Mine rock burst early warning, prevention and control method, device, equipment and medium

Abstract

The invention relates to the technical field of coal mine safety, in particular to a mine rock burst early warning, prevention and control method, device and equipment and a medium, and the method comprises the steps: obtaining multi-source monitoring data based on a high-precision positioning method; performing feature extraction on the multi-source monitoring data to obtain deep feature data; constructing a network prediction model, and training the network prediction model by using the deep feature data to obtain a trained network prediction model; and constructing a three-dimensional cellular automaton model of the coal and rock mass fracture process, and simulating a prevention and control strategy by using the three-dimensional cellular automaton model of the coal and rock mass fracture process to obtain a final prevention and control strategy. By integrating multi-source data such as coal and rock mass microstructure parameters, geological conditions and mining processes and combining machine learning and dynamic weight distribution technologies, accurate prediction and active prevention and control of rock burst risks are realized, and the problems of low early warning precision, poor real-time performance and weak adaptability in the prior art are solved.

Description

Technical Field

[0001] The present application relates to the field of coal mine safety technology, and in particular to a mine rock burst warning and prevention method, device, equipment and medium. Background Art

[0002] Rock burst is one of the common dynamic disasters in deep mining, which seriously threatens the safety of miners' lives and mine production efficiency. At present, rock burst monitoring and early warning technologies at home and abroad mainly rely on the static analysis of a single physical field (such as microseismicity, stress or ground sound), which is difficult to fully reflect the multi-field coupling characteristics of energy, stress, vibration field, etc. in the process of coal and rock mass fracture, resulting in insufficient early warning accuracy. Traditional early warning models are based on fixed thresholds or empirical formulas and cannot dynamically adapt to the dynamic changes of different geological conditions and mining processes, especially in small sample impact events, with a high rate of missed reports. In addition, there are delays in data processing and early warning issuance, making it difficult to guide the implementation of on-site prevention and control measures in a timely manner. Some mines rely on imported monitoring equipment, and the cost of a single set of equipment and maintenance is high.

[0003] In summary, how to design an efficient and accurate rock burst risk prediction method is an urgent problem that needs to be solved. Summary of the Invention

[0004] The present application aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the first purpose of this application is to propose a mine rock burst warning and prevention method to solve the problems of low warning accuracy, poor real-time performance, and weak adaptability of existing technical means.

[0006] The second object of this application is to provide a device.

[0007] The third objective of this application is to provide an electronic device.

[0008] The fourth object of this application is to provide a computer-readable storage medium.

[0009] To achieve the above objectives, the first embodiment of the present application proposes a mine rock burst warning and prevention method, comprising:

[0010] Based on high-precision positioning methods, multi-source monitoring data is obtained;

[0011] Extracting features from the multi-source monitoring data to obtain deep-level feature data;

[0012] Constructing a network prediction model based on a Bayesian optimization algorithm, training the network prediction model using the deep feature data, and dynamically adjusting the network prediction model based on an estimation method to obtain a trained network prediction model;

[0013] Using the trained network prediction model to predict the mine to be predicted, and obtaining a first prevention and control strategy;

[0014] A three-dimensional cellular automaton model of the coal-rock mass fracture process is constructed, and the first prevention and control strategy is simulated using the three-dimensional cellular automaton model of the coal-rock mass fracture process to obtain a final prevention and control strategy.

[0015] Preferably, extracting features from the multi-source monitoring data to obtain deep-level feature data includes:

[0016] Performing denoising on the multi-source monitoring data to obtain denoised data;

[0017] Performing time synchronization processing on the denoised data using a dynamic time warping algorithm to obtain synchronized data;

[0018] Normalizing the synchronized data and constructing spatiotemporal features;

[0019] The spatiotemporal features are extracted using a deep neural network to obtain deep feature data.

[0020] Preferably, the performing time synchronization processing on the denoised data using a dynamic time warping algorithm to obtain synchronized data includes:

[0021] The dynamic time warping algorithm is used to construct the optimal time mapping path of each data sequence in the denoised data, the cumulative distance matrix is solved by the dynamic programming method, the matching path with the minimum cumulative distance is determined, and the synchronized data is obtained based on the matching path with the minimum cumulative distance.

[0022] Preferably, the normalizing the synchronized data and constructing spatiotemporal features includes:

[0023] The synchronization data is normalized, and the calculation formula is:

[0024]

[0025] Among them, x max is the maximum value of historical data, x min Minimum value of historical data;

[0026] Map the normalized data and construct spatiotemporal features.

[0027] Preferably, the extracting the spatiotemporal features using a deep neural network to obtain deep-level feature data includes:

[0028] The spatiotemporal features are input into a deep neural network, and different convolution kernels in the deep neural network are used to perform sliding convolution on the data to extract local features in the data and obtain deep feature data.

[0029] Preferably, the network prediction model is constructed based on the Bayesian optimization algorithm, the network prediction model is trained using the deep feature data, and the network prediction model is dynamically adjusted based on an estimation method to obtain the trained network prediction model, which includes:

[0030] Based on the dynamic time warping method, the Bayesian optimization algorithm is used to set a comprehensive objective function and calculate the spatiotemporal consistency index. The calculation formula is:

[0031] L(θ)=α·F1+β·precision+γ·recall+δ

[0032] Among them, F1, precision, and recall are evaluation performances in different dimensions, α, β, and γ are weight coefficients, and δ is the spatiotemporal consistency index;

[0033] Based on the spatiotemporal consistency index, a Bayesian network prediction model including a causal chain is constructed, and the network prediction model is trained using the deep feature data. Based on the inference mechanism of the Bayesian network, the weights are dynamically adjusted to obtain a trained network prediction model.

[0034] Preferably, constructing a three-dimensional cellular automaton model of the coal-rock mass fracture process, simulating the first prevention and control strategy using the three-dimensional cellular automaton model of the coal-rock mass fracture process, and obtaining the final prevention and control strategy includes:

[0035] A three-dimensional cellular automaton model of the coal-rock fracture process is constructed, and the coal-rock mass is divided into cellular units. The prevention and control effect of the first prevention and control strategy is simulated through digital twin technology, and the first prevention and control strategy is automatically adjusted to obtain the final prevention and control strategy.

[0036] To achieve the above objectives, the second embodiment of the present application provides a mine rock burst warning and prevention device, comprising:

[0037] The data acquisition module acquires multi-source monitoring data based on high-precision positioning methods;

[0038] A feature extraction module extracts features from the multi-source monitoring data to obtain deep feature data;

[0039] A training module constructs a network prediction model based on a Bayesian optimization algorithm, trains the network prediction model using the deep feature data, and dynamically adjusts the network prediction model based on an estimation method to obtain a trained network prediction model;

[0040] A prediction module, using the trained network prediction model to predict the mine to be predicted, and obtain a first prevention and control strategy;

[0041] The simulation module constructs a three-dimensional cellular automaton model of the coal-rock mass fracture process, and uses the three-dimensional cellular automaton model of the coal-rock mass fracture process to simulate the first prevention and control strategy to obtain the final prevention and control strategy.

[0042] To achieve the above-mentioned purpose, a third embodiment of the present application provides an electronic device, comprising: a processor, and a memory communicatively connected to the processor;

[0043] The memory stores computer-executable instructions;

[0044] The processor executes the computer-executable instructions stored in the memory to implement any of the above methods.

[0045] To achieve the above-mentioned purpose, the fourth embodiment of the present application proposes a computer-readable storage medium, including computer-executable instructions stored in the computer-readable storage medium, and the computer-executable instructions are used to implement any of the methods described above when executed by a processor.

[0046] The present application provides a mine rock burst warning and prevention method, which comprehensively reflects the multi-field coupling characteristics in the coal-rock fracture process through multi-source data fusion and deep neural network feature mining, and significantly improves the warning accuracy. The dynamic time warping (DTW) algorithm is used to solve the time asynchrony problem of multi-source sensor data, ensure the accuracy of feature extraction, and reduce the delay in data processing and warning issuance. The Bayesian optimization algorithm is used to dynamically adjust the weights of each monitoring parameter to adapt to the dynamic changes of different geological conditions and mining processes, especially in small sample impact events with a low missed reporting rate. Through digital twin technology to simulate the prevention and control effect, a full-process closed loop of "monitoring-warning-prevention and control-verification-optimization" is realized, and the prevention and control plan is dynamically optimized to improve the prevention and control effect.

[0047] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through practice of the present application. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0049] Figure 1 This is a flow chart of a first specific embodiment of a mine rock burst warning and prevention method provided by the present invention;

[0050] Figure 2This is the flow chart of the DTW time alignment algorithm;

[0051] Figure 3 Flowchart for dynamic adjustment of Bayesian network structure and weights;

[0052] Figure 4 This is a structural block diagram of a mine rock burst warning and prevention device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] The core of the present invention is to provide a mine rock burst warning and prevention and control method, device, equipment and medium, which realizes precise time and space alignment through dynamic time regularization, dynamic weight adaptation based on Bayesian optimization, and digital twin verification combined with cellular automation to form a full-process intelligent prevention and control system.

[0054] In order to enable those skilled in the art to better understand the present invention, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0055] Please refer to Figure 1 , Figure 1 This is a flow chart of a first specific embodiment of a mine rock burst warning and prevention method provided by the present invention; the specific operating steps are as follows:

[0056] Step S101: Acquire multi-source monitoring data based on a high-precision positioning method;

[0057] Step S102: extracting features from multi-source monitoring data to obtain deep feature data;

[0058] De-noising is performed on multi-source monitoring data to obtain de-noised data;

[0059] The denoised data are processed in time synchronization using the dynamic time warping algorithm to obtain synchronized data;

[0060] In one embodiment, a dynamic time warping algorithm is used to construct the optimal time mapping path of each data sequence in the denoised data, the cumulative distance matrix is solved by a dynamic programming method, the matching path with the minimum cumulative distance is determined, and the synchronized data is obtained based on the matching path with the minimum cumulative distance.

[0061] Normalize the synchronized data and construct spatiotemporal features;

[0062] In one embodiment, the synchronization data is normalized, and the calculation formula is:

[0063]

[0064] Among them, x max is the maximum value of historical data, x min Minimum value of historical data;

[0065] Map the normalized data and construct spatiotemporal features.

[0066] Use deep neural networks to extract spatiotemporal features and obtain deep feature data;

[0067] In one embodiment, the spatiotemporal features are input into a deep neural network, and different convolution kernels in the deep neural network are used to perform sliding convolution on the data to extract local features in the data and obtain deep-level feature data.

[0068] Step S103: constructing a network prediction model based on the Bayesian optimization algorithm, training the network prediction model using deep feature data, and dynamically adjusting the network prediction model based on an estimation method to obtain a trained network prediction model;

[0069] Based on the dynamic time warping method, the Bayesian optimization algorithm is used to set a comprehensive objective function and calculate the spatiotemporal consistency index. The calculation formula is:

[0070] L(θ)=α·F1+β·precision+γ·recall+δ

[0071] Among them, F1, precision, and recall are evaluation performances in different dimensions, α, β, and γ are weight coefficients, and δ is the spatiotemporal consistency index;

[0072] Based on the spatiotemporal consistency index, a Bayesian network prediction model containing a causal chain is constructed, and the network prediction model is trained using deep feature data. Based on the inference mechanism of the Bayesian network, the weights are dynamically adjusted to obtain a trained network prediction model.

[0073] Step S104: using the trained network prediction model to predict the mine to be predicted, and obtaining a first prevention and control strategy;

[0074] Step S105: constructing a three-dimensional cellular automaton model of the coal-rock mass fracture process, and using the three-dimensional cellular automaton model of the coal-rock mass fracture process to simulate the first prevention and control strategy to obtain the final prevention and control strategy.

[0075] A three-dimensional cellular automaton model of the coal-rock fracture process is constructed, and the coal-rock mass is divided into cellular units. The prevention and control effect of the first prevention and control strategy is simulated through digital twin technology, and the first prevention and control strategy is automatically adjusted to obtain the final prevention and control strategy.

[0076] This embodiment provides a mine rock burst warning and prevention method, which comprehensively reflects the multi-field coupling characteristics in the coal-rock fracture process through multi-source data fusion and deep neural network feature mining, and significantly improves the warning accuracy. The dynamic time warping (DTW) algorithm is used to solve the time asynchrony problem of multi-source sensor data, ensure the accuracy of feature extraction, and reduce the delay in data processing and warning issuance. The Bayesian optimization algorithm is used to dynamically adjust the weights of each monitoring parameter to adapt to the dynamic changes of different geological conditions and mining processes, especially in small sample impact events with a low missed reporting rate. The prevention and control effect is simulated by digital twin technology, and the full process closed loop of "monitoring-warning-prevention and control-verification-optimization" is realized, and the prevention and control plan is dynamically optimized to improve the prevention and control effect.

[0077] Based on the above embodiment, this embodiment describes a mine rock burst warning and prevention method, as follows:

[0078] (1) Data collection;

[0079] It integrates more than 10 types of monitoring data such as electromagnetic radiation, microseismicity, stress, and drill cuttings, and combines it with underground high-precision positioning technology to achieve real-time collection of spatiotemporal dynamic data on the working face.

[0080] (2) Multi-source data fusion and feature extraction;

[0081] Data denoising: The raw multi-source monitoring data (electromagnetic radiation, microseismic, and stress data) contains significant noise. Filtering algorithms (such as Kalman filtering and wavelet filtering) are first used to perform preliminary denoising. For electromagnetic radiation data, which is susceptible to underground electromagnetic interference, wavelet filtering is used to remove high-frequency noise while preserving the signal's key features. A spatiotemporal alignment algorithm is used to eliminate data noise, and deep neural networks are used to automatically extract precursory features of coal and rock fractures (such as energy fluctuations and stress gradient anomalies).

[0082] Time synchronization and time offset elimination: The dynamic time warping (DTW) algorithm is used to address the time inconsistency problem of data collected by different sensors. Taking microseismic monitoring data and stress monitoring data as an example, there may be a recording deviation between the time of microseismic events and the time of stress changes. The DTW algorithm controls the time synchronization error within ±5ms by constructing the optimal time mapping path between the microseismic energy release sequence and the stress change sequence. During the calculation process, a distance metric function is defined, such as the Euclidean distance matrix, to represent the difference between corresponding points in two time series;

[0083] D(i,j)=|x 微震 (i)-x 应力 (j)|

[0084] The cumulative distance matrix is solved by dynamic programming to determine the matching path with the minimum cumulative distance:

[0085] G(i,j)=D(i,j)+min[G(i-1,j),G(i,j-1),G(i-1,j-1)]

[0086] And based on the minimum cumulative distance path:

[0087] γ=(γ1,γ2,…,γ K )

[0088] Achieve nonlinear alignment of time series to ensure accurate extraction of the spatiotemporal coupling characteristics of microseismic events and stress mutations.

[0089] In one embodiment, Figure 2 As shown, input two time series that need to be aligned;

[0090] Create a matrix that stores the distance between each pair of points in the two sequences.

[0091] Compute the distance between each pair of corresponding points in the two sequences using Euclidean distance or another appropriate distance metric.

[0092] By using dynamic programming methods, the cumulative distance matrix is filled and the minimum cumulative distance between each pair of points is found.

[0093] Find a path in the cumulative distance matrix with the smallest cumulative distance. This path is the optimal alignment path.

[0094] According to the optimal alignment path, the aligned time series is output.

[0095] Data normalization: Normalize the denoised and time-aligned data. Use the normalization formula:

[0096]

[0097] Among them, x max is the maximum value of historical data, x min Minimum value of historical data;

[0098] Map the data to the [0,1] interval. For example, for stress monitoring data, find the maximum value x in its historical data. max and the minimum value x min , the real-time monitored stress data x is normalized to make monitoring data of different types and magnitudes comparable, which is convenient for subsequent feature extraction and model training.

[0099] Constructing spatiotemporal features: Data after data processing contains rich spatiotemporal information. Based on the data after spatiotemporal alignment, construct spatiotemporal features. Taking the spatiotemporal series data of the working face advancement position and velocity and the microseismic source location parameters as an example, combined with the working face advancement velocity and the position change of the microseismic source in space, calculate the rate of change of microseismic events at different spatial locations over time. For example, the source movement velocity can be solved using the following formula:

[0100]

[0101] At the same time, the spatiotemporal entropy of the stress gradient is calculated to measure the dynamic evolution of the stress concentration area, forming a three-dimensional spatiotemporal characteristic vector including the time offset rate and spatial overlap. These spatiotemporal characteristics can more comprehensively reflect the dynamic changes of the coal rock mass before the impact rock pressure occurs.

[0102] Deep neural network feature mining: Deep neural networks (such as convolutional neural networks (CNNs)) are used to automatically extract features that indicate coal and rock fracture precursors. The spatiotemporally aligned and normalized data is fed into the CNN. The convolutional layers in the CNN perform sliding convolutions on the data using different convolution kernels to extract local features.

[0103] In one embodiment, for the spatiotemporal distribution characteristics of electromagnetic radiation data, the convolutional layer can capture the spatial variation trend of electromagnetic radiation intensity, the temporal aggregation characteristics of abnormal electromagnetic radiation events, etc.

[0104] The pooling layer reduces the dimensionality of the features extracted by the convolutional layer, retaining key features and reducing the amount of data. Through multiple layers of convolution and pooling, deep-level features such as energy fluctuations and stress gradient anomalies that can indicate precursors to coal and rock fractures are ultimately extracted, providing a more accurate basis for subsequent rock burst risk assessment. However, the comparative files may not have adopted such a comprehensive approach that incorporates spatiotemporal information during data processing and feature extraction, failing to fully tap into the potential information within the data.

[0105] (3) Dynamic weight adaptive early warning model;

[0106] The dynamic time warping (DTW) multi-sensor data synchronization method is adopted, and the Bayesian optimization algorithm is used to set a comprehensive objective function and calculate the spatiotemporal consistency index:

[0107] L(θ)=α·F1+β·precision+γ·recall+δ

[0108] Among them, F1, precision, and recall are evaluation performances in different dimensions, α, β, and γ are weight coefficients, and δ is the spatiotemporal consistency index, which is defined as the spatial overlap ratio between the microseismic source and the stress concentration area and can be calculated using the following formula:

[0109]

[0110] Here, F1, precision, and recall evaluate model performance from different dimensions. Dynamically adjusting the weight coefficients α, β, γ, and δ through Bayesian optimization can better balance the model's precision and recall in identifying small-sample events. During training, historical data from over 100 mines (including rock burst event samples) was combined to continuously optimize the objective function.

[0111] This method uses Bayesian network reinforcement learning to dynamically adjust the weights of each monitoring parameter and output the dynamic weight adjustment results. The specific process is as follows:

[0112] Network structure construction: Based on the physical meaning of multi-source monitoring data (such as electromagnetic radiation, microseismic, stress, etc.) and the correlation in actual monitoring, a Bayesian network structure containing the causal chain of "ground stress field → microseismic energy → electromagnetic radiation" is constructed. Stress changes may affect the release of microseismic energy, and electromagnetic radiation anomalies may also be related to coal and rock fractures. Based on these relationships, the connection method of network nodes and edges is determined. When a sudden increase in gas pressure is detected (node C), the gas parameter weight is dynamically increased (for example, from 0.15 to 0.3) through the conditional probability P (shock risk | A, B, C), which is smarter than the static weight allocation in the existing technology (the weight is fixed at 0.2).

[0113] Parameter Learning: After preprocessing and feature extraction, various data collected during the coal mining process are input into the Bayesian network for parameter learning. The conditional probability distribution of each node in the network is determined using maximum likelihood estimation or Bayesian estimation methods.

[0114] Dynamic weight adjustment: During real-time monitoring, when new monitoring data arrives, the probability distribution of each node is updated based on the inference mechanism of the Bayesian network. For example, if the amount of microseismic energy released suddenly increases, the network will recalculate the probabilities related to other monitoring data nodes based on the learned conditional probability relationship. Based on these updated probabilities, the weights of each monitoring parameter are dynamically adjusted. If the contribution of a certain monitoring parameter to the judgment of rock burst risk increases under the current state, its weight will be increased accordingly; otherwise, it will be reduced. This dynamic weight adjustment based on the Bayesian network can more flexibly and accurately reflect the changes in the importance of each monitoring parameter under different geological conditions and mining processes, while the comparative documents may not involve such a detailed process of dynamic adjustment based on actual monitoring data.

[0115] (4) Dynamic optimization of proactive prevention and control decisions;

[0116] Cellular automaton model construction: This paper establishes a three-dimensional cellular automaton (CA) model of the coal-rock mass fracture process, divides the coal-rock mass into 1m×1m×1m cell units, and defines the fracture rules:

[0117] σunit ≥ σc·(1-fracturing damage factor) → unit rupture

[0118] Among them, the fracturing damage factor is calculated in real time through the hydraulic fracturing pressure-time curve. The error between the simulation results and the on-site microseismic positioning is ≤3m, which significantly improves the accuracy compared with the traditional empirical formula (error ≥10m).

[0119] Closed-loop control process: Digital twin technology is used to simulate prevention and control effectiveness. When the digital twin simulation shows that the stress concentration zone elimination rate is less than 80%, it automatically triggers strategy adjustments (such as reducing the fracturing hole spacing from 8m to 5m). This forms a complete closed loop of "monitoring-early warning-prevention-verification-optimization", enabling intelligent generation and dynamic adjustment of rock burst prevention and control plans. However, the disclosed invention patent only provides early warning output and does not include subsequent dynamic optimization of prevention and control strategies.

[0120] The embodiment of the present invention provides a mine rock burst warning and prevention method, which comprehensively reflects the multi-field coupling characteristics in the coal-rock fracture process through multi-source data fusion and deep neural network feature mining, and significantly improves the warning accuracy. The dynamic time warping (DTW) algorithm is used to solve the time asynchrony problem of multi-source sensor data, ensure the accuracy of feature extraction, and reduce the delay in data processing and warning issuance. The Bayesian optimization algorithm is used to dynamically adjust the weights of each monitoring parameter to adapt to the dynamic changes of different geological conditions and mining processes, especially in small sample impact events with a low missed reporting rate. The prevention and control effect is simulated through digital twin technology, and the full process closed loop of "monitoring-warning-prevention and control-verification-optimization" is realized, and the prevention and control plan is dynamically optimized to improve the prevention and control effect.

[0121] Based on the above content, this embodiment uses specific data to briefly describe this method, as follows:

[0122] (1) Data collection

[0123] Collect and organize various types of data during the coal mining process, including coal and rock microstructure parameter monitoring data, geological conditions data, mining process data, rock burst historical data, etc., and establish a database.

[0124] The specific monitoring parameters are shown in Table 1:

[0125]

[0126]

[0127] Table 1 Monitoring parameters list

[0128] (2) Data processing

[0129] After preprocessing (denoising and normalization), the raw data is fed into the feature engineering module to generate time series features (such as mean, variance, and frequency domain energy). XGBoost is used to perform feature importance analysis and screen for sensitive parameter combinations.

[0130] (3) Model training

[0131] Based on historical data from more than 100 mines (including samples of rock burst events), an LSTM-CNN model was trained and the loss function was optimized to F1-score to improve the ability to recognize small sample events.

[0132] (4) Model Verification

[0133] On-site monitoring equipment collects data in real time and feeds this data back to the intelligent support system during implementation. The system then evaluates the effectiveness of the prevention and control plan. For example, the warning range is reduced from 50 meters to 30 meters, the response time of prevention and control measures is shortened by 40%, the incidence of rock burst accidents is reduced by 65%, or the stress concentration within the coal and rock mass is significantly alleviated, the propagation of microcracks is effectively controlled, and the frequency of rock burst is significantly reduced. If the results do not meet these requirements, the prevention and control strategy is adjusted in a timely manner and the prevention and control plan is regenerated to ensure the effectiveness and reliability of rock burst prevention and control.

[0134] Geological data: burial depth 600m, coal seam dip angle 15°, adjacent to F3 fault (drop 5m);

[0135] Mining data: daily advancement of the working face is 3.2m, and the initial support force of the support is 25MPa;

[0136] Monitoring data: The daily energy release suddenly increased to 8×10 6 J, the earthquake source is concentrated 20-40 m in front of the working face; the borehole stress gauge shows a vertical stress of 28 MPa (in-situ rock stress 18 MPa); the intensity is >500 μV / m for 3 consecutive hours, and the pulse frequency shows a downward trend;

[0137] Fusion analysis: Through the space-time alignment algorithm, it was found that the stress concentration area and the microseismic energy release area overlapped spatially. Combined with the electromagnetic radiation frequency attenuation characteristics, it was determined to be a strong impact danger zone, triggering the hydraulic fracturing prevention and control instructions.

[0138] In addition, based on the present invention, causal reasoning (such as Bayesian networks) can also be used to identify the causal relationship between monitoring data, establish a superposition model of the mining disturbance stress field and the geological structure stress field, or construct a chain early warning indicator system of "mining disturbance-energy accumulation-microfracture-impact manifestation".

[0139] This embodiment further provides a rock burst active prevention and control system, which is characterized by comprising:

[0140] Integrated well-ground monitoring device, distributed electromagnetic radiation sensor (working frequency band 0.3-3GHz), three-component microseismic detector (positioning accuracy ±1.5m), industrial Ethernet communication module (data transmission delay <50ms).

[0141] Please refer to Figure 4 , Figure 4 This is a structural block diagram of a mine rock burst warning and prevention device provided by an embodiment of the present invention; the specific device may include:

[0142] The data acquisition module 100 acquires multi-source monitoring data based on a high-precision positioning method;

[0143] The feature extraction module 200 extracts features from the multi-source monitoring data to obtain deep feature data;

[0144] A training module 300 constructs a network prediction model based on a Bayesian optimization algorithm, trains the network prediction model using the deep feature data, and dynamically adjusts the network prediction model based on an estimation method to obtain a trained network prediction model;

[0145] The prediction module 400 uses the trained network prediction model to predict the mine to be predicted and obtain a first prevention and control strategy;

[0146] The simulation module 500 constructs a three-dimensional cellular automaton model of the coal-rock mass fracture process, and uses the three-dimensional cellular automaton model of the coal-rock mass fracture process to simulate the first prevention and control strategy to obtain a final prevention and control strategy.

[0147] A mine rock burst warning and control device of this embodiment is used to implement the aforementioned mine rock burst warning and control method. Therefore, the specific implementation method of a mine rock burst warning and control device can be seen in the embodiment part of a mine rock burst warning and control method in the previous text. For example, the data acquisition module 100, the feature extraction module 200, the training module 300, the prediction module 400, and the simulation module 500 are respectively used to implement steps S101, S102, S103, S104, and S105 in the aforementioned mine rock burst warning and control method. Therefore, its specific implementation method can refer to the description of the corresponding embodiments of each part, and will not be repeated here.

[0148] In order to implement the above embodiments, the present application also proposes an electronic device, comprising: a processor, and a memory communicatively connected to the processor; the memory stores computer-executable instructions; the processor executes the computer-executable instructions stored in the memory to implement the method provided by the above embodiments.

[0149] In order to implement the above embodiments, the present application also proposes a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the methods provided by the above embodiments.

[0150] In order to implement the above embodiments, the present application also proposes a computer program product, including a computer program, which implements the methods provided by the above embodiments when executed by a processor.

[0151] The collection, storage, use, processing, transmission, provision and disclosure of user personal information involved in this application are in compliance with relevant laws and regulations and do not violate public order and good morals.

[0152] It is important to note that personal information collected from users should be used for legitimate and reasonable purposes and should not be shared or sold beyond these legitimate uses. Furthermore, such collection / sharing should be conducted only after receiving the user's informed consent, including but not limited to notifying the user to read the user agreement / user notice and sign an agreement / authorization that includes the relevant user information before using the feature. Furthermore, any necessary steps must be taken to safeguard and secure access to such personal information and ensure that others with access to personal information comply with its privacy policy and procedures.

[0153] This application contemplates providing implementations that allow users to selectively block the use or access of personal information data. Specifically, this disclosure contemplates providing hardware and / or software to prevent or block access to such personal information data. Risks can be minimized by limiting data collection and deleting data once it is no longer needed. Furthermore, where applicable, such personal information can be de-identified to protect user privacy.

[0154] In the descriptions of the foregoing embodiments, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are mutually inconsistent.

[0155] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of such features. Throughout the description of this application, "plurality" means at least two, for example, two, three, etc., unless otherwise specifically defined.

[0156] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0157] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0158] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0159] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0160] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0161] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A mine rock burst warning and prevention method, characterized in that: include: Based on high-precision positioning methods, multi-source monitoring data is obtained; Extracting features from the multi-source monitoring data to obtain deep-level feature data; Constructing a network prediction model based on a Bayesian optimization algorithm, training the network prediction model using the deep feature data, and dynamically adjusting the network prediction model based on an estimation method to obtain a trained network prediction model; Using the trained network prediction model to predict the mine to be predicted, and obtaining a first prevention and control strategy; A three-dimensional cellular automaton model of the coal-rock mass fracture process is constructed, and the first prevention and control strategy is simulated using the three-dimensional cellular automaton model of the coal-rock mass fracture process to obtain a final prevention and control strategy.

2. The mine rock burst warning and prevention method according to claim 1, characterized in that: The feature extraction of the multi-source monitoring data to obtain deep feature data includes: Performing denoising on the multi-source monitoring data to obtain denoised data; Performing time synchronization processing on the denoised data using a dynamic time warping algorithm to obtain synchronized data; Normalizing the synchronized data and constructing spatiotemporal features; The spatiotemporal features are extracted using a deep neural network to obtain deep feature data.

3. The mine rock burst warning and prevention method according to claim 2, characterized in that: The performing time synchronization processing on the denoised data using a dynamic time warping algorithm to obtain synchronized data includes: The dynamic time warping algorithm is used to construct the optimal time mapping path of each data sequence in the denoised data, the cumulative distance matrix is solved by the dynamic programming method, the matching path with the minimum cumulative distance is determined, and the synchronized data is obtained based on the matching path with the minimum cumulative distance.

4. The mine rock burst warning and prevention method according to claim 2, characterized in that: The normalizing the synchronized data and constructing spatiotemporal features includes: The synchronization data is normalized, and the calculation formula is: Among them, x max is the maximum value of historical data, x min Minimum value of historical data; Map the normalized data and construct spatiotemporal features.

5. The mine rock burst warning and prevention method according to claim 2, characterized in that: The extracting of the spatiotemporal features by using a deep neural network to obtain deep-level feature data includes: The spatiotemporal features are input into a deep neural network, and different convolution kernels in the deep neural network are used to perform sliding convolution on the data to extract local features in the data and obtain deep feature data.

6. The mine rock burst warning and prevention method according to claim 1, characterized in that: The method of constructing a network prediction model based on a Bayesian optimization algorithm, training the network prediction model using the deep feature data, and dynamically adjusting the network prediction model based on an estimation method to obtain a trained network prediction model includes: Based on the dynamic time warping method, the Bayesian optimization algorithm is used to set a comprehensive objective function and calculate the spatiotemporal consistency index. The calculation formula is: L(θ)=α·F1+β·precision+γ·recall+δ Among them, F1, precision, and recall are evaluation performances in different dimensions, α, β, and γ are weight coefficients, and δ is the spatiotemporal consistency index; Based on the spatiotemporal consistency index, a Bayesian network prediction model including a causal chain is constructed, and the network prediction model is trained using the deep feature data. Based on the inference mechanism of the Bayesian network, the weights are dynamically adjusted to obtain a trained network prediction model.

7. The mine rock burst warning and prevention method according to claim 1, characterized in that: The three-dimensional cellular automaton model of the coal-rock mass fracture process is constructed, and the three-dimensional cellular automaton model of the coal-rock mass fracture process is used to simulate the first prevention and control strategy to obtain the final prevention and control strategy. A three-dimensional cellular automaton model of the coal-rock fracture process is constructed, and the coal-rock mass is divided into cellular units. The prevention and control effect of the first prevention and control strategy is simulated through digital twin technology, and the first prevention and control strategy is automatically adjusted to obtain the final prevention and control strategy.

8. A mine rock burst warning and control device, characterized in that: include: The data acquisition module acquires multi-source monitoring data based on high-precision positioning methods; A feature extraction module extracts features from the multi-source monitoring data to obtain deep feature data; A training module constructs a network prediction model based on a Bayesian optimization algorithm, trains the network prediction model using the deep feature data, and dynamically adjusts the network prediction model based on an estimation method to obtain a trained network prediction model; A prediction module, using the trained network prediction model to predict the mine to be predicted, and obtain a first prevention and control strategy; The simulation module constructs a three-dimensional cellular automaton model of the coal-rock mass fracture process, and uses the three-dimensional cellular automaton model of the coal-rock mass fracture process to simulate the first prevention and control strategy to obtain the final prevention and control strategy.

9. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images