling step length;

[0128] p is the length of the precursor pattern sequence, and the precursor pattern sequence set W of the constructed i-th sensor i like Figure 2 As shown, it can be expressed as:

[0129]

[0130] Where: q is the total number of precursor pattern sequences;

[0131] S1.3: In view of the differences in the degree of danger of different mining areas under the same microseismic energy / magnetic field / stress or frequency, directly inputting the original data into the model can easily lead to the model being unable to adapt to the specific conditions of each mining area, showing the problem of insufficient generalization. To solve this problem, this method standardizes the data of each sensor in the precursor pattern sequence, converts numerical data such as microseismic energy / magnetic field / stress, frequency, etc. into graded information, and uses grades instead of specific values ​​as model input, thereby effectively improving the adaptability and prediction performance of the model under different mining conditions.

[0132] Taking the maximum energy and frequency of microseismic events as an example, different levels can be divided according to specific needs under different coal mine conditions. Table 1 shows examples of the classification of microseismic energy and frequency in two coal mines. For coal mines that have not yet been mined, the initial classification standards can be formulated by statistically analyzing the historical data of other working faces of the coal mine, and the classification standards can be appropriately adjusted after accumulating sufficient data.

[0133] Table 1 Classification information of different coal mines

[0134]

[0135] In addition, the definition of hazard level may vary among coal mines. For example, as shown in Table 2, different hazard level labels need to be set according to the actual situation of the coal mine and used as classification labels in subsequent model training to improve the prediction accuracy of the model in a variety of application scenarios.

[0136] Table 2 Classification of labels for different coal mine hazard levels

[0137]

[0138] S2. In the rock burst prediction module, Transformer is used as the core framework to process the precursor pattern sequence after level division. The rock burst prediction module mainly includes input embedding and position encoding layer, Transformer encoder and fully connected layer. Each module works together to finally predict the probability of rock burst level.

[0139] S2.1: Input embedding and position encoding layers are:

[0140] Input embedding maps the input sequence (precursor pattern sequence) to a high-dimensional space, forming a fixed length suitable for model processing. For each input segment x i , the embedding vector e is obtained by linear transformation i :

[0141] e i =W e x i +b e (6)

[0142] Where: W e and b e are the weights and biases of the embedding layer;

[0143] Since the Transformer itself does not have the ability to process position information, the timing information is introduced through the position encoding layer, which uses sine and cosine functions to generate position information:

[0144]

[0145] Where: pos represents the sequence position;

[0146] a is the dimension index;

[0147] d model is the embedding dimension;

[0148] The sequence obtained after the input embedding plus the position encoding layer can be expressed as:

[0149] z0=[e1+PE1,e2+PE2,...,e N +PE N ] (9)

[0150] S2.2: Transformer encoder is specifically:

[0151] The Transformer encoder is the core part of the entire network, which is used to extract the global features of the precursor pattern sequence. The module is composed of multiple encoders stacked together. Each encoder includes a multi-head self-attention mechanism, a feedforward neural network, and a residual connection and normalization. The output of the encoder is a high-dimensional feature representation containing complex relationships between different time segments.

[0152] S2.2.1: Multi-head self-attention mechanism

[0153] Multi-head self-attention mechanism Figure 3 , Figure 4 As shown, it is used to calculate the display weights between sequence segments, so as to dynamically capture the time dependency and cross-modal correlation of the rock burst precursor pattern and effectively mine the potential feature patterns. The multi-head attention mechanism includes the self-attention mechanism and the multi-head mechanism;

[0154] The self-attention mechanism generates a query vector, a key vector, and a value vector for each input sequence z0, and calculates the similarity weight through a dot product operation (MatMul):

[0155] Q=z0W Q ,K=z0W K ,V=z0W V (10)

[0156] Where: W Q , W K , W V is a learnable weight matrix;

[0157] The similarity weights are scaled by the dot product operation and then normalized by softmax:

[0158]

[0159] Where: dk is the dimension of the key vector, which is used to prevent gradient instability caused by excessive dot product values.

[0160] In order to enhance the feature extraction capability of the model, multiple heads are used to calculate attention in parallel, and each head has an independent W Q , W K , W V :

[0161] MultiHead(Q,K,V)=Concat(head1,...,head h )W O (12)

[0162] Where:

[0163]

[0164] h is the number of attention heads, W O is the output linear transformation matrix;

[0165] S2.2.2: Feedforward Neural Networks

[0166] Feedforward neural network Figure 5 As shown in the figure, the features output by the self-attention mechanism are transformed nonlinearly to further improve the expressiveness of the model. After the multi-head attention mechanism, the feature vector of each position passes through two layers of fully connected networks and adds the ReLU activation function:

[0167] FFN(x)=W2(ReLU(W1x+b1))+b2 (14)

[0168] Where: W1 and b1 are the weights and biases of the first fully connected network layer;

[0169] W2 and b2 are the weights and biases of the second fully connected network layer;

[0170] S2.2.3: Residual Connections and Normalization

[0171] To avoid the problem of gradient vanishing and gradient exploding, residual connections and layer normalization are added after each sublayer, such as Figure 6 As shown:

[0172] Output=LayerNorm(x+SubLayer(x)) (15)

[0173] Where: LayerNorm(·) is the layer normalization calculation;

[0174] SubLayer(x) is the output of a multi-head attention or feed-forward neural network;

[0175] S2.3: The fully connected layer is as follows:

[0176] The high-dimensional feature representation Z generated by the Transformer encoder L As the input of the fully connected layer, after one or more layers of linear transformation, the probability distribution of the danger level of rock burst is finally output, and the Softmax activation function can be used to output the level with the highest probability as the prediction result:

[0177] p c =softmax(W d (x)+b d ) (16)

[0178] Where: W d and b d are the weights and biases of the first fully connected layer;

[0179] p c is the predicted probability that the rock burst category is predicted to be c.

[0180] S3. In the hazard level determination module, the comprehensive index method is used to independently evaluate the hazard level of mining information data and geological structure data, and combined with the hazard probability results output by the rock burst prediction module, the information entropy weight calculation method based on the time window design is used to comprehensively evaluate the overall hazard level of rock burst. Specifically, for the three parts of data (mining information, geological structure, prediction results), the contribution ratio of each part in the hazard level is allocated through the weight division method, so as to achieve a comprehensive prediction of the hazard level.

[0181] First, the rock burst hazard level RL is normalized to the [0,1] interval, as shown in Table 3. The mining information, geological structure, and prediction results are also divided into the [0,1] interval to facilitate the final assessment of the rock burst hazard;

[0182] Table 3 Rock burst hazard levels

[0183] Hazard Level Danger level RL corresponding range none 0≤RL<0.25 weak 0.25≤RL<0.5 middle 0.5≤RL<0.75 powerful 0.75≤RL<1

[0184] S3.1: The classification criteria for mining information data using the comprehensive index method are shown in Table 4, and the specific criteria for classification of some factors can be modified accordingly according to actual conditions;

[0185] Table 4 Mining information classification criteria

[0186]

[0187]

[0188] Mining information data impact factor W e The calculation is as follows:

[0189]

[0190] S3.2: The rules for dividing geological structure data using the comprehensive index method are shown in Table 5;

[0191] Table 5 Geological structure division criteria (a) Geological structure division criteria under the influence of geological data

[0192]

[0193]

[0194] (b) Geological structure classification criteria under the influence of mining data

[0195]

[0196]

[0197] The geological structure analysis of the impact of the above geological data and mining data using the comprehensive index method:

[0198] Geological data: Mining data:

[0199] Influencing factor of geological structure data W g Select the maximum value of geological data and mining comprehensive index, that is:

[0200] W g =max{W g1 , W g2} (19)

[0201] S3.3: Traditional deep learning models usually use the maximum value of the model output category probability as the prediction result. When the maximum probability value is high, the model's credibility in its output result is relatively high; however, when the probability values ​​of multiple categories are close, the model's judgment on category attribution may be uncertain, thereby reducing the reliability of the prediction result. To address this problem, the present invention proposes a method that comprehensively considers the maximum probability value of the model output and the impact rock pressure hazard level. First, the corresponding hazard level (none, weak, medium, strong) is determined based on the maximum probability value output by the model, and the maximum probability value is further divided into five sub-level ranges; then, the range of the determined hazard level (as shown in Table 3) is divided into 5 refined hazard degree values ​​corresponding to the five sub-level ranges of the maximum probability value; finally, based on the range of the maximum probability value, the hazard degree value is determined as the deep learning data influencing factor W m , in order to improve the accuracy and practicality of the prediction.

[0202] Therefore, the present invention proposes a method that comprehensively considers the maximum probability of model output and the rock burst hazard level. First, the hazard level (none, weak, medium, strong) is determined by the maximum probability value output by the model. Then, the maximum probability value is divided into 5 levels. At the same time, the 4 hazard level ranges are divided into 5 levels corresponding to the maximum probability value ranges, so as to output the hazard level value as the deep learning data influencing factor W. m , further improving the accuracy and practicality of the prediction.

[0203] Analysis shows that the maximum value of the probability of the model outputting the danger level (p c ) max The range is (0.25, 1]. In order to fully reflect the different levels of danger, the present invention constructs different deep learning data impact factors W according to the distribution characteristics of the model output probability. m , the specific classification standards are detailed in Table 6. Through this classification method, each hazard level can not only reflect the probability output results of the model, but also effectively improve the classification accuracy of the hazard level, thereby achieving a more reliable risk assessment of rock burst.

[0204] Table 6 Output criteria for deep learning data impact factors

[0205]

[0206] In order to further comprehensively evaluate the danger level of rock burst, the present invention proposes a method for calculating information entropy weight based on time window design. e , geological structure data influence factor W g , Deep learning data impact factor W m , using weight division to determine the weight α of each part e , α g , α m , satisfying α e +α g +α m =1.

[0207] First, considering that the weights should change dynamically with the mining process, the same time window as the precursor pattern sequence is used to count the three types of data, and the probability distribution of each type of data is calculated:

[0208]

[0209] Among them, W k (l) represents the impact factor of mining information data, geological structure data, and deep learning data of the lth sample, and b represents the total number of samples.

[0210] Then, calculate the information entropy of each type of data:

[0211]

[0212] Where E k represents the information entropy of the kth type of data, and ln(b) is the normalized coefficient of entropy.

[0213] Therefore, the weight calculation formula for each part is:

[0214]

[0215] Among them, α e , α g , α m W is the impact factor of the information data e , geological structure data influence factor W g , Deep learning data impact factor W m The weight of .

[0216] Finally, the risk level RL of the predicted time period is calculated as:

[0217] RL=α e W e +α g W g +α m W m (twenty three)

[0218] Table 3 is used to determine the danger level (none, weak, medium, strong) in the forecast time period, and to predict the rock burst danger.

Claims

1. A method for constructing a large rock burst prediction model based on multimodal data, characterized in that: It includes a multimodal data acquisition and preprocessing module, a rock burst prediction module and a hazard level determination module. The specific steps are as follows: S1. In the multimodal data acquisition and preprocessing module, first, a multimodal data set is constructed by collecting data of different modes; then, the multimodal data set is preprocessed to construct a precursor pattern sequence for model training; finally, according to the characteristics of different mining areas, the precursor pattern sequence is converted into a corresponding grade form and the corresponding rock burst hazard grade label is divided; S2. In the rock burst prediction module, Transformer is used as the core framework to process the precursor pattern sequence after level division. The rock burst prediction module mainly includes input embedding and position encoding layer, Transformer encoder and fully connected layer. Each module works together to finally predict the probability of rock burst level. S3. In the hazard level determination module, the comprehensive index method is used to independently evaluate the hazard level of mining information data and geological structure data, and the overall hazard level of rock burst is comprehensively evaluated in combination with the hazard probability results output by the rock burst prediction module; specifically, for the three parts of data, namely mining information, geological structure and prediction results, their contribution ratios in the comprehensive index are allocated through the weight division method, thereby achieving a comprehensive prediction of the hazard level.

2. The method for constructing a large rock burst prediction model based on multimodal data according to claim 1, characterized in that: The multimodal data set in step S1 is formed by fusion of dynamic data consisting of sensor system data and mining information data and static data consisting of geological structure data; The sensor system data is collected in real time through high-precision sensors arranged reasonably in the mine, including microseismic monitoring waveform data, geoacoustic waveform data, rock stress waveform data and electromagnetic signal waveform data; the microseismic monitoring waveform data is collected through the microseismic sensor array arranged in the mine to capture the vibration signal caused by the rock stress change; the geoacoustic waveform data is collected through the geoacoustic sensors arranged in the mine to capture the tiny sound fluctuations of the rock mass and reflect the dynamic changes of the rock layer stress; Rock stress waveform data collects dynamic changes in stress in the rock formations of the mine through stress sensors; electromagnetic signal waveform data monitors the changes in electromagnetic signals in the rock formation during stress in real time through electromagnetic sensors installed in the mine; through the joint application of multiple sensing systems, high-frequency collection of multi-dimensional data is achieved, providing multi-angle information for rock burst prediction; Mining information data is used to describe the mining status of the mine. As the mining process continues to change, it is crucial for the risk assessment and prediction of rock burst. Mining information data includes the closest distance between the mining location and the opening of the cutter eye and the non-aligned area of ​​the stop line of the "knife handle" irregular working face or multiple working faces. The shortest distance W between the mining position and the "square" area of ​​the goaf of the working face e 2 , the shortest distance between the mining location and the intersection area of ​​the "triangle" tunnel Mining speed and the closest distance between the mining location and the structural features around the mine, including the closest distance between the mining location and the fault Closest distance to the fold The closest distance to the goaf Coal seam thickness change rate at mining location Geological structure data is used to describe the geological factors of mining mines and evaluate the overall rock burst hazard level of coal mines before mining; geological structure data includes geological data and mining data; among them, geological data includes the historical number of rock burst occurrences W1 1 , Mining depth W1 2 , the distance between the hard and thick rock layer in the overlying fracture zone and the coal seam W1 3 , roof rock thickness characteristic parameter W1 4 , the concentration degree of structural stress in the mining area W1 5 , uniaxial compressive strength of coal W1 6 The elastic energy index of coal W1 7 ; Mining data includes the degree of decompression of the protective layer W2 1 , the horizontal distance between the working face and the coal pillars left over from the upper protective layer mining Relationship with adjacent goaf Working surface strength Section coal pillar width Thickness of bottom coal The distance from the goaf when excavating into it The distance from the goaf when advancing towards it Distance from fault Distance from folds and the distance to the coal seam phase change zone 3. The method for constructing a large rock burst prediction model based on multimodal data according to claim 2, characterized in that: The specific method of step S1 is as follows: S1.1: First, the raw data of the sensor system data is preprocessed. For the microseismic monitoring waveform data and ground sound waveform data, the low-frequency or high-frequency background noise is removed by using the bandpass filtering method; For rock stress waveform data, outlier detection methods are used to remove data deviations caused by sensor errors or environmental interference; For electromagnetic signal data, wavelet transform is used to perform denoising and extract effective electromagnetic signal components; S1.2: Secondly, the denoised raw sensor data is converted into a format. Specifically, the microseismic monitoring waveform data and the ground sound waveform data are converted into time-energy format data; Convert rock mass stress waveform data into time-stress format data; Convert electromagnetic signal waveform data into time-magnetic field data; Through this step, high-quality multimodal data that meets the needs of model training and prediction can be generated, providing reliable input support for subsequent modeling and analysis. i Recorded as Then the jth data of the i-th sensor is expressed as: Where: d i represents the i-th sensor system data set; T i j It is represented as the time corresponding to the jth data of the i-th sensor; It is represented as the energy / stress / magnetic field corresponding to the jth data of the i-th sensor; Using k fixed time windows to count sensor system data, the number of data is n, then the i-th sensor system time window sequence data set is It is expressed as: Statistical analysis of time window series data sets The obtained sensor data set is recorded as U, and the k-th time window data record of the i-th sensor is It is expressed as: Where: is the number of the kth time window of the i-th sensor; is the maximum energy / stress / magnetic field in the kth time window; is the average energy / stress / magnetic field in the kth time window; f i k is the frequency of energy / stress / magnetic field in the kth time window; According to the sensor data set U, the precursor pattern sequence w is constructed, and the precursor pattern sequence of the e-th sensor of the i-th sensor is expressed as Where: g is the sampling step length; p is the length of the precursor pattern sequence, and the precursor pattern sequence set W of the constructed i-th sensor i It is expressed as: Where: q is the total number of precursor pattern sequences; S1.3: The sensor data in the precursory pattern sequence were standardized. The microseismic energy / magnetic field / stress and frequency numerical data were converted into graded information, and the grades were used instead of specific values ​​as model input, which effectively improved the adaptability and prediction performance of the model under different mining conditions.

4. The method for constructing a large rock burst prediction model based on multimodal data according to claim 2, characterized in that: The input embedding and position encoding layers of step S2 are specifically: Input embedding maps the precursor pattern sequence to a high-dimensional space, forming a fixed length suitable for model processing. For each input segment x i , the embedding vector e is obtained by linear transformation i : e i =W e x i +b e (6) Where: W e and b e are the weights and biases of the embedding layer; Since the Transformer itself does not have the ability to process position information, the timing information is introduced through the position encoding layer, which uses sine and cosine functions to generate position information: Where: pos represents the sequence position; a is the dimension index; d model is the embedding dimension; The sequence representation obtained after the input embedding plus the position encoding layer is: z0=[e1+For1,e2+For2,…,and N +PE N ](9) The Transformer encoder is specifically: The Transformer encoder is the core part of the entire network, which is used to extract the global features of the precursor pattern sequence. The module is composed of multiple encoders stacked inside. Each encoder includes a multi-head self-attention mechanism, a feedforward neural network, and a residual connection and normalization. The output of the encoder is a high-dimensional feature representation containing complex relationships between different time segments, as follows: The multi-head self-attention mechanism is used to calculate the display weights between sequence segments, so as to dynamically capture the temporal dependency and cross-modal correlation of rock burst precursor patterns and effectively mine potential feature patterns. The multi-head attention mechanism includes the self-attention mechanism and the multi-head mechanism. The self-attention mechanism generates a query vector, a key vector, and a value vector for each input sequence z0, and calculates the similarity weight through a dot product operation: Q=z0W Q ,K=z0W K ,V=z0W V (10) Where: W Q , W K , W V is a learnable weight matrix; The similarity weights are scaled by the dot product operation and then normalized by the softmax activation function: Where: d k is the dimension of the key vector; In order to enhance the feature extraction capability of the model, multiple heads are used to calculate attention in parallel, and each head has an independent W Q , W K , W V : MultiHead(Q,K,V)=Concat(head1,...,head h )W O (12) Where: h is the number of attention heads, W O is the output linear transformation matrix; The feedforward neural network performs a nonlinear transformation on the features output by the self-attention mechanism. After the multi-head attention mechanism, the feature vector of each position passes through two layers of fully connected networks and adds the ReLU activation function: FFN(x)=W2(ReLU(W1x+b1))+b2(14) Where: W1 and b1 are the weights and biases of the first fully connected network layer; W2 and b2 are the weights and biases of the second fully connected network layer; Residual connections and normalization Add residual connections and layer normalization after each sublayer: Output=LayerNorm(x+SubLayer(x))(15) Where: LayerNorm(·) is the layer normalization calculation; SubLayer(x) is the output of a multi-head attention or feed-forward neural network; The fully connected layer is specifically: The high-dimensional feature representation Z generated by the Transformer encoder L As the input of the fully connected layer, after one or more layers of linear transformation, the probability distribution of the danger level of rock burst is finally output, and the Softmax activation function is used to output the level with the highest probability as the prediction result: p c =softmax(W d (x)+b d )(16) Where: W d and b d are the weights and biases of the first fully connected layer; p c is the predicted probability that the rock burst category is predicted to be c.

5. The method for constructing a large rock burst prediction model based on multimodal data according to claim 2, characterized in that: The details of step S3 are as follows: First, the rock burst hazard level RL is normalized to the interval [0,1], and the mining information, geological structure, and prediction results are also divided into the interval [0,1] to facilitate the final assessment of the rock burst hazard. The mining information data is divided into criteria using the comprehensive index method, and the specific criteria for classification of some factors can be modified accordingly according to actual conditions; Mining information data impact factor W e The calculation is as follows: The geological structure data is divided into rules using the comprehensive index method, and the geological structure analysis of the above geological data and mining data is as follows: Geological data: Mining data: Geological structure data influence factor W g Select the maximum value of geological data and mining comprehensive index, that is: W g =max{W g1 ,W g2 }(19) Deep learning data impact factor W m To output the danger level value, the danger level is first determined by the maximum probability value output by the model, and then the maximum probability value is divided into 5 levels. At the same time, the 4 danger level ranges are divided into 5 levels corresponding to the maximum probability value range, which is used as the deep learning data influencing factor W. m ; Analysis shows that the maximum value of the probability of the model outputting the danger level (p c ) max The range is (0.25, 1]. In order to fully reflect the different levels of danger, the present invention constructs different deep learning data impact factors W according to the distribution characteristics of the model output probability. m ,Through this classification method, each hazard level can not only reflect the ,probability output results of the model, but can also effectively improve the ,accuracy of the classification of the hazard degree, thereby achieving a more ,reliable risk assessment of rock burst; The mining information data influence factor W obtained above is e , Geological structure data influence factor W g , Deep learning data impact factor W m , using weight division to determine the weight α of each part e , α g , α m , satisfying α e +α g +α m =1; First, considering that the weights should change dynamically with the mining process, the same time window as the precursor pattern sequence is used to count the three types of data, and the probability distribution of each type of data is calculated: Where: W k (l) represents the impact factor of mining information data, geological structure data, and deep learning data of the lth sample; b represents the total number of samples; Then, calculate the information entropy of each type of data: Where: E k Represents the information entropy of the k-th category of data; ln(b) is the normalized coefficient of entropy; Therefore, the weight calculation formula for each part is: Where: α e , α g , α m W is the impact factor of the information data e , geological structure data influence factor W g , Deep learning data impact factor W m The weight of Finally, the risk level RL of the predicted time period is calculated as: RL=a e W e +a g W g +a m W m (23).

Images