Karst tunnel collapse disaster microearthquake precursor information identification method

Abstract

The invention relates to the technical field of geological disaster forecasting, in particular to a karst tunnel collapse disaster microearthquake precursor information identification method. According to the technical scheme, the method comprises the steps of constructing a micro-seismic monitoring system fusing multi-source data, carrying out processing and analysis based on machine learning, establishing a multi-factor coupling analysis and early warning model, and carrying out real-time early warning and dynamic feedback. According to the method, noise can be accurately removed through wavelet transformation, key features of the micro-seismic signals can be reserved, the data quality is improved, multi-dimensional features including frequency, amplitude, phase and the like are extracted from the micro-seismic signals, multi-dimensional feature extraction is achieved, and therefore the accuracy of capturing precursor information related to collapse disasters is improved; in addition, comprehensive classification and judgment are carried out by utilizing a machine learning algorithm and combining multiple factors such as geological conditions and construction progress, so that the recognition result is more in line with the actual situation, misjudgment caused by a single factor is avoided, and the recognition accuracy and efficiency are improved.

Description

Technical Field

[0001] The invention relates to the technical field of geological disaster prediction, and in particular to a method for identifying microseismic precursor information of karst tunnel collapse disaster. Background Art

[0002] With the continuous development of transportation infrastructure construction, the number of tunnel projects in karst areas is increasing. Karst geology is complex and uncertain, and there are a large number of special geological phenomena such as caves, dissolution cracks and groundwater activities. In the construction process of karst tunnels, landslide disasters are one of the most serious engineering accidents, which will not only cause construction progress delays and cost increases, but also endanger the lives of construction workers. Therefore, how to effectively predict and prevent karst tunnel landslide disasters has become a key issue that needs to be solved in the engineering field. Since karst tunnel landslides are often the result of the combined action of multiple factors, such as geological structure, groundwater activity, rock mechanical properties and engineering construction, most of the existing identification methods only focus on the analysis of a single factor or a few factors, lacking comprehensive consideration and in-depth research on microseismic precursor information under the coupling of multiple factors, and it is difficult to fully and accurately reflect the breeding process of landslide disasters. Therefore, we propose a method for identifying microseismic precursor information of karst tunnel landslide disasters. Summary of the invention

[0003] The purpose of the present invention is to propose a method for identifying microseismic precursor information of karst tunnel collapse disasters in order to address the problem that the existing identification methods in the background technology lack comprehensive consideration and in-depth research on microseismic precursor information under the coupling of multiple factors, and are difficult to fully and accurately reflect the breeding process of landslide disasters.

[0004] The technical solution of the present invention is a method for identifying microseismic precursor information of karst tunnel collapse disaster, comprising the following steps:

[0005] Build a microseismic monitoring system that integrates multi-source data;

[0006] Processing and analyzing the microseismic signals collected by the microseismic detection system based on machine learning to extract multi-dimensional features of the microseismic signals;

[0007] Based on the multi-dimensional characteristics of microseismic signals, a multi-factor coupling analysis and early warning model is established to analyze the influence mechanism and degree of different factors on microseismic precursor information, and determine the early warning indicators and their thresholds;

[0008] Real-time early warning and dynamic feedback: Based on the early warning indicators and their thresholds, dynamic assessment and early warning are carried out on the collapse risk of karst tunnels.

[0009] Optionally, the construction of a microseismic monitoring system integrating multi-source data comprises the following steps:

[0010] Distributed optical fiber sensors are laid along the tunnel lining, vault, and sidewalls, and point sensors are arranged in areas with complex geological structures, high-risk rock crushing, as well as in tunnel entrances and exits and fault fracture zones to form a monitoring network.

[0011] Geological data, groundwater data and construction data are collected synchronously, where geological data includes rock mechanics parameters and geological structure information, groundwater data includes water level changes and water flow velocity, and construction data includes excavation progress and support parameters.

[0012] Optionally, the processing and analysis based on machine learning includes the following steps:

[0013] Using wavelet transform to perform denoising and filtering preprocessing on the microseismic signals collected by the microseismic detection system;

[0014] Extract the amplitude, frequency, energy, and duration features of the signal to construct a feature data set;

[0015] A support vector machine is used to train and optimize the model by taking the feature data set as input and whether it is microseismic precursor information of landslide disaster as output label;

[0016] The weighted average method is used to fuse and integrate multiple trained machine learning models.

[0017] Optionally, when using wavelet transform to perform denoising and filtering preprocessing on microseismic signals, the original microseismic signal is assumed to be f(t), the wavelet basis function ψ(t) is selected, and the original signal is subjected to wavelet decomposition to obtain the wavelet coefficients at different scales a and displacements b:

[0018]

[0019] The wavelet coefficients are processed by setting the threshold λ. for:

[0020] when hour,

[0021] when hour, Reconstruct the processed wavelet coefficients to obtain the denoised signal:

[0022]

[0023] in,

[0024]

[0025] in, is the Fourier transform of ψ(t);

[0026] Construct a feature data set, the amplitude feature includes the maximum amplitude of the signal and the average amplitude Where T is the duration of the signal;

[0027] The frequency characteristics are obtained by Fourier transforming the signal Get the spectrum of the signal and then calculate the main frequency where ω p is the angular frequency corresponding to the frequency component with the largest amplitude in the spectrum, and the frequency bandwidth Δf = f h -f t , f h and f t They are the high-frequency and low-frequency boundary frequencies corresponding to a certain energy proportion in the signal spectrum;

[0028] The energy feature uses the energy of the signal

[0029] The duration characteristic is the time interval T0 from the start to the end of the recording signal;

[0030] In model training and optimization, the feature dataset {x i ,y i}(where x i is the feature vector, y i ∈{-1,1} indicates whether it is precursor information), and the optimization function is constructed as follows:

[0031]

[0032] Among them, the constraint y i (ω·x i +b)≥1-ξ i ,ξ i ≥0(i=1,2,…,n), where ω is the normal vector of the hyperplane, b is the bias term, C is the penalty parameter, ξ i is the slack variable, and the optimal hyperplane parameter ω is obtained by solving * and b * , thus determining the classification model f(x)=sign(ω * ·x+b * ), during the training process, the cross-validation method is used to adjust the parameters to improve the generalization ability of the model;

[0033] The weighted average method is used to fuse and integrate multiple trained machine learning models. k Prediction results for test samples, final prediction results where α i It is model M iThe weights can also be determined based on the model performance evaluation indicators, integrating the prediction information of multiple models.

[0034] Optionally, the cross-validation method adopts a K-fold cross-validation method, and the validation process includes the following steps:

[0035] Dataset division: Assume that the microseismic signal feature dataset contains n samples and randomly divide it into K mutually exclusive subsets F1, F 2, …,F K ,satisfy D represents the entire data set, and The number of samples in each subset is

[0036] Training and validation process and performance index calculation: For k = 1, 2, ..., K rounds, select k The K-1 subsets outside of the above are used as training sets Train k =∪ i≠k F i , F k As the validation set Val k , using the training set Train k Train the machine learning model, and the trained model is denoted as M k , for classification tasks, if the model M k In the validation set Val k The above prediction result is i represents the sample number, j represents the category number, and the true label is y ij , the accuracy calculation formula is as follows:

[0037]

[0038] Where II(·) is an indicator function, which takes the value 1 when the condition in the brackets is met, otherwise it takes the value 0. k | indicates the validation set Val k The number of samples, for regression tasks, if the predicted value is The true value is y i , the mean square error calculation formula is as follows:

[0039]

[0040] Parameter adjustment and model selection: Set multiple values ​​for the model parameters to be adjusted. For each parameter value θ, repeat the training and verification process for K rounds to obtain the average performance index under the parameter value, including the average accuracy or the average mean square error. The calculation formulas are:

[0041]

[0042]

[0043] Select the optimal parameter value θ based on the average performance index opt , so that for the classification task For regression tasks To improve the generalization ability of the model.

[0044] Optionally, the establishment of a multi-factor coupling analysis and early warning model includes the following steps:

[0045] Construct a multi-factor coupling analysis model based on physical mechanics principles and numerical simulation methods;

[0046] According to the interaction between geological structure, groundwater, rock mechanical properties and engineering construction factors, the stress-strain state and microseismic activity of the rock and soil around the tunnel under different working conditions are simulated by data simulation method;

[0047] The influence of different factors on microseismic precursor information is analyzed through a multi-factor coupling analysis model, and the early warning indicators and their thresholds are determined.

[0048] Optionally, the numerical simulation method adopts the finite element method or the discrete element method, and the constitutive relationship of the rock and soil body in the simulation process adopts the Mohr-Coulomb criterion or the Drucker-Prager criterion, and its stress-strain relationship expression is: {σ}=[D]{ε}, where {σ} is the stress vector, [D] is the elastic matrix, and {ε} is the strain vector.

[0049] Optionally, the real-time warning and dynamic feedback include:

[0050] Establish a data transmission system to transmit multi-source data from the monitoring site to the data processing center in real time;

[0051] Use cloud computing or edge computing technology to quickly process and analyze data in the data processing center;

[0052] Dynamic assessment and early warning are carried out on the collapse risk of karst tunnels. When the microseismic precursor information exceeds the early warning threshold, an early warning signal is issued and fed back to the construction site.

[0053] Optionally, the data transmission system adopts wireless transmission technology or wired transmission technology, and the data transmission protocol follows a specific industrial standard protocol, including the Modbus protocol or the OPC UA protocol. The application of cloud computing or edge computing technology in data processing includes data distributed storage, parallel computing task allocation, and resource scheduling optimization.

[0054] Optionally, during the construction process, as the tunnel construction progresses and geological conditions change, the parameters in the early warning model are updated and optimized, including adjusting the early warning indicator thresholds and retraining the machine learning model weights to adapt to the needs of identifying microseismic precursor information of karst tunnel collapse disasters in different construction stages and geological environments.

[0055] In summary, the present application includes at least one of the following beneficial technical effects:

[0056] The present invention can accurately remove noise through wavelet transform, retain key features of microseismic signals, improve data quality, lay a foundation for subsequent accurate identification of precursor information, and extract multi-dimensional features including frequency, amplitude, phase, etc. from microseismic signals, realize multi-dimensional feature extraction, and comprehensively characterize the characteristics of microseismic events, thereby improving the accuracy of capturing precursor information related to landslide disasters and reducing the possibility of missed reports and false reports;

[0057] The present invention uses a machine learning algorithm to automatically learn and identify the microseismic patterns of different types of landslide disasters from a large amount of microseismic data. At the same time, it also combines geological conditions, construction progress and other factors for comprehensive classification and judgment, so that the recognition results are more in line with the actual situation, avoiding misjudgment caused by a single factor, and improving the accuracy and efficiency of recognition;

[0058] The present invention processes and analyzes the collected microseismic data in real time, and quickly gives the identification results of precursor information, which meets the needs of real-time monitoring and timely early warning during karst tunnel construction. It also has strong dynamic adaptability and can continuously update and optimize the identification model as the tunnel construction progresses and the geological conditions change, so as to always maintain accurate identification of microseismic precursor information and effectively cope with the complex and changeable construction environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flow chart of a method for identifying microseismic precursor information of karst tunnel collapse disasters according to the present invention is provided. DETAILED DESCRIPTION

[0060] The technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.

[0061] The components of the embodiments of the present invention generally described and shown in the drawings herein may be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but merely represents selected embodiments of the present invention.

[0062] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making any creative work shall fall within the scope of protection of the present invention.

[0063] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0064] It should be noted that the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In this specification, the schematic representation of the above terms does 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.

[0065] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0066] like Figure 1 As shown, the present invention proposes a method for identifying microseismic precursor information of karst tunnel collapse disaster, including the following implementation steps:

[0067] 1. Constructing a microseismic monitoring system integrating multi-source data:

[0068] Sensor layout: Distributed fiber optic sensors are laid along the tunnel lining, vault, and sidewalls, with a monitoring node set every 10 meters to ensure continuous monitoring of the microseismic situation of the entire tunnel section; at the same time, point sensors are arranged in areas with complex geological structures (including the vicinity of proven faults and large-scale karst cave-dense areas), high-risk areas for rock crushing (including areas with serious development of joints and fissures determined through preliminary geological surveys), and key locations at tunnel entrances and exits and fault fracture zones. A total of 500 distributed fiber optic sensor nodes and 100 point sensors are arranged to form an all-round, multi-level monitoring network;

[0069] Data collection: Geological data, groundwater data and construction data are collected simultaneously. For geological data, cores are drilled at different locations along the tunnel to test rock mechanics parameters, including compressive strength, elastic modulus and Poisson's ratio. A total of 100 sets of rock mechanics parameter data are obtained. Geological structure information maps are drawn through geological radar detection to clarify the location and scale of faults and large caves. For groundwater data collection, 20 water level sensors and 15 water flow velocity sensors are set at different depths in the tunnel to monitor groundwater level changes and water flow velocity in real time. Data is collected every 30 minutes. Construction data is automatically recorded by the construction recording system, including the progress of each excavation and support parameters (including anchor length, installation time, thickness and strength grade of shotcrete).

[0070] 2. Process and analyze the microseismic signals collected by the microseismic detection system based on machine learning to extract the multi-dimensional features of the microseismic signals;

[0071] Data preprocessing: Wavelet transform is used to perform denoising and filtering preprocessing on the collected microseismic signals. Assuming the original microseismic signal is f(t), the wavelet basis function ψ(t) is selected, and the original signal is decomposed by wavelet to obtain the wavelet coefficients at different scales a and displacements b:

[0072]

[0073] The wavelet coefficients are processed by setting the threshold λ = 0.3. for:

[0074] when hour,

[0075] when hour, Reconstruct the processed wavelet coefficients to obtain the denoised signal:

[0076]

[0077] in,

[0078]

[0079] in, is the Fourier transform of ψ(t);

[0080] Feature extraction: Extract the amplitude, frequency, energy, and duration features of the signal to construct a feature data set. The amplitude feature includes the maximum amplitude of the signal. and the average amplitude Where T is the duration of the signal;

[0081] Frequency characteristics are obtained by Fourier transforming the signal Get the spectrum of the signal and then calculate the main frequency where ω p is the angular frequency corresponding to the frequency component with the largest amplitude in the spectrum, and the frequency bandwidth Δf = f h -f t , f h and f t They are the high-frequency and low-frequency boundary frequencies corresponding to a certain energy proportion in the signal spectrum;

[0082] The energy feature uses the energy of the signal

[0083] The duration feature is the time interval T0 from the start to the end of the recording signal;

[0084] Model training and optimization: The support vector machine (SVM) algorithm is used for model training. During model training and optimization, the feature data set {x i ,y i}(where x i is the feature vector, y i ∈{-1,1} indicates whether it is precursor information), and the optimization function is constructed as follows:

[0085]

[0086] Among them, the constraint y i (ω·x i +b)≥1-ξ i ,ξ i ≥0(i=1,2,…,n), where ω is the normal vector of the hyperplane, b is the bias term, C is the penalty parameter, ξ i is the slack variable, and the optimal hyperplane parameter ω is obtained by solving * and b * , thus determining the classification model f(x)=sign(ω * ·x+b * ), during the training process, the cross-validation method is used to adjust the parameters to improve the generalization ability of the model;

[0087] The cross-validation method uses a 5-fold cross-validation method to randomly divide the data set into 5 mutually exclusive subsets F1, F 2, F 3, ,F 4, F5, for k = 1, 2, ..., 5 rounds, select all the k The other four subsets are used as training sets Train k =∪ i≠k F i , F k As the validation set Val k , using the training set Train k Train the SVM model, and the trained model is denoted as M k , for classification tasks, if the model M k In the validation set Val k The above prediction result is i represents the sample number, j represents the category number, and the true label is y ij , the accuracy calculation formula is as follows:

[0088]

[0089] For different C values, repeat the above 5-fold cross-validation process to obtain the average validation set performance index corresponding to each C value. The calculation formula is:

[0090]

[0091] Compare the average accuracy under different C values;

[0092] Model fusion: Use weighted average method to fuse and integrate multiple trained machine learning models. k Prediction results for test samples, final prediction results where α i It is model M i The weights can also be determined based on the model performance evaluation indicators, integrating the prediction information of multiple models.

[0093] 3. Establish a multi-factor coupling analysis and early warning model based on the multi-dimensional characteristics of microseismic signals, analyze the influence mechanism and degree of different factors on microseismic precursor information, and determine the early warning indicators and their thresholds;

[0094] Model construction: A multi-factor coupling analysis model is constructed based on the finite element method, considering the interaction between geological structure, groundwater, rock mechanical properties and engineering construction factors. In the model, the constitutive relationship of the rock and soil mass adopts the Mohr-Coulomb criterion, and its stress-strain relationship expression is: {σ}=[D]{ε}, where {σ} is the stress vector, [D] is the elastic matrix, and {ε} is the strain vector. A finite element model is established according to parameters such as the actual size of the tunnel, geological structure and material properties, and the tunnel and the surrounding rock and soil mass are divided into 10,000 units.

[0095] Simulation analysis and determination of early warning indicators: Simulate the stress-strain state and microseismic activity law of the rock and soil around the tunnel under different working conditions: set the groundwater level to change gradually from low to high (a total of 10 water level change levels are set, ranging from 1 meter to 10 meters in depth), and change the elastic modulus of the rock (set 5 different elastic modulus values, from 10GPa to 50GPa) and the excavation progress of the tunnel (simulate different stages from the initial to the middle and late stages of excavation, including 1000 meters of initial excavation, 2000 meters of mid-term excavation, and 2000 meters of late excavation). Through simulation, it is found that when the groundwater level rises faster than 0.5m / d and the energy change rate of the microseismic signal in a specific frequency band (80-120Hz) exceeds 30%, the risk of landslides increases significantly. A warning model is constructed by integrating multiple early warning indicators. A weighted comprehensive evaluation method is used to assign different weights to each early warning indicator according to its importance to the landslide disaster. The actual monitoring value of each indicator is compared with the corresponding threshold and normalized, and then weighted summed according to the weight to obtain a comprehensive early warning index.

[0096] 4. Real-time early warning and dynamic feedback: Based on the early warning indicators and their thresholds, dynamic assessment and early warning of the collapse risk of karst tunnels are carried out;

[0097] Data transmission and processing: A data transmission system is established to transmit multi-source data from the monitoring site to the data processing center in real time. The data transmission system adopts wired Ethernet transmission technology and follows the Modbus protocol. A data acquisition terminal is set up in the tunnel. After the data collected by the sensor is summarized and preliminarily processed, the data is transmitted to the data processing center outside the tunnel via Ethernet. The data processing center uses cloud computing technology to quickly process and analyze the data. The data is distributed and stored on 5 cloud storage nodes. Parallel computing task allocation is used to decompose the data processing tasks to 10 computing nodes for simultaneous processing. Resource scheduling optimization dynamically allocates computing resources, storage resources, and network resources according to task requirements and resource usage.

[0098] Early warning and feedback: Based on the real-time monitoring data and the calculation results of the early warning model, the risk of karst tunnel collapse is dynamically evaluated and warned. At a certain moment, the groundwater level is monitored to rise at a rate of 0.6m / d, and the energy change rate of microseismic signals in a specific frequency band is 35%. The data is input into the early warning model to calculate the comprehensive early warning index of 0.7, which exceeds the set early warning threshold of 0.6. The data processing center immediately sends out an early warning signal, and an audible and visual alarm is issued by the sound and light alarm device in the data processing center. At the same time, the early warning information is sent to the mobile phone of the construction site manager through the SMS notification system. The content of the SMS includes the warning location, the degree of danger, and the possible cause of the collapse. In addition, a warning pop-up window is popped up on the monitoring terminal at the construction site, showing detailed warning information and response suggestions. The construction site personnel respond according to the early warning. The construction is suspended upon receiving the warning information, and personnel are evacuated to a safe area. Emergency support is provided to key parts of the tunnel. Support measures include increasing the number of anchor rods and improving the strength of shotcrete. As the construction progresses and geological conditions change, the parameters in the early warning model are continuously updated and optimized. When the tunnel excavation enters a new geological area, the values ​​of the rock mechanics parameters in the model are adjusted according to the geological exploration data of the new area. The model parameters related to the support are adjusted according to the actual effect of the support structure during the construction process. In addition, the machine learning model is retrained according to the newly collected microseismic data and actual collapse cases, and the model weights are optimized, so that the early warning model can always adapt to the needs of identifying microseismic precursor information of karst tunnel collapse disasters in different construction stages and geological environments, and continue to provide accurate and reliable early warning services.

[0099] The present invention can accurately remove noise through wavelet transform, retain key features of microseismic signals, improve data quality, lay a foundation for subsequent accurate identification of precursor information, and extract multi-dimensional features including frequency, amplitude, phase, etc. from microseismic signals, realize multi-dimensional feature extraction, and comprehensively characterize the characteristics of microseismic events, thereby improving the accuracy of capturing precursor information related to landslide disasters and reducing the possibility of missed reports and false reports;

[0100] The present invention uses a machine learning algorithm to automatically learn and identify the microseismic patterns of different types of landslide disasters from a large amount of microseismic data. At the same time, it also combines geological conditions, construction progress and other factors for comprehensive classification and judgment, so that the recognition results are more in line with the actual situation, avoiding misjudgment caused by a single factor, and improving the accuracy and efficiency of recognition;

[0101] The present invention processes and analyzes the collected microseismic data in real time, and quickly gives the identification results of precursor information, which meets the needs of real-time monitoring and timely early warning during karst tunnel construction. It also has strong dynamic adaptability and can continuously update and optimize the identification model as the tunnel construction progresses and the geological conditions change, so as to always maintain accurate identification of microseismic precursor information and effectively cope with the complex and changeable construction environment.

[0102] The above specific embodiments are only several optional embodiments of the present invention. Based on the technical solutions of the present invention and the relevant inspirations of the above embodiments, those skilled in the art can make various alternative improvements and combinations to the above specific embodiments.

Claims

1. A method for identifying microseismic precursor information of karst tunnel collapse disaster, characterized in that: The following steps are involved: Build a microseismic monitoring system that integrates multi-source data; Processing and analyzing the microseismic signals collected by the microseismic detection system based on machine learning to extract multi-dimensional features of the microseismic signals; Based on the multi-dimensional characteristics of microseismic signals, a multi-factor coupling analysis and early warning model is established to analyze the influence mechanism and degree of different factors on microseismic precursor information, and determine the early warning indicators and their thresholds; Real-time early warning and dynamic feedback: Based on the early warning indicators and their thresholds, dynamic assessment and early warning are carried out on the collapse risk of karst tunnels.

2. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 1, characterized in that: The construction of a microseismic monitoring system integrating multi-source data comprises the following steps: Distributed optical fiber sensors are laid along the tunnel lining, vault, and sidewalls, and point sensors are arranged in areas with complex geological structures, high-risk rock crushing, as well as in tunnel entrances and exits and fault fracture zones to form a monitoring network; Geological data, groundwater data and construction data are collected synchronously, where geological data includes rock mechanics parameters and geological structure information, groundwater data includes water level changes and water flow velocity, and construction data includes excavation progress and support parameters.

3. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 1, characterized in that: The processing and analysis based on machine learning includes the following steps: Using wavelet transform to perform denoising and filtering preprocessing on the microseismic signals collected by the microseismic detection system; Extract the amplitude, frequency, energy, and duration features of the signal to construct a feature data set; A support vector machine is used to train and optimize the model by taking the feature data set as input and whether it is microseismic precursor information of landslide disaster as output label; The weighted average method is used to fuse and integrate multiple trained machine learning models.

4. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 3 is characterized in that: include: When using wavelet transform to denoise and filter the microseismic signal, the original microseismic signal is assumed to be f(t), the wavelet basis function ψ(t) is selected, and the original signal is decomposed by wavelet to obtain the wavelet coefficients at different scales a and displacements b: The wavelet coefficients are processed by setting the threshold λ. for: when hour, when hour, Reconstruct the processed wavelet coefficients to obtain the denoised signal: in, in, is the Fourier transform of ψ(t); Construct a feature data set, the amplitude feature includes the maximum amplitude of the signal and the average amplitude Where T is the duration of the signal; The frequency characteristics are obtained by Fourier transforming the signal Get the spectrum of the signal and then calculate the main frequency where ω p is the angular frequency corresponding to the frequency component with the largest amplitude in the spectrum, and the frequency bandwidth Δf = f h -f t , f h and f t They are the high-frequency and low-frequency boundary frequencies corresponding to a certain energy proportion in the signal spectrum; The energy feature uses the energy of the signal The duration characteristic is the time interval T0 from the start to the end of the recording signal; In model training and optimization, the feature dataset {x i ,y i }(where x i is the feature vector, y i ∈{-1,1} indicates whether it is precursor information), and the optimization function is constructed as follows: Among them, the constraint y i (ω·x i +b)≥1-ξ i ,ξ i ≥0(i=1,2,…,n), where ω is the normal vector of the hyperplane, b is the bias term, C is the penalty parameter, ξ i is the slack variable, and the optimal hyperplane parameter ω is obtained by solving * and b * , thus determining the classification model f(x)=sign(ω * ·x+b * ), during the training process, the cross-validation method is used to adjust the parameters; The weighted average method is used to fuse and integrate multiple trained machine learning models. k Prediction results for test samples, final prediction results where α i It is model M i The weights are determined according to the model performance evaluation indicators, and the prediction information of multiple models is integrated.

5. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 4, characterized in that: The cross-validation method adopts K-fold cross-validation method, and the validation process includes the following steps: Dataset division: Assume that the microseismic signal feature dataset contains n samples and randomly divide it into K mutually exclusive subsets F1, F 2, …,F K ,satisfy D represents the entire data set, and The number of samples in each subset is Training and validation process and performance index calculation: For k = 1, 2, ..., K rounds, select k The K-1 subsets outside of the above are used as training sets Train k =∪ i≠k F i , F k As the validation set Val k , using the training set Train k Train the machine learning model, and the trained model is denoted as M k , for classification tasks, if the model M k In the validation set Val k The above prediction result is i represents the sample number, j represents the category number, and the true label is y ij , the accuracy calculation formula is as follows: Where II(·) is an indicator function, which takes the value 1 when the condition in the brackets is met, otherwise it takes the value 0. k | indicates the validation set Val k The number of samples, for regression tasks, if the predicted value is The true value is y i , the mean square error calculation formula is as follows: Parameter adjustment and model selection: Set multiple values ​​for the model parameters to be adjusted. For each parameter value θ, repeat the training and verification process for K rounds to obtain the average performance index under the parameter value, including the average accuracy or the average mean square error. The calculation formulas are: Select the optimal parameter value θ based on the average performance index opt , so that for the classification task For regression tasks To improve the generalization ability of the model.

6. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 1, characterized in that: The establishment of the multi-factor coupling analysis and early warning model comprises the following steps: Construct a multi-factor coupling analysis model based on physical mechanics principles and numerical simulation methods; According to the interaction between geological structure, groundwater, rock mechanical properties and engineering construction factors, the stress-strain state and microseismic activity of the rock and soil around the tunnel under different working conditions are simulated by data simulation method; The influence of different factors on microseismic precursor information is analyzed through a multi-factor coupling analysis model, and the early warning indicators and their thresholds are determined.

7. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 5, characterized in that: The numerical simulation method adopts the finite element method or the discrete element method. During the simulation process, the constitutive relationship of the rock and soil body adopts the Mohr-Coulomb criterion or the Drucker-Prager criterion, and its stress-strain relationship expression is: {σ}=[D]{ε}, where {σ} is the stress vector, [D] is the elastic matrix, and {ε} is the strain vector.

8. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 1, characterized in that: The real-time early warning and dynamic feedback include: Establish a data transmission system to transmit multi-source data from the monitoring site to the data processing center in real time; Use cloud computing or edge computing technology to quickly process and analyze data in the data processing center; Dynamic assessment and early warning are carried out on the collapse risk of karst tunnels. When the microseismic precursor information exceeds the early warning threshold, an early warning signal is issued and fed back to the construction site.

9. The method for identifying microseismic precursor information of karst tunnel collapse disaster according to claim 7, characterized in that: The data transmission system adopts wireless transmission technology or wired transmission technology, and the data transmission protocol follows a specific industrial standard protocol, including the Modbus protocol or the OPC UA protocol. The application of cloud computing or edge computing technology in data processing includes data distributed storage, parallel computing task allocation, and resource scheduling optimization.

10. A method for identifying microseismic precursor information of karst tunnel collapse disaster according to any one of claims 1 to 9, characterized in that: During the construction process, as the tunnel construction progresses and geological conditions change, the parameters in the early warning model are updated and optimized, including adjusting the early warning indicator thresholds and retraining the machine learning model weights to adapt to the needs of identifying microseismic precursor information of karst tunnel collapse disasters in different construction stages and geological environments.

Images