Tunnel rock burst type prediction method and system based on multi-dimensional mechanism fusion, and medium

Abstract

The application discloses a tunnel rock burst type prediction method and system based on multi-dimensional mechanism fusion, and a medium; relates to the technical field of tunnel rock burst; dynamic acquisition of multi-dimensional basic information in the tunnel construction process; according to the microseismic information fusion matrix tensor inversion, the first proportion of different rupture modes of the microseismic event is obtained, and the second proportion of different rupture modes of each acoustic emission event is obtained by energy evolution of acoustic emission information; the dynamic damage weight of different rupture modes is comprehensively obtained; based on the dynamic damage weight and the multi-dimensional basic information, spatial clustering analysis is carried out, a rock burst type probability model of different rock burst types is constructed, and the probability of different rock burst types is predicted; the scheme combines real-time multi-dimensional basic information to construct a rock burst type probability model, realizes dynamic discrimination and rock burst type prediction of the rock burst type, and provides technical support for deep-buried tunnel construction safety.

Description

Technical Field

[0001] This invention relates to the field of tunnel rockburst technology, specifically to a method, system, and medium for predicting tunnel rockburst types based on multi-dimensional mechanism fusion. Background Technology

[0002] Rockbursts are dynamic disasters caused by the sudden release of energy from high-stress rock masses in deeply buried underground engineering projects. They are complex in type and highly destructive. Existing rockburst early warning technologies mostly rely on single microseismic monitoring data, making it difficult to accurately distinguish between different types of rockbursts, such as strain-type and tectonic-type rockbursts, resulting in a lack of targeted protective measures.

[0003] In addition, traditional methods have the following limitations: (1) The data dimension is singular, relying only on the statistics of microseismic events and ignoring key information such as acoustic emission spectrum and structural geometry, which cannot fully analyze the rupture mechanism; (2) Static analysis is lagging, based on the statistics of the proportion of damage in a fixed time window, which makes it difficult to capture the spatiotemporal evolution law of rockburst; (3) The model has poor universality: the existing prediction model is not adaptable to complex geological conditions and lacks physical mechanism support, and has weak interpretability; (4) Passive early warning mode: it cannot link construction equipment to achieve active prevention and control, and the risk response efficiency is low.

[0004] In recent years, although multi-source data fusion and machine learning methods have been introduced into the field of rockburst prediction, challenges such as insufficient multi-sensor collaboration, inadequate dynamic feature extraction, and lack of physical constraints still exist. Summary of the Invention

[0005] The technical problem this invention aims to solve is that while traditional methods of multi-source data fusion and machine learning have been introduced into the field of rockburst prediction, they still face challenges such as insufficient multi-sensor collaboration, inadequate dynamic feature extraction, and lack of physical constraints, resulting in defects in the accuracy of rockburst type prediction. This invention aims to provide a method, system, and medium for predicting tunnel rockburst types based on multi-dimensional mechanism fusion. It improves upon traditional rockburst prediction techniques by integrating a network of multi-source data monitoring systems, including microseismic, acoustic emission, and ground-penetrating radar, to collect multi-dimensional basic information, constructing a probabilistic model of rockburst types, and achieving dynamic identification and proactive prevention of rockburst types, thus providing technical assurance for the safety of deep-buried tunnel construction.

[0006] This invention is achieved through the following technical solution:

[0007] This solution provides a tunnel rockburst type prediction method based on multi-dimensional mechanism fusion, including:

[0008] A multi-source data monitoring system is constructed in the tunnel construction disturbance zone to dynamically collect multi-dimensional basic information during the tunnel construction process. The multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area.

[0009] The first proportion of different rupture modes of microseismic events is derived from the fusion moment tensor of microseismic information. The second proportion of different rupture modes of each acoustic emission event is obtained by energy evolution of acoustic emission information. The dynamic damage weights of different rupture modes are generated by combining the first and second proportions.

[0010] Based on dynamic damage weights and multidimensional basic information, spatial clustering analysis is performed to obtain a comprehensive judgment model for different rockburst types;

[0011] A rockburst type probability model is constructed based on a comprehensive judgment model. The multidimensional basic information of the microseismic event to be predicted is input into the rockburst type probability model to calculate the probability of different rockburst types.

[0012] A further optimized scheme is to deduce the first proportion of different rupture modes of microseismic events based on the fusion moment tensor of microseismic information; including the following method:

[0013] The source of the microseismic event is obtained, and the source is approximated by the moment tensor analysis method to obtain the equivalent matrix containing the moment tensor.

[0014] The moment tensor in the equivalent force matrix is ​​regarded as the interaction of pure double couple, compensated linear vector dipole and isotropic component, so as to establish the correlation R between the moment tensor and the rupture mode;

[0015] The rupture mode of each microseismic event is determined based on the correlation R and the double couple component of the moment tensor.

[0016] The percentage of different rupture modes was calculated and used as the first percentage.

[0017] A further optimized scheme is that the correlation R between the moment tensor and the rupture mode includes:

[0018] ;

[0019] m i *= M i - tr ( M ) / 3;

[0020] in, tr ( M ) represents the moment tensor M traces; m i * Represents the eigenvalues ​​of a partial tensor; M i Representing the moment tensor M The i Each feature value.

[0021] A further optimized scheme involves obtaining a second proportion of different breakup modes for each acoustic emission event by performing energy evolution on the acoustic emission information; including the following method:

[0022] First, the acoustic emission information is bandpass filtered, and then the short-time Fourier transform of the bandpass filtered acoustic emission information is performed to obtain the acoustic emission time spectrum.

[0023] Extract the main frequency component from the acoustic emission time spectrum: Configure different identification frequency peak energy ranges for different fracture modes, and statistically calculate the proportion of identification frequency peaks in different identification frequency peak energy ranges in the acoustic emission time spectrum as the second proportion.

[0024] A further optimized scheme involves generating dynamic damage weights for different fracture modes by combining the first and second proportions; including the following method:

[0025] The microseismic information weights are determined based on the inversion confidence level during the first proportion inversion process, and the acoustic emission information weights are determined based on the deviation threshold of the identified peak energy proportions during the second proportion evolution process. The microseismic information weights and acoustic emission information weights are then combined to obtain the comprehensive weight W. 综合 :

[0026] ;

[0027] Among them, W MS N represents the weight of microseismic information; MS W represents the total number of microseismic events. AE N represents the weight of acoustic emission information. AE Indicates the total number of acoustic emission events;

[0028] Set a statistical period, and within the statistical period, calculate the proportion of different failure modes based on comprehensive weights, and dynamically calculate the dynamic failure weight of different failure modes according to time windows.

[0029] A further optimized solution involves performing spatial clustering analysis based on dynamic damage weights and multidimensional basic information to obtain a comprehensive judgment model for different rockburst types; including the following methods:

[0030] Each microseismic event is considered as a microseismic cluster. The spatial clustering areas of each microseismic cluster are identified, and the positional relationship between the microseismic cluster and the structural plane and fault is used as the spatial determination criterion for different rockburst types.

[0031] The proportion of different fracturing modes is determined based on dynamic failure weights, which serve as the criteria for determining the fracturing mode of different rockburst types.

[0032] The ratio of rock mass strength to in-situ stress or the angle between the direction of the maximum principal stress and the orientation of the structural plane are used as mechanical criteria for different rockburst types.

[0033] A further optimized solution is to construct a rockburst type probability model based on a comprehensive judgment model; including the following methods:

[0034] A graph neural network is constructed with each microseismic cluster as a graph node; the edge weights of the graph neural network are determined by the spatial distance between microseismic clusters and whether the microseismic clusters are connected to the structural surface; the output layer probability distribution of the graph neural network is as follows:

[0035] ;

[0036] Where P(k) represents the probability of rockburst type k; e represents the natural base; zk represents the logit value of rockburst type k; j represents the index of rockburst type (from 1 to 3); K represents the total number of rockburst types (3 types: structural, strain, and mixed); and zj represents the logit value of index j.

[0037] A further optimized scheme is that the rockburst types include: tectonic rockburst, strain-induced rockburst, and hybrid rockburst; and the fracture modes include: tensile fracture, shear fracture, and compression fracture.

[0038] This solution also provides a tunnel rockburst type prediction system based on multi-dimensional mechanism fusion, used to implement the aforementioned tunnel rockburst type prediction method based on multi-dimensional mechanism fusion. The system includes:

[0039] The data acquisition module is used to construct a multi-source data monitoring system in the tunnel construction disturbance zone to dynamically acquire multi-dimensional basic information during the tunnel construction process. The multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area.

[0040] The inversion and evolution module is used to invert the first proportion of different rupture modes of microseismic events based on the fusion moment tensor of microseismic information, and to perform energy evolution on acoustic emission information to obtain the second proportion of different rupture modes of each acoustic emission event; and to generate dynamic damage weights for different rupture modes by combining the first and second proportions.

[0041] The analysis module is used to perform spatial clustering analysis based on dynamic damage weights and multidimensional basic information to obtain a comprehensive judgment model for different rockburst types.

[0042] The calculation module is used to construct a rockburst type probability model based on the comprehensive judgment model. It inputs the multidimensional basic information of the microseismic event to be predicted into the rockburst type probability model to calculate the probability of different rockburst types.

[0043] This solution also provides a computer-readable medium having a computer program stored thereon, which, when executed by a processor, can implement the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion as described above.

[0044] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0045] 1. The tunnel rockburst type prediction method, system and medium based on multi-dimensional mechanism fusion provided by this invention improves the traditional rockburst prediction technology by integrating a multi-source data monitoring system network of microseismic, acoustic emission and ground-penetrating radar to collect multi-dimensional basic information, and constructing a rockburst type probability model by combining real-time multi-dimensional basic information, so as to realize dynamic identification and prediction of rockburst type, and provide technical guarantee for the safety of deep buried tunnel construction.

[0046] 2. The present invention provides a method, system, and medium for predicting tunnel rockburst types based on multi-dimensional mechanism fusion; it extracts the proportion of fracture modes through moment tensor inversion (microseismic) and energy evolution analysis (acoustic emission), unifying data from different sensors to the "fracture mechanism" dimension, avoiding the bias caused by simple numerical superposition;

[0047] 3. The present invention provides a method, system, and medium for predicting tunnel rockburst types based on multi-dimensional mechanism fusion; based on microseismic moment tensor inversion and acoustic emission energy evolution, the dynamic weights of different fracture modes are calculated in real time (e.g., shear fracture accounts for 60% → high rockburst risk), and the weights are updated with the construction process to reflect the cumulative effect of rock mass damage; geological parameters such as rock mass strength participate in spatial clustering, and spatial clustering analysis is performed based on the location of microseismic events and dynamic weights to avoid misjudgment of isolated points and ensure that the prediction results are consistent with geological conditions. Attached Figure Description

[0048] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:

[0049] Figure 1 This is a schematic diagram of the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.

[0051] Traditional methods combining multi-source data fusion and machine learning have been introduced into the field of rockburst prediction, but challenges such as insufficient multi-sensor collaboration, inadequate dynamic feature extraction, and lack of physical constraints still exist, leading to deficiencies in the accuracy of rockburst type prediction. Therefore, this solution provides the following embodiments to address the aforementioned technical problems:

[0052] Example 1: This example provides a tunnel rockburst type prediction method based on multi-dimensional mechanism fusion, such as... Figure 1 As shown, it includes:

[0053] Step 1: Construct a multi-source data monitoring system in the tunnel construction disturbance zone to dynamically collect multi-dimensional basic information during the tunnel construction process; the multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area;

[0054] The multi-source data monitoring system integrates a network of microseismic sensors, acoustic emission sensors, and geological sensors (3D laser scanners) to collect multi-dimensional basic information in real time during tunnel construction. Specifically, the microseismic sensor array consists of at least six triaxial microseismic sensors (covering the transverse, longitudinal, and vertical directions) evenly distributed along the tunnel cross-section within a range of 30-50m behind the tunnel face, ensuring three-dimensional capture of microseismic events across the entire cross-section. The acoustic emission sensors are deployed in potential fracture zones (such as structural plane intersections and high-stress concentration areas), with one sensor spaced every 10m, for a total of 4-6 sensors. The 3D laser scanner uses a track-mounted ground-penetrating radar installed behind the tunnel face, which advances to scan the geological structure within a 20m range ahead as excavation progresses. Every 20m of construction advance, the sensors 30m ahead of the tunnel face are moved forward in batches to maintain coverage of the construction disturbance area. A fiber optic + 5G redundant network is used to transmit microseismic, acoustic emission, and 3D laser scan information back in real time.

[0055] Further geological information of the tunnel face area includes: fault distribution information, geometric characteristics of structural planes, and rock mass mechanical parameters, as shown in Table 1:

[0056] Table 1 Geological information of the working face area

[0057]

[0058] Step 2: Based on the fusion moment tensor of microseismic information, the first proportion of different rupture modes of microseismic events is derived, and the energy evolution of acoustic emission information is performed to obtain the second proportion of different rupture modes of each acoustic emission event; the dynamic failure weights of different rupture modes are generated by combining the first and second proportions; the rupture modes include: tensile rupture, shear rupture and compression rupture.

[0059] The method for retrieving the first proportion of different rupture modes of microseismic events based on the fusion moment tensor of microseismic information includes:

[0060] The sources of microseismic events are obtained, and the moment tensor analysis method is used to make an equivalent approximation of the sources, resulting in an equivalent effectiveness matrix containing the moment tensor; the equivalent effectiveness matrix is ​​expressed as: u=GM In the formula, u The vector representing far-field displacement. M It is a moment tensor. G Let be the Green's function. If we consider the earthquake source as a point source in time, then the generation of microfractures can be viewed as the moment tensor acting over time. t Afterwards, x The displacement generated at that point. The equivalent matrix is ​​then expressed as:

[0061] ;

[0062] in, x k Components representing spatial coordinates; x 0 Indicates the initial location of the earthquake source; t 0 Indicates the time when the epicenter event occurred; M ij Represents the components of the moment tensor; G ij Represents the components of the Green's function; u i (x,t) represents treating the earthquake source as a point source in time, with the generation of microfractures considered as the moment tensor acting over time. t Afterwards, x The equivalent matrix of the displacement generated at that location;

[0063] In actual field conditions, the formation of micro-fractures is a continuous process, and the displacement is a function of time. When only considering... p When the far-field displacement occurs, we have:

[0064] ;

[0065] In the formula: u i It is relative to the first i The first movement of the P wave on each axis, r i This represents the radial component from the earthquake source to the sensor. r p This represents the lateral component from the earthquake source to the sensor. r q This represents the longitudinal component from the earthquake source to the sensor. x 0 Indicates the initial location of the earthquake source; R It is the distance from the epicenter. ρ It is the density of the rock.

[0066] Expanded to ;

[0067] in, Indicates the epicenter number n Each decomposition mode corresponds to a Green's function of the b-source component under the q-field component; q=1,2,3; b=1,2,3; n=1,2,...,N;

[0068] The moment tensor in the equivalent force matrix is ​​regarded as the interaction of pure double couple, compensated linear vector dipole and isotropic component, so as to establish the correlation R between the moment tensor and the rupture mode;

[0069] ;

[0070] M DC Indicates a pure two-force couple; M CLVD Indicates a compensated linear vector dipole; M ISO X represents the coefficient of the isotropic component; Y represents the coefficient of the compensated linear vector dipole; Z represents the coefficient of the isotropic component.

[0071] The relationship R between the moment tensor and the rupture mode includes:

[0072] ;

[0073] m i *= M i - tr ( M ) / 3;

[0074] in, tr ( M ) represents the moment tensor M traces; m i * Represents the eigenvalues ​​of a partial tensor; M i Representing the moment tensor M The i Each feature value.

[0075] The rupture mode of each microseismic event is determined based on the correlation R and the double couple component of the moment tensor.

[0076] The results for the correlation R and the two-couple components of the moment tensor C DC (The relative faulting mechanism representing shear fracture of rock mass or fault) will generate different criteria, and the corresponding failure mode judgments are shown in Table 2:

[0077] Table 2 Damage Mode Judgment

[0078] Mechanical state <![CDATA[ C DC values]]> value Rupture mode Pure tension 0% -100% Tension fracture Shear-tensioning 40% -43% Tension fracture Pure cut 60% -30% Shear fracture Shear-compression 100% 0% Shear fracture Shear-compression 60% 30% Compression fracture Shear-compression 40% 43% Compression fracture Pure compression 0% 100% Compression fracture

[0079] The percentage of different rupture modes was calculated and used as the first percentage.

[0080] The method for obtaining the second proportion of different breakup modes for each acoustic emission event by performing energy evolution on acoustic emission information includes:

[0081] First, the acoustic emission information is bandpass filtered (20kHz-1MHz) to remove low-frequency noise. Then, the acoustic emission information after bandpass filtering is subjected to short-time Fourier transform to obtain the acoustic emission time spectrum.

[0082] Extract the main frequency component from the acoustic emission time spectrum: Configure different identification frequency peak energy ranges for different fracture modes, and statistically calculate the proportion of identification frequency peaks in different identification frequency peak energy ranges in the acoustic emission time spectrum as the second proportion.

[0083] For example, calculate the spectral energy range and identify the dominant frequency peak (e.g., high frequency peak > 100kHz, low frequency peak < 50kHz).

[0084] The characteristics of tensile fracture include: a high proportion of high-frequency energy (>70%), corresponding to brittle crack propagation;

[0085] The characteristics of shear fracture include: a high proportion of low-frequency energy (>60%), corresponding to frictional slip.

[0086] The proportion of high-frequency spectrum energy is: The proportion of low-frequency energy is: If α > 0.7, it is marked as tension-dominant; if β > 0.6, it is marked as shear-dominant.

[0087] The dynamic damage weights for different fracture modes are generated by combining the first and second proportions; including the method:

[0088] For microseismic events, when the microseismic sensor detects a waveform amplitude exceeding a threshold, event recording is triggered. The moment tensor is used to calculate the rupture type (tensile, shear, compression) in real time. Each microseismic event outputs a failure mode label and confidence level (e.g., shear percentage > 60% is labeled "shear"). For acoustic emission events, the dominant frequency energy ratio (high frequency / low frequency) is extracted using Fast Fourier Transform (FFT); high frequency energy percentage > 70% is labeled as tensile rupture; low frequency energy percentage > 60% is labeled as shear rupture; intermediate states are labeled as mixed rupture (compression or a combination of tensile and shear).

[0089] The microseismic information weight is determined based on the inversion confidence level in the first proportion inversion process (e.g., when the inversion confidence level is >60%, the microseismic information weight = 0.8; in practical applications, the specific weight should be set based on experience). The acoustic emission information weight is determined based on the deviation of the peak energy proportion from the threshold in the second proportion evolution process (e.g., when α > 0.7, the weight = 0.7; in practical applications, the specific weight should be set based on experience). The microseismic information weight and the acoustic emission information weight are combined to obtain the comprehensive weight W. 综合 :

[0090] ;

[0091] Among them, W MS N represents the weight of microseismic information; MS W represents the total number of microseismic events. AE N represents the weight of acoustic emission information. AE Indicates the total number of acoustic emission events;

[0092] Set a statistical period, and within the statistical period, calculate the proportion of different failure modes based on comprehensive weights, and dynamically calculate the dynamic failure weight of different failure modes according to time windows.

[0093] The percentages of different fracture modes are as follows:

[0094]

[0095]

[0096]

[0097] Among them, P 张拉 Indicates the proportion of tension-induced fracture mode; P 剪切 P represents the proportion of shear fracture modes; 压缩 N represents the proportion of compression fracture modes; 张拉 N represents the total number of tension fracture mode events in a single rockburst; 剪切 N represents the total number of shear fracture modes in a single rockburst event; 压缩 N represents the total number of compression fracture mode events in a single rockburst; 总 Represents the total of all break mode events;

[0098] Mapping the percentage to a weighting coefficient, the dynamically disrupted weight is:

[0099]

[0100]

[0101]

[0102] Among them, W张拉 W represents the dynamic failure weight of the tension fracture mode; 剪切 W represents the dynamic failure weight of the shear fracture mode; 压缩 The dynamic failure weight of the compression fracture mode is represented by α, β, and γ, which are the tensile empirical coefficient, shear empirical coefficient, and compression empirical coefficient, respectively (e.g., α=1.2 in strain-type rockburst and β=1.5 in tectonic rockburst).

[0103] Step 3: Based on dynamic damage weights and multidimensional basic information, perform spatial clustering analysis to obtain a comprehensive judgment model for different rockburst types; the rockburst types include: tectonic rockburst, strain-induced rockburst, and mixed rockburst; this step specifically includes the following methods:

[0104] Each microseismic event is considered as a microseismic cluster. The spatial clustering areas of each microseismic cluster are identified, and the positional relationship between the microseismic cluster and the structural plane and fault is used as the spatial determination criterion for different rockburst types.

[0105] Specifically, for each microseismic cluster, the distance D between its geometric center and the nearest structural surface is calculated. If D ≤ 2m and the cluster density > 5 events / m³, it is marked as a "structural surface-related cluster".

[0106] High-roughness (roughness JRC>12) structural surfaces have high frictional resistance and are prone to accumulating elastic strain energy, triggering microseismic events of shear fracture.

[0107] Establish a linear regression model for the proportion of shear fracture: ;in Indicates the proportion of shear fracture; shear fracture proportion; 'a' represents the first coefficient, and 'b' represents the second coefficient;

[0108] Structural surfaces with steep dip angles (dip angle v > 60°) are prone to shear slip, triggering shear fracture; structural surfaces with gentle dip angles (dip angle v < 30°) are mainly prone to tensile fracture, triggering tensile fracture. The type of rockburst is determined based on roughness and dip angle. If the dip angle v > 60° and the proportion of shear fracture > 60%, it is identified as a tectonic rockburst.

[0109] The density of microseismic clusters within the fault zone is significantly higher than that in the surrounding rock area (density difference > 2 times); shear failure accounts for a high proportion (> 70%) near strike-slip faults; and tension failure accounts for a high proportion (> 50%) in normal fault areas.

[0110] The proportion of different fracturing modes is determined based on dynamic failure weights, which serve as the criteria for determining the fracturing mode of different rockburst types.

[0111] The ratio of rock mass strength to in-situ stress (SSR) or the angle between the direction of the maximum principal stress (σ1) and the strike of the structural plane is used as the mechanical criteria for different rockburst types.

[0112] Specifically, in areas where the rock mass strength to in-situ stress ratio (SSR) is low (SSR<2), the rock mass is prone to brittle failure (triggering strain-induced rockburst); in areas where the rock mass strength to in-situ stress ratio (SSR>4), energy accumulation is slow, with shear failure controlled by structural planes as the main cause (triggering tectonic rockburst). Regarding the direction of in-situ stress and microseismic distribution, if the angle between the direction of the maximum principal stress (σ1) and the strike of the structural plane is <30°, the structural plane is easily activated, triggering tectonic rockburst.

[0113] The comprehensive determination results of rockburst type mechanism are shown in Table 3:

[0114] Table 3 Comprehensive Determination of Rockburst Types and Mechanisms

[0115] Rockburst type Spatial conditions Destruction Mode Mechanical conditions Structural rockburst Microseismic clusters are distributed along structural planes or faults (distance ≤ 2m). Shearing percentage > 60% <![CDATA[SSR>4 or σ1 direction is nearly parallel to the orientation of the structural plane. Strain-type rockburst Microseismic clusters are scattered and far from structural surfaces. Tensioning ratio > 50% <![CDATA[The SSR < 2 or the σ1 direction intersects the strike of the structural plane at a large angle.]]> Hybrid rockburst Microseismic clusters are partly located near structural planes and partly located within intact rock masses. Tension, shear, and compression failures all accounted for more than 30% of the total failure rate. \

[0116] Step 4: Construct a rockburst type probability model based on the comprehensive judgment model. Input the multidimensional basic information of the microseismic event to be predicted into the rockburst type probability model to calculate the probability of different rockburst types.

[0117] The method for constructing a rockburst type probability model based on a comprehensive judgment model includes:

[0118] The rockburst type probability model in this scheme is a physical information graph neural network hybrid driven model, which is obtained by fusing graph neural networks with physical constraints (such as elasticity equations and energy conservation). Its input three-dimensional features are microseismic event rupture type (tension / shear / compression), energy release rate (dE / dt), and acoustic emission spectrum energy ratio; the edge features are microseismic event spatial distance, structural surface roughness (JRC), and rock mass strength (UCS).

[0119] A graph neural network is constructed with each microseismic cluster as a graph node; the graph node features include failure mode, energy, and spectral parameters.

[0120] The edge weights of the graph neural network are determined by the spatial distance between microseismic clusters and whether the microseismic clusters are connected to the structural surface (e.g., the closer the spatial distance and the stronger the correlation between the microseismic clusters and the structural surface, the higher the edge weights); the probability distribution of the output layer of the graph neural network is as follows:

[0121]

[0122] Where P(k) represents the probability of rockburst type k; e represents the natural base; zk represents the logit value of rockburst type k; j represents the index of rockburst type (from 1 to 3); and K represents the total number of rockburst types (3 types: structural, strain, and mixed).

[0123] Rebuild the graph data and run the model every 5 minutes following the steps above to ensure that the prediction results are updated in real time as the construction progresses.

[0124] Finally, based on the calculated probability of different rockburst types, the construction equipment is linked to achieve proactive risk prevention and control. The probability of different rockburst types is converted into prevention and control instructions, and the construction equipment is linked to reduce risks.

[0125] Example 2: This example provides a tunnel rockburst type prediction system based on multi-dimensional mechanism fusion, used to implement the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion described in Example 1. The system includes:

[0126] The data acquisition module is used to construct a multi-source data monitoring system in the tunnel construction disturbance zone to dynamically acquire multi-dimensional basic information during the tunnel construction process. The multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area.

[0127] The inversion and evolution module is used to invert the first proportion of different rupture modes of microseismic events based on the fusion moment tensor of microseismic information, and to perform energy evolution on acoustic emission information to obtain the second proportion of different rupture modes of each acoustic emission event; and to generate dynamic damage weights for different rupture modes by combining the first and second proportions.

[0128] The analysis module is used to perform spatial clustering analysis based on dynamic damage weights and multidimensional basic information to obtain a comprehensive judgment model for different rockburst types.

[0129] The calculation module is used to construct a rockburst type probability model based on the comprehensive judgment model. It inputs the multidimensional basic information of the microseismic event to be predicted into the rockburst type probability model to calculate the probability of different rockburst types.

[0130] Example 3: This example provides a computer-readable medium storing a computer program, which, when executed by a processor, can implement the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion as described in Example 1; specifically, the following steps are performed:

[0131] Step 1: Dynamically collect multi-dimensional basic information during tunnel construction; the multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area;

[0132] Step 2: Based on the fusion moment tensor of microseismic information, the first proportion of different rupture modes of microseismic events is derived. The energy evolution of acoustic emission information is performed to obtain the second proportion of different rupture modes of each acoustic emission event. The dynamic damage weights of different rupture modes are generated by combining the first and second proportions.

[0133] Step 3: Based on dynamic damage weights and multidimensional basic information, perform spatial clustering analysis to obtain a comprehensive judgment model for different rockburst types;

[0134] Step 4: Construct a rockburst type probability model based on the comprehensive judgment model. Input the multidimensional basic information of the microseismic event to be predicted into the rockburst type probability model to calculate the probability of different rockburst types.

[0135] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A tunnel rockburst type prediction method based on multi-dimensional mechanism fusion, characterized in that, include: A multi-source data monitoring system was constructed in the tunnel construction disturbance zone to dynamically collect multi-dimensional basic information during the tunnel construction process. The multidimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the working face area; The first proportion of different rupture modes of microseismic events is derived from the fusion moment tensor of microseismic information. The second proportion of different rupture modes of each acoustic emission event is obtained by energy evolution of acoustic emission information. The dynamic damage weights of different rupture modes are generated by combining the first and second proportions. Based on dynamic failure weights and multidimensional basic information, spatial clustering analysis is performed to obtain a comprehensive judgment model for different rockburst types. Specifically, the method includes: treating each microseismic event as a microseismic cluster, identifying the spatial clustering areas of each microseismic cluster, and using the positional relationship between the microseismic cluster and the structural plane and fault as the spatial judgment condition for different rockburst types; determining the proportion of different fracturing modes based on dynamic failure weights as the fracturing mode judgment condition for different rockburst types; and using the ratio of rock mass strength to in-situ stress or the angle between the direction of the maximum principal stress and the strike of the structural plane as the mechanical judgment condition for different rockburst types. A rockburst type probability model is constructed based on a comprehensive judgment model. The multidimensional basic information of the microseismic event to be predicted is input into the rockburst type probability model to calculate the probability of different rockburst types. The method for constructing a rockburst type probability model based on a comprehensive judgment model includes: A graph neural network is constructed with each microseismic cluster as a graph node; the edge weights of the graph neural network are determined by the spatial distance between microseismic clusters and whether the microseismic clusters are connected to the structural surface; the output layer probability distribution of the graph neural network is as follows: ； Where P(k) represents the probability of rockburst type k; e represents the natural base; zk represents the logit value of rockburst type k; j represents the index of rockburst type; K represents the total number of rockburst types; and zj represents the logit value of index j. The rockburst types include: tectonic rockburst, strain-induced rockburst, and mixed rockburst; the fracture modes include: tensile fracture, shear fracture, and compression fracture.

2. The tunnel rockburst type prediction method based on multi-dimensional mechanism fusion according to claim 1, characterized in that, The first proportion of different rupture modes of microseismic events is derived by fusing the moment tensor with microseismic information; Including methods: The source of the microseismic event is obtained, and the source is approximated by the moment tensor analysis method to obtain the equivalent matrix containing the moment tensor. The moment tensor in the equivalent force matrix is ​​regarded as the interaction of pure double couple, compensated linear vector dipole and isotropic component, so as to establish the correlation R between the moment tensor and the rupture mode; The rupture mode of each microseismic event is determined based on the correlation R and the double couple component of the moment tensor. The percentage of different rupture modes was calculated and used as the first percentage.

3. The tunnel rockburst type prediction method based on multi-dimensional mechanism fusion according to claim 2, characterized in that, The relationship R between the moment tensor and the rupture mode includes: ； m i *= M i - tr ( M ) / 3； in, tr ( M ) represents the moment tensor M traces; m i * Represents the eigenvalues ​​of a partial tensor; M i Representing the moment tensor M The i Each feature value.

4. The tunnel rockburst type prediction method based on multi-dimensional mechanism fusion according to claim 1, characterized in that, The method for obtaining the second proportion of different breakup modes for each acoustic emission event by performing energy evolution on acoustic emission information includes: First, the acoustic emission information is bandpass filtered, and then the short-time Fourier transform of the bandpass filtered acoustic emission information is performed to obtain the acoustic emission time spectrum. Extract the main frequency component from the acoustic emission time spectrum: Configure different identification frequency peak energy ranges for different fracture modes, and statistically calculate the proportion of identification frequency peaks in different identification frequency peak energy ranges in the acoustic emission time spectrum as the second proportion.

5. The tunnel rockburst type prediction method based on multi-dimensional mechanism fusion according to claim 4, characterized in that, The combined first and second proportions generate dynamic damage weights for different fracture modes; Including methods: The microseismic information weights are determined based on the inversion confidence level during the first proportion inversion process, and the acoustic emission information weights are determined based on the deviation threshold of the identified peak energy proportions during the second proportion evolution process. The microseismic information weights and acoustic emission information weights are then combined to obtain the comprehensive weight W. 综合 : ; Among them, W MS N represents the weight of microseismic information; MS W represents the total number of microseismic events. AE N represents the weight of acoustic emission information. AE This indicates the total number of acoustic emission events; Set a statistical period, and within the statistical period, calculate the proportion of different failure modes based on comprehensive weights, and dynamically calculate the dynamic failure weight of different failure modes according to time windows.

6. A tunnel rockburst type prediction system based on multi-dimensional mechanism fusion, characterized in that, The system is used to implement the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion as described in any one of claims 1-5, the system comprising: The data acquisition module is used to construct a multi-source data monitoring system in the tunnel construction disturbance zone to dynamically acquire multi-dimensional basic information during the tunnel construction process. The multi-dimensional basic information includes: microseismic information, acoustic emission information, three-dimensional laser scanning information, and geological information of the tunnel face area. The inversion and evolution module is used to invert the first proportion of different rupture modes of microseismic events based on the fusion moment tensor of microseismic information, and to perform energy evolution on acoustic emission information to obtain the second proportion of different rupture modes of each acoustic emission event; and to generate dynamic damage weights for different rupture modes by combining the first and second proportions. The analysis module is used to perform spatial clustering analysis based on dynamic damage weights and multidimensional basic information to obtain a comprehensive judgment model for different rockburst types. The calculation module is used to construct a rockburst type probability model based on the comprehensive judgment model. It inputs the multidimensional basic information of the microseismic event to be predicted into the rockburst type probability model to calculate the probability of different rockburst types.

7. A computer-readable medium having a computer program stored thereon, characterized in that, The computer program, when executed by a processor, can implement the tunnel rockburst type prediction method based on multi-dimensional mechanism fusion as described in any one of claims 1-5.

Images