Intelligent surrounding rock stability monitoring method and system based on artificial intelligence

Abstract

The invention relates to the technical field of underground engineering monitoring, and discloses a surrounding rock stability intelligent monitoring method and system based on artificial intelligence, and the method comprises the steps: collecting initial index data in real time through a micro sensor, carrying out the noise reduction of the initial index data, and removing abnormal data and missing data through data cleaning; carrying out normalization processing to obtain preprocessed data; the method comprises the following steps: constructing a sensor spatial topological graph through a graph convolutional network to extract spatial features, capturing a long-term dependency relationship of a deformation rate by adopting a multi-head attention mechanism to extract time sequence features, and performing fusion processing to obtain a fusion feature vector; inputting the fusion feature vector into a prediction model to obtain a deformation prediction value and a stability evaluation result of the surrounding rock; when the stability state of the surrounding rock is unstable, a risk early warning signal is automatically sent out; according to the method, the precision and reliability of surrounding rock stability prediction are improved, and powerful technical support is provided for stability management of underground engineering surrounding rocks.

Description

Technical Field

[0001] The present invention relates to the technical field of underground engineering monitoring, and in particular to an intelligent monitoring method and system for surrounding rock stability based on artificial intelligence. Background Art

[0002] Underground projects, such as tunnels and mine tunnels, occupy an important position in infrastructure construction. Surrounding rock stability is a key factor in the safety of underground projects, and its monitoring is crucial to ensuring the construction and operation safety of the project. Traditional underground project surrounding rock stability monitoring methods mainly rely on manual inspections and single sensor measurements, which have problems such as low monitoring efficiency, poor real-time performance, and limited data processing capabilities. Summary of the Invention

[0003] In order to solve the above problems, the present invention designs an intelligent monitoring method and system for surrounding rock stability based on artificial intelligence.

[0004] A first aspect of the present invention provides an artificial intelligence-based intelligent monitoring method and system for surrounding rock stability, the method comprising the following steps: Using micro sensors to collect initial indicator data in real time, where the initial indicator data at least includes the thickness of the shotcrete layer, the anchor stress and the surrounding rock deformation; The denoising autoencoder is used to denoise the initial indicator data, and abnormal and missing data are removed through data cleaning. The preprocessed data is obtained after normalization. Based on the preprocessed data, a sensor spatial topology map is constructed through a graph convolutional network to extract spatial features. A multi-head attention mechanism is used to capture the long-term dependency of deformation rates and extract temporal features. After fusion processing, a fused feature vector is obtained. The fused feature vector is input into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; When the stability of the surrounding rock reaches instability, a risk warning signal will be automatically issued to remind relevant personnel to take corresponding support and reinforcement measures.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the denoising autoencoder is used to perform denoising on the initial indicator data, abnormal data and missing data are removed by data cleaning, and preprocessed data is obtained after normalization, including: The initial indicator data is input into the denoising autoencoder, and controllable noise is injected into the initial indicator data through Gaussian noise perturbation, random masking and noise intensity control, and the denoised data is generated through the encoding-decoding process; A sliding window is constructed to identify outliers in the noise-reduced data through unsupervised learning. If the outlier is a single point anomaly, the weighted average of the mean before and after the outlier and the mean of the three adjacent sensors during the same period is used for repair. If the outlier is an anomaly of multiple consecutive points, spatial interpolation of adjacent sensors is used to replace it. For short-term missing data, cubic spline interpolation is used to fill in the gaps. For long-term missing data, a time series prediction model is established using historical data to predict and fill in the gaps. The cleaned indicator data is normalized using the minimum-maximum normalization method to obtain the preprocessed data.

[0006] Optionally, in a second implementation of the first aspect of the present invention, based on the preprocessed data, a sensor space topology map is constructed through a graph convolutional network to extract spatial features, a multi-head attention mechanism is used to capture the long-term dependency of the deformation rate to extract temporal features, and a fused feature vector is obtained after fusion processing, including: Based on the deployment location and network connection relationship of the sensors, a sensor space topology map is constructed. Based on the sensor space topology map, a convolutional network is used to extract spatial features to obtain spatial features. Calculate the deformation rate sequence and input it into the multi-head attention mechanism. Multiple parallel attention heads simultaneously focus on the dependencies at different time intervals. Each attention head calculates the weighted sum of Q, K, V, and the attention weight to capture the long-term dependency characteristics of the deformation rate at different time scales. The spatial feature vector output by the graph convolutional network and the temporal feature vector output by the multi-head attention mechanism are concatenated by dimension to obtain a fused feature vector.

[0007] Optionally, in a third implementation of the first aspect of the present invention, extracting spatial features using a graph convolutional network based on the sensor spatial topology graph to obtain spatial features includes: The input layer organizes the preprocessed data into a node feature matrix, calculates the adjacency matrix and degree matrix according to the sensor position based on the sensor space topology graph, and adds self-loops; The graph convolution layer normalizes, propagates, linearly transforms, and activates the data in the input layer. It learns the spatial dependencies between sensors by aggregating the features of nodes and their adjacent nodes, and obtains the final spatial features through multi-layer graph convolution operations.

[0008] Optionally, in a fourth implementation of the first aspect of the present invention, the prediction model adopts a dual-channel modeling architecture, the data-driven channel adopts a spatiotemporal Transformer network, and the physical engine channel is simulated based on the finite element method.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, inputting the fused feature vector into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock includes: Position encoding is added to the fused feature vector, and the encoded fused feature vector is input into the encoder layer of the spatiotemporal Transformer network. The multi-head self-attention module captures the spatial and temporal dependencies of the features. The decoder layer generates the surrounding rock deformation prediction value based on the feature vector output by the encoder, and obtains the surrounding rock deformation prediction value sequence of the data-driven channel. Obtain a pre-established 3D finite element model of the surrounding rock. Set excavation boundary conditions and support conditions based on the current engineering status. Calculate the stress, strain, and displacement distribution of the surrounding rock under the current working conditions using the finite element method. Obtain the physical simulation of the surrounding rock deformation results. Extract the deformation prediction value corresponding to the sensor position from the simulation results. The weighted fusion method is used to fuse the results of the data-driven channel and the physical engine channel to obtain the final prediction value of surrounding rock deformation; The final predicted value of surrounding rock deformation is compared with the preset stability threshold, and the stability state of the surrounding rock is graded and evaluated to obtain the stability assessment result.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, comparing the final predicted value of surrounding rock deformation with a preset stability threshold, performing a graded assessment on the stability state of the surrounding rock, and obtaining a stability assessment result includes: When the predicted value of surrounding rock deformation is less than the stability threshold, the surrounding rock is judged to be in a stable state; When the predicted value of surrounding rock deformation is between the stability threshold and the instability threshold, it is judged to be in a basically stable state; When the predicted value of surrounding rock deformation is greater than or equal to the instability threshold, it is judged to be in an unstable state.

[0011] A second aspect of the present invention provides an artificial intelligence-based intelligent monitoring method and system for surrounding rock stability, the system comprising: A data acquisition module is used to collect initial indicator data in real time using micro sensors, wherein the initial indicator data at least includes the thickness of the sprayed layer, the stress of the anchor rod and the deformation of the surrounding rock; The data processing module is used to perform noise reduction processing on the initial indicator data using a noise reduction autoencoder, remove abnormal data and missing data through data cleaning, and obtain pre-processed data after normalization; The feature extraction module is used to extract spatial features based on the preprocessed data by constructing a sensor spatial topology map through a graph convolutional network. It also uses a multi-head attention mechanism to capture the long-term dependencies of the deformation rate and extract temporal features. After fusion processing, it obtains a fused feature vector. The prediction module is used to input the fused feature vector into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; The early warning module is used to automatically issue a risk warning signal when the stability of the surrounding rock reaches instability, reminding relevant personnel to take corresponding support and reinforcement measures.

[0012] The third aspect of the present invention provides an artificial intelligence-based intelligent monitoring device for surrounding rock stability, which includes a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory to enable the artificial intelligence-based intelligent monitoring device for surrounding rock stability to execute each step of the artificial intelligence-based intelligent monitoring method for surrounding rock stability as described in any of the above items.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed by a processor, implement the various steps of the intelligent monitoring method for surrounding rock stability based on artificial intelligence as described in any of the above items.

[0014] In the technical solution provided by the present invention, micro sensors are used to collect initial indicator data in real time, and noise reduction processing is performed on the initial indicator data using a noise reduction autoencoder. Abnormal data and missing data are removed through data cleaning, and preprocessed data is obtained after normalization processing; based on the preprocessed data, a sensor space topology map is constructed through a graph convolutional network to extract spatial features, and a multi-head attention mechanism is used to capture the long-term dependency of the deformation rate to extract the time series features, and a fusion feature vector is obtained after fusion processing; the fusion feature vector is input into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; when the stability state of the surrounding rock reaches instability, a risk warning signal is automatically issued to remind relevant personnel to take corresponding support and reinforcement measures; the present invention realizes efficient processing and precise analysis of monitoring data, can automatically extract key features in the data, improves the accuracy and reliability of surrounding rock stability prediction, improves the accuracy and effectiveness of risk warning, and provides strong technical support for the stability management of surrounding rock in underground engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Various other advantages and benefits will become apparent to those skilled in the art by reading the following detailed description of the preferred embodiment.The accompanying drawings are only for the purpose of illustrating the preferred embodiment and are not to be considered as limiting the present invention.

[0016] Figure 1 A flowchart of an intelligent monitoring method for surrounding rock stability based on artificial intelligence provided by an embodiment of the present invention; Figure 2A schematic diagram of the structure of an intelligent monitoring system for surrounding rock stability based on artificial intelligence provided by an embodiment of the present invention; Figure 3 A schematic structural diagram of an intelligent monitoring device for surrounding rock stability based on artificial intelligence provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The terms "first," "second," "third," "fourth," and so forth (if any) in the description and claims of the present invention and in the accompanying drawings are used to distinguish similar items and are not necessarily used to describe a particular order or sequential sequence. It should be understood that the terms used in this manner are interchangeable under appropriate circumstances, so that the embodiments described herein can be practiced in an order other than that illustrated or described herein. Furthermore, the terms "including," "comprising," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product, or device that includes a series of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, product, or device.

[0018] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 The embodiment of the present invention provides a flowchart of an intelligent monitoring method for surrounding rock stability based on artificial intelligence, which specifically includes the following steps: Step 101: Using micro sensors to collect initial indicator data in real time, wherein the initial indicator data at least includes the thickness of the sprayed layer, the anchor stress, and the surrounding rock deformation; In this embodiment, at engineering sites such as tunnels or mines, a grid-based deployment approach is adopted based on the surrounding rock structure characteristics and geological conditions. Sensors are placed at key locations such as the tunnel vault, haunch, and foot of the vault, as well as at the sides, top, and bottom of mine tunnels. The sensor spacing is determined based on the preliminary assessment of the surrounding rock stability and is generally 5-10 meters. In areas with complex geological conditions or the presence of faults or fracture zones, the sensor layout is increased, with spacing shortened to 3-5 meters to ensure the comprehensiveness and representativeness of the monitoring data. During sensor installation, the sensors must be positioned and fixed strictly in accordance with the design requirements to ensure close contact between the sensors and the surrounding rock surface, avoiding data collection errors caused by improper installation. A sensor archive is also established, recording information such as the installation location, model, and number of each sensor to facilitate subsequent maintenance and data management. Step 102: Use a denoising autoencoder to perform denoising on the initial indicator data, remove abnormal data and missing data through data cleaning, and obtain pre-processed data after normalization; In this embodiment, the initial indicator data is input into the denoising autoencoder, and controllable noise is injected into the initial indicator data through Gaussian noise perturbation, random masking and noise intensity control, and the denoised data is generated through the encoding-decoding process; a sliding window is constructed, and outliers in the denoised data are identified through unsupervised learning. If the outlier is a single-point anomaly, the weighted average of the mean before and after the outlier and the mean of the three adjacent sensors during the same period is taken for repair; if the outlier is an anomaly of multiple consecutive points, spatial interpolation of adjacent sensors is used to replace it; short-term missing data is supplemented by cubic spline interpolation, and long-term missing data is supplemented by using historical data to establish a time series prediction model to predict and supplement the missing data; the cleaned indicator data is normalized by the minimum-maximum normalization method to obtain preprocessed data.

[0019] In this embodiment, a three-layer fully connected denoising autoencoder is constructed, which includes an input layer and a hidden layer. The hidden layer includes an encoding layer and a decoding layer. The number of neurons in the input layer is consistent with the dimension of the monitoring indicator. The encoding layer adopts a two-layer dimensionality reduction structure and uses ReLU as the activation function to achieve data feature compression. The decoding layer has a symmetrical two-layer dimensionality increase structure, and the output layer uses a linear activation function to restore the original data dimension.

[0020] In this embodiment, abnormal data identification uses a combination of statistical methods and machine learning to calculate statistical quantities such as the mean and standard deviation of each indicator data. A reasonable threshold range is set, such as the mean ± 3 times the standard deviation. Data exceeding the threshold is considered abnormal data. At the same time, machine learning algorithms such as isolation forest and local outlier factor are used to detect outliers in the data to improve the accuracy of abnormal data identification. Abnormal data is processed based on its characteristics. If the abnormal data is a single point anomaly, the mean or median of the data at adjacent time points is used for interpolation and replacement. If the abnormal data is multiple points in a row, the on-site sensor status and actual project conditions are combined to determine whether it is a sensor failure or data transmission error. If it is a sensor failure, it is repaired or replaced in a timely manner, and the abnormal data segment of the sensor is eliminated in subsequent data processing. For short-term missing data, since data at adjacent time points usually have strong continuity and linear change trends, interpolation methods are suitable for completion. Linear interpolation assumes that the two valid data points before and after the missing point change linearly, and calculates the value of the missing position by the slope of the line connecting the two points. Cubic spline interpolation further optimizes smoothness. It constructs a piecewise cubic polynomial curve to ensure the continuity of function values, first-order derivatives, and second-order derivatives in adjacent intervals, avoiding sudden changes that may be caused by linear interpolation. It is particularly suitable for indicators with physical continuity such as surrounding rock deformation. For example, when the surrounding rock deformation rate is stable within a certain period of time, cubic spline interpolation can more accurately fit the gradual change process at the missing moment. The core advantages of these two methods are that they rely on local adjacent data, have low computational complexity, can quickly recover high-frequency data that is missing for a short time, and retain the trend characteristics of the original signal. They are suitable for short-term missing scenarios caused by temporary sensor signal interruption, data transmission delay, etc. When data is missing for a long period, simple interpolation methods are unable to capture the data's long-term trends, cyclical, or nonlinear variations. Therefore, a time series prediction model is needed to predict and complete the missing data. First, based on complete historical data, a monitoring series typically spanning 1-7 days prior to the missing point is selected. The data's statistical characteristics, such as mean, variance, and autocorrelation, are analyzed to select an appropriate model, such as ARIMA or LSTM. Taking the ARIMA model as an example, the data must first be stabilized, the model order determined using the autocorrelation and partial autocorrelation functions, and the fitting parameters estimated using maximum likelihood. Finally, the values ​​for the missing time period are extrapolated based on the fitted model. For surrounding rock deformation data with significant nonlinear characteristics, LSTM neural networks are more advantageous. They use memory cells to capture dependencies in long time series. For example, using deformation data 24 hours prior to the missing point as input, the model is trained to learn the mapping between time and deformation, thereby predicting missing values ​​for the next several hours. The key to the model prediction method is to utilize the overall pattern of historical data rather than relying solely on local adjacent points. It can effectively handle long-term data loss caused by sensor failure, network interruption, etc. The completed results not only conform to short-term change trends, but also reflect the long-term evolution of surrounding rock stability, such as the influence of cyclic loads and progressive deformation of geological structures. In engineering applications, it is usually necessary to adjust model parameters through cross-validation and evaluate prediction accuracy through indicators such as root mean square error to ensure the reliability of the completed data.

[0021] Step 103: Based on the preprocessed data, a sensor spatial topology map is constructed through a graph convolutional network to extract spatial features. A multi-head attention mechanism is used to capture the long-term dependency of the deformation rate to extract temporal features. After fusion processing, a fused feature vector is obtained. In this embodiment, a sensor space topology map is constructed based on the deployment location and network connection relationship of the sensors. A graph convolutional network is used to extract spatial features based on the sensor space topology map to obtain spatial features. The deformation rate sequence is calculated and input into a multi-head attention mechanism. Multiple parallel attention heads simultaneously focus on the dependencies at different time intervals. Each attention head calculates the weighted sum of Q, K, V and the attention weight to capture the long-term dependency characteristics of the deformation rate at different time scales. The spatial feature vector output by the graph convolutional network and the temporal feature vector output by the multi-head attention mechanism are concatenated by dimension to obtain a fused feature vector.

[0022] In this embodiment, the input layer organizes the preprocessed data into a node feature matrix, calculates the adjacency matrix and degree matrix according to the sensor position based on the sensor spatial topology map, and adds self-loops; the graph convolution layer normalizes, propagates, linearly transforms and activates the data of the input layer, learns the spatial dependency relationship between sensors by aggregating the features of nodes and their adjacent nodes, and obtains the final spatial features through multi-layer graph convolution operations.

[0023] In this embodiment, a sensor space topology graph is constructed based on the deployment locations and network connection relationships of the sensors. In the graph, nodes represent sensors, and edges represent the spatial adjacency relationships between sensors. The weight of the edge is determined by the inverse of the distance between sensors. The closer the distance, the greater the weight, to reflect the spatial correlation between sensors.

[0024] In this embodiment, a self-loop is a special edge in graph theory, which refers to a node directly connected to itself through an edge, forming a closed loop. In an undirected graph, a self-loop is represented by an edge whose two endpoints are the same node; in a directed graph, it is a directed edge starting from a node and returning to itself. For example, in a sensor space topology graph, if a sensor not only interacts with adjacent devices but also needs to model its own state, this self-association relationship can be represented by a self-loop.

[0025] Step 104: Input the fused feature vector into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; In this embodiment, the prediction model adopts a dual-channel modeling architecture, the data-driven channel adopts a spatiotemporal Transformer network, and the physical engine channel is simulated based on the finite element method.

[0026] In this embodiment, position encoding is added to the fused feature vector, and the encoded fused feature vector is input into the encoder layer of the spatiotemporal Transformer network. The spatial and temporal dependencies of the features are captured by the multi-head self-attention module. The decoder layer generates a surrounding rock deformation prediction value based on the feature vector output by the encoder, and obtains a sequence of surrounding rock deformation prediction values ​​of the data-driven channel; a pre-established three-dimensional finite element model of the surrounding rock is obtained, and the excavation boundary conditions and support conditions are set according to the current engineering status. Based on the finite element method, the stress, strain and displacement distribution of the surrounding rock under the current working conditions are calculated to obtain the surrounding rock deformation results of the physical simulation, and the deformation prediction value corresponding to the sensor position is extracted from the simulation results; the results of the data-driven channel and the physical engine channel are fused using a weighted fusion method to obtain the final surrounding rock deformation prediction value; the final surrounding rock deformation prediction value is compared with a preset stability threshold, and the stability state of the surrounding rock is graded and evaluated to obtain a stability evaluation result.

[0027] In this embodiment, when the predicted value of surrounding rock deformation is less than the stability threshold, the surrounding rock is determined to be in a stable state; when the predicted value of surrounding rock deformation is between the stability threshold and the instability threshold, it is determined to be in a basically stable state; when the predicted value of surrounding rock deformation is greater than or equal to the instability threshold, it is determined to be in an unstable state.

[0028] In this embodiment, the spatiotemporal Transformer network structure includes an input layer, an encoder layer, and a decoder layer. The input layer takes the fused features as input, which enters the encoder layer after position encoding. The encoder layer is composed of multiple multi-head self-attention modules and a feedforward neural network. The self-attention mechanism captures the spatial and temporal dependencies of the fused features to predict the deformation trend of the surrounding rock. The decoder layer generates the predicted value of the surrounding rock deformation based on the feature vector output by the encoder. The model training adopts the supervised learning method, with historical fusion features and the corresponding actual values ​​of surrounding rock deformation as training data. The loss function adopts the mean square error, and the optimizer selects the Adam optimizer. The model parameters are adjusted through the back propagation algorithm to minimize the error between the predicted value and the actual value. In this embodiment, the physical engine channel uses numerical simulation of the rock mass constitutive equation based on the finite element method. Based on engineering geological survey data, a three-dimensional finite element model of the surrounding rock is established. The model is meshed using tetrahedron or hexahedron elements. The mesh density is determined based on the complexity of the surrounding rock and the required calculation accuracy, and mesh density is increased in key locations. Based on the lithology and mechanical properties of the surrounding rock, an appropriate rock mass constitutive equation, such as the elastic constitutive equation or the elastoplastic constitutive equation, is selected. For rock masses with well-developed joints, a discrete element model or a contact mechanics model that considers joint surfaces is used. An initial geostress field is applied, and the initial geostress is determined based on geological exploration data and field measurement results. Then, the engineering excavation process is simulated, considering the effects of factors such as excavation sequence and support measures on the stress and deformation of the surrounding rock. Finally, the stress, strain, and displacement distribution of the surrounding rock under different working conditions are calculated to obtain the physical simulation results of the surrounding rock deformation.

[0029] Step 105: When the stability state of the surrounding rock reaches instability, a risk warning signal is automatically issued to remind relevant personnel to take corresponding support and reinforcement measures.

[0030] In this embodiment, when the surrounding rock stability is assessed as unstable, a risk warning signal is automatically triggered. Warning signals include audible alarms, flashing lights, text message notifications, and platform pop-up notifications, ensuring that relevant personnel receive timely warning information. The warning signal includes detailed information such as the warning time, warning location, stability level, and predicted deformation, allowing relevant personnel to quickly understand the dangerous situation. When a stable state warning is received, the normal monitoring frequency and inspection system are maintained, and monitoring equipment and support structures are regularly inspected and maintained. When a basic stable state warning is received, the monitoring frequency is increased to once every five minutes, and on-site inspections are strengthened. Surrounding rock deformation and the working status of support structures are closely observed, and targeted support and reinforcement plans are developed to prepare for construction. When an unstable state warning is received, the emergency plan is immediately activated, operations in the relevant area are halted, and personnel are evacuated to a safe area. Professional technicians and construction teams are notified, and emergency measures are implemented according to pre-set support and reinforcement measures, such as increasing the number of anchor rods and cables and using shotcrete reinforcement. During this process, surrounding rock deformation and stability are continuously monitored until the dangerous situation is under control.

[0031] See also Figure 2 , a schematic structural diagram of an intelligent monitoring system for surrounding rock stability based on artificial intelligence provided by an embodiment of the present invention, the system includes: A data acquisition module is used to collect initial indicator data in real time using micro sensors, wherein the initial indicator data at least includes the thickness of the sprayed layer, the stress of the anchor rod and the deformation of the surrounding rock; The data processing module is used to perform noise reduction processing on the initial indicator data using a noise reduction autoencoder, remove abnormal data and missing data through data cleaning, and obtain pre-processed data after normalization; The feature extraction module is used to extract spatial features based on the preprocessed data by constructing a sensor spatial topology map through a graph convolutional network. It also uses a multi-head attention mechanism to capture the long-term dependencies of the deformation rate and extract temporal features. After fusion processing, it obtains a fused feature vector. The prediction module is used to input the fused feature vector into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; The early warning module is used to automatically issue a risk warning signal when the stability of the surrounding rock reaches instability, reminding relevant personnel to take corresponding support and reinforcement measures.

[0032] In this embodiment, a micro sensor unit is pre-integrated between the shotcrete and the anchor support system to form a structure-sensing integrated support system. The module is mainly composed of a high-strength steel mesh, a reinforced fiber concrete layer and multiple micro strain and displacement sensing units, and is firmly installed on the surrounding rock surface through pre-buried anchor rods. After the construction is completed, the internal sensor elements can realize long-term online monitoring of key indicators such as the thickness of the shotcrete layer, anchor stress, and surrounding rock deformation, and transmit the data to the back-end platform in real time through a wireless network to realize intelligent diagnosis of the support status and risk warning. It has both flexible and rigid support characteristics, is suitable for temporary or permanent support needs under different geological conditions, significantly improves the prediction and control capabilities of the surrounding rock stability of the cavern group, and has good prospects for engineering promotion and application.

[0033] Figure 3 This is a schematic diagram of the structure of an AI-based intelligent monitoring device for surrounding rock stability provided by an embodiment of the present invention. This AI-based intelligent monitoring device 300 can vary significantly depending on configuration or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors), memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) storing application programs 333 or data 332. The memory 320 and storage medium 330 may be either transient or persistent storage. The program stored in the storage medium 330 may include one or more modules (not shown), each of which may include a series of instructions and operations within the AI-based intelligent monitoring device for surrounding rock stability 300. Furthermore, the processor 310 may be configured to communicate with the storage medium 330, allowing the AI-based intelligent monitoring device for surrounding rock stability 300 to execute the series of instructions and operations stored in the storage medium 330 to implement the method provided in the above embodiment.

[0034] The artificial intelligence-based intelligent monitoring device 300 for surrounding rock stability may further include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input and output interfaces 360, and / or one or more operating devices 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. It will be understood by those skilled in the art that Figure 3The structure of the artificial intelligence-based intelligent monitoring equipment for surrounding rock stability shown does not constitute a limitation on the computer equipment provided by the present invention, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0035] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions. When the instructions are executed on a computer, the computer executes the various steps of the artificial intelligence-based intelligent monitoring method for surrounding rock stability provided in the above-mentioned embodiments.

[0036] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices, and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0037] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0038] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention, and such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent monitoring method for surrounding rock stability based on artificial intelligence, characterized in that: The method comprises the following steps: Using micro sensors to collect initial indicator data in real time, where the initial indicator data at least includes the thickness of the shotcrete layer, the anchor stress and the surrounding rock deformation; The denoising autoencoder is used to denoise the initial indicator data, and abnormal and missing data are removed through data cleaning. The preprocessed data is obtained after normalization. Based on the preprocessed data, a sensor spatial topology map is constructed through a graph convolutional network to extract spatial features. A multi-head attention mechanism is used to capture the long-term dependency of deformation rates and extract temporal features. After fusion processing, a fused feature vector is obtained. The fused feature vector is input into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; When the stability of the surrounding rock reaches instability, a risk warning signal will be automatically issued to remind relevant personnel to take corresponding support and reinforcement measures.

2. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 1, characterized in that: The denoising autoencoder is used to perform denoising on the initial indicator data, abnormal data and missing data are removed through data cleaning, and pre-processed data is obtained after normalization, including: The initial indicator data is input into the denoising autoencoder, and controllable noise is injected into the initial indicator data through Gaussian noise perturbation, random masking and noise intensity control, and the denoised data is generated through the encoding-decoding process; A sliding window is constructed to identify outliers in the noise-reduced data through unsupervised learning. If the outlier is a single point anomaly, the weighted average of the mean before and after the outlier and the mean of the three adjacent sensors during the same period is used for repair. If the outlier is an anomaly of multiple consecutive points, spatial interpolation of adjacent sensors is used to replace it. For short-term missing data, cubic spline interpolation is used to fill in the gaps. For long-term missing data, a time series prediction model is established using historical data to predict and fill in the gaps. The cleaned indicator data is normalized using the minimum-maximum normalization method to obtain the preprocessed data.

3. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 1, characterized in that: Based on the preprocessed data, the sensor space topology map is constructed through a graph convolutional network to extract spatial features. The multi-head attention mechanism is used to capture the long-term dependency of the deformation rate to extract temporal features. After fusion processing, a fused feature vector is obtained, including: Based on the deployment location and network connection relationship of the sensors, a sensor space topology map is constructed. Based on the sensor space topology map, a convolutional network is used to extract spatial features to obtain spatial features. Calculate the deformation rate sequence and input it into the multi-head attention mechanism. Multiple parallel attention heads simultaneously focus on the dependencies at different time intervals. Each attention head calculates the weighted sum of Q, K, V, and the attention weight to capture the long-term dependency characteristics of the deformation rate at different time scales. The spatial feature vector output by the graph convolutional network and the temporal feature vector output by the multi-head attention mechanism are concatenated by dimension to obtain a fused feature vector.

4. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 3, characterized in that: The spatial feature extraction is performed using a graph convolutional network based on the sensor spatial topology map to obtain spatial features, including: The input layer organizes the preprocessed data into a node feature matrix, calculates the adjacency matrix and degree matrix according to the sensor position based on the sensor space topology graph, and adds self-loops; The graph convolution layer normalizes, propagates, linearly transforms, and activates the data in the input layer. It learns the spatial dependencies between sensors by aggregating the features of nodes and their adjacent nodes, and obtains the final spatial features through multi-layer graph convolution operations.

5. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 1, characterized in that: The prediction model adopts a dual-channel modeling architecture, the data-driven channel uses a spatiotemporal Transformer network, and the physical engine channel is simulated based on the finite element method.

6. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 5, characterized in that: The fused feature vector is input into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock, including: Position encoding is added to the fused feature vector, and the encoded fused feature vector is input into the encoder layer of the spatiotemporal Transformer network. The multi-head self-attention module captures the spatial and temporal dependencies of the features. The decoder layer generates the surrounding rock deformation prediction value based on the feature vector output by the encoder, and obtains the surrounding rock deformation prediction value sequence of the data-driven channel. Obtain a pre-established 3D finite element model of the surrounding rock. Set excavation boundary conditions and support conditions based on the current engineering status. Calculate the stress, strain, and displacement distribution of the surrounding rock under the current working conditions using the finite element method. Obtain the physical simulation of the surrounding rock deformation results. Extract the deformation prediction value corresponding to the sensor position from the simulation results. The weighted fusion method is used to fuse the results of the data-driven channel and the physical engine channel to obtain the final prediction value of surrounding rock deformation; The final predicted value of surrounding rock deformation is compared with the preset stability threshold, and the stability state of the surrounding rock is graded and evaluated to obtain the stability assessment result.

7. The method for intelligent monitoring of surrounding rock stability based on artificial intelligence according to claim 6, characterized in that: The final predicted deformation value of the surrounding rock is compared with the preset stability threshold, and the stability state of the surrounding rock is graded and evaluated to obtain the stability evaluation result, including: When the predicted value of surrounding rock deformation is less than the stability threshold, the surrounding rock is judged to be in a stable state; When the predicted value of surrounding rock deformation is between the stability threshold and the instability threshold, it is judged to be in a basically stable state; When the predicted value of surrounding rock deformation is greater than or equal to the instability threshold, it is judged to be in an unstable state.

8. An intelligent monitoring system for surrounding rock stability based on artificial intelligence, characterized in that: The system includes: A data acquisition module is used to collect initial indicator data in real time using micro sensors, wherein the initial indicator data at least includes the thickness of the sprayed layer, the stress of the anchor rod and the deformation of the surrounding rock; The data processing module is used to perform noise reduction processing on the initial indicator data using a noise reduction autoencoder, remove abnormal data and missing data through data cleaning, and obtain pre-processed data after normalization; The feature extraction module is used to extract spatial features based on the preprocessed data by constructing a sensor spatial topology map through a graph convolutional network. It also uses a multi-head attention mechanism to capture the long-term dependencies of the deformation rate and extract temporal features. After fusion processing, it obtains a fused feature vector. The prediction module is used to input the fused feature vector into the prediction model to obtain the deformation prediction value and stability assessment result of the surrounding rock; The early warning module is used to automatically issue a risk warning signal when the stability of the surrounding rock reaches instability, reminding relevant personnel to take corresponding support and reinforcement measures.

9. An intelligent monitoring device for surrounding rock stability based on artificial intelligence, characterized in that: The artificial intelligence-based intelligent monitoring device for surrounding rock stability includes a memory and at least one processor, wherein the memory stores instructions; the at least one processor calls the instructions in the memory so that the artificial intelligence-based intelligent monitoring device for surrounding rock stability executes each step of the artificial intelligence-based intelligent monitoring method for surrounding rock stability as described in any one of claims 1 to 7.

10. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by the processor, the various steps of the intelligent monitoring method for surrounding rock stability based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

Images